10 1002@job 1829

Journal of Organizational Behavior, J. Organiz. Behav.
(2012)
Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/job.1829
Research Article
Employee silence motives: Investigation of
dimensionality and development of measures
CHAD T. BRINSFIELD*
Opus College of Business, University of St. Thomas, Minneapolis, Minnesota U.S.A.
Summary In four studies, I examine the motives for employee silence. In Study 1, I examine open-ended survey
responses to determine the nature and scope of silence motives. Study 2 develops measures of these motives
and explores their factor structure. Study 3 renes the measures and provides conrmatory evidence of factor
structure. Study 4 examines relationships between the new measures and related factors (employee voice,
psychological safety, neuroticism, extraversion). Results indicate that six dimensions of silence motives
(ineffectual, relational, defensive, difdent, disengaged, and deviant) emerged from the data, which can be
reliably measured and provide incremental value for understanding and assessing employee silence. Copyright
2012 John Wiley & Sons, Ltd.
Keywords: employee silence; organizational silence; employee voice
Introduction
Employees reluctance to speak up about work-related matters has been linked to many important individual and
organizational outcomes (e.g., decreased innovation, Argyris & Schn, 1978; failure to address ethical transgressions,
Clapham & Cooper, 2005; stress and depression, Cortina & Magley, 2003; lower commitment and job satisfaction,
Vakola & Bouradas, 2005). Over the years, a considerable number of different constructs have been developed to
describe different forms and facets of employees willingness or unwillingness to speak up about important issues,
situations, or concerns in the workplace (e.g., voice, whistle-blowing, spiral of silence, employee silence; Graham,
1986; Morrison & Milliken, 2000; Noelle-Neumann, 1974; Pinder & Harlos, 2001; Van Dyne & LePine, 1998). These
constructs differ in communicative direction (upward, lateral, downward, or external), level of analysis (individual,
group, or organization), the type of situation or concern to which one may speak up or remain silent (e.g., idea for
improvement, ethical transgression, dissatisfaction), and whether the primary focus is on verbal expression or
suppression (for reviews, see Brinseld, Edwards, & Greenberg, 2009; Morrison, 2011). In addition, a considerable
number of different constructs have been examined as antecedents of employees willingness or unwillingness to
speak up at work (e.g., psychological safety, implicit voice theories, diffusion of responsibility, organizational climate,
trust; Detert & Edmondson, 2011; LePine & Van Dyne, 1998, 2001; Morrison & Milliken, 2000; Van Dyne, Ang, &
Botero, 2003).
There is little doubt that this prior research has yielded many important insights regarding employees communicative
behaviors. Despite the wide array of different constructs and prior research, scholars contend that there is still much
we do not know about the nature and scope of situations where employees have something meaningful to say, but
also feel compelled to remain silent (Detert & Edmondson, 2011; Morrison, 2011). Scholars have suggested a number
of issues with prior research, which may account for decits in our understanding, including (i) prior research has
often assumed that what is understood about speaking up behavior (i.e., voice) fully applies to intentional silence
*Correspondence to: Chad T. Brinseld, Opus College of Business, University of St. Thomas, Minneapolis, Minnesota, U.S.A. E-mail: cbrinseld
@stthomas.edu
Received 6 January 2012

