Sie sind auf Seite 1von 88

Congressional Media Strategy: House Members, Rhetorical Strategies and Press

Releases.
Political Science Distinction Thesis, 2012-2013
Brittany R. Legasey



































2
On February 24
th
, 2011 in Houston Texas, a fire destroyed the home-based daycare run
by Jessica Tata. Tata returned to the daycare from a shopping trip to find the home ablaze and
the children still inside. Four toddlers died as a result of the fire, and three others were injured.
Initially, Tata was not charged, but new developments soon made prosecutors reevaluate Tatas
role in the situation. Tata fled the country for Nigeria a few days after the fire occurred.
Congresswoman Sheila Jackson Lee, of Houstons 18
th
Congressional District, distributed a
press release on March 8
th
clarifying her stance on the situation and the efforts she has made in
attempt to bring Tata to justice. Congresswoman Jackson Lees press release said:
I therefore urgently request that the State Department and the Department of Justice
officially send additional U.S. law enforcement officers to Nigeria, and that they engage
the Nigerian government and urgently ask for their immediate and intensified efforts to
find, detain and arrest this alleged perpetrator who has left 4 dead babies in Harris County
and some 3 other babies fighting for their lives while their parents and loved ones suffer
in a state of excruciating anguish. (2011)
When Tata returned to the U.S., she was arrested and charged with four counts of felony
murder, four counts of abandoning a child and one count of reckless injury to a child. In 2012,
Tata was convicted of felony murder and sentenced to 80 years in prison.
After distributing the press release on March 8
th
2011, Rep. Jackson Lee was featured in
eight different news stories about Tatas return to the U.S. Each of Houstons four network news
stations produced at least one news report about Tatas dramatic return featuring Rep. Jackson
Lee. Houstons newspaper, the Houston Chronicle, printed two news articles about Tatas return
featuring Jackson Lee. By distributing one press release, Rep. Jackson Lee successfully
3
associated herself with Tatas extradition process. She made herself a primary source for all of
the local news coverage surrounding the extradition process.
Interestingly enough, the location of Tatas home-based daycare was not found in
Jackson Lees Congressional district. Rep. Jackson Lee had no obvious association with any of
the victims of the fire, Tata, or the daycare; yet she was the only member of Houstons nine-
member Congressional Delegation that played any public role in the incident. However, Rep.
Jackson Lees press release was written to highlight the emotional plight of those impacted by
the fire. Rep. Jackson Lee personalized the issue by focusing on the emotional hardship that the
victims and their families were suffering while Tata was evading due process; she did this while
asserting that the United States government should take action on behalf of these victims. At the
end of the press release, Rep. Jackson Lee also highlighted her personal connection to the
situation. It said, In the wake of the tragedy, Congresswoman Jackson Lee visited the site of the
gruesome fire, and she commemorated her visit by leaving behind flowers for the victims and
their families.
1
By articulating her personal experience with the aftermath of the fire and forging
these emotional ties, Rep. Jackson Lee was able to associate herself to the situation through these
newfound personal ties and her policy efforts in Washington. Rep. Jackson Lee often speaks out
about local or national issues pertaining to the justice system and she often makes the news
commenting about these issues.
Members of the United States House of Representatives work in Washington D.C., but
ultimately answer to the finite portion of the population that elected them. Constituents
perception of their member of congress is based on what information is available, either from the
member of congress directly or from the local media outlets whose job it is to cover them. While
members of congress can control any direct communication they have with their constituents,

1
Congresswoman Sheila Jackson Lee, Congressional Office, 2011.
4
such as through direct mail or town hall meetings, indirect communication through media outlets
is somewhat beyond their control. Members of congress do try to influence the news making
process and they do so through the implementation of a media strategy. Strategy implies direct
and planned decision making employed in an attempt to yield a specific outcome. Members of
congress look to yield as much positive media coverage as possible. They also seek to make the
news on topics that of are particular importance to them. Members of congress have three main
interests in garnering positive news coverage: reelection, making policy, and bolstering their
professional reputation.
2
Thus, whether or not members of congress can influence the news
making process would be important because the ability to influence making process would
signify that they are able to increase their chances of reelection, promote their policy and bolster
their professional reputation.
But how much influence do members of congress have over the news making process?
And which media strategies most effectively influence the news making process?
This project seeks to evaluate congressional media strategies. The relationship between
members of the U.S. House of Representatives and the local media outlets that cover them
provides ample opportunity for an evaluation of congressional rhetoric because the nature of the
constituent-congressperson relationship is repeatedly renegotiated every two years in an election.
Members of the U.S. House of Representatives represent a very specific subset of the population
in a very specific geographic location. Each member of congress also functions within a specific
market of public communication institutions dedicated to covering the member of Congress
and reporting any developments to those who seek information from within the media market.

2
Mayhew, David. 1974. Congress: The Electoral Connection. New Haven, CT: Yale University Press. pp. ix.

5
This project will evaluate the congressional media news making process, focusing
particularly on the use of rhetorical strategies used in pursuit of media coverage. Personalization
and policy assertion can be seen as alternative rhetorical strategies employed by members of
Congress when they communicate. I developed the terms personalization and policy assertion
based on a close reading of the press releases used for the data analysis section of this paper. The
use of these strategies is pronounced when members of Congress communicate directly with the
press through press releases. An evaluation of different rhetorical strategies and their relative
effectiveness would be of interest to public relations professionals, political decision makers, and
media professionals.
For the purposes of this analysis, personalization and policy assertion will have narrow
definitions. Personalization or issue personalization is classified as any sentence that articulates:
a personal or emotional reaction to the issue in question; an association between the issue in
question and some facet of the authors life or personal opinions; potential consequences on a
community member or a community (i.e. the community or community member receives the
action of legislation); or an expression of want or desire (unrelated to a piece of legislation).
Policy assertion or assertion is defined as any sentence that articulates: information about a piece
of legislation that involves the author (i.e. the author sponsored a piece of legislation or supports
it); commentary on legislation that involves the author; the potential consequences or impact of a
piece of legislation (generally without direct individuals that receive the consequences of the
legislation); normative policy assertions using imperatives (i.e. we need, we must,) or
expression of want or desire related to a policy position or a piece of legislation. I developed
these definitions myself based on a close reading of the population of press releases distributed
by members of the Houston Congressional Delegation during the 112
th
session of congress. I
6
thought it was appropriate to develop these definitions and focus my research on the distinction
between these two rhetorical strategies in congressional communication because the existing
literature was void of any discussion about the congressional news making process and the
rhetorical strategies employed by members of congress.
Members of congress are ideal test subjects when focusing on media strategy because
congressional elections are held every two years and members of congress are responsible to a
definitive, localized constituency. In order to win reelection every two years, members of
Congress must reach this localized constituency and effectively communicate a persuasive
argument, one that is persuasive enough to garner enough electoral turnout. Thus, for members
of Congress, media strategy is an implicit component of congressional campaign strategy, and
since U.S. House campaigns are somewhat continual, for members of the U.S. House media
strategy is constantly a priority. With elections held every two years there is less downtime in-
between campaign cycles and media decisions during any point of a congressional session
become a component of a member of congress media strategy.
Ultimately, media practitioners decide when a member of Congress makes the news.
Journalists and editors decide news content based on the concept of newsworthiness. How these
decision-makers interpret Congressional news, or employ commentary from members of
Congress in local or national news stories, largely determines what information constituents
receive about their representatives.
How effective are congressional media strategies? When answering this question, a
number of other questions arise, such as: how often are members of congress in the news, and on
what topics are they newsworthy; how often do they employ different rhetorical strategies; and
7
do the news stories correspond with press releases? By answering these questions, I should be
able to discern which rhetorical strategy is more effective at cultivating local media attention.
This paper seeks to examine congressional media strategy by preforming a quantitative
analysis using a sample of congressmen from the U.S. House of Representatives. The analysis
examines the entire 112
th
session of congress and content analysis will be done on the
congressional input to the news making process, press releases, as well as the output of news
stories. The data gleaned from the content analysis will provide the resources to present a
descriptive background about the congressional news making process and it will also be used to
run a probit analysis to test the effectiveness of both personalization and policy assertion.
This paper seeks to evaluate the effectiveness of congressional media strategy. It also
seeks to determine whether issue personalization is a more effective rhetorical strategy than
policy assertion within the context of congressional media strategy. First, I present the origins of
this topic and the purpose of answering such questions. I will then present the existing literature
about this topic, highlighting its limitations and laying the theoretical foundations for my
hypothesis. I also utilize first-person perspectives to provide background information as to how
the congressional news making process currently functions through the use of press releases and
utilizes issue personalization and policy assertion. I then present my hypothesis, that issue
personalization is more effective in yielding media coverage than is policy assertion. From there,
I demonstrate how I conceptualized each of the rhetorical strategies and describe my research
methods. My findings are organized into two parts, representing how press releases can incite
news coverage through either direct or indirect causality. The indirect causality section makes
use of descriptive data to showcase what members of congress are doing and how their actions
are being reciprocated by the local news media. The direct causality section presents different
8
ways for evaluating the effectiveness of the press releases and the rhetorical strategies used
within them. I present four models: the first using dummy variables to represent the rhetorical
strategies; the second using proportional values for the real use of the rhetorical strategies; the
third which uses calculated probabilities to assess the effectiveness of each combination of
rhetorical strategies; and the fourth which reverts back to the use of dummy variables to test the
interaction effect between those issues identified as newsworthy and the various rhetorical
strategies. I then reflect upon my research in the conclusion and make suggestions for future
research.
Research Origins
An inquiry into local media strategy is necessary because members of the United States
House of Representatives are locally elected and serve at the federal level of government, thus
they need to articulate their work in Washington D.C. in a way that speaks to those back home in
their district. Members of congress advocate for and vote on national changes, changes that may
or may not be relevant to their local constituents, but in order to maintain their positions
members of congress need to maintain frequent interaction with their constituents in order to
make their legislative successes known. The ways in which local media cover members of
congress is important in shaping how the constituents perceive the work that is being done on
their behalf in Washington D.C. and a positive portrayal by local media is often critical to
maintaining office.
It is important to study both the decisions made by the members of congress and the
actions of the press because these individuals are the primary decision makers and practitioners
of media strategy. The purpose of a strategy is to achieve a particular end and a strategys
usefulness is determined by its effectiveness at achieving these ends. Thus, strategy needs to be
9
tried, tested, and proven successful. A media strategy that is tried, tested, and proven successful
would be of interest to both legislators and their counterparts in local media because it would
expose the Congressional news making process itself. Legislators have an interest in which
strategies were most successful, so that they could use these strategies to suit their own
legislative interests or to manipulate the local media in their favor.
Exploring how effective the members of the U.S. House of Representatives are at
controlling their media image and maintaining their personal brand is important because
explorations of this nature should be representative of how political agenda setting works from a
procedural standpoint. Any procedures that are informative to how a member of congress
creates a personal brand, a brand that is used to influence their portrayal in the news media, is
important because these actions may be characteristics of a media strategy.
Theory and Rational
Most research on congressional news making focuses little on strategy, but rather, it
focuses either on how news media cover politicians or how politicians create and act on an issue
agenda. The work of literature that is of most importance to this project is Congress, the Press,
and Political Accountability by R. Douglas Arnold. In his book, Arnold (2004) reports on
empirical studies he carried out regarding the volume of newspaper coverage for all members of
congress during a two-year term, how newspapers cover legislatives and how legislatives take
positions. Arnold finds that geographic location is a large determinant of how much information
about their representative citizens are exposed to. Arnold also finds that the volume of news
coverage changes depending on the number of representatives representing the newspapers
market.
3
Drawing from these conclusions, Arnold found that a majority of the articles written

3
Arnold, R. Douglas. 2004. Congress, the Press, and Political Accountability. Princeton, NJ: Princeton Press. May.
Pp. 62
10
about members of congress focus on the representatives role in national policy making and that
most stories approach the issue neutrally, only ten percent containing criticisms of the
representatives performance.
4

Arnold also does extensive research on how newspapers cover legislative position taking
and how the legislators vote in congress. Arnolds research design regarding volume of coverage
and the content of this coverage largely inspired how I designed and carried out my analysis of
local news content. While Arnolds research on congressional press coverage and its
implications is quite thorough, he largely disregards the news making process, focusing more on
the press output than the member of congress input.
In a work similar to Arnolds, C. Danielle Vinsons Local Media Coverage of Congress
and its Members looks at how members of congress are covered by local media outlets by
preforming content analysis on a number of different news sources from multiple geographic
locations in order to draw conclusions about congressional news in general. Vinson found that
the congruence between a members district boundaries and the media market that covers the
member of congress is a good indication of how much coverage a member of congress will
receive.
5
In this sense, a media market includes both the print and broadcast news organizations
within one geographic area such as a city and its suburbs. The congruence between media market
and congressional district also has an impact on a members media strategy. Vinson explained,
"Even if all congressional offices were equal in their media relations strategies, one should not
necessarily expect uniform results. the congruence between media markets and congressional
districts may render such strategies all but useless.
6
Vinson also found that news coverage
increased with market-district congruence and that senators, congressional leaders and women

4
Arnold, R. Douglas. 2004. Pp. 89
5
Vinson, Danielle. 2002. Through Local Eyes: Local Media Coverage of Congress. Creskill, NJ: Hampton. pp. 101.
6
Vinson, Danielle. 2002. pp. 101.
11
were more newsworthy.
7
She also found that members of Congress vulnerable to electoral defeat
and members who had been in office longer were less likely to be covered.
8

Vinson also tested whether the media directly used press releases, but she had mixed
results. The medias use of press releases varied by news medium and it also varied by the type
of media market.
9
Vinson compared the press releases that she had access to with the news
coverage from the same period of time, and she found that, publication of members statements
in most cases was hit or miss, even in weekly papers. More specifically, six of the
representatives received no coverage incited by news releases in any of the media she evaluate,
four of these six members were from low-congruence markets.
10
Six of the members she studied
distributed press releases that did receive attention, but she found that television news stations
did not utilized any of the press releases in their coverage.
11
She also found that newspapers
tended to extract quotes from the press releases that were about major, national issues but
ignored those that were focused on local issues.
12
Vinsons analysis of the impact of press
releases on the news making process is largely incomplete, considering that she studied only four
weeks of news coverage and press release input. Within her findings, there was a large variation
in whether the press release was picked up by the local media or not. While studying whether or
not the press release directly incites a news story is important, Vinson largely disregards the use
of press releases for the construction of an issue agenda, which would not be directly measured
by whether or not the press release incites a news story, but rather it would be measured by

7
Vinson, Danielle. 2002. pp. 44.
8
Vinson, Danielle. 2002. pp. 44.
9
Vinson, Danielle. 2002. pp. 96.
10
Vinson, Danielle. 2002. pp. 96-97.
11
Vinson, Danielle. 2002. pp. 97.
12
Vinson, Danielle. 2002. 97
12
comparing the issues discussed in the press releases to the issues on which the member of
congress is covered by local media.
Justin Grimmer (2010) created a model to discern the political rhetoric of press releases
originating from U.S. Senate offices. He was able to analyze which topics were the most
preeminent across all press releases and how these topics relate to each individual Senator,
allowing him to pinpoint the Senators expressed agendas. Grimmer found that local newspapers
regularly use press releases; making press releases an important interlocutor between Senators,
the mass media, and constituents.
13
Grimmer presents an interesting, yet complex, method for
analyzing press releases. His focus was on press releases as a means of assessing the Senators
expressed agendas, which highlighted the importance of issues and the role that they play in
communicating a specific agenda to the press. Grimmers research method demonstrates how
useful press releases can be when assessing Congressional communication.
In Making Laws and Making News, Timothy E. Cook (1989) examines the relationship
between members of Congress and the news industry. He describes the relationship between
members of Congress and the news media as a constant negotiation and renegotiation because
each group holds important negotiating powers over the other.
14
This is important, as it shows
that neither group, members of congress or media professionals, has primary control over the
outcome of the news making process.
Yiannakis (1982) found that members of congress communication styles were a function
of district and member characteristics.
15
The findings of Arnold (2004), Vinson (2003), and