Copyright 2012 John Wiley & Sons, Ltd. Revised 1 August 2012, Accepted 22 August 2012
C. T. BRINSFIELD
(Tangirala & Ramanujam, 2008; Van Dyne et al., 2003); (ii) research that has specically focused on silence, has mainly
been conceptual or qualitative (e.g., Milliken, Morrison, & Hewlin, 2003; Morrison & Milliken, 2000), although there
have been a few notable exceptions (e.g., Tangirala & Ramanujam, 2008); and (iii) prior research has largely focused on
employees wanting to avoid the risks associated with speaking up to the exclusion of other motives for remaining silent
(e.g., Detert & Edmondson, 2011). I discuss next the problematic nature of these issues and the resultant gaps in our
knowledge.
Prior research that has assumed that what is understood about voice fully applies to silence is problematic because
an absence of intentional silence is not necessarily indicated by the presence of voice behavior (Van Dyne et al.,
2003). In terms of Gestalt psychology (Kohler, 1992), this prior research has mainly considered voice as gure
and silence as ground. However, consider a person who is naturally outgoing. This person often may express
themselves, and others may rate them high on measures of voice, but there also could be a host of situations to which
they remain silent for a variety of different reasons. However, the extent, nature, and implications of their silence
may not surface with assessments of voice or its antecedents.
Conceptual and qualitative research focused on intentional silence has been crucial in advancing our knowledge
(e.g., Morrison & Milliken, 2000; Van Dyne et al., 2003). This research has laid the foundation for the consideration
of voice and silence as distinct constructs and the recognition that there may be meaningful distinctions between
forms of silence based on different motives (Pinder & Harlos, 2001). However, any single method of scientic
inquiry involves trade-offs in the resulting theorys simplicity, generalizability, and accuracy (Thorngate, 1976).
Subsequently, there have been numerous calls for quantitative research to test and rene these prior conceptual
and qualitative efforts (e.g., Edwards & Greenberg, 2009).
Much of the prior research has focused on the risks associated with speaking up and has explicitly or implicitly
placed psychological safety as a mediator between antecedent variables and voice behavior (Detert & Edmondson,
2011). However, there is also evidence to suggest that silence often may be based on motives other than risk
avoidance (e.g., Morrison, Wheeler-Smith, & Kamdar, 2011; Van Dyne et al., 2003). This primary focus on silence
as a tactic to avoid negative consequences associated with speaking up has left many unresolved questions. For
example, what exactly is the nature of these other motives for silence and how pervasive are they? What types of
antecedents and outcomes are associated with different motives for silence?
Most conceptualizations and denitions of employee silence characterize it as the intentional withholding of
seemingly meaningful information, including questions, concerns, and suggestions (Tangirala & Ramanujam,
2008). Hence, employee silence does not refer to the unintentional failure to communicate that might result
from mindlessness or having nothing to say (Van Dyne et al., 2003). Translating this conceptualization of silence
into a valid construct that can reliably be measured has been difcult, but important strides have been made.
For example, Vakola and Bouradas (2005) measured organizational silence by assessing climate factors such
as employees perception of management and supervisors attitudes toward silence, communication opportunities,
and extent of expressions of disagreement. Tangirala and Ramanujam (2008) adapted items from the proposed
(but not empirically validated) measure of silence by Van Dyne et al. (2003) to assess the extent to which
employees intentionally withheld ideas, concerns, questions, or information about patient safety in their
workgroup. Detert and Edmondson (2011) empirically assessed employees taken-for-granted beliefs about
when and why speaking up at work is risky or inappropriate (i.e., implicit voice theories). Morrison et al.
(2011) investigated group voice climate, which consists of shared beliefs about whether speaking up is safe and
whether it is likely to be effective.
These prior efforts have led to many important insights and have been helpful in advancing our understanding of
silence in the workplace. However, none of the previous efforts to quantitatively investigate silence have examined
the nature, pervasiveness, or implications of different forms of silence. Hence, gaps in our knowledge remain, and
important questions such as those previously discussed remain unresolved. In sum, investigating voice behavior and
inferring employee silence is problematic because the presence of voice is not necessarily equivalent to an absence
of intentional silence. Assessing the behavior of silence is inadequate because silence behavior by itself is not
necessarily indicative of a conscious desire to remain silent. Assessing a general intention to remain silent is a
Copyright 2012 John Wiley & Sons, Ltd. J. Organiz. Behav. (2012)
DOI: 10.1002/job
EMPLOYEE SILENCE MOTIVES
step in the right direction, but does not account for the potentially diverse meanings that may underlie a desire to
remain silent (Van Dyne et al., 2003). The approach recently introduced by Detert and Edmondson (2011; i.e.,
implict voice theories) is promising, but limits silence as a response to the risks associated with speaking up.
In an attempt to resolve gaps in our knowledge related to the nature, pervasiveness, and implications of different
forms of silence, some scholars have conceptualized it as a multidimensional construct based on distinct underlying
motives (Pinder & Harlos, 2001; Van Dyne et al., 2003). They suggest that it is the underlying motives that give
meaning to the behavior of silence and distinguish it from simply an absence of voice. They also suggest that it is
the motives that account for the phenomenological differences between different forms of silence. Hence, a variety
of different motives have been proposed in the literature (e.g., defensive, prosocial, acquiescence; Pinder & Harlos,
2001; Van Dyne et al., 2003), but there has been no attempt to empirically examine the nature, scope, and implications
of different silence motives.
Pinder (1998) dened work motivation as a set of energetic forces that originates both within as well as beyond
an individuals being, to initiate work-related behavior, and to determine its form, direction, intensity, and duration
(p. 11). Building on the prior research on employee silence motives, and using Pinders denition as a guide, in this
research, silence motives represent the energetic forces that provide the impetus for intentional silence. The form is
the underlying reason (e.g., to protect oneself, acquiescence), and the direction is toward intentional silence.
Moreover, the motives themselves are not necessarily unique to silence. For instance, the same motive could be
an impetus for silence or voice (e.g., a person could be motivated to remain silent or speak up on the basis of a desire
to defend oneself). Rather, it is when a specic motive is coupled with an intention to remain silent that it represents
the phenomenological experience of employee silence. In addition, silence motives are distinct from antecedents in
that motives are integral to the experience of employee silence, whereas potential antecedents (e.g., leadership style,
organizational climate) are more psychologically distal.
In summary, the direct examination of silence motives should enhance our understanding of situations where
employees have something meaningful to say, but remain silent. Silence motives represent a way to quantify
different forms of silence and should serve as a catalyst toward the examination of silence for a wider range of reasons
than typically has been the case. Therefore, to advance our understanding of the nature, scope, and implications of
employee silence motives, I conducted four studies. Study 1 is an inductive investigation of the content domain and
motives for employee silence. In Study 2, I develop measures of these motives and conduct an exploratory factor
analysis (EFA) to investigate their factorial structure. Study 3 further tests construct validity through conrmatory factor
analysis (CFA) of the resultant dimensions. Study 4 begins explication of the nomological network and investigates the
incremental value of empirically examining the motives for employee silence.
Study 1
Whetten (1989) suggested that authors be comprehensive, without ignoring parsimony, when rst developing a
construct as the ideas will be trimmed and rened over time. Therefore, in this study, I conduct a broad and inductive
exploration of the different reasons employees have for remaining silent in response to important issues, situations,
or concerns at work. In addition, I examine two boundary constraints that have been inconsistently specied in
previous research, but which have the potential to shape its meaning and function. The rst is the nature of the focal
issue, situation, or concern to which employees remain silent. For example, Pinder and Harlos (2001) constrained
employee silence as taking place in response to a perceived injustice. Van Dyne et al. (2003) constrained it to the
intentional withholding of ideas, information, and opinions for constructive ways to improve work. Milliken et al.
(2003) in an exploratory study identied eight different types of issues to which employees remained silent
(e.g., concerns about a colleagues or supervisors competence, problems with organizational processes, harassment,
or abuse). The second constraint examined is the target of the silence. Although much of the previous research
constrains employee silence toward a superior within the organization (Morrison, 2011), others have examined
DOI: 10.1002/job
C. T. BRINSFIELD
silence targeted at workgroup members (e.g., Tangirala & Ramanujam, 2008), whereas still others have left the
target broader such as persons who are perceived to be capable of effecting change or redress (Pinder & Harlos,
2001, p. 334).
Sample
Respondents consisted of undergraduate business students with at least two years work experience (n = 123,
68 percent response rate), full-time MBA students (n = 105, 84 percent), and evening MBA students at a midwestern
university (n = 165, 68 percent), and front-line and administrative employees of a midwestern manufacturing
company (n = 68, 59 percent). A total of 461 (69 percent) responded, and 438 (95 percent) reported at least one
incident to which they had intentionally remained silent at work. I recruited students via e-mail, and the human
resource department of the participating organization recruited organization members by sending an e-mail on my
behalf. The average age of respondents was 32 years, the average work experience was 11 years, 84 percent were
Caucasian, and 40 percent were female. Respondents spanned multiple functions and hierarchical levels in a wide
range of industries.
Design and procedure