13
Grimmer, Justin. 2010. A Bayesian Hierarchical Topic Model for Political Texts: Measuring Expressed Agendas
in Senate Press Releases. Political Analysis, 18(1), 1. Pp. 6
14
Cook, Timothy. 1989. Making laws and making news: Media strategies in the US House of Representatives.
Washington, DC: Brookings. pp. 31.
15
Yiannakis, Diana Evans. 1982. House members communication styles: Newsletter and press releases. Journal
of Politics 44(4). pp.1049.
13
Yiannakis, as well as similar qualitative studies about congressional communication strategy by
Fenno (1978) and Mayhew (1974), suggest that geographic characteristics of the members
district influences the news making process. In studying the congressional news making process
further, it is thus necessary to control for geographic characteristics.
The related literature looks at two different sides of the same coin, these sides being
congressional communication and congressional news coverage. Neither topic assesses the
coins value or examines how congressional communication and congressional news coverage
go together. Congressional communication can be seen as an input and congressional news
coverage can be seen as an output. Congressional news making can be seen as a larger process,
and research into the intricacies of this larger process would be informative as to how the process
works and what can be done to alter the output. An evaluation of this nature would have practical
implications, as its results would demonstrate what types of strategies employed within
individual press releases would be more effective in increasing the amount of news coverage a
member of congress receives. The analyses performed by Grimmer(2010) and Arnold(2004) can
be combined in a way that would allow for investigation into the characteristics of press releases,
so that these characteristics could be compared to the media coverage during the same period of
time. Such an analysis would allow for an evaluation of the news making process from the
perspectives of both the legislators and the news professionals.
Broadcast news stations and newspapers are businesses and news editors and producers
make decisions on content with a business model in mind. In line with this logic, Lance Bennett
(2007) defines four information biases that influence news content personalization and
dramatization being two of these biases. Personalization, as an information bias, drives news
coverage of humanized stories of success and failure, largely devaluing the importance of the
14
larger structural phenomena behind the individualized stories.
16
Bennett describes this tendency,
saying, The reasons for this are numerous, from the journalists fear that probing analysis will
turn off audiences to the relative ease of telling the human-interest side of a story as opposed to
explaining deeper causes and effects.
17
Reporting from a personalized angle is aimed at
cultivating audience attention and adding ease to the news making process.
Bennetts analysis of dramatization as an information bias is also relevant because
dramatization plays off personalized content. Dramatization as an information bias is the
journalistic tendency to favor stories or narratives as a format rather than analysis.
18
Bennett
describes this information bias, saying, News dramas emphasize crisis over continuity, the
present over the past or future, and the personalities at their center. News dramas downplay
complex policy information, the workings of government institutions, and the bases of power
behind the central characters.
19
Bennett argues that these biases have a large impact on the news
making process, which causes personalized stories to be more abundant than policy related
stories.
20
The discussions of newsworthiness within journalism textbooks also demonstrate the
presence of Bennetts information biases. Lloyd and Guzzo (2009) highlight people as a
newsworthy characteristic, saying, [p]eople are more interesting to read about than programs
and policies.
21


16
Bennett, W. Lance. 2007. News: The Politics of Illusion, 7th ed. New York: Pearson Longman. Chapter 2 (News
Content: Four Information Biases that Matter), pp. 40
17
Bennett, W. Lance. 2007. pp. 40
18
ibid.
19
Bennett, W. Lance. 2007. pp. 41
20
Bennett, W. Lance. 2007. pp. 41
21
Lloyd, Robert, and Glenn Guzzo. 2009. Writing and reporting the news as a story. Pearson Education, Inc., pp.
27.

15
While Bennett describes how these information biases take shape within the media, the
literature is largely void of any mention of personalizations role in the news making process.
The literature does not differentiate between legislators use of personalization or assertion as
tools to further their media coverage related goals. The literature also did not provide any
distinction between personalization or assertion and their role in the news making process. This
paper serves as an original contribution to the study of congressional communication.
From a procedural standpoint, legislators and their staffs should be able to play into the
information biases identified by Bennett to increase the likelihood that their press releases
become news. First-person perspectives provided by a congressional communications director
and a newspaper editor from Houston, Texas work to support this claim.
Rep. Sheila Jackson Lees Communications Director Mike McQuerry offered insight that
suggests that congressional staff often consider the implications of a press release when deciding
how to frame an issue. McQuerry said, It is important to think about the impact of the press
release on the public, and so there is virtually always a degree of personalization involved in a
press release.
22
McQuerry focuses on the potential media output presented by each press release
and for him, personalizing issues is a means for associating these issues with the public. When
asked which rhetorical strategy he found more effective at meeting his press related goals,
personalization or policy assertion, McQuerry said that he finds personalization more effective.
He said, A policy position can only go so far, since everyone can have an opinion. Making the
connection to how the issues affects constituents is more important.
23
For McQuerry and Rep.
Jackson Lee, personalization is not only a means to garner media coverage; it is a means by
which Rep. Jackson Lee connects with her constituents.

22
McQuerry, Mike. Personal communication. January 31
st
, 2013.
23
McQuerry, Mike. Personal communication.
16
Since press releases can be seen as an attempt to connect with constituents by way of
media outlets, the role of the media outlets in this process is of vital importance. McQuerry said
that policy positions with a local angle are most likely to become news. He said, Local media
outlets are often concerned about the members position on issues that have a direct affect on her
constituents.
24
This positions contrasts Vinsons finding that press releases about local media
tended to ignore press releases written about local issues. McQuerrys conception of what local
news outlets look for also drastically contrasts what the Houston Chronicles Washington Bureau
Chief Richard Dunham identified as the type of information about local members that he looks
for. Since the Houston Chronicle is a major newspaper, it caters to a geographically diverse
audience.
25
Policy related to constituents was not of primary concern for Dunham, but rather he
focuses on timely national and local issues in general, neglecting to associate them with any
particular audience. Dunham said, A comment or analysis on breaking news is the most likely.
Input on important national issues before Congress is another. We also look for issues of
particular local interest such as energy, NASA/space/science, medical research and
transportation.
26
Dunham also identified issues that would be of interest to Houstons Latino
and African American readers, saying that the Chronicle would be likely to cover issues such as
voting rights, education, discrimination, redistricting and diversity.
27

McQuerry and Dunham share very similar views about the use and implications of issue
personalization. Dunham identified personalization as the rhetorical strategy most likely to
create news. He said, Yes, for better or for worse, personalizing an issue yields better results in
modern-day journalism. Readers (or editors?) are looking for a connection between their elected

24
McQuerry, Mike. Personal communication.
25
Dunham, Richard. Personal communication. December, 17
th
, 2012.
26
Dunham, Richard. Personal communication.
27
Dunham, Richard. Personal communication.
17
officials and the issues that matter to them.
28
Dunham also highlighted the importance of policy
assertion, saying that policy assertion made in a timely fashion in relation to breaking news
events yields good results.
29

Based on this theory and rationale, members of congress who represent districts within
the same media market as other members of congress should receive less news coverage than
members of congress whose district is congruent with a media market. It is unclear from past
research what kind of role press releases play in the news making process, but based on the first-
person accounts press releases appear to be a means by which members of congress attempt to
make news and attempt directly connect with constituents by way of the local media outlets.
Based on Dunhams first-person account, it appears that press releases play a role. Dunham
indicated that press releases indirect inform or incite the political content he creates for the Texas
on the Potomac website about 20% of the time. Logically, the employing personalization within
a press release should thus increase the likelihood that the press release plays a role inciting a
news story.
Hypothesis
I argue that issue personalization is an effective media strategy in that local media outlets
would cover this type of political messaging most of the time. A complementary argument that I
will make is that policy assertion would not be an effective media strategy in that local media
outlets would not cover this type of political messaging most of the time. Since personalization
and policy assertion tend to occur simultaneously, it would be impossible to directly test these
arguments. By combining the logic used in formulating these arguments, I formulated a

28
Dunham, Richard. Personal communication.
29
Dunham, Richard. Personal communication.
18
generalized hypothesis: personalization is more effective in yielding media coverage than is
policy assertion.
Personalization an issue would be a good strategy for members of congress because it
would put the issue within a specific framework for their constituents to understand.
Personalizing issues would allow members of congress to forge connections with their
constituents by way of the issues that they are personalizing, because it would demonstrate that
they have some sort of personal relationship with the issue they are communicating about. Those
constituents who also have a personal relationship or found their position on particular issue with
personal evidence would be likely to feel some connection with the legislator who personalizes
that issue. While personalizing an issue is helpful when looking to connect to constituents, it
would also be helpful when communicating with the media. When communicating with the
media through the distribution of press releases, personalizing an issue would demonstrate that
the legislator has a personal stake in an issue and give the press a reason to pay attention to what
they are saying about that issue.
Since members of congress serve at the federal level of government but advocate for
individuals from very specific geographic regions, personalizing issues would be a means by
which members of congress could translate the issues within the national debate to make them
relevant to the unique neighborhoods and communities within their district. Personalizing issues
would also highlight the consequences and impacts of certain policies on particular individuals
or groups within the district. Personalizing issues in this way would also make national issues
more relevant to ones constituents, putting a local face on those constituents who would either
benefit or be harmed by a particular issue or policy. For issues that are highly technical such as
19
the federal budget, personalization would also be a means by which to simplify the issue into
situations and scenarios that the average constituent could understand.
In contrast, policy assertions can be provided by anyone. The assertion would not be
expected to have the same effectiveness as personalization would because assertions are
commonplace for members of congress. Creating and enacting policy is what members of
congress are sent to Washington to do, thus policy assertion in itself is not as newsworthy as
issue personalization.
We should expect issue personalization to generate more media coverage than policy
assertion because of the format and numerous biases inherent in the mass media, characteristics
that make media coverage more apt to cover personalized, dramatic narratives rather than
straightforward policy assertions, especially during campaign seasons. During campaign
seasons, congressional incumbents or congressional candidates are looking to win the votes of
constituents and to do so, candidates or incumbents often attempt to relate their own personal
experiences to issues of importance to their platform. Anecdotes about personal experiences of
constituents also commonplace during campaigns, as these anecdotes are used as a tool to
connect personal struggles with larger issues that a candidate or incumbent looks to address.
Campaigns in themselves are dramatic because they are contest with high political stakes that
arouses much media fanfare. Candidates and incumbents can also use campaign advertisements
as a platform from which to attack their opponent personally or about a policy issue. Often, the
campaign advertisements are produced to dramatize the claims being made and to highlight the
potential impacts of the claim on specific individuals or groups of constituents.
Journalists gravitate towards stories that have a personalized angle and offer a clean
storyline due to the fact there are many members of Congress and each member promotes their
20
own stories and angles.
30
The tendency to gravitate towards stories with personalized angles and
with clean storylines, meaning that there is an obvious beginning, middle and end, would be
especially true for media markets that service an area in which there are a number of
congressional representatives.
When a member of congress personalizes an issue, they are framing that issue in a
context that speaks to the information biases that journalists look for. Providing local media
outlets story ideas or issues packaged in a personalized frame should increase the likelihood that
these stories or issues become news content, because the framing speaks to the biases journalists
are trained to look for when evaluating newsworthiness. Based on the news biases identified by
Bennett and the first-person perspective provided by McQuerry and Dunham, we should expect
personalization to be a more effective rhetorical strategy than policy assertion, in that
personalization incites a greater volume of news coverage than policy assertion does.
Research Design
To specifically test the hypotheses and address the larger questions at hand, namely the
relationship between local media and congressional representatives, I will use the data
accumulated through the content analysis to conduct a large-n study. Content analysis is an
appropriate data collection method because its goal is to produce counts of key categories, and
measurements of the amounts of other variables.
31
Preforming content analysis will allow me to
measure the presence of both personalization and policy assertion in congressional press
releases. Ultimately, the data gleaned from content analysis regarding personalization and policy

30
IIyengar, Shanto and Jennifer A. McGrady. 2007. Media Politics: A Citizens Guide. New York:
W.W. Norton. Chapter 7 (Going Public: Governing through the Media), pp. 186.
31
Neuendorf, Kimberly A. 2002. The Content Analysis Guidebook. Thousand Oaks, CA: Sage Publications, Inc., pp.
14.
21
assertion in press releases will become independent variables used in an evaluation of the news
making process with a dependent variable derived from the content analysis of news stories that
measures whether or not there was a corresponding news story that followed a press release.
I preform content analysis on 1215 press releases sent out by eight members of the
United States House of Representatives during the 112
th
session of Congress (January 5
th
, 2011
January 3
rd
, 2013). The press releases were taken from each member of Congress website,
where they are posted shortly after being sent out to the press. In one case, a significant
proportion of the press releases were not available on the website and I received them directly
from the congressional office. When preforming content analysis on the press releases, I used the
sentence as my unit of measurement. I coded each sentence as one of four possible rhetorical
strategies and also categorized the entire press release into one of seventeen issue areas. These
issue areas were determined during the coding process, by evaluating which issues were
discussed and then formulating groups to categorize those specific issues under a more
generalized issue title. For example, if a press release were written about the Keystone XL
Pipeline, it would be coded as an Energy/Environment issue and if a press release was written
about the national deficit, it was coded as a Budget issue. By coding the press releases in such
a fashion, I was able to generate a data set that allowed for analysis at the press release level, by
member of Congress, by issue area, or as a total for all members of the sample.
I also utilized 1048 local news stories and I preformed content analysis on all of the news
stories that involved the eight members of Congress during the 112
th
session. These news stories
consisted of 635 newspaper articles and 414 broadcast news clips. These new stories are the
population of news stories produced about eight members of Congress during the entire 112
th

session. The newspaper articles were located using LexisNexis news search, which allowed me
22
to locate all of the stories that mentioned a member of Congress within the specified time period
and in a particular newspaper. The broadcast news stories, produced by four independent local
news stations, were not archived on LexisNexis and had to be accessed from each individual
news stations website. The content available through these online archives varied with each
individual news station, some stations making fewer clips available than others. I utilized
whichever news clips were available on the online archives that fell within the dates of interest
for this study. When preforming content analysis on the newspaper articles and broadcast TV
segments, I coded for: the primary and secondary issues addressed; the context of which the
member of congress is mentioned (i.e. are they the main topic, do they comment on an issue, or
are they just mentioned) any signifiers of issue personalization or policy assertion; as well as
generic attributes such as length, author, date released, and placement within the newspaper or
news hour. Similar to the content analysis performed on the press releases, I will categorize each
news story into one of seventeen issue categories. The sentence will not be the unit of analysis
used for news stories, but rather each individual news story will be a unit of analysis from which
the descriptive data regarding issue, role, and use of rhetorical devices is gleaned.
This analysis uses the Houston, Texas region as a sample population of members of the
U.S. House of Representatives and local media outlets. Focusing on one geographic region
allows for an intricate investigation into the relationships between individual members of
congress and the local media outlets that cover them because it accounts for regional factors that
are hard to control for. For example, by using one geographic region, it is not necessary to take
into account regional political differences, differences in the media environment, or differences
in issue selection when comparing members of Congress. Houston, Texas is the fifth largest
media market in the United States and the largest city in the state of Texas. With a population of
23
2.1 million people, it boasts a diverse mixture of urban, suburban and rural areas. Being one of
the United States largest metropolises, Houston has a well-established media market, including
one major newspaper, The Houston Chronicle, the ninth largest newspaper in the U.S. with a
circulation of close to 326,000 on weekdays, 340,000 on Saturdays and 1.07 million on
Sundays.
32
The Houston Chronicle services much of the area surrounding the city of Houston by
covering both regional and state news. The Chronicle has a bureau in Austin, Texas as well as
Washington, D.C. Houston is also the home to a local affiliate of each of the four major
television networks, ABC (KTRK-13), CBS (KHOU-11), Fox (KRIV-26) and NBC (KPRC-2).
Houston has one of the largest congressional delegations in the United States, boasting
nine members of Congress. This analysis will use eight of these members of Congress as a
sample population. The members of congress this study will focus on are: Rep. Ted Poe (TX-2),
Representative John Culberson (TX-7), Rep. Kevin Brady (TX-8), Rep. Al Green (TX-9), Rep.
Michael McCaul (TX-10), Rep. Sheila Jackson-Lee (TX-18), Rep. Pete Olson (TX-22), and Rep.
Raymond Gene Green (TX-29).