I administered a web-based survey asking respondents to think of up to four incidents when they intentionally
remained silent in response to an important issue, situation, or concern at work. Considering that the motives for
silence may vary according to the type of issue, situation, or concern an employee is remaining silent in response
to, as well as the type of target their silence is directed toward, rst, I asked respondents to describe the nature of
each issue, situation, or concern to which they intentionally remained silent (silence incident). Next, I asked them
to provide a brief description of the person(s) to whom they were intentionally silent (silence target). Next, I asked
them to provide a description of the reason(s) why they intentionally remained silent in response to each incident
(silence reason). Collecting information about silence incidents and targets, in addition to reasons, provided
important insights regarding the content domain and boundary constraints of the construct, and helped guide the
latter three studies with regard to these considerations.
I conducted content coding of responses using standard practices for qualitative data analysis (Miles & Huberman,
1994). This entailed a progressive process wherein I analyzed the responses and developed and applied exemplar
statements (short non-contextual descriptions) to each response. This enabled uniform descriptions of similar themes
to be applied to responses and therefore reduced the number of responses which otherwise would consist of only
minor syntactical or other non-substantive differences. Since each respondent provided separate open-ended
responses regarding the incident, target, and reason for silence, I developed and applied separate exemplar statements
for each of these three questions. Throughout this phase, two additional colleagues who had experience conducting
qualitative research checked the evolving exemplar lists and application of exemplar statements to responses.
We discussed issues regarding objectivity, clarity, and consistency and rened the exemplars as deemed necessary.
Analysis and results

Incidents to which employees remain silent
Respondents (N = 438) provided 1240 open-ended responses (I asked each respondent to provide up to four
incidents). From these 1240 responses, I initially created 84 unique exemplars to describe the various types of incidents
to which respondents reported remaining silent (e.g., treated unfairly by co-workers, disagreement with business
decision, witnessed sexual harassment, co-worker not doing their fair share, idea for process improvement). Next,
DOI: 10.1002/job
I provided lists to four independent judges (doctoral students in management) of the 84 silence incident exemplars, and I
asked them to group them into like categories. The four judges and I then met to evaluate commonalities and differences
in the groupings. From this process, a consensus emerged that the nine broader silence incident categories indicated in
Table 1 accurately represented the 84 silence incident exemplars. To assess the reliability of the nine categories relative
to the original 1240 open-ended responses, I asked three different colleagues to assign one of the categories to a random
sample of 10 percent of the original responses. I then calculated Fleiss kappa to determine agreement among the three
judges (Fleiss, 1971). The analysis revealed a k = 0.63. k of 0.70 or greater is generally thought to be sufcient for
psychological measurement (Fleiss, Levin, & Paik, 2004), but the level of agreement is expected to decrease as the
number of categories increases. Moreover, during a follow-up investigation of categories with low agreement, it was
apparent that some of the original responses described incidents which spanned more than one category, but raters
had been instructed to assign the one best category to a response. Therefore, considering the multifaceted nature of some
of the original responses and the large number of categories, I considered the resultant kappa to be acceptable. Also
indicated in Table 1 is the reported frequency of the different types of incidents. The most common type of silence
incident was unfair treatment, which accounted for 21 percent of the total responses. Someone else behaving unethically
was the second most common incident type (17.6 percent). Concerns about a co-workers competence or performance
(16.8 percent), and operational process concern and/or idea for improvement (13.3 percent) also were relatively
common silence incidents. As indicated, a number of other types of incidents also were reported, but their relative
frequencies were considerably less (Table 1).
Targets of silence
Following the same analytic procedure that was used to develop exemplars of incidents to which employees remain
silent, exemplars of nine different types of targets of silence emerged from the responses (Table 2). Because only
nine different exemplars of targets emerged, I did not ask independent judges to group these into similar categories,
as I did with the more numerous silence incident exemplars. However, to assess the reliability of the nine target
exemplars, I also asked the same three judges who coded the sample for silence incidents to assign one of the nine
target exemplars to a random sample of 10 percent of the original responses. The analysis revealed a k = 0.83. As
indicated, employees may remain silent to wide array of different types of targets. In fact, about every type of
conceivable workplace target appears represented in this categorization; however, some are much more frequent targets
of silence than others. For example, upper management and direct supervisor were the two most frequently reported
targets and together account for nearly 69 percent of the responses. This is not surprising considering the potential
harm associated with speaking up to people in positions of power. Co-workers also were a relatively common target
(17 percent), which is not surprising considering the prevalence of these relationships. As indicated, a number of other
types of targets also were reported, but their relative frequencies were considerably less (Table 2).
Table 1. Study 1: Silence incident types.

Types of incidents to which respondents reported remaining silent Silence incidents reported % of total
Experienced unfair treatment 260 21.0

Someone else behaving unethically 218 17.6
Concerns about a co-workers competence or performance 208 16.8
Operational process concern and/or idea for improvement 165 13.3
Disagreement or concerns with company policies or decisions 81 6.5
Personal performance issue 77 6.2
Concerns about supervisor or management competence 74 6.0
Someone else being treated unfairly 44 3.5
Personal career issue or concern 24 1.9
Unclear 89 7.2
Total 1240 100
DOI: 10.1002/job
C. T. BRINSFIELD
Table 2. Study 1: Silence targets.

Types of targets to which respondents reported remaining silent Silence targets reported % of total
Upper management 504 40.6

Direct supervisor 346 27.9
Co-worker 211 17.0
Team members 38 3.1
Everyone 31 2.5
Outside authorities 19 1.5
Subordinate 15 1.2
Customer 10 0.8
Outside consultant 8 0.6
Unclear 58 4.7
Total 1240 100
Motives for remaining silent

From the 1240 open-ended responses, respondents provided 2277 statements regarding reasons for remaining silent.
There are more statements regarding reasons for silence than the 1240 individual open-ended responses because
many respondents provided more than one reason per response. Exemplars of 59 different motives for silence
emerged from the 2277 statements regarding reasons for silence. To assess the reliability of the 59 exemplars relative
to the original 1240 open-ended responses, I asked three independent judges to assign one of the 59 exemplars to a
random sample of 10 percent of the original responses. The analysis revealed a k = 0.73. I did not group the
exemplars of these motives into broader categories at this point but rather I used them as items to determine
dimensionality through exploratory and CFAs in Studies 2 and 3, respectively. On the basis of the responses in this
study, the conceptual range of motives for remaining silent is quite varied, and potentially more nuanced than
accounted for in the existing literature. The most frequently reported silence reason exemplar was I did not think
it would do any good to speak up. Other frequently reported reasons included to avoid conict with another
individual, due to fear of retaliation, management did not appear interested in hearing about these types of issues,
I did not want to get involved, and I did not want to be viewed as a complainer (see Table 3 for a complete list of
silence reason exemplars and their reported frequencies). As indicated in Table 3, items seemingly related to silence as a
tactic to avoid the risks associated with speaking up (a primary focus of prior research; e.g., Detert & Edmondson, 2011)
accounted for only a moderate proportion of the total responses.
Summary of Study 1
Although it has been over a decade since employee silence was introduced to the management sciences as a
multidimensional construct (Pinder & Harlos, 2001), to date, there has been only modest empirical examination
(Edwards & Greenberg, 2009). Therefore, the purpose of this study was to conduct a broad and inductive
investigation of the content domain and motives for employee silence as a step toward advancing our empirical
and theoretical understanding of the construct. In this study, 438 or 95 percent of the 461 respondents reported at
least one time when they intentionally remained silent at work. These responses spanned a wide variety of different
types of incidents and targets to which employees remained silent and furthers our understanding of the scope and
pervasiveness of silence in regard to these factors. In addition, respondents reported remaining silent for a wide
variety of different reasons. In this regard, these ndings build upon prior conceptual work on employee silence
motives (Pinder & Harlos, 2001; Van Dyne et al., 2003) and provide the groundwork for a more comprehensive
and systematic study of the phenomenon.
DOI: 10.1002/job
Table 3. Study 1, silence reason exemplars, and Study 2, EFA pattern matrix.
Silence factor/dimension (Study 2)
Reported 1 2 3 4 5 6
frequency
Item Item description (Study 1) Deviant Relational Defensive Difdent Ineffectual Disengaged
1 To get even with another 2 0.98 0.01 0.02 0.01 0.05 0.04
person
2 To purposefully harm 1 0.98 0.01 0.04 0.07 0.02 0.03
another individual
3 To retaliate against the 3 0.94 0.03 0.07 0.01 0.03 0.05
organization
4 To purposefully harm the 3 0.91 0.03 0.05 0.03 0.04 0.05
organization
5 To make management 2 0.84 0.02 0.07 0.04 0.10 0.04
look bad
6* I did not want to go 17 0.46 0.06 0.01 0.12 0.05 0.37
beyond the minimum
required of me
7 I didnt want to harm my 32 0.01 0.93 0.10 0.23 0.08 0.20
relationship with another
individual
8 I did not want to create 56 0.04 0.93 0.03 0.16 0.03 0.12
tension with co-worker
9 To avoid conict with 153 0.04 0.91 0.08 0.18 0.05 0.06
another individual
10 To protect my relationship 77 0.04 0.88 0.01 0.11 0.03 0.15
with another individual
11 To avoid hurting 29 0.17 0.74 0.12 0.03 0.09 0.16
someones feelings
12* I did not want others to 24 0.01 0.55 0.04 0.53 0.01 0.18
think negatively of me
13* To protect another person 67 0.20 0.52 0.02 0.03 0.14 0.35
from harm
14* I did not want to be viewed 80 0.01 0.51 0.15 0.46 0.20 0.15
as a complainer
15* To protect my image or 61 0.08 0.44 0.05 0.39 0.01 0.05
reputation
16* I did not want to be viewed 53 0.09 0.44 0.12 0.38 0.09 0.17
as causing problems
17* I didnt want others to 2 0.01 0.23 0.05 0.03 0.05 0.00
know I was bothered by
incident
18* I was planning on leaving 15 0.04 0.15 0.10 0.02 0.01 0.07
the organization soon
19* I only recently became 4 0.04 0.09 0.07 0.03 0.08 0.08
aware of the situation/
incident
20 I felt it was dangerous to 13 0.09 0.02 0.91 0.14 0.00 0.06
speak up
21 To protect myself from 19 0.05 0.05 0.87 0.06 0.09 0.15
harm
22 I felt it was risky to speak 17 0.08 0.01 0.87 0.01 0.03 0.10
up
(Continues)
DOI: 10.1002/job
C. T. BRINSFIELD
Table 3. (Continued)
frequency
23 I believed that speaking up 90 0.06 0.06 0.84 0.01 0.08 0.12