32
Hearst Corporation. Houston chronicle. [cited 12/15/2012 2012]. Available
from http://www.hearstmediahouston.com/?id=houston-chronicle.

24

These representatives all have a stake in the Houston area media market, as they
represent some geographic area covered by the market. This study will not include
representative Ron Paul (TX-14) because Rep. Paul was considerably engaged in the national
media due to his Presidential campaign and he also did not be seek re-election for Texas 14
th

congressional district, suggesting that his media strategy operated under unique circumstances.
His presence in the national spotlight did not fit well with this analysis, as his presidential
campaign would have impacted when and for what reason he made the local news.
Operationalization
Drawing from the works of Arnold and Grimmer, I use press releases as the primary
source of communication between members of congress and the press and compare this primary
source to the news media produced by local news outlets. The press releases were used to
25
capture the direct interaction between members of congress and local media outlets. Press
releases serve as the primary means by which members of Congress communicate with media
outlets. While press releases are the primary means by which members of Congress
communicate with media outlets, this communication is one way and does not provide insight to
the decisions that are made by news media. The news content produced by local news media
provides an assessment of how the local media responded to the communication from the
members of Congress, by either writing a story about an issue communicated in a press release,
or ignoring the issue.
Conceptualizations
In preforming content analysis, each sentence was placed into one of four categories and
thus creating four rhetorical variables used in Congressional communication. These variables
were: issue personalization, policy assertion, both issue personalization and policy assertion, and
information.
Issue personalization is classified as any sentence that articulates: a personal or emotional
reaction to the issue in question; an association between the issue in question and some
facet of the authors life or personal opinions; potential consequences on a community
member or a community (i.e. the community or community member receives the action
of legislation); or an expression of want or desire (unrelated to a piece of legislation).
Policy assertion is defined as any sentence that articulates: information about a piece of
legislation that involves the author (i.e. the author sponsored a piece of legislation or
supports it); commentary on legislation that involves the author; the potential
consequences or impact of a piece of legislation (generally without direct individuals that
receive the consequences of the legislation); normative policy assertions using
26
imperatives (i.e. we need, we must,) or expression of want or desire related to a
policy position or a piece of legislation.
Issue personalization and Policy assertion (hybrid) is defined as any sentence that
contains characteristics of from both the issue personalization and policy assertion
categories. It normally would articulate a reaction or personal statement from the
member of congress along with specific information regarding a piece of legislation.
Information is defined as any sentence that: does not contain any of the elements from the
issue personalization or policy assertion categories. Generally, sentences coded as
information tend provide background information. The information provided may
present opinionated or biased information, but is phrased in a factual way.
Important note on sentence order: sentence order was often a factor in determining what
category a specific sentence was classified as. The press releases were read as a narrative would
be. In order for a sentence regarding a specific piece of legislation to be classified as a policy
assertion, there first needed to be a sentence that explained the member of congress relation to
that legislation (i.e. did he or she sponsor it, co-sponsor it, vote for, etc.). Any sentences that
described a piece of legislation but came before the sentence that specifically connected this
legislation to the particular member of congress was not considered policy assertion, but rather it
was considered information. In another sense, the member of Congress who produced the press
release needed to rhetorically associate him or herself with a piece of legislation before it was
considered a policy assertion by these terms. This is applicable only when legislation is being
discussed, as a member of Congress can make a generalized policy assertion as any time.
Measuring effectiveness
27
The data gleaned through content analysis and the interpretation of this data has allowed
me to draw conclusions regarding the effectiveness of the rhetorical strategies employed in the
press releases. The effectiveness of the rhetorical strategies is compared using two causal
explanations: indirect and direct causality. Indirect causality would be generating media content
in specific issue areas based on a well defined and articulated issue agenda in that said issue area.
An issue agenda is defined as a single issue category or multiple issue categories which a
member of Congress has involved him or herself. It can be identified by evaluating a member of
congress committee assignments, legislative priorities, and for the purpose of this experiment,
the number of press releases distributed. In terms of press releases, the issue agenda would be
determined by identifying the most frequently used issue categories.
By comparing a member of congress self-declared issue agenda, as stated on their
official websites, to their actual output of press releases, I was able to determine which issues
were integral parts of each member of congress issue agenda. It was then possible to compare
the press release input to the output of news stories categorized in those issue areas
previously defined as part of the issue agenda. If there were correlation between a members
self-declared issue agenda, press release input by issue and media output by issue, then a
members issue agenda would be considered strong.
Direct causality would be when a press release directly incites a related news story.
Direct causality is helpful when looking at the impact of personalization, policy assertion, and
the hybrid of the two rhetorical strategies because it would allow for measurement of how
effective each strategy was at generating media content directly. By comparing press releases
and media output directly, I was able to test whether or not issue personalization and policy
assertion generate news output.
28
During the 112
th
session of Congress, there were several major news events that dominated the
headlines. These events are summarized in the news events section of the appendix, which also
shows how each story was categorized into one of seventeen different issue categories. These
news events will be used during the analysis as specific examples to complement the quantitative
analysis.
The majority of these stories originated independently, in that the members of congress
did not play a direct role in inciting the news storys initial development. Thus, press releases
distributed about these stories were reactions. Only in the Chad Holley case and the Metro
Expansion news events were members of Congress implicitly involved in the storys
development.

Data analysis:
When evaluating Congressional media strategy, it is important to keep in mind why
members of Congress want to make news. As discussed earlier, House members want to be
reelected every two years, they want to legislate on issues that are important to them and they
want to bolster their reputation and improve name recognition. Members of Congress indicate
which issues are most important to them in a number of ways, for instance, most members of
Congress have a dedicated issues page on their official website. They also indicate which
issues they find the most important by sponsoring/cosponsoring legislation, serving on related
committees or generally, talking a lot about those issues. In answering the question are
congressional media strategies effective? one first needs to consider whether members of
congress are attempting to make news about those issues that they find important. One would
expect that a member of Congress who self-identifies Taxes as being an important issue would
29
spend a considerable amount of time talking about budgetary issues in press releases and that this
member would be featured in news stories about budgetary issues. If the member who identifies
taxes as being an important issue does not allocate time to discuss budgetary issues in his or
her press releases and is not featured in news stories about budgetary issues, it is safe to assume
that budgetary issues and taxes are not of high importance to that member of Congress. One of
the first and most important characteristics of an effective media strategy would thus be
coherence between ones self-declared issues and personal platform (ones issue agenda) and
ones communications record on these issues. Coherence between ones issue agenda that one
communicates about in press releases can be considered an effective issue agenda.
Using content provided on each member of Congress official website denoting which
issues him or her finds important, I compared these self-declared issues, matched them to one
of the seventeen issue categories of use for this analysis and provided a percentage for each issue
category to convey how much of each members time was spent on the given issues. Self-
declared issues that made up less than five percent of the total press release distribution were left
out of this chart.
30

Reps. Brady, Culberson and Poe were quite good at articulating their issue agendas, as
their press release output matches those issues they made note of on their websites. Reps. Gene
Green, Jackson Lee, McCaul and Olson do fairly well at articulating their issue agendas, but
have neglected at least one issue category that they frequently communicate about through press
releases. Important to note, that constituency as an issue category is not necessarily a standard
issue category that a representative would be expected to communicate about. The constituency
issue category encompasses issues or events that are specifically local and concern ones district.
Since members of Congress represent a specific district, it is somewhat implied that district level
issues would be important to them. Justice is a similar issue category to that of constituency
in that it pertains situations within the justice system and these situations were usually local or
regional in scope. Since members of Congress are lawmakers, it is clear why they would want to
Self Declared Important Issues Member
Brady Economic Recovery Economy/Jobs* 26%
Spending Reform
Your Taxes
Budget* 31%
Budget
Culberson Budget Budget* 17%
Medicare Healthcare* 9%
Veterans Veterans/Military* 9%
Health Care Healthcare 7% A. Green
G. Green Energy
Environment
Energy/Environment* 22%
Energy/Environment*
Health Healthcare or Other 9%
Foreign Aairs Foreign Policy* 15% Jackson-Lee
Homeland Security National Security 5%
Homeland Security National Security* 44% McCaul
Energy Energy/Environment* 26% Olson
NASA Space 14%
Foreign Aairs and Terrorism Foreign Policy*/National Security 18%
Poe
Energy Energy/Environment* 12%
Unarticulated & Frequent issue categories
*one of members most frequent press release issue categories
see appendix for full issue agenda/press release comparison
Justice 17%, Foreign Policy 9%
Constituency 18%
Transportation 9%
Foreign Policy 17%
Constituency 14%
Associated Issue Category/% of Press Releases
Self-Declared Issue Agenda compared to Press Release distribution
31
be vocal about how laws are enforced. Similar to constituency, it is somewhat implied that
issues regarding the justice system would be important to them.
Drawing from the issue categories the members of congress self-identified on their
websites, the absence of constituency as an issue category is important to note. One possible
reason for its absence is that members of congress attempt to communicate with their
constituents through issues that can be both national and local, such as the categories
transportation, or energy/environment. Rep. Al Green failed to articulate on the issue page of
his website any issue categories that corresponded with his press release output. All of the issue
categories he referenced on his website made up less than five percent of his total press release
distribution, suggesting that the topics he listed are not quite as important as his website might
suggest.
Using the information provided by the members of Congress and comparing it to their
press release distribution provides us with a general understanding about each members issue
agenda. When the issues that members listed on the issue page of their websites did not
correspond with the issues they made use of in their press releases, their issue agenda was
formulated by using the most frequent issue categories they made use of when producing their
press releases. In these cases, such as the case of Rep. Al Green, the primary and secondary
components of his issue agenda were created by using the most frequent and second most
frequent issue category that he made use of when producing his press releases. Each member of
Congress issue agenda was then judged as strong, moderate or weak based on the coherence
between their self-declared issue agenda and the top two issues that were most frequently
featured in their press releases.

32
Member Issue Agenda vis--vis issue categories Strength of Issue Agenda
Brady Budget, Economy/Jobs Strong
Culberson Budget, Health Care Strong
A. Green Justice, Foreign Policy* Weak
G. Green Energy/Environment Moderate
Jackson Lee Foreign Policy, Transportation Moderate
McCaul National Security, Foreign Policy Moderate
Olson Energy/Environment, Space Moderate
Poe Foreign Policy, Energy/Environment Strong

Press releases are a very common form of political communication and they are
specifically used to communicate with the media. Since many members of Congress use press
releases as a news-making tool, the releases themselves are useful units of analysis from which
to draw conclusions and make comparisons about political media strategy. Press releases have
no standard format, but they tend to be written in a similar fashion to that of the typical news
story, which clearly demarcates a title, date, location, headline and contact information.
Congressman Benjamin Butler who represented Massachusetts 5
th
, 6
th
and 7
th
districts during
the late nineteenth century is credited as having invented the press release, which was formatted
as a lengthy, self-prepared opinion.
33
Clearly, members of Congress have had a long history of
using press releases, as they continue to use them in attempt to garner the medias attention. It is
thus necessary to evaluate the descriptive qualities of press releases in order to gauge exactly
what a member of Congress is doing in attempt to garner that media attention and make news.

33
Cook, Timothy. 1989. Making laws and making news: Media strategies in the US House of
Representatives. Washington, DC: Brookings. pp. 20.
33
Descriptive qualities that provide information regarding the frequency of distribution, quantity,
issue subject, and use of rhetorical strategies will shed light on the congressional logic behind the
news-making process.
Figure 3 provides a general overview of how many press releases each congressional
office of interest to this study distributed during the 112
th
session of Congress. By evaluating the
total number of press releases put out, one can get a general understanding of how frequently
these members of Congress chose to distribute press releases. There is much variation between
these frequencies, as Rep. Jackson Lee put out almost three hundred press releases during the
two-year term and Rep. Al Green put out only forty-six press releases during that same period of
time. At this rate, Rep. Jackson Lee put out a new press release about every two and a half days,
where as Rep. Al Green distributed one about every sixteen days. Of course, the press releases
were not distributed so systematically, but framing the frequency with a systematic evaluation of
the highest and lowest distributor of press releases from the sample population of members of
Congress demonstrates that members of Congress distribute press releases quite frequently.

34

Delving a bit deeper into the content and strategy employed within the press releases,
figure 4 displays the use of the four different rhetorical strategies by member of Congress. These
figures were calculated by taking the total number of sentences in which each member of
congress made use of one of the four rhetorical categories and then dividing by the total number
of sentences to get the percentage. Information was universally the most used rhetorical strategy,
but members of Congress use information at variant rates. For instance, Rep. McCaul provides
information about sixty-four percent of the time, where Rep. Al Green only provides information
about thirty-four percent of the time. The hybrid strategy employing both issue personalization
and policy assertion in the same sentence was by far the least used rhetorical strategy, whereas
policy assertion and personalization are used at variant rates. All eight members of congress
Brady Culberson A. Green G. Green Jackson Lee Olson Poe 0
50
100
150
200
250
300
McCaul
Figure 3: Total Press Release Output
Member of Congress Total #
Rep. Culberson
Rep. A. Green
Rep. G. Green
Rep. McCaul
Rep. Jackson Lee
Rep. Olson
Rep. Poe
171
46
76
291
140
116
141
Top Category
Budget (29)
Justice (8)
Energy/Environment (16)
Foreign Policy (44)
National Security (48)
Energy/Environment (30)
Foreign Policy (25)
2nd Category
Healthcare (16)
Foreign Policy (4)
Constituency (15)
Transportation (27)
Foreign Policy (24)
Constituency (16)
Energy/Environment (17)
Rep. Brady 234 Budget (70) Economy/Jobs (61)
35
used policy assertion more frequently than personalization. The differences between members of
Congress were not found to be statistically significant.
34



Although member differences in use of the rhetorical strategies were not found to be
significant, the usage the rhetorical strategies by party affiliation were found to be significant.
Figure 5 displays how the Democrats and Republicans in the Houston Congressional Delegation
made use of the four different rhetorical strategies. When evaluating the use of the different
rhetorical strategies from a party perspective, there are differences between how members of
Congress employ these rhetorical strategies. Democrats assert policy, personalize and use the
hybrid strategy more often than Republicans. Republicans use information far more than
Republicans do on average, providing informational sentences void of any rhetorical strategy the

34
by way of an ANOVA test
Personalization
Policy Assertion
Personalization & Policy Assertion
Information
Figure 4: Press Release Content
54%
12%
29%
5%
Rep. Brady
Press Release Content by Member of Congress
53%
12%
31%
4%
Rep. Culberson
47%
13%
6%
34%
Rep. A. Green
34%
29%
16%
21%
Rep. G. Green
46%
14%
33%
7%
Rep. Jackson Lee
48%
12%
35%
5%
Rep. McCaul
64%
24%
10%
2%
Rep. Olson
50%
12%
33%
5%
Rep. Poe
52%
16%
23%
11%
36
majority of the time. These results were found to be statistically significant, with a P-value of
0.000
!
.
35


It is also important to consider that members of congress dedicate differing amounts of
time to the production of press releases. Figure 6 graphically conveys each member of congress
average number of sentences. The average press release for all members was found to be about
twelve sentences. Rep. McCaul tended to be the lengthiest in his press releases, with an average
of about fourteen sentences. To contrast, Rep. Culberson tended to keep his press releases brief,
with an average of only nine sentences.