may negatively impact my
career
24 I was afraid of adverse 59 0.10 0.03 0.79 0.10 0.02 0.06
consequences (e.g., being
criticized, losing my job)
25 Due to fear of retaliation 90 0.03 0.25 0.71 0.04 0.05 0.01
26* I have previously 14 0.06 0.04 0.63 0.22 0.08 0.16
witnessed others
experience negative
consequences as a result of
speaking up
27* I have previously 23 0.10 0.05 0.51 0.20 0.12 0.19
experienced negative
consequences as a result of
speaking up
28* I felt it was safer to do 15 0.09 0.16 0.43 0.12 0.07 0.21
nothing
29* I believed that time would 15 0.01 0.01 0.17 0.13 0.07 0.03
handle the situation
30* I was too proud to speak 4 0.08 0.08 0.15 0.01 0.04 0.10
up about the incident
31 I did not feel condent 37 0.02 0.11 0.01 0.83 0.00 0.12
enough to speak up
32 To avoid embarrassing myself 45 0.07 0.08 0.04 0.78 0.07 0.06
33 I was unsure what to say 71 0.02 0.07 0.03 0.65 0.01 0.28
34 I felt insecure 47 0.01 0.17 0.20 0.65 0.05 0.12
35* I was uncertain 76 0.02 0.07 0.00 0.62 0.03 0.29
36* I did not want to draw 26 0.13 0.34 0.11 0.62 0.07 0.01
attention to myself
37 I did not want to appear 65 0.19 0.00 0.12 0.62 0.04 0.01
incompetent
38* Other people involved 5 0.01 0.17 0.05 0.57 0.08 0.40
knew more about the
situation than I did
39* I did not fully understand 25 0.18 0.27 0.01 0.50 0.04 0.45
the issue
40* I was unsure who to speak 8 0.03 0.02 0.03 0.23 0.10 0.00
up to
41* I had nothing new to add 3 0.02 0.18 0.00 0.21 0.12 0.05
to discussion
42* I was too busy at the time 12 0.11 0.04 0.15 0.19 0.02 0.07
to speak up
43* I wanted co-worker to 4 0.02 0.05 0.09 0.13 0.07 0.01
learn from their mistake
44 I did not believe my 29 0.06 0.06 0.09 0.12 1.00 0.03
concerns would be
addressed
(Continues)
DOI: 10.1002/job
Table 3. (Continued)
frequency
45 Management did not 88 0.05 0.01 0.09 0.15 0.95 0.09

appear interested in
hearing about these types
of issues
46 No one was interested in 30 0.01 0.10 0.07 0.10 0.91 0.01
taking appropriate action
47 I did not feel I would be 21 0.02 0.00 0.12 0.17 0.72 0.12
taken seriously
48 I did not think it would do 230 0.19 0.02 0.03 0.04 0.59 0.31
any good to speak up
49* I had previously spoken up 14 0.34 0.05 0.11 0.12 0.52 0.06
about similar issues, but it
did not do any good
50* I have witnessed others 37 0.29 0.02 0.17 0.09 0.50 0.07
speak up about similar
issues, but it did not do
any good
51* I wanted to handle the 5 0.09 0.02 0.16 0.11 0.18 0.01
situation myself
52 The issue did not 28 0.13 0.19 0.07 0.10 0.04 0.63
personally affect me
53 I did not care what 35 0.35 0.01 0.08 0.10 0.02 0.56
happened
54 I did not want to get 88 0.08 0.24 0.08 0.16 0.01 0.54
involved
55* I did not care about the 15 0.35 0.04 0.10 0.10 0.28 0.40
organization
56* It is not my responsibility 55 0.09 0.01 0.04 0.26 0.15 0.38
to speak up
57* I did not think it is worth 55 0.13 0.19 0.11 0.18 0.33 0.36
the effort to speak up
58* I believed someone else 79 0.14 0.00 0.15 0.27 0.02 0.32
should speak up
59* I was taught not to 4 0.04 0.06 0.05 0.02 0.08 0.11
complain
Eigen values 20.25 4.73 3.43 2.52 1.97 1.53
Explained variance % 34.22 8.02 5.81 4.28 3.34 2.60
Note: To eliminate the need for two lengthy tables, Table 3 includes both the complete list of silence reason exemplars derived in Study 1, and the
EFA pattern matrix derived in Study 2.
Study 2
Hinkin (1998) suggested that measures can be developed either deductively (logical partitioning or classication
from above) or inductively (grouping or classication from below). Hinkin argues that the inductive approach is
most appropriate for conducting exploratory research on abstract phenomena when the conceptual basis for a
DOI: 10.1002/job
C. T. BRINSFIELD
construct may not result in easily identiable dimensions. Therefore, in this study, I created Likert-type scale items
based on the silence reason exemplars inductively derived in Study 1 and conducted an EFA to investigate their
underlying dimensionality.
Sample
Respondents consisted of undergraduate business students with at least two years work experience (n = 134,
89 percent response rate), full-time MBA students (n = 87, 92 percent), and executive and evening MBA students
at a midwestern university (n = 109, 87 percent). A total of 330 (89 percent) responded. On the basis of the reviews
of EFA sample size guidelines (Preacher & MacCallum, 2002), I deemed this sample size to be sufcient. I recruited
students via e-mail and offered them a small amount of extra course credit for their participation. The average age of
respondents was 27 years, the average work experience was six years, 77 percent were Caucasian, and 61 percent
were female. This sample was distinct from Study 1, and respondents spanned multiple functions and hierarchical
levels in a wide range of industries.

I administered a web-based survey with the following instructions:
Employees may be confronted with important issues, events, or concerns at work to which they may desire to
remain silent. The underlying reasons for remaining silent may be quite varied. For the following set of questions,
please indicate the extent to which you typically experience the stated reasons for wanting to remain silent in
response to important issues, events, or concerns in your current job (if presently employed), or your most recent
job (if not presently employed).
Immediately following the instructions, I provided the measures of employee silence, which consisted of the
prex I wanted to remain silent followed by an appropriate transitional term (e.g., because, due to), and I then
concluded with the exemplar statements derived in Study 1. I presented items in random order and assessed them
using a 7-point Likert-type scale ranging from 1 = not at all to 7 = to a very large extent.