35
difference of two proportions test
Personalization Policy Assertion
Hybrid
Information
10%
20%
30%
40%
50%
60%
70%
lssue ersonallzauon
ollcy Asseruon
lssue ersonallzauon and
ollcy Asseruon
lnformauon
Sentence Type
Personalization
Policy Assertion
Personalization & Policy Assertion
Information
Republicans Democrats
11.03%
26.76%
4.42%
57.79%
13.23%
34.19%
6.20%
46.38%
Press Release Content by Party Affliation
Figure 5:
37

When evaluating descriptive information regarding press release output, it is important to
consider how the member of Congress has chosen to allocate time and resources. It is possible to
evaluate how members of Congress allocate time and resources when distributing press releases
by identifying categories of press releases that serve no obvious purpose. One such category of
press releases, which was somewhat of a reoccurring phenomenon, was the occasional holiday
dedicated press release. Figure 7 provides a the numerical breakdown of how often during the
112
th
session of Congress these eight members of Congress sent out holiday related press
releases and it also displays what percentage of the total number of press releases holiday press
releases made up for each member of Congress. Holiday press releases are not prevalent, but
Brady Culberson A. Green G. Green Jackson-Lee McCaul Olson Poe
Member of Congress Average # of Sentences
Rep. Brady
Rep. Culberson
Rep. A. Green
Rep. G. Green
Rep. McCaul
Rep. Jackson-Lee
Rep. Olson
Rep. Poe
11
9
12
11
13
14
13
11
0
2
4
6
8
10
12
14
16
Total Average 12
Figure 6: Average Number of Sentences by Member of Congress
38
they do exist. They appear to be a misallocation of resources, in that under no circumstances do
holiday related press releases lead to news output. Since press releases are sent to media outlets,
these holiday releases may be seen as a showing of good faith and thanks to the media outlets
that cover them. Holiday press releases sometimes contain generalized policy assertions
regarding things such as national security, the military, and foreign policy. These generalized
policy assertions may be helpful in maintaining an issue agenda, in that a press release sent out
on the anniversary of September 11
th
, calling for increased national security would be supportive
of an issue agenda centered on national security.
36
But not all holiday press releases serve this
purpose. Some appear useless, as Rep. Culberson put out a release entitled, Happy
Thanksgiving! that included no body message.
Six of the eight in sample members of Congress dedicate less than five percent of their
press releases to holidays. Seven of the eight in the sample dedicate less than ten percent of their
releases to holiday. Rep. Culberson is a major outlier, dedicating almost twenty percent of his
total press release output to holidays.

Since the purpose of a press release is to incite media interest and draw attention, it is
thus necessary to evaluate how local media outlets cover members of Congress. In evaluating
media content, it becomes clear that certain members of the sample make the news more often

36
this would be a good place to quote a press release
Figure 7: Holiday Press Release Totals
Member of Congress Total #
Rep. Brady
Rep. Culberson
Rep. A. Green
Rep. G. Green
Rep. McCaul
Rep. Jackson-Lee
Rep. Olson
Rep. Poe
3
14
8
1
11
2
2
4
Percentage of Total Releases
2.65%
19.72%
6.84%
0.89%
4.70%
1.23%
1.96%
2.92%
39
than their counterparts in the Houston delegation. The results are provided in Figure 8. Rep.
Jackson-Lee made the news the most, playing a role in a total of 234 news stories, mirroring her
role in Congress as one of the most talkative and outspoken members. Rep. Michael McCaul,
who in November of 2012 was chosen to become Chairman of the Committee on Homeland
Security, played a role in 162 news stories. The majority of media coverage about Rep. McCaul
was centered around issues of national security, mirroring his rising stature during the 112
th

Congress as a member of the Committee on Homeland Security. National security is the major
issue of which McCaul focuses his issue agenda and his presence in the news demonstrates that
he is relatively successful with this issue agenda, as he played a role in a total of eighty stories
relating to national security. Rep. McCauls second most frequent category of which he played a
role in news coverage was on issues relating to his constituency, as he played a role in a total of
only 18 stories. In comparison, Rep. Jackson-Lee was most newsworthy when it came to local
issues, as she played a role in forty-eight stories about local issues. Jackson-Lees presence in
stories relating to her constituency indicates that she plays a more direct role in her district,
advocating on behalf of her constituents, being involved in local projects and attending
community events. In contrast, McCauls issue agenda puts a lot of emphasis on the role he
plays in the formulation of U.S. national security policy, an issue that is (usually) far removed
from the district. Based on these results, Reps. McCaul, Olson, G. Green and Brady would have
effective issue agendas, because at least one issue category component of their issue agenda is
also the category in which they receive the most news coverage.
40

Since this analysis looks at both print and television news coverage, it is necessary to
break down the media coverage and evaluate the frequency of which each member of congress
was covered by either medium. Television news has limited space to which it can feature
members of Congress and would generally cover members in a less technical way than print
news would. Based on the issue categories, I found that there is a difference between which
issues each medium focuses on when featuring members of Congress. For six of the eight
members in the sample, their top issue category in which they are featured in print news was not
the same as their top issue category in which they were featured in broadcast news. Rep. McCaul
plays a role in television news about issues of national security the most often, being featured in
a total of forty-seven television news stories. National security is also the issue of which he
Brady
Culberson
A. Green G. Green Jackson Lee Olson Poe McCaul
0
50
100
150
Figure 8: Media Coverage Totals
200
250
Media Coverage Totals
Member of Congress Total #
Rep. Brady
Rep. Culberson
Rep. A. Green
Rep. G. Green
Rep. McCaul
Rep. Jackson-Lee
Rep. Olson
Rep. Poe
113
71
117
112
234
162
102
137
Top Category
Budget (21)
Space (11), Tranportation (11)
Constituency (20)
Energy/Environment (23)
Constituency (48)
National Security (80)
Space (23)
Justice (17)
2nd Category
Healthcare (10)
Constituency (6)
Justice (17)
Constituency(11)
Justice (31)
Constituency (18)
Energy/Environment (16)
National Security (15)
41
plays a role the most often in print news, as he played a role in thirty-three newspaper articles.
Rep. Jackson Lee plays a role in print news concerning local issues the most often, as she is
featured in twenty-four newspaper articles about local issues. She also plays a role in broadcast
news concerning local issues the most often, as she is featured in twenty-one broadcast news
stories about local issues (although she is also featured in twenty-one television news stories
about issues of justice as well).
Figure 9 breaks down the media coverage data by medium, displaying each member of
Congress strongest issue area for both newspaper and broadcast news coverage. For Rep.
Culberson, the news event concerning Houston Metros expansion into his district altered the
amount of print coverage he received about transportation. Rep. Culberson played a role in print
news nine times in articles about Metros expansion, leaving only two stories that were not
related to Metros expansion to make up Culbersons maximum of eleven new stories
categorized as transportation. Culbersons major issue category for broadcast news was space,
with all four of the stories being about the space shuttle retirement news event. For Rep.
Culberson, this analysis shows that the majority of the news content about him was the result of
the major news events that took place. Since neither of these issues related to issues of budget or
health care (Rep. Culbersons issue agenda), Rep. Culbersons media strategy can be considered
ineffective, as the majority of the media coverage he receives was not related to the issues that he
was primarily concerned with in his press releases.
As a point of comparison, Rep. Bradys strongest category for print news were budgetary
issues, whereas on television news he was strongest with issues that related to the economy or
jobs. The issue categories of budget and economy/jobs are similar, dealing with finances and
economics, but the fact that there is a divergence between his print and television coverage
42
suggests that some issues may be more or less suitable for each medium, for instance, economy
and jobs related news is more suitable to televisions format because of its dramatic and timely
nature, whereas budgetary issues tend to be more technical and easily explained by print news.

Members of congress themselves are newsworthy when it comes to two distinct issue
categories: Congress and elections. For this reason, media coverage of these two issue categories
was separated from the general presentation of news output of the other thirteen issue categories.
The results are displayed in figure 10. In stories about Congress generally, such as stories about
partisanship, political gridlock, or congressional wealth, all members of Congress are
newsworthy. Since these media outlets are focus primarily on local or regional news, the local
Brady Culberson A. Green G. Green Jackson Lee Olson Poe McCaul 0
20
40
60
80
100
120
140
Media Totals by Medium
Member of Congress Totals
Rep. Brady
Rep. Culberson
Rep. A. Green
Rep. G. Green
Rep. McCaul
Rep. Jackson-Lee
Rep. Olson
Rep. Poe
91
57
59
84
119
85
71
69
Top Category
Budget (19)
Tranportation (11)
Constituency (16)
Energy/Environment (21)
Constituency (24)
National Security (33)
Energy/Environment (16)
National Security (8)
Newspaper
Totals
22
14
58
28
115
77
31
68
Top Category
Economy/Jobs (5)
Space (4)
Budget (13)
Space (5)
National Security (47)
Space (13)
Justice (11)
Broadcast
Constituency (21), Justice (21)
Figure 9:Media Coverage by Medium
43
delegates tend to be the focus of the article. The same can be said about elections, since all
members of Congress are up for election every two years, local members of Congress are all
featured in stories conveying information about the elections, the stakes, the opponents, and the
public sentiment. Since 2012 was a Presidential election year, the Houston delegation also
frequently played a role in articles that had to do with the national elections and their role in
these articles often provided a regional or partisan angle. Since their newsworthiness is inherent
when it comes to issues of congress and elections, members of Congress who have staked out a
definitive role for themselves in the local delegation, by establishing an issue agenda or
possessing authority in a committee, tend to make news on these issues more so than those that
do not. Another important characteristic when analyzing the newsworthiness of these two issue
areas is political ambition. Rep. McCaul was somewhat vocal about his political ambitions, as
many speculated he was going to pursue Texas Senate seat when the opportunity arose after
Senator Kay Bailey Hutchison announced her retirement in 2012.
44

When evaluating Congressional media coverage, it is also important to examine the role
members of Congress play within each news story. A visual breakdown of how members of the
sample are featured in the local media is provided in Figure 11. A feature article about a
member of Congress has greater significance than does a short blurb mentioning how that
member attended a community event. Members of Congress provide commentary on local or
national events most frequently, which allows them to voice their opinions in connection to an
issue that they have made part of their issue agenda. Members of Congress who are very active
in the community and who work to maintain a strong relationship with their constituency tend to
be mentioned in the news equally as much as they provide commentary in the news. Where as
Brady Culberson A. Green G. Green Jackson Lee Olson
Poe
McCaul
0
5
10
15
20
25
30
Figure 10: Media Coverage of Congress and Elections
Media Coverage of Elections & Congress
Member of Congress Congress
Rep. Brady
Rep. Culberson
Rep. A. Green
Rep. G. Green
Rep. McCaul
Rep. Jackson-Lee
Rep. Olson
Rep. Poe
17
8
11
16
25
20
12
16
Election
10
10
6
4
9
12
17
12
Media Coverage of Elections & Congress
Congress Election
10
10
6
4
9
12
17
12
45
Rep. McCaul with his strong national security issue agenda, primarily provides commentary on
news stories. Both Reps. Culberson and Al Green are mentioned far more than they comments
on developments, but Rep. Al Green takes a very active role in the community and interacts far
more with his constituency than does Rep. Culberson. It is not often that members of Congress
are newsworthy enough to merit their own story, but Rep. McCaul was most frequently the focus
of a new story, with headline news stories making up about eight percent of his total news
coverage. He made the headlines a record thirteen times during this two-year period. A few
other members of Congress were not very far behind him, with headline news stories making up
about eight percent of Reps. Culberson, Al Green and Jackson Lees total coverage. Rep.
Culberson, who is mentioned more than he provides commentary, made headlines relatively
often compared to other members of Congress who weld considerable more political clout and
have a more effective issue agenda, such as Rep. Brady. Rep. Culbersons frequent appearance
in the headlines was due to his opposition to the Metro rail expansion, which again is another
indication that he has been ineffective at establishing his issue agenda with local media outlets.
46

When determining whether or not press releases yielded media coverage, it is useful to
make a comparison between the number of press releases about particular issues and the number
of news stories produced on those same issues. Within this comparison, one can test the
effectiveness of the issue agenda that a member of Congress developed through the distribution
of press releases. Figure 12 graphically displays this comparison. Rep. McCaul was the only
member of the sample who received a greater percentage of his total news coverage about
national security than he himself devoted to national security in his press releases. Rep. Gene
Greens news coverage of energy and environment issues is slightly less than the percentage of
which he released press releases about energy or the environment. For the remaining six
members of the sample, their percentage of the press release input regarding their issue agenda
was considerably lower than the percentage of total media coverage that they received.

Figure 11: Role of Members of Congress in Newspaper Articles


Mention
Comment
Headline
Average Role
48%
47%
5%
Rep. Poe
48% 49%
3%
Rep. Brady
46%
51%
3%
Rep. Culberson
32%
61%
7%
Rep. A. Green
22%
71%
7%
Rep. G. Green
66%
32%
2%
Rep. Jackson Lee
43%
50%
7%
Rep. McCaul
65%
27%
8%
Rep. Olson
52%
42%
6%
47
When performing the same comparison for the secondary issue of each members issue
agenda, the results were similar to that of the first comparison. Only Rep. Al Greens foreign
policy related news coverage received a higher percentage of his total news coverage, than did
his press release input about issues of foreign policy. The remaining seven members of the
sample input a higher percentage of their total number of press releases about their secondary
issue than they received output about that same issue in the news media. For particular
members, such as Rep. McCaul who devoted a considerable amount to maintaining his national
security issue agenda, the other issues that he focused on had far less traction in the news media
than did national security.


48




20%
32%
10%
12%
14%
16%
18%
22%
24%
26%
28%
30%
34%
36%
40%
42%
44%
46%
48%
50%
52%
20%
32%
10%
12%
14%
16%
18%
22%
24%
26%
28%
30%
34%
36%
40%
42%
44%
46%
48%
50%
52%
0%
2%
4%
6%
8%
0%
2%
4%
6%
8%
Member of Congress
Rep. Brady
Rep. Culberson
Rep. A. Green
Rep. G. Green
Rep. McCaul
Rep. Olson
Rep. Poe
Top Category
Budget
Budget
Justice
Energy/Environment
National Security
Energy/Environment
Foreign Policy
% of total input
29.91%
16.96%
17.39%
21.05%
34.29%
25.86%
17.73%
Press Release Input
% of output
18.58%
7.04%
14.53%
20.54%
49.38%
15.69%
10.22%
Media Output
Most Freequent Press Release Category Compared to Categorys Media Output
20%
10%
12%
14%
16%
18%
22%
24%
26%
28%
30%
20%
10%
12%
14%
16%
18%
22%
24%
26%
28%
30%
0%
2%
4%
6%
8%
0%
2%
4%
6%
8%
Second Most Frequent Press Release Category Compared to Categorys Media Output
Member of Congress
Rep. Brady
Rep. Culberson
Rep. A. Green
Rep. G. Green
Rep. McCaul
Rep. Jackson Lee
Rep. Olson
Rep. Poe
Top Category % of total input
Press Release Input
% of output
6.18%
2.82%
11.11%
9.82%
6.84%
4.94%
5.88%
5.11%
Media Output
Economy/Jobs
Healthcare
Foreign Policy
Constituency
Transportation
Foreign Policy
Constituency
Energy/Environment
26.07%
9.36%
8.70%
19.74%
19.28%
17.14%
13.79%
12.06%
Figure 12: Press Release Input Compared to Media Output by Top Issue Categories
% ofPress Releases % of Media Output
% ofPress Releases % of Media Output
Rep. Jackson Lee Foreign Policy 15.12% 2.56%
49
Direct Causality
To determine whether or not a press release directly incited a news story, I created a new
variable called hit that served as my dependent variable. This variable was created by
combining aspects of the data gleaned from the content analysis of the press releases to that of
the media coverage. A press release of a given issue category was coded as 1 if a news story
categorized in the same issue category appeared within seven days from the press release issue
date. If there was not a corresponding news story within those seven days, the press release was
coded as 0. This process was automated using Stata for fifteen of the seventeen issue
categories. The issue categories constituency and other were coded manually, as these
categories encompassed a broad range of story topics, unlike the other fifteen issue categories.
Hit, as the dependent variable, was compared against independent variables in four different
models. The independent variables in the first model were dummy variables assigned to each
rhetorical strategy. When a press release had at least one sentence that was categorized as
personalization, the dummy variable for personalization was coded one, and when there were not
any sentences in a press release that were categorized as personalization, the dummy variable
was coded as zero. Probit analysis with these dummy variables allowed for an evaluation of
whether the presence of each rhetorical strategy had an impact on the likelihood that press
release would incite a news story. The independent variables in the second model were the
relative proportional values of each rhetorical strategy. Probit analysis with the proportional
values as the independent variables allowed for an evaluation of whether the presence, as well as
the amount of each rhetorical strategy, impacted the likelihood that a press release would incite a
news story role. For these first two models, I was able to control for each rhetorical strategy by
including an independent variable representing each of the three rhetorical strategies into each
50
model.
37
In these two models, I was able to control for other attributes that might alter the
likelihood that a press release would directly incite a news story, such as party affiliation,
individual member effects, sentence length of the press release, and issue category.
The third model proceeds with the dummy independent variables used in model one, as
the independent variables used in model two were found to be inconclusive, and it uses these
independent variables to calculate each press releases predicted probability of getting a hit.
Model three presents an analysis of these predicted probabilities, examining how the predicted
probability of getting a hit changes with the presence, or lack of presence, of each rhetorical
strategy. Model three examines the predicted probability of a press release getting a hit in each
issue category.
38

The fourth model proceeds with the dummy independent variables and builds off the
results from model three regarding the predicted probability of press releases from each issue
category. Specifically, model four focuses on issue areas in which personalization worked to
increase the predicted probability of getting a hit by a large amount. A new variable was created
to capture the interaction effect between those issue areas and the presence of either
personalization or assertion. Similarly, a variable was also created to capture the interaction
between personalization and assertion, as model three found this interaction to be quite effective.
Those issues that were determined the most likely to receive a hit were also combined into a new
independent variable, which was run in a probit analysis with the dummy independent variables
to reassess each rhetorical strategys effectiveness of garnering a hit while controlling for these
special issue categories.