To examine the underlying factor structure, I conducted an EFA using SPSS v.17 (SPSS Inc., Chicago, IL) with
maximum likelihood extraction and oblique rotation. I chose oblique rotation because the silence factors were not
expected to be completely independent of one another. I provide the EFA pattern matrix of the silence reasons in
Table 3. The scree plot of the eigenvalues revealed a natural bend or break point with six data points above the
break. I provide the eigenvalues and percentage of explained variance for each of the six factors at the bottom of
Table 3. Together, the six factors explained 58.27 percent of the total variance. I based judgments regarding which
items best represented a particular factor on the loading weight on a primary factor and the extent to which an item
loaded on only one factor. I used a loading weight of at least 0.50 on the primary factor and a loading on no more
than 0.32 (which equates to approximately 10 percent overlapping variance) on another factor as general guidelines
(Tabachnick & Fidell, 2001). In addition, items 26, 27, 35, and 50 met the retention criteria, but I did not retain them
for redundancy and parsimony considerations. In Table 3, retained items selected as loading together on a particular
factor have their factor scores indicated with bold italics. Items indicated with an asterisk were not retained. I
developed preliminary labels which represent the underlying nature of each factor through discussion and consensus
with the same two colleagues who helped check the evolving exemplar list in Study 1. I have inserted these labels
DOI: 10.1002/job
into the appropriate column headings in Table 3, and I discuss next the dimensions of employee silence they
represent.
Deviant silence
Silence as a form of deviant workplace behavior has received little attention in the management literature. However,
anecdotal evidence suggests that this behavior does exist and may have negative consequences. For example, a U. K.
employee won 800 000 in damages from Deutsche Bank after suing the organization for psychiatric injury
following a sustained period of bullying and harassment, which included the intentional withholding of important
information (Martin & Woollard, 2009). From a research perspective, Gruys and Sackett (2003) in an investigation
of the dimensionality of counterproductive work behavior identied Misuse of Information as a distinct dimension.
This dimension contained ve related items, one of which (intentionally fail to give supervisor or co-worker
necessary information) is conceptually similar to deviant silence. However, their other four items (e.g., destroy
or falsify records or documents) appear distinct from deviant silence. As indicted in Table 3, items loading on
deviant silence accounted for only 11 or 0.48 percent of the 2277 reasons provided in Study 1. On the basis of
anecdotal and sparse empirical evidence, one could speculate that deviant silence is more pervasive than reported
in Study 1. This may be due to respondents reluctance to report such deviant behavior or respondents paradigmatic
assumptions around the topic of silence precluded them from thinking about this form of silence in an open-ended
question. Therefore, one could further speculate that the responses may be more substantial if Likert-type questions
were used. Given the plausibility of these concerns, and the general exploratory nature of this research, I decided to
retain deviant silence as a dimension for further investigation in the subsequent two studies.
Relational silence
I labeled the second factor as relational silence, and it accounted for 347 or 15.24 percent of the total responses in
Study 1. On the basis of the nature of the items that comprise this dimension, it appears that people often remain
silent because they do not want to harm a relationship or general relational concerns. Researchers also have
previously found that employees may remain silent for relational oriented reasons. For example, Milliken et al.
(2003) found that 27.5 percent of the 40 respondents they interviewed reported fear of damaging a relationship as
a reason for silence at work. This dimension also appears partly related to the prosocial motives for silence proposed
by Van Dyne et al. (2003), which they conceptualized as proactive and other-oriented on the basis of altruism and
cooperation. However, the items comprising relational silence are not necessarily reective of altruism. Instead, the
general relational orientation of these items could be based on deeper motives of self-interest, but it also could be
argued that at some level, all of the motives could be tied back to self-interest. Moreover, the low cross-loadings
of these items suggest that respondents do make a distinction between these items and items loading on other
ostensibly more self-interested silence motives. In addition, Van Dyne et al. also conceptualized prosocial silence
as protecting propriety knowledge to benet the organization, which is distinct from relational silence.
Defensive silence
I labeled the third factor as defensive silence, and items loading on this factor accounted for 288 or 12.65 percent of
the total responses provided in Study 1. This dimension of employee silence has previously been proposed in the
literature and continues to be a common focus of research on silence. Pinder and Harlos (2001) used the term
quiescent silence to describe deliberate omission on the basis of fear of the consequences associated with speaking
up. Morrison and Milliken (2000) emphasized the emotion of fear as a key motivator of organizational silence. This
dimension is also consistent with psychological safety (Edmondson, 1999) as a critical precondition for speaking up
in work contexts. Moreover, Milliken et al. (2003) found that 22.5 percent of respondents reported fear of retaliation
or punishment as a reason for not speaking up about concerns or problems in the workplace. Van Dyne et al. (2003)
also proposed this dimension of employee silence and dened it as withholding relevant ideas, information, or
opinions as a form of self-protection, based on fear (p. 1367). This dimension is also similar to the type of work-
place silence that Detert and Edmondson (2011) investigated with regard to implicit voice theories. As indicated in
DOI: 10.1002/job
C. T. BRINSFIELD
Table 3, items comprising this dimension are primarily based on fear of extrinsic consequences associated with
speaking up. However, fear of extrinsic consequences appears to be distinct from the more internally focused
trepidation indicated by the next dimension.
Difdent silence
I labeled the fourth factor indicated in Table 3 as difdent silence. Merriam-Webster denes difdent as hesitant in
acting or speaking through lack of self-condence. Items loading on this factor accounted for 265 or 11.64 percent
of the total responses provided in Study 1. This dimension is composed of items related to ones insecurities, self-
doubt, or uncertainties regarding the situation or what to say. There may be some conceptual overlap between this
dimension and defensive silence to the extent that this dimension is also related to remaining silent to avoid negative
outcomes. With difdent silence, however, the nature of the negative outcomes appear more internally focused (e.g.,
to avoid embarrassing myself), whereas the nature of the negative outcomes associated with defensive silence are
more extrinsic in nature (e.g., afraid of losing my job). Furthermore, the absence of nontrivial cross loading of items
provides support for this dimension as a distinct form of silence. Although not previously proposed as a dimension
of employee silence, this type of reason has previously been proposed in related areas of inquiry. For example,
Noelle-Neumann (1974) suggested that feelings of self-doubt may discourage people from expressing ideas that fail
to conform to public opinion, thus perpetuating spirals of silence wherein one opinion erroneously becomes
predominant. LePine and Van Dyne (1998) found that self-esteem had a positive inuence on peoples expressive
behavior. Similarly, LePine and Van Dyne (2001) demonstrated that neuroticism (a personality trait described as
anxious, depressed, angry, embarrassed, emotional, worried, or insecure; Barrick & Mount, 1991) is negatively
related to voice behavior. Clearly, more work is needed to determine the discriminant validity of this form of silence.
However, considering its pervasiveness, it represents a potentially important topic for future investigation.
Ineffectual silence
I labeled the fth factor indicated in Table 3 as ineffectual silence, and items loading on this dimension accounted for
398 or 17.48 percent of the total responses in Study 1. Items loading on this dimension are related to the general
belief that speaking up would not be useful in effecting change relative to the focal issue, situation, or concern.
Related reasons for voice and silence have previously been proposed in the literature. For example, Morrison and
Milliken (2000) conceptualized a climate of silence as characterized, in part, by the shared belief that speaking up
about problems in the organization is not worth the effort. Pinder and Harlos (2001) conceptualized employee
acquiescence as implying a deeply felt acceptance of organizational circumstances, feelings of resignation, giving
up hope for improvement, a taking-for-granted of the situation, and limited awareness that alternatives exist. Van Dyne
et al. (2003) dened acquiescent silence as withholding relevant ideas, information, or opinions, based on resigna-
tion (p. 1366). However, the label ineffectual better represents the distinct nature of this dimension as the items
loading on this factor are more focused on the belief that speaking up would not positively affect the situation, whereas
common denitions of acquiescence are more focused on feelings of resignation or reluctant agreement.
Disengaged silence
Kahn (1990) dened disengagement as the uncoupling of selves from work roles (p. 694). Drawing from Kahns
denition, I labeled the sixth factor indicted in Table 3 as disengaged silence. This dimension accounted for 151
or 6.63 percent of the total responses provided in Study 1. Previous research has also found a relationship between
engagement-related constructs and voice. For example, Graham and Van Dyne (2006) found that experienced
signicance of work was predictive of voice. ODriscoll, Pierce, and Coghlan (2006) reported that organization-based
psychological ownership was positively related to voice. Tyler and Blader (2003) found that organizational identica-
tion also was positively related to voice. This dimension also partially overlaps with the acquiescent dimension of
silence proposed by Van Dyne et al. (2003), but as previously discussed, their conceptualization of acquiescent silence
also includes feelings of low self-efcacy to make a difference. In contrast, the ndings of this study indicate that silence
based on feeling unable to make a difference, and silence based on disengagement are empirically distinct.
DOI: 10.1002/job
Summary of Study 2
Scholars have proposed that employee silence is a multidimensional construct based on different underlying motives,
but there has been little empirical substantiation regarding the nature of these dimensions. This study builds upon
and extends this line of reasoning through exploratory investigation of the factor structure of the silence reason
exemplars derived in Study 1. On the basis of this investigation, it appears that distinct dimensions of motives for
employee silence do exist. However, the results indicate that the dimensionality is more varied than the silence
dimensions proposed in the literature (e.g., Pinder & Harlos, 2001; Van Dyne et al., 2003). In this study, six distinct
silence dimensions emerged. The defensive dimension is similar to what Pinder and Harlos (2001) referred to as
quiescent silence (i.e., deliberate omission based on fear) and has been the main focus of subsequent research. However,
other dimensions also emerged, which have not been directly investigated in research on employee silence but also
appear to be common (ineffectual, relational, difdent, disengaged). As indicated, a deviant silence dimension also emerged,
but it appears to be much less common. However, as discussed earlier, this may partly be due to methodological reasons.
Study 3
One of the weaknesses of EFA is its inability to quantify the goodness of t of the resulting factor structure. In
contrast, CFA enables assessment of the quality of the factor structure by statistically testing the signicance
of the overall model and of item loadings on factors. This affords a stricter interpretation of the underlying
dimensionality than does EFA (Hinkin, 1998). Therefore, in this study, I conducted a CFA of the silence items
retained in Study 2 to further test and rene construct validity.
Sample
Respondents consisted of undergraduate business students with at least two years work experience (n = 180, 73
percent response rate) and evening MBA students at a midwestern university (n = 102, 65 percent), and front-
line and administrative employees of a midwestern manufacturing company (n = 229, 55 percent). A total of 511
(62 percent) responded. For CFA, recommendations concerning sample size vary considerably. For example,
Schwab (1980) recommended using at least 10 times as many respondents as items. For this study, there were
17.6 times as many respondents as items. I recruited students via e-mail, and the human resource department of
the participating organization recruited organization members by sending an e-mail on my behalf. The average
age of respondents was 31 years, the average work experience was eight years, 79 percent were Caucasian, and
56 percent were female. This sample was distinct from those of Studies 1 and 2, and respondents spanned multiple
functions and hierarchical levels in a wide range of industries.