37
Although data was also collected regarding the presence and amount of text categorized as information,
information was not included in the analysis.
38
Issue categories immigration, trade, and veterans/military were exempt from this analysis because these
categories predicted failure perfectly.
51
Hit as a dependent variable
Members of Congress receive a hit, meaning that one of their press releases directly
incited a news story within seven days of the press release issue date, relatively infrequently.
During the 112
th
session of Congress, twenty-nine was the maximum number of hits achieved by
a member of this sample. Getting a hit is thus somewhat of a rare occasion, as it happened less
than a quarter of the time that a press release is distributed for all members of this sample and
less than ten percent of the time for two of the members. Figure 13 displays the number of hits
each member of congress received as a percentage of the total number of press releases they
distributed during the 112
th
Congress. Rep. Culberson received the least amount of hits, only
directly inciting six news stories over the course of the two-year session. Only four percent of
Rep. Culbersons press releases helped him to directly cultivate news coverage. In comparison,
Reps. Brady, Jackson Lee, McCaul and Poe all incited more than twenty news stories. Rep.
Jackson Lee distributed far more press releases than did the other seven members of the sample,
so while she directly incited twenty-four news stories, those twenty-four instances made up only
eight percent of her total press release output. Rep. Gene Green the most success and was the
most effective at directly inciting news coverage, as about twenty-two percent of the press
releases he distributed received a hit.

52

When breaking these results down by issue category, the results were reflective of which
members were the most effective at directly inciting news coverage. Figure 14 displays the
number of hits received my each issue category. The issue categories national security and
budget received the most hits, which was to be expected since Reps. Brady and McCaul received
the most hits and their issue agendas focused on budget and national security respectively. Also
not surprising were the number of hits received by the categories energy/environment and
constituency, as Rep. Poe received a considerable number of hits and he had a strong
energy/environment issue agenda. Although Rep. Jackson-Lee had a foreign policy related issue
agenda, the constituency issue area received a considerable number of hits because of her as
her most frequented category of news output was constituency.
Brady Culberson A. Green G. Green Jackson Lee McCaul Olson Poe
0
5%
10%
15%
20%
25%
Figure 13: Hitsas a Perentage of Total Number of Press Releases by Member
of Congress
Member of Congress %
Rep. Brady
Rep. Culberson
Rep. A. Green
Rep. G. Green
Rep. McCaul
Rep. Jackson Lee
Rep. Olson
Rep. Poe
12%
4%
20%
22%
8%
21%
15%
16%
53

Evaluating the distribution of the dependent variable hit demonstrates that receiving a hit
is a fairly rare occurrence when considered in the context of the total number of press releases
distributed by each member. Since each member of the sample has a fairly low rate at which
they receive hits, a rhetorical strategys ability to increase or decrease the likelihood that a press
release directly incites news coverage would be important because with such low success rates,
strategies that would increase the success rate by even a few percentage points would have
tangible effects.
Model 1:
Model one
39
uses dummy variables for each of the independent variables. These
variables are named: personalization, assertion and hybrid, representing the three rhetorical
strategies that were coded for during the data collection process. In this model, the presence of
personalization and assertion both increased the likelihood that the press release would incite a
news story. However, the difference between personalization and assertion was not found to be

39
See appendix Tables 1-9

Figure 14: Total Hits by Issue Category


B
u
d
g
e
t
C
o
n
g
r
e
s
s
C
o
n
s
t
i
t
u
e
n
c
y
E
c
o
n
o
m
y
/
J
o
b
s
E
d
u
c
a
t
i
o
n
E
l
e
c
t
i
o
n
s
E
n
e
r
g
y
/
E
n
v
i
r
o
n
m
e
n
t
F
o
r
e
i
g
n

P
o
l
i
c
y
H
e
a
l
t
h

C
a
r
e
I
m
m
i
g
r
a
t
i
o
n
J
u
s
t
i
c
e
N
a
t
i
o
n
a
l

S
e
c
u
r
i
t
y
S
p
a
c
e
T
r
a
d
e
T
r
a
n
s
p
o
r
t
a
t
i
o
n
V
e
t
e
r
a
n
s
/
M
i
l
i
t
a
r
y
Budget (26)
Congress (4)
Constituency (19)
Economy/Jobs (8)
Education (1)
Election (1)
Energy/Environment (23)
Foreign Policy (11)
Health Care (8)
Immigration (0)
Justice (8)
National Security (28)
Space (8)
Trade (0)
Transportation (6)
Veterans/Military (0)
0
5
10
15
20
25
30
54
statistically significant, with a P-Value of 0.826. Assertion increased the likelihood that a press
release would incite a news story just slightly, as the marginal effect of assertion increased more
from zero to one than did the marginal effect of personalization.
40
The hybrid strategy generally
worked to decrease the likelihood of getting a hit. In describing the marginal effects, assertion
increased the probability of receiving a hit by about 1%, where as personalization increased the
probability of getting a hit by about 1.5%. In this model, we can be about 83% confident that the
effect of personalization is greater than zero, where as we can be about 82% confident that the
effect of assertion is greater than zero. However, based on their Z-Scores, we cannot distinguish
any of the coefficients representing the rhetorical strategies from zero.
Table 2. Dummy independent variables
Variable Coefficient Standard Error Z P>|z|
Personalization 0.13378 0.09726 1.38 0.169
Assertion 0.16642 0.12433 1.34 0.181
Hybrid -0.10376 0.09713 -1.07 0.285
Constant -1.33036 0.13568 -9.80 0.000
N= 1214

Proceeding with this same set of variables, I controlled for the individual member effects.
Members of congress themselves may be influencing how much news coverage they receive;
impacting how often they directly incite news stories from their press releases. This model thus
helps to weed out the influence of member characteristics such as name recognition, tenure in
office and likeability, making it possible to evaluate the impact of each of the rhetorical
strategies without these personal characteristics that might be confounding the results calculated
for each the rhetorical strategies. When controlling for the influence of each individual member,
it used the first member, in this case Rep. Brady, as a baseline. I found that Reps. Culberson and
Jackson-Lee had a negative influence on likelihood of getting a hit. This means that by just

40
Marginal effects for this probit analysis are listed in Table 22 of the appendix
55
being himself, Rep. Culberson reduced the likelihood of directly inciting a news story with a
press release. The same can be said about Rep. Jackson Lee, but the negative impact she had was
much less than that of Rep. Culberson. In comparison, being Reps. McCaul, Gene Green, Al
Green, Olson or Poe increased the likelihood of getting a hit. Based on the Z scores, the
coefficients for each of the rhetorical strategies could not be distinguished from zero with high
confidence. Reps. Culberson McCaul, and Gene Green had statistics that were found to be
distinguishable from zero with high confidence.
Table 3. Dummy independent variables controlling for member
Variable Coefficient Standard Error Z P>|z|
Personalization 0.10376 0.10035 1.03 0.301
Policy Assertion 0.14163 0.12824 1.10 0.269
Hybrid -0.14061 0.10306 -1.36 0.172
N= 1214
Table 4. Individual member effects
Member Coefficient Standard error Z P>|z|
Culberson -0.61747 0.21160 -2.92 0.004
A. Green 0.36249 0.24178 1.50 0.134
G. Green 0.42604 0.19364 2.20 0.028
Jackson-Lee -0.21867 0.15062 -1.45 0.147
McCaul 0.33712 0.16078 2.10 0.036
Olson 0.14302 0.17875 0.80 0.424
Poe 0.22379 0.16824 1.33 0.183

Next, I supplemented the original three indepdent variables with one additional dummy
variable representing each members party affiliation. This model controled for party, adding an
independent dummy variable representing affiliation with the Democratic Party. This new
variable was called democrat, as press releases put out by members affiliated with the
Democratic party were assigned a 1 and those put out by Republicans were assigned a 0.
Democrat was found to be statistically significant with a P-value of 0.001. I found that being a
democrat generally decreased the likelihood of receiving a hit. When controlling for party, the
marginal effects of personalization and assertion remain somewhat constant with the results from
56
the first probit.
41
The main difference in the marginal effects of this model was that using
assertion tended to decrease the likelihood of receiving a hit by about 1.2%. Calculating the
marginal effects allowed for an evaluation of the marginal effects of being a Democrat or a
Republican. I found that democrats were about 1.5% less likely to receive a hit than were
republicans. For a member of congress who distributes many press releases like Rep. Jackson
Lee, this would equate to a loss of about five news stories. Importantly, Democrat was also
found to be distinguishable from zero with ninty-five percent confidence.
Table 5. Controlling for Party
Variable Coefficient Standard Error Z P>|z|
Personalization 0.12809 0.09804 1.31 0.191
Assertion 0.17279 0.12485 1.38 0.166
Hybrid -0.11038 0.09774 -1.13 0.259
Democrat -0.32823 0.09523 -3.45 0.001
Constant -1.12220 0.14835 07.56 0.00
N= 1214

Next, I supplemented the original three independent dummy variables with a new
independent variable representing the number of sentences in the press releases. This effectively
controlled for length, allowing me to evaluate how sentence length altered the impact of each of
the rhetorical strategies The difference between personalization and assertion in this model was
not found to be significant, with a P-value of 0.942. Generally, Sentence length increased the
likelihood of receiving a hit. In this probit analysis, none of the four independent variables were
found to be distinguishable from zero with high confidence.
Table 6. Controlling for sentence length
Variable Coefficient Standard Error Z P>|z|
Personalization 0.107819 0.09908 1.09 0.277
Assertion 0.11864 0.12858 0.92 0.356
Hybrid -0.12122 0.09190 -1.23 0.217
Sentence Length 0.00826 0.00548 1.51 0.131
Constant -1.36826 0.13777 -9.93 0.00

41
The marginal effects for this probit can be found in Table 23 of the appendix
57
Lastly, I ran the original probit analysis with the three independent dummy variables and
controlled for the issue categories. Three of the categories were not included as control variables
because they predicted failure perfectly and including them in the probit analysis would have
limited the number of cases analyzed. When controlling for the fourteen remaining issue
categories, Personalization was found to increase the likelihood that a press release would
directly incite a news story, where assertion was found to decrease the likelihood that a press
release would directly incite a news story. Assertion was not statistically significant, with a P-
value of 0.993, where personalization was close to the 0.05 P-value needed to achieve 95%
confidence with a P-value of 0.070. The difference between personalization and assertion was
not found to be significant, with a P-value of 0.276. The rhetorical strategies were also not found
to be distinguishable from zero with high confidence as their Z-scores were less than 1.96.
However, the control variable for each issue category was found to be distinguishable from zero
with high confidence and all fifteen issue areas had a P-value of less than 0.05. This suggests
that the issue topic of a press release is an important characteristic when determining a press
releases likelihood of receiving a hit. All of the issue categories were found to have coefficients
that were positive and since they were all found to be positive, comparing the coefficients and
the marginal effects of these issue categories as independent variables provides provide a general
sense as to which issue categories were the most effective at increasing the likelihood of
receiving a hit. Press releases categorized as justice or national security tended to increase the
likelihood of receiving a hit the most with a coefficients of 2.06 and 2.02 respectively. Based on
the marginal effects of each of the control issue category variables, being categorizes as an issue
pertaining to justice increased the likelihood of getting a hit by about 38%, where as being
categorized as an issue of national security increased the likelihood of getting a hit by about
58
35%.
42
Other issue categories which were more likely to receving a hit (relative to categories
with lower likelihoods) were election, with a coefficient of 1.83, energy/environment with a
coefficient of 1.77, and constituency with a coefficient of 1.7. Based on the marginal effects of
each of the control issue category variables, being categorized as election increased the
likelihood of getting a hit by about 37% being categorized as energy/environment increased the
likelihood of getting a hit by about 33% and being categorized as constituency increased the
likelihood of getting a hit by about 43%. With coefficients around 1.5, the issue categories space,
transportation and budget were also increased the likelihood of receiving a hit by a relatively
large ammount. The marginal effects of these three categories were: 33%, 32% and 27.9%
respectively. The remaining four issue categories healthcare, economy/jobs, education and
foreign policy were the least likely to receive a hit with coefficients around 1.1.
Table 7. Controlling for issue
Variable Coefficient Standard Error Z P>|z|
Personalization 0.19363 0.10686 1.81 0.070
Assertion -0.00127 0.14898 -0.01 0.993
Hybrid -0.68070 0.10833 -0.63 0.530
Budget 1.52315 0.28357 5.37 0.00
Congress 1.47676 0.39173 3.77 0.00
Constituency 1.71796 0.29634 5.80 0.00
Economy/Jobs 1.06067 0.31407 3.38 0.001
Education 1.12922 0.57225 1.97 0.048
Election 1.83316 0.73899 2.48 0.013
Energy/Environment 1.77465 0.29415 6.03 0.000
Foreign Policy 1.14764 0.30242 3.79 0.000
Healthcare 1.18218 0.31946 3.70 0.000
National Security 2.01599 0.29368 6.86 0.000
Justice 2.05654 0.366483 5.61 0.000
Space 1.58734 0.33797 4.70 0.000
Transportation 1.51969 0.35728 4.25 0.000
Constant -2.60263 0.28466 -9.14 0.000


42
The marginal effects for this probit can be found in Table 25 of the appendix
59
In conclusion, in the first probit model one found that personalization and assertion both
increased the likelihood of receiving a hit. However, the independent variables in the first probit
were found to be statistically the same as zero. When controlling for individual member effects,
the independent variables for the rhetorical strategies were still found to be statistically the same
as zero. Personalization and assertion still worked to increase the likelihood of receiving a hit
when controlling for member effects. Reps. Culberson and Jackson Lee decreased the likelihood
of receiving a hit, while Reps. Gene Green, Al Green, McCaul, Poe and Olson all increased the
likelihood of receiving a hit.
43
When controlling for party affiliation, the Democrat control
variable was also found to be unequal from zero. Being a Democrat tended to decrease the
likelihood, as the three Democrats in the sample were about 1.5% less likely to receive a hit than
were the Republicans in the sample. Controlling for sentence length did not have any tangible
impact on the effects of the three rhetorical strategies. When controlling for issue, assertion no
longer worked to increase the likelihood of getting a hit, while personalization was the only
rhetorical strategy that did so. All of the issue category control variables were found to be
different from zero. The categories justice, national secuirty, election, energy/environment, and
constituency increased the likelihood of getting a hit the most.
44

The hybrid indepdent dummy variable was not discussed because in each analysis its
coefficient was found to be negative in each analysis and it was found to be statistically equal to
zero. This could be for various reasons, one being that the hybrid strategy was employed the
least often which meant that there were less cases for which the analysis to generate the statistics
regarding the hybrid strategy


43
List provided in descending order, with Rep Gene Green increasing the likelihood of receiving a hit the most.
44
List provided in descending order, with Justice increasing the likelihood of receiving a hit the most.
60
Model 2:
The second model
45
used proportions as the value for each of the independent variables:
personalization, assertion and hybrid. For each of the probit analyses run with these variables, all
of the coefficients for the rhetorical strategies were negative and none of these variables were
found to be statistically different from zero. This suggests this model did not capture the effects
of each independent variable as hoped. The sentence level coding did not capture the rhetorical
strategies at work and for this reason, the findings using proportional variables were laregly
inconclusive.
The analysis done using the proportional independent variables ran the same probit
analyses as model one, substituting the dummy variables for the proportional variables. Two of
the analyses demonstrated limited results, these two analyses being the control for party and the
control for member effects.
When controlling for individual member effects, personalization was found to increase
the likelihood of getting a hit more so than assertion but again, both coefficients were negative
indicating that they both worked to decrease the likelihood of getting a hit. As for the individual
member results, model two expands upon the results from model one in that four of the seven
members of the sample, for which the coefficients were calculated, had coefficients that were
found to be distinguishable from zero with high confidence. Being Reps. Al Green, Gene Green,
and McCaul and increased the likelihood of getting a hit, while being Rep. Culberson decreased
the likelihood of getting a hit.
In this analysis, being Al Green would increase the likelihood of getting a hit the most a
finding quite different from the results found during the model one analysis analysis with the
dummy variables, where Rep. Gene Green increased the likelihood of getting a hit the most.

45
See appendix Tables 10-17
61
Looking back to the total numbers of hits by member, Rep. Al Green had the second fewest
number hits overall making these results quite compelling and suggesting that his use of the
rhetorical strategies and the degree to which he employs each rhetorical strategy may be
important.
Table 8. Controlling for member
Variable Coefficient Standard Error Z P>|z|
Personalization -0.28235 0.27918 -1.01 0.312
Assertion -0.30307 0.22289 -1.36 0.174
Hybrid -1.04808 0.53311 -1.97 0.049
Table 9. Individual member effects
Member Coefficient Standard Error Z P>|z|
Culberson -0.61882 0.21252 -2.91 0.004
A. Green 0.48461 0.24727 1.96 0.050
G. Green 0.44623 0.19352 2.31 0.021
Jackson-Lee -0.18362 0.15154 -1.21 0.226
McCaul 0.33207 0.16128 2.06 0.039
Olson 0.13918 0.17897 0.78 0.437
Poe 0.24551 0.16864 1.46 0.145
N: 1213

When controlling for the effect of party affiliation, the effectiveness of each rhetorical
strategy was largely inconclusive but the Democrat control variable was found to be
distinguishable from zero with high confidence, with a Z-Score of 3.22. Being a Democrat
decreased the likelihood that of receiving a hit.
Table 10. Controlling for party
Variable Coefficient Standard Error Z P>|z|
Personalization -0.23527 0.26923 -0.87 0.382
Assertion -0.32560 0.21304 -1.53 0.126
Hybrid -0.69586 0.47986 -1.45 0.147
Democrat -0.30907 0.09597 -3.22 0.001
Constant -0.77918 0.11248 -6.93 0.00
N= 1213

In conclusion, the analyses conducted in model two produced few statistically significant
results. This model was useful in demonstrating that sentence level data did not produce results
62
that could be interpreted and explained. This suggests that the rhetorical strategies may be taking
effect at a larger level of analysis, such as the paragraph or even the press release level. This
also suggests that the amount that each rhetorical strategy was employed may not be as important
as the employment of the presence of strategy in general.
Model 3:
While an analysis based on marginal effects and statistical significance may be useful in
order to understand the statistical importance of those concepts being tested, an analysis of the
calculated probabilities for each of those concepts being tested would address my hypothesis
more directly. Since I am attempting to make comparisons between various rhetorical strategies,
it is helpful to evaluate the effectiveness of these rhetorical strategies by examining the impact
each strategy has on a press releases predicted probability of getting a hit. Stata was used to
automate the prediction process, leaving me with a predicted probabiltiy for each press release.
These individual press release results were then analyzed as a whole to calcualte the overall
probability that a hit will result from a press release with the baseline press release being one that
does not employ any of the three rhetorical strategies. With this baseline, it was then possible to
compare how the predicted probability changed as the implemnation of rhetorical strategies
changed. I then proceeded to repeat this process within eleven of the fourteen issue categories
used as controls in models one and two.
Overall, the probability of getting a hit without employing one of the three rhetorical
strategies was found to be .092. Employing both personalization and assertion in the same
release was found to be the most effective strategy, as it increased the probability of getting a hit
to .152, an increase of about 6%. In general, employing all of the rhetorical strategies together
was also very helpful, as it increased the likelihood of getting a by about 3.7% to a probability of
63
.128. Using only personalization was found to have a probability of .116, compared to using
only assertion, which had a probability of .112. Using only the hybrid rhetorical strategy
generally decreased the likelihood of getting a hit - its probability was found to be .076, a
decrease of about 1.6%.
Table 11. Calculated Probabilities
Rhetorical Strategy Probability Frequency Percent
Hybrid only 0.07577 11 0.91
No Strategies 0.0917 48 3.95
Personalization and Hybrid 0.0967 53 4.37
Assertion and Hybrid 0.1025 141 11.61
Assertion only 0.1122 257 21.17
Personalization only 0.1157 117 9.64
All strategies 0.1284 246 20.26
Personalization and Assertion 0.1515 341 28.09

Overall, personalization only and both personalization and assertion were found to be the
most effective rhetorical strategies, in that for each of the issue categories a press release with
either option had the highest probability. For every issue category, the probabilities of these two
options (to be called the effective options) were separated by less than a percentage point. This
analysis shows that employing either option would be the most effective strategy for any given
issue category. This analysis also demonstrates that using policy assertion alone tends to
decrease the likelihood of getting a hit, as the probability of only employing assertion is less than
the probability of getting a hit when only information has been provided for each issue category
(to be called the baseline).
Press releases categorized as justice tended to be the most likely to directly incite news
coverage. Although there was no probability for the baseline for issues of justice, press releases
in each the other four rhetorical options had higher probabilities than did all of the other issue
categories. While issues of justice were the most likely to get a hit, issues of national security
64
displayed the largest percentage change from the baseline to either personalization only or both
personalization and assertion. Issues of national security that employed either option were 6.8%
more likely to get a hit than the baseline option without any of the rhetorical strategies. Issues of
energy/environment showed a 5.9% change between the baseline and the effective options and
issues of constituency displayed a 5.6% change between the baseline and the effective options.
In contrast, press releases categorized as economy/jobs were found to have the lowest
probability of directly inciting news stories, with a baseline probability of .06 and a percent
change to the effective options of only 2.7%. Issues of foreign policy and healthcare were also
found to be less likely to incite news stories in compared to the other issues. Press releases
written about issues of foreign policy had a baseline probability of 0.07 and a percent change of
only 3.1 % from the baseline to the effective options. Healthcare had a baseline probability of
0.08 and a percent change of 3.2% from the baseline to the effective options.
Table 12. Calculated Probabilities by issue category
46

Issue category Baseline Only
Personalization
% Only
Assertion
% Both % All %
Justice 0.36225 0.29207 0.36177 0.33659
National Security 0.27872 0.34716 6.84 0.27830 0.00 0.34669 6.79 0.32192 4.32
Energy/Envir 0.20384 0.26293 5.91 0.20348 0.00 0.26251 5.87 0.24081 3.70
Constituency 0.18817 0.24477 5.66 0.18783 0.00 0.24437 5.62 0.22351 3.53
Space 0.15498 0.20564 5.07 0.15468 0.00 0.20528 5.03 0.18647 -6.3
Budget 0.14019 0.18785 4.77 0.13991 0.00 0.18751 4.73 0.15827 1.81
Congress 0.13011 0.17561 4.55 0.12984 0.00 0.17528 4.52 0.15827 2.82
Healthcare 0.07774 0.10995 3.22 0.07755 0.00 0.10971 3.20 0.09746 1.97
Foreign Policy 0.07284 0.10359 3.08 0.07266 0.00 0.10336 3.05 0.09164 1.88
Economy/Jobs 0.06154 0.08878 2.72 0.06139 0.00 0.08857 2.70 0.07814 1.66

Employing all three rhetorical strategies was also found to be helpful, but not as helpful
as employing only personalization or both personalization and assertion. Employing all three

46
Issue categories: education, transportation, and election were left out of this analysis because the calculated
probabilities for these issue areas could be calculated for less than three of the eight possible rhetorical strategy
combinations.
65
rhetorical strategies was most helpful for issues of national security, as the option employing all
strategies was found to be 4.3% greater than the baseline.
In conclusion, the calculated probabilities of each combination of the three rhetorical
strategies helps to demonstrate which strategies have the highest probability of receiving a hit.
This model found that employing only personalization or employing both personalization and
assertion were the most effective options, in that they increased the probability of getting a hit
the most in comparison to those other combinations. Using only personalization or using both
personalization and assertion will henceforth be referred to as the effective options. I was also
able to calculate the probabilities for each combination of rhetorical strategies within each issue
category and found that the issue categories justice and national security had the highest
probability of receiving a hit. Thus, employing either of the effective options in either or these
two issue categories in a press release would substantially increase the probability that a member
of congress would directly incite a news story. Justice was found to be the most effective issue
category and economy/jobs was found to be the least effective issue category.
Model 4:
Model three found that five of the ten possible issue categories were more likely to get
hits than the remaining five. Model four builds on this finding, by creating a new dummy
variable to represent those issues that generally have a higher probabilty of directly inciting news
coverage than other issues do. Controlling for these issues will allow for an evaluation of how
the rhetorical strategies function generally and it will also provide results that are somewhat
more generalized as some of the super issues are issues that are very important to Houston, but
may not be for other regions of the country. These super issues are: justice, national security,
energy/environment, constituency, and space. While justice and constituency are issues that are
66
going to be important to public represntatives serving anywhere in the United States, Houston
Texas has a special relationship with the issues areas: national security, energy/environment and
space. Houstons is relatively close to the U.S. border and it also has the largest U.S. port for
international commerce, characteristics which make national security an important issue for
Houston. A large part of Houstons economy is based in energy and it recognized worldwide for
its focus on oil and natural gas. The Houston region is also the home of the Johnson Space
Center, which used to employ many people and served as a source of pride for the region.
Controlling for these issues should provide a more generalized understanding of how the
rhetorical strategies function independently from the issue categories.
Table 13: Controlling for Super issue
Variable Coefficient Standard Error Z P>|z|
Personalization 0.12654 0.10082 1.26 0.209
Assertion 0.20273 0.12900 1.57 0.116
Hybrid -0.45639 0.10085 -0.45 0.651
Superissue 0.75942 0.09726 7.81 0.00
Constant 1.65889 0.14939 -11.10 0.00
N=1214

When controlling for the super issues, personalization and assertion still work to
increase the likelihood of getting a hit. In this probit analysis, the rhetorical strategies were not
found to be distinguishable from zero with high confidence based on their respective Z-Score.
The super issues control variable was found to be distinguishable from zero with high
confidence based on its Z-Score. Generally, being a super issue increased the likelihood of
getting a hit and it did so more than the presence of the rhetorical strategies. Being a super issue
increased the likelihood of receiving a hit by about 5% based on the marginal effect of the super
issue control variable, and while controlling for the super issues, assertion was found to increase
67
the likelihood of receiving a hit by about 4% and personalization was found to increase the
likelihood by about 3%.
47

Since model 3s results suggested that the interaction between personalization and
assertion was important, I next ran a probit analysis with personalization, assertion, the
interaction between personalization and assertion and the super issues control variable.
48
This
analysis provides an idea of how effective the interaction of the rhetorical strategies is when
controlling for the super issues.
Table 14. Controlling for the interaction between Personalization and Assertion and for super issue
Variable Coeficient Standard Error Z P>|z|
Personalization 0.05453 0.27399 0.20 0.842
Assertion 0.13483 0.25453 0.53 0.596
Personalization*Assertion 0.08180 0.29469 0.28 0.781
Superissue 0.76491 0.09709 7.88 0.00
Constant -1.61708 0.24252 -6.67 0.00
N=1214
When controlling for both the super issue category and the interaction between
personalization and assertion, personalization and assertion both work to increase the likelihood
of receiving a hit. However, we can not distinguish these effects from zero with high confidence
based on their Z-scores. To get the effect of interaction term, the coefficient provided by the
analysis should be subtracted from the combination of the coefficients personalization and
assertion. This value was found to be 0.10756, which suggests that when personalization and
assertion are both present in a press release it increases the likelihood of receiving a hit. The
interaction between personalization and assertion was found to be more effective than only
personalizing but less effective than only using assertion.
Next, I ran a similar probit, this time controlling for the interaction between superissues
and personalization as well as the interaction between the interaction term for personalization

47
Calculated using the marginal effects of each rhetorical strategy, table 33 is provided in the appendix.
48
The hybrid strategy was left out of this analysis because of the inconclusiveness of its effects
68
and assertion and the superissues category. This analysis would allow for a comparison within
the context of those super issues between those two most effective options identified in model
three.
Table 15. Controlling for the interaction between personalization and assertion, super issues, the
interaction between personalization and super issues and the interaction between personalization,
assertion and super issues
Variable Coeficient Standard Error Z P>|z|
Personalization 0.07393 0.30921 0.24 0.811
Assertion 0.13652 0.25411 0.54 0.591
Personalization*Assertion 0.03008 0.32666 0.09 0.927
Super issues 0.72843 0.16305 4.47 0.00
Personalization*Superissue -0.02845 0.30924 -0.09 0.927
Personalization*Assertion*Superissue 0.10756 0.29586 0.36 0.716
Constant -1.60349 0.24679 -6.50 0.00
N=1214

In this analysis, the results for personalization and assertion provide the predicted
effects of personalization for those issues not categorized as superissues. Based on this analysis,
assertion should be more effective in increasing the likelihood of receiving a hit than
personalization. We also have a predicted effect for the interaction between personalization and
assertion, the effective option of having both personalization and asseriton in a press release. For
those issues not categorized as super issues, the predicted effect of having both personalization
and assertion would be found by combining personalization and assertion and subtracting
personalizaton*assertion. The coefficient for having both personalization and assertion in a
press release was thus found to be 0.18037, suggesting that this option increases the likelihood of
receiving a hit for those issues not categorized as a super issues more so than personalizaton or
assertion do independently. To compare the relative impacts of the two most effective options
identified my model three, the predicted effects for super issues need to be calculated for both
options. To calculate the predicted effects of super issues that were only personalized, we need
to combine personalization and personalization*superissue to get a value of 0.04548. For the
69
predicted effect of having both personalization and assertion in a press release about a super
issue, we need to combine both personalization and assertion, subtract the interaction term
personalization*assertion and add personalization*assertion*superissue to get a value of
0.28793. This suggests that employing both personalization and assertion in the same press
release is more likely to directly incite a news story than only employing personalization.
While it is helpful to gauge the effectiveness between our two most effective options, it is
also helpful to compare the predicted effects of personalization and assertion to evaluate which
rhetorical strategy is more effective. I ran a probit controlling for the super issues, the interaction
between the super issues and personalization and the interaction between the super issues and
assertion.
Table 16. Controlling for super issues, the interaction between personalization and super issues, and the
interaction between assertion and super issues
Variable Coefficient Standard Error Z P>|z|
Personalization 0.09783 0.13095 0.75 0.455
Assertion 0.12294 0.17117 0.72 0.473
Super issue 0.57894 0.28136 2.06 0.040
Personalization*superissue 0.07402 0.20493 0.36 0.718
Assertion*superissue 0.16601 0.25653 0.65 0.518
Constant -1.59155 0.18312 -8.69 0.00
N=1214

This analysis provided us with two predicted effects for personalization and assertion.
The first predicted effect is the personalization variable, which provides the effect of
personalization on issues not classified as superissues. The second predicted effect of
personalization is the effect of personaliztion on those superissues, which is found by combining
personalization and superissue. This new coefficient was found to be 0.67677. The same
can be said about assertion, the predicted effect of assertion on superissues being found 0.70188.
To compare the effects of personalization and assertion specifically on superissues, this is done
70
by combining personalization and personalization*superissue and comparing it to the
combination of assertion and assertion*superissue. The effect of personalization on
superissues was found to be 0.17185 where the effect of assertion on superissues was found to be
0.28895. This suggests that assertion would increase the likelihood of getting a hit on
superissues more so than personalization would.
I then tested the interaction between the individual super issues and the rhetorical
strategies, as done with the superissue control. This will allow for an evaluation of whether
personalization or assertion was more effective within the context of specific issue categories.
Table 17: Controlling for Justice, Interaction between Personalization and Justice and the Interaction
between Assertion and Justice.
Variable Coefficient Standard Error Z P>|z|
Personalization 0.13432 0.09900 1.36 0.175
Assertion 0.14770 0.12524 1.18 0.238
Justice 1.01393 0.94770 1.07 0.285
Personalization*Justice -0.20599 0.57366 -0.36 0.720
Assertion*Justice -0.21937 0.84172 -0.26 0.794
Constant -1.37298 0.13622 -10.08 0.00
N=1214

For issues of justice, personalization has a predicted effect of 1.14825 and assertion has a
predicted effect of 1.6163. While both strategies work to increase the likelihood of getting a hit,
the predicted effect of assertion is greater than that of personalization. This shows that for issues
of justice, employing either rhetorical strategy increases the likelihood of getting a hit by quite a
lot as compared to issues not concerned with justice. To compare the relative impacts of
personalization and assertion, this is done by combining the coefficients for personalization
and personalization*justice to the combination of assertion and assertion*justice. The
relative impacts of personalization and assertion on issues of justice were found to be the same at
71
a value of -0.07167. This suggests that the relative impact of both rhetorical strategies is the
same for issues of justice.

Table 18: Controlling for National Security, Interaction between Personalization and the Interaction
between National Security and Assertion and National Security.
Variable Coefficient Standard Error Z P>|z|
Personalization 0.14464 0.10508 1.38 0.169
Assertion 0.10480 0.12826 0.82 0.414
National Security 0.86622 0.58634 1.48 0.140
Personalization*National
Security
-0.00033 0.29984 -0.00 0.999
Assertion*National
Security
-0.15083 0.56787 -0.27 0.791
Constant -1.39904 0.14056 -9.95 0.00
N=1214

Similar to the probit run with the superissues, in this analysis there are two predicted
effects for both personalization and assertion. For national security issues, personalization has a
predicted effect of 1.01086 and assertion has a predicted effect of .97102. The predicted effects
for issues not concerend with national security are presented by the personalization and
assertion variables While both strategies work to increase the likelihood of getting a hit,
personalization is more effective than assertion for issues of national security. This shows that
for national security issues, employing either rhetorical strategy increases the likelihood of
getting a hit by quite a lot. To compare the relative impacts of personalization and assertion, this
is done by combining the coefficients for personalization and personalization*national
security to the combination of assertion and assertion*national security. The relative
impacts of personalization and assertion on issues of national security were found to be 0.14431
for personalization and -0.04603 for assertion. This suggests that personalizing issues of
national security is more effective than providing policy assertions.


72
Table 19: Controlling for Space, Interaction between Personalization and Space and the Interaction
between Assertion and Space.
Variable Coefficient Standard Error Z P>|z|
Personalization 0.11908 0.09881 1.21 0.228
Assertion 0.19119 0.12903 1.48 0.138
Space 0.41241 0.64274 0.64 0.521
Personalization*Space 0.28625 0.65070 0.44 0.660
Assertion*Space -0.53582 0.50369 -1.06 0.287
Constant -1.39080 0.13978 -9.95 0.00
N=1214

Again there are two predicted effects for both personalization and assertion. For issues of
space, personalization has a predicted effect of 0.51349 and assertion has a predicted effect of
.6036. The predicted effect for non-space issues is demonstrated by the personalization and
assertion variables. While both strategies work to increase the likelihood of getting a hit,
assertion is more effective than personalization for issues of space. This shows that for issues of
space, employing either rhetorical strategy increases the likelihood of getting a hit by quite a lot
as compared to issues not concerend with space. To compare the relative impacts of
personalization and assertion, this is done by combining the coefficients for personalization
and personalization*space to the combination of assertion and assertion*space. The
relative impacts of personalization and assertion on issues of national security were found to be
0.40533 for personalization and -0.34463 for assertion. This suggests that personalizing issues
of space is more effective than providing policy assertions.
Table 20: Controlling for Constituency, Interaction between Personalization and Constituency and the
Interaction between Assertion and Constituency.
Variable Coefficient Standard Error Z P>|z|
Personalization 0.09368 0.10080 0.93 0.353
Assertion 0.17266 0.14010 0.123 0.218
Constituency 0.10936 0.44691 0.24 0.807
Personalization*Constituency 0.18255 0.42168 0.43 0.665
Assertion*Constituency 0.46416 0.34421 1.35 0.177
Constant -1.38841 0.14913 -9.31 0.00
N=1214

73
For issues of constituency, personalization has a predicted effect of .020304 and assertion
has a predicted effect of 0.28202. While both strategies work to increase the likelihood of
getting a hit, assertion should be more effective than personalization for issues of constituency.
This shows that for issues of constituency, employing either rhetorical strategy increases the
likelihood of getting a hit as compared to issues not concerend with constituency. To compare
the relative impacts of personalization and assertion, this is done by combining the coefficients
for personalization and personalization*constituency to the combination of assertion and
assertion*national security. The relative impacts of personalization and assertion on issues of
constituency were found to be 0.27623 for personalization and 0.63682 for assertion. This
suggests that providing policy assertions about issues of constituency was more effective than
personalizing.
Table 21: Controlling for Energy/Environment, Interaction between Personalization and
Energy/Environment and the Interaction between Assertion and Energy/Environment.
Variable Coefficient Standard Error Z P>|z|
Personalization 0.14018 0.10457 1.34 0.180
Assertion 0.12200 0.12664 0.88 0.376
Energy/Environment 0.37638 0.68855 0.55 0.585
Personalization*Energy 0.08985 0.29652 0.30 0.762
Assertion*Energy 0.01941 0.67568 0.03 0.977
Constant -1.37311 0.13916 -9.87 0.00
N=1214

For issues of energy/environemnt, personalization has a predicted effect of .51656 and
assertion has a predicted effect of 0.49838. While both strategies work to increase the likelihood
of getting a hit, personalization should be more effective than personalization for issues of
energy/environment. This shows that for issues of energy/environment, employing either
rhetorical strategy increases the likelihood of getting a hit as compared to issues not concerend
with energy/environment. To compare the relative impacts of personalization and assertion, this
74
is done by combining the coefficients for personalization and personalization*energy to the
combination of assertion and assertion*energy. The relative impacts of personalization and
assertion on issues of energy/environment were found to be 0.23003 for personalization and
0.14141 for assertion. This suggests that personalizing issues of energy or environment is more
effective than assertion in that it increases the likelihood of getting a hit more.
In conclusion, model four demonstrated that the interaction between personalization and
assertion was more effective than only employing personalization when controlling for the super
issues. When controlling for the interaction between personalization and assertion, the super
issues, the interaction between the super issues and personalization and the interaction
between the super issues and the interaction between personalization and assertion, employing
both personalization and assertion was identified as being more effective than either
personalization or assertion independnetly for those issues not categorized as super issues. That
same probit analysis also demonstrated that the interaction between personalization and assertion
was more effective for the super issues than was only personalization. When comparing the
effects of personalization and assertion in the context of the super issues, assertion was
identified as the more effective strategy. When evaluating whether personalization or assertion
was more effective for each of the individual super issues, the results were varied.
Personalization was found to be the more effective strategy when evaluating the interaction
effects between the rhetorical stragies and the issue categories national security, space and
energy/environment. Where as for issue categories justice and constituncy, assertion was found
to be the more effective strategy when evaluating the interaction effects between the rhetorical
strategies and these two issue categores. This suggests that personalization may be the most
effective when it is used to personalize an issue that isnt already personal or unique to the
75
district. Those issue categories justice and constituency deal primarily with local issues, and
would inherently have a local, more personalized angle than would the more generalized issue
categories that deal with a policy debate, such as national security. These findings support the
findings from model three, in that the issue category of a press release plays an important role in
determining that press releases likelihood of directly inciting news coverage.
Evaluating My Hypothesis
Drawing from the results presented in the Direct Causality section, I have evidence to
support my original hypothesis, that personalization is more effective than assertion in yeilding
media coverage. Models one and two were generally did not provide evidence that could be
used to prove or disprove my hypothesis because the difference between personalization and
asssertion was found to be statistically insignificant in each of the analyses. Also, the statistics
calculated for personaliztion and assertion in models one and two were not distinguishable from
zero with high confidence based on their Z-Scores. I believe model three provided the best
scenario for directly comparing the effectivness of personalization and assertion because it
provided the calculated probability for each combination of rhetorical strategies. Model three
found that employing only perosnalization in a press release tended to increase the likelihood of
directly inciting news coverage as compared to the baseline, where as employing only assertion
in a press release tended to decrease the likelihood of directly inciting news coverage as
compared to the baseline. With this evidence, I failed to reject my hypothesis. However, in
model three I also found that employing both personalization and assertion tended to increase the
likelihood of directly inciting news coverage compared to the baseline, a scenario that my
hypothesis failed to consider. Model four expands upon these results, suggesting that for the
combination of personalization and assertion was more effective for super issues than was
76
personalization. I also compared the effectiveness of personalization and assertion for each of
the five individual super issues, finding that the interaction between personalization and each
issue was only more effective than assertion for issues of national security, energy/environment,
and space. This being, for issues of justice and constituency assertion was found to be more
effective. The results from model four suggest that the results are more complex than model
three indicates, showing that assertion may be more effective for particular issues. But in
general, the findings from model three support my hypothesis, while the statistics regarding the
rhetorical strategies in models one, two and four were not distinguishable from zero with high
confidence.
Conclusion
Are congressional media strategies effective? There is no simple answer to this question,
because each member of congress employs a different strategy within the context of a specific
geographic location. For the most part, those legislators who had a strong issue agenda centered
around one or two legislative issues tended to receive more media coverage about those issues.
Although it is important to note that having a strong issue agenda isnt causally related to a high
frequency of news coverage about that particular issue, because there are other factors that are
difficult to quantify, such as having a leadership role in committee, tenure in office and general
name recognition that may be intervening and causing a member to receive more coverage than
they otherwise would of. As demonstrated in model three, certain issues are more likely to get
covered than others, so those members of congress who have chosen a super issue as part of
their issue agenda increase the probability that they will receive news coverage when distributing
press releases about those issues. Having a strong issue agenda seems to help the members of
77
congress of interest to this study receive more frequent news coverage about that issue,
especially when they are the only legislator in their geographic region focusing on that issue.
While this study only looked at eight members of Congress from one geographic
location, its finding can be generalized to suggest that any member of congress could
strategically utilize the two most effective options to further increase the likelihood they will
receive news coverage. However, the findings of this project are also limited, as the analysis
was conducted using data derived from only eight members of congress. By focusing narrowly
on one geographic area, my findings were somewhat limited to this geographic area. For
evaluating local media strategy, having a sample that was limited in size and geographic might
have been a benefit, because it captured some of the location specific characteristics such as the
locally important issue categories. Doing a broad, all encompassing study of a similar nature
would have provided results that were more generalized, but these results would have been
useful when explaining specific media markets, areas or individuals. I coded all of the primary
sources myself by hand, which may have worked to reduce the accuracy of my results.
To expand upon my findings, I would recommend that similar studies are done using
different geographic locations and different sized media markets. It would be beneficial to
compare the findings presented in this analysis of the Houston delegation and the Houston media
market to similar analyses done on other markets. Such research would complement the findings
of Vinson (2003) and Arnold (2004) and allow for further evaluation as to how the amount of
media coverage of members of congress changes with the location and media market size. Such
research would also allow for comparison between location and market size and it would
differentiate which aspects of strategy are broad and can be applied anywhere from those
characteristics that are media market specific or region specific. This project also only focused
78
on very specific rhetorical strategies defined within the very narrow context of press releases.
Further research could delve deeper into the rhetorical aspects of these strategies or even identify
other potential strategies at work. My analysis in model two was largely inconclusive,
suggesting that, for the type of probit analysis that I used, the sentence was not an appropriate
unit of analysis. Further research could use paragraphs as a unit of analysis, or focus more
broadly on each press release as a whole similar to what I did in models one, three and four.
The inconclusiveness of model two demonstrates the need for further research to be conducted to
better understand how the amount of which a rhetorical strategy was employed impacted the
effectiveness of the overall strategy. Congressional media strategy could also be researched by
evaluating the members of congress themselves and assessing the logic behind the decisions they
make regarding local media. It would also be helpful for a direct evaluation of this nature to be
conducted on local media professionals as well to asses their decision making process. This
would allow for further insight into the logic behind the usage of various rhetorical strategies.
Overall, the findings of this project may have practical implications for both members of
congress and media professionals. Members of Congress could make use of the effective options,
employing only personalization or both personalization and assertion, to increase the probability
of directly inciting a news story with a press release. Members could also distribute press
releases about those issues that were found to have a higher probability of getting a hit. For
media professionals, knowing about which strategies were found to be the most likely to incite
news coverage would simplify how press releases are evaluated. Knowing that there are
rhetorical strategies that increase the likelihood of inciting a news story might make media
professionals less likely to take the statements directly from press releases, for media
79
professionals would be aware that the statements were written with the intention of receiving
news coverage.





































80
Appendix

Model 1

Difference between personalization and assertion was not found to be significant with a P-Value of 0.826

Table 22. Marginal effects of dummy independent variables
Variable Dy/dx Standard Error Z P>|z| X
Personalization 0.02404 0.0171 1.41 0.160 0
0.02596 0.01857 1.40 0.162 1
Assertion 0.03052 0.02137 1.43 0.153 0
0.03167 0.02223 1.42 0.154 1
Hybrid -0.01593 0.01458 -1.09 0.275 0
-0.02305 0.0214 -1.08 0.281 1

Difference between personalization and assertion was not found to be significant with a p-value of 0.804

Difference between personalization and assertion was not found to be significant, p-value of the
difference was 0.764

Table 23. marginal effects of dummy independent variables contorlling for party
Variable Dy/dx Standard
Error
Z P>|z| x
Personalization 0.02920 0.0219 1.33 0.183 0
0.02128 0.01618 1.32 0.186 1
Assertion 0.04032 0.02765 1.46 0.145 0
0.02788 0.01884 1.48 0.139 1
Hybrid -0.02202 0.01917 -1.15 0.251 0
-0.02132 0.0187 -1.14 0.254 1
Republican -0.05742 0.02016 -2.85 0.004 0
Democrat -0.07190 0.02261 -3.18 0.001 1
The difference between personalization and assertion was not found to be significant with a value of
0.942

Table 24. Marginal effects of dummy independent variables controlling for sentence length
Variable Dy/dx Standard
Error
Z P>|z| x
Personalization 0.01814 0.01651 1.10 0.272 0
0.01907 0.01756 1.09 0.277 1
Assertion 0.02010 0.02111 0.95 0.341 0
0.02085 0.02193 0.95 0.342 1
Hybrid -0.01744 0.01374 -1.27 0.204 0
-0.02465 0.01963 -1.26 0.209 1
Sentence Length 0.00129 0.00085 1.52 0.128 0
0.00156 0.00093 1.67 0.094 5
(average) 0.00170 0.00112 1.51 0.131 11.784
0.00175 0.00119 1.47 0.143 15
0.00183 0.0013 1.41 0.160 20
81

The difference between personalization and assertion was not found to be significant with a P-value of
.276

Table 25. Controlling for issue - marginal effects
Variable Dy/dx Standard Error Z P>|z| X
Personalization 0.00337 0.00278 1.21 0.225 0
0.02663 0.01459 1.83 0.068 1
Assertion -0.00001 0.00201 -0.01 0.993 0
-0.00020 0.02349 -0.01 0.993 1
Hybrid -0.00084 0.00114 -0.59 0.588 0
-011219 0.01774 -0.63 0.527 1
Budget 0.13556 0.04211 3.22 0.001 0
0.41475 0.08858 4.68 0.00 1
Congress 0.12548 0.06972 1.80 0.072 0
0.44350 0.14562 3.05 0.002 1
Constituency 0.18354 0.0506 3.63 0.00 0
0.51657 0.09873 5.23 0.00 1
Economy/Jobs 0.05692 0.02718 2.09 0.036 0
0.26640 0.09912 2.69 0.007 1
Education 0.06569 0.07133 0.92 0.357 0
0.31118 0.21331 1.46 0.145 1
Election 0.21618 0.21047 1.03 0.304 0
0.58687 0.26403 2.22 0.026 1
Energy/Environment 0.19921 0.05797 3.44 0.001 0
0.53025 0.09536 5.56 0.000 1
Foreign Policy 0.06821 0.02953 2.31 0.021 0
0.29165 0.09479 3.08 0.002 1
Healthcare 0.07311 0.03562 2.05 0.040 0
0.31438 0.10691 2.94 0.003 1
National Security 0.27410 0.06872 3.99 0.000 0
0.62046 0.08862 7.00 0.00 1
Justice 0.28788 0.10455 2.75 0.006 0
0.65705 0.11219 5.86 0.00 1
Space 0.15036 0.06225 2.42 0.016 0
0.48147 0.12071 3.99 0.00 1
Transportation 0.13479 0.0649 2.08 0.038 0
0.45652 0.12987 3.52 0.00 1

Table 26. Controlling for interaction between personalization and assertion
Variable Coefficient Standard Error Z P>|z|
Personalization 0.17010 0.26676 0.64 0.524
Assertion 0.19788 0.24891 0.80 0.426
Hybrid -0.10415 0.09716 -1.07 0.284
Interaction -0.04191 0.28628 -0.15 0.884
Constant -1.3582 0.23450 -5.79 0.00
N=1214
82
Table 27. Marginal effects of probit controlling for interaction between personalization and assertion
Variable Dy/dx Standard error Z P>|z| x
Personalization 0.03020 0.0449 0.67 0.501 0
0.03221 0.4536 0.71 0.478 1
Assertion 0.03576 0.04099 0.87 0.383 0
0.03684 0.04076 0.90 0.366 1
Hybrid -0.01538 0.01452 -1.06 0.290 0
-0.02309 0.02135 -1.08 0.279 1
Interaction -0.00646 0.04117 -0.16 0.875 0
-0.00898 0.0626 -0.14 0.886 1

The difference between personalization and assertion was not found to be significant, with a P-value of
.855

Model 2

Table 28. Proportions as independent variables
Variable Coefficient Standard Error Z P>|z|
Personalization -0.28283 0.26802 -1.06 0.291
Assertion -0.40931 0.20950 -1.95 0.051
Hybrid -0.70806 0.47874 -1.48 0.139
Constant -0.94196 0.10045 -9.38 0.000
Pseudo R2: 0.0065 N=1213
Table 29. Marginal effects of Proportions as independent variables
Variable Dy/dx Standard Error Z P>|z| X
Personalization -0.07240 0.07306 -0.99 0.322 0
-0.02771 0.01465 -1.89 0.059 1
Assertion -0.10478 0.06163 -1.70 0.089 0
-0.04010 0.1535 -2.61 0009 1
Hybrid -0.18126 0.1286 -1.41 0.159 0
-0.05402 0.00927 -5.83 0.00 1


Table 30. Marginal effects of proportional independent variables controlling for party
Variable Dy/dx Standard Error Z P>|z| x
Personalization -0.06929 0.08272 -0.84 0.402 0
-0.00600 0.00737 -0.81 0.415 1
Assertion -0.09588 0.06844 -1.40 0.161 0
-0.00831 0.00979 -0.85 0.396 1
Hybrid -0.20493 0.14705 -1.39 0.163 0
-0.01776 0.0152 -1.17 0.243 1
Republican -0.07969 0.02588 -3.08 0.002 0
Democrat -0.01136 0.01477 -0.77 0.442 1

Table 31. Proportional independent variables controlling for sentence length
Variable Coefficient Standard Error Z P>|z|
Personalization -0.20204 0.27438 -0.74 0.462
Assertion -0.38672 0.21150 -1.83 0.067
83
Hybrid -0.66876 0.48206 -1.39 0.165
Sentence Length 0.00814 0.00527 1.54 0.123
Constant -1.06191 0.12806 -8.29 0.00
Table 32. Marginal effects of proportional independent variables controlling for sentence length.
Variable Dy/dx Standard Error Z P>|z| x
Personalization -0.04587 0.06622 -0.69 0.489 0
Assertion -0.08779 0.05655 -1.55 0.121 0
Hybrid -0.15182 0.11651 -1.30 0.193 0
Sentene Length 0.00185 0.00106 1.74 0.082 0
Table 33. Controlling for member effects
Variable Coefficient Standard Error Z P>|z|
Personalization 0.14741 0.31344 0.47 0.638
Personalization -0.39087 0.24305 -1.61 0.108
Hybrid -0.62011 0.54105 -1.15 0.252
Budget 1.55930 0.28674 5.43 0.00
Congress 1.48130 0.39322 3.77 0.00
Constituency 1.69240 0.29840 5.67 0.00
Economy/Jobs 1.07269 0.31728 3.38 0.001
Education 1.14982 0.57227 2.01 0.045
Election 1.79475 0.73064 2.46 0.014
Energy/Envir 1.78791 0.29638 6.03 0.001
Foreign Policy 1.17381 0.30432 3.86 0.00
Healthcare 1.23540 0.32108 3.85 0.00
National Security 2.01762 0.29627 6.81 0.00
Justice 2.12333 0.36791 5.77 0.00
Space 1.58698 0.34021 4.66 0.00
Transportation 1.58699 0.36115 4.39 0.00
Constant -2.39278 0.28065 -8.53 0.00

Model 4: Interaction effects

Table 34. Marginal effects when controlling for super issues
Variable Dy/dx Standard error Z P>|z| X
Personalization 0.01415 0.01108 1.28 0.201 0
0.04007 0.03162 1.64 0.101 1
Assertion 0.024105 0.01406 1.17 0.086 0
0.06248 0.0381 1.64 0.101 1
Hybrid -0.00443 0.02939 -0.46 0.647 0
-0.01527 0.03361 -0.45 0.650 1
Superissue 0.13563 0.02939 4.61 0.00 0
0.18448 0.03033 6.08 0.00 1


Example of a coded Press Release
Personalization highlighted in red
Assertion highlighted in blue
Hybrid highlighted in yellow
Information not highlighted
84

jacksonlee.house.gov
Jackson-Lee
Contact: Michael McQuerry 202-225-7080


CONGRESSWOMAN SHEILA JACKSON LEE CALLS FOR JANITOR NEGOTIATIONS TO
START UP NEXT WEEK

WASHINGTON, Jul 27 -
Congresswoman Sheila Jackson Lee released the following statement regarding Renewed Negotiations in
Janitors' Strike:
The Houston area Janitors engaged in negotiations are our neighbors and friends. All they are asking for
are fair wages for a fair day of work. I have met senior citizens who are making less $9 and hour. Houston
cannot be engaged in denying those that clean major corporate buildings where billions of dollars are
made from earning a fair pay. I support the efforts of the janitors to negotiate peacefully, to strike
peacefully and to engage in civil disobedience peacefully.
Houstons economy owes a great deal to the Houston business community for their contributions as
successful, innovative, and responsible corporate citizens. Their high quality products and services, as
well as your community leadership and philanthropy helps make Houston one of the best cities in the
nation to live and work. Accordingly, I am very proud to represent the Houston corporate community in
Congress.
The enduring power of the American Dream is the promise that hard work and perseverance will bring
prosperity and a better life for American workers and their families. Unfortunately, for millions of
minimum wage workers the American Dream is illusory. In the past few weeks the janitors have made
their demands and withheld their labor. I applaud all of those involved in the negotiations because we
could not let this dispute stain the soil of economic opportunity and social growth.
At many junctures in our nation's history, people of good will have courageously and nonviolently
refused to cooperate with injustice by engaging in nonviolent protest and civil disobedience. These
include the Boston Tea Party, Harriet Tubman's Underground Railroad, and the fight for women to gain
the right to vote. In 1961, civil rights activists engaged in Freedom Rides to Deep South. These were
caravans of African Americans and whites who traveled far and wide to protest Jim Crow segregation.
Inspired by the Freedom Riders, janitors from across the country have traveled to Houston to fight
injustice by engaging in nonviolent civil disobedience.
Their cause was just. Houston janitors work hard mopping, dusting, scrubbing, and polishing thousands
of offices each and every day. I look forward to working with all sides to develop a successful result in
this process. Houston is a major metropolis and cannot be the only city that has not resolved this
difference.
###
85

Coded as: Constituency
Print
Close
Highway Funding Extended Two Years

June 29, 2012
Brady not satisfied but applauds more local control, fewer delays in road projects
Washington, DC Today the U.S. House of Representatives voted overwhelmingly to extend federal
highway, transit, and other surface transportation programs for two years - through September 30,
2014. The measure also included provisions to extend low-interest college student loans.
I fought for a five year highway bill that also removed roadblocks to more American-made energy
production. Our Texas highway officials really wanted more long-term certainty to give the green light to
needed roadway projects, said Brady. Nonetheless, the bill cuts red tape and returns more local control
to Texas which will shorten the approval process and get our road projects underway in half the time,
which is what weve been insisting upon for a long time. Overall, this bill is a step in the right direction
for Texas.
The Highway Bill includes several long overdue reforms, like provisions that eliminate red tape and
streamline the process for completing federally-funded transportation projects. It also consolidates dozens
of federal programs, which Brady sees as another step in the right direction.
Categorized as: Budget

News events
86


Major News Events in Houston, TX During the 112th Session
Story Synopsis Categorized as
Congressional Actors &
their involvement
Story Dates
Houstons failed bid for
a retired space shuttle
As the home of the Johnson Space Center, Houston
was being considered as a location for one of three
space shuttles to be retired in 2011. Lawmakers
grew worried in March when those in Washington
indicated that Ohio was being considered as a
potential location and those in the Houston Delegation
publically pressured the NASA administrator to select
Houston. NASA annouced the retirement location
for the shuttles Discovery, Atlantis and Endeavor in
April, 2011 and New York City and Washington D.C.
being chosen over Houston. The Houston Delegation
lobbied against the decision into September, 2011.
Space Brady (4)
Culberson (6)
A. Green (3)
G. Green (8)
Jackson-Lee (3)
McCaul (3)
Olson (16)
Poe (6)
March 2011-
October 2011
Chad Holleys arrest
and evidence of police
brutality
Chad Holley was a local high school student who
was arrested as bugrlary suspect and later found
guilty. During his trial, video evidence surfaced
showing that he was brutally beaten during his
arrest by several members of the Houston Police
Department. Rep. Sheila Jackson-Lee played a role
in distributing the video to local media . The trial of
the police offcers sparked a debate about single race
juries , as completly white juries acquited one of the
police offcers involved in the incident.
Justice February 2011-
May 2012
A. Green (9)
Jackson-Lee (9)
Houston Metros
expansion into the 7th
congressinal district
Rep. Culberson came out in strong opposition
after Houston Metro released plans for expansion.
Culberson opposed Metros expansion into his
district, particulary in the Richmond and Post Oak
area because Metro did not hold a public referendum
concerning its expansion into this area, as it did
with other areas of the city. The Richmond and Post
Oak expansion was a late addition to expansion
plans. Culberson sought to prevent the expansion by
inserting language that would block federal funding
for Metros expansion into the Richmon and Post
Oak area. His manuevering was ultimately sucessful,
as the House adopted the spending measure to which
he attached the language to block the funding.
Transportation April 2011-
August 2012
Culberson (9)
A. Green (1)
G. Green (2)
Jackson-Lee (2)
Medicare fraud in
Houston
The Houston Chronicle exposed the misuse of
Medicare funding in October of 2011. The papers
intial probe and discovery of fraud initiated several
other investigations, which revealed that several other
members of the community had been committing
Medicare fraud. In response to these fndings, Rep.
Brady called for a congressional hearing into the
fraud and more generally, into the allocation of
Medicare funds to the area. Soon after Brady called
for this assesment, arrests were made in a $90 million
Medicare fraud.
Health Care October-
December 2011
Brady (3)
Culberson (2)
Jackson-Lee (1)
Olson (1)
Story Synopsis Categorized as
Congressional Actors &
their involvement
Story Dates
Allegations of religious
censorship at Houstons
National Cemetary
Accusations surfaced that offcials from the
Department of Veterans Affairs had censored
religious speech, including the word God at
Houston National Cemetary. Rep. Poe acted on these
allegations by attending a ceremony and witnessing
the censorship, he became one of the major actors
in the story and eventually called for the Cemetary
director to be fred.
Veterans/Military Culberson (1)
McCaul (3)
Olson (1)
Poe (7)
June-July 2011
Reaction and local angle
of Tuscon shooting
U.S. Rep Gabrielle Giffords was shot and critically
injured on January 8th, 2011 in a suburb of Tuscon,
AZ. This incident prompted a debate about gun
violence, a discussion about congressional security
measures as well as commentary on the incident.
Other Brady (3)
A. Green (3)
G. Green (8)
Jackson-Lee (4)
McCaul (2)
Olson (6)
Poe (8)
Janurary 2011
87
Works Cited:

Arnold, R. Douglas. 1992. The logic of congressional action. New Haven, CT: Yale University Press.

Arnold, R. Douglas. 2004. Congress, the Press, and Political Accountability. Princeton, NJ: Princeton
Press.

Bennett, W. Lance. 2007. News: The Politics of Illusion, 7th ed. New York: Pearson Longman. Chapter 2
(News Content: Four Information Biases that Matter), pp. 36-73.

Congresswoman Sheila Jackson Lee, Congressional Office, (2011). Congresswoman Sheila Jackson Lee
Calls on the State Department and the Department of Justice to Engage more Resources on the
Ground to Find Ms. Tata [Press release].

Cook, Timothy. 1988. Press secretaries and media strategies in the House of Representatives: Deciding
whom to pursue. American Journal of Political Science 32(4):104769.

Cook, Timothy. 1989. Making laws and making news: Media strategies in the US House of
Representatives. Washington, DC: Brookings.

Druckman, James and Michael Parkin. 2005. The Impact of Media Bias: How Editorial Slant Affects
Voters. Journal of Politics.

Dunham, Richard. Personal communication. December, 17
th
, 2012.

Fenno, Richard. 1973. Congressmen in Committees. Boston: Little Brown and Company.

Fenno, Richard. 1978. Home Style: House Members in their Districts. Boston: Addison Wesley.

Gans, Herbert J. 1979. Deciding Whats News. New York: Pantheon. Chapter 2 (Values in the News),
pp. 39-55.

Grimmer, Justin. (2010). A Bayesian Hierarchical Topic Model for Political Texts: Measuring Expressed
Agendas in Senate Press Releases. Political Analysis, 18(1), 1.

Iyengar, Shanto and Jennifer A. McGrady. 2007. Media Politics: A Citizens Guide. New York:
W.W. Norton. Chapter 7 (Going Public: Governing through the Media), pp. 167-196.
Lloyd, Robert, and Glenn Guzzo. 2009. Writing and reporting the news as a story. Pearson Education,
Inc.
Mayhew, David. 1974. Congress: The Electoral Connection. New Haven, CT: Yale University Press.

McCombs, Maxwell. 2004. Setting the Agenda: The Mass Media and Public Opinion. Cambridge: Polity.

McQuerry, Mike. Personal communication. January 31
st
, 2013.
Neuendorf, Kimberly A. 2002. The Content Analysis Guidebook. Thousand Oaks, CA: Sage Publications,
Inc.
88



Niven, David. 2002. Tilt? The Search for Media Bias. Westport, Conn.: Praeger. Chapter 3 (The Jury is
Still Out: Academic Evidence on Media Bias), pp. 51-71.

Schaffner, Brian. 2006. Local news coverage and the incumbency advantage in the US house.
Legislative Studies Quarterly 31(4):491511.

Schiller, Wendy. 2000. Partners and rivals: Representation in US Senate delegations. Princeton, NJ:
Princeton University Press.

Sulkin, Tracy. 2005. Issue Politics in Congress. Cambridge: Cambridge University Press.

Vinson, Danielle. 2002. Through Local Eyes: Local Media Coverage of Congress. Creskill, NJ: Hampton.

Yiannakis, Diana Evans 1982. House members communication styles: Newsletter and press releases.
Journal of Politics 44(4):104971.

Das könnte Ihnen auch gefallen