I administered a web-based survey that randomly presented one of the two following scenarios to respondents:
(a) think of a time at work when you either witnessed or experienced an unfair or unethical situation, or (b) think
of a time at work when you encountered a signicant operational problem or opportunity for improving a
particular work process or function.
I based the specication of these two particular scenarios on prior research (e.g., Milliken et al., 2003) and the
ndings of Study 1 where employees reported frequently remaining silent to both of these types of incidents.
DOI: 10.1002/job
C. T. BRINSFIELD
In addition, the apparent conceptual distinction between these two types of incidents enabled investigation of
construct validity across meaningfully different types of incidents.
Depending on which random scenario was presented, I then presented the respondents with the appropriate
instructions:
Below are different reasons employees may have for wanting to remain silent to their supervisor or management
in response to (a) an unfair or unethical situation at work, or (b) encountering an operational problem or
opportunity for improvement at work. Please indicate the extent to which you experienced each of the following
reasons for wanting to remain silent to your supervisor or management in response to the situation you described
above.
Next, I presented items retained from Study 2 in random order. The reason that the target of silence was con-
strained to supervisor or management was based on the ndings of Study 1 in which these targets emerged as
the most common (Table 2).

Factor structure
I conducted CFA using LISREL 8.8 (Scientic Software International, Inc., Skokie, IL; Jreskog & Srbom, 1993). I
used the maximum likelihood estimation with the raw data as input. I rst compared the t of six different factor
structures. In this analysis, I included all responses (i.e., both scenarios). In the rst model, all items were indicative
of one larger factor. The second model was a two-factor model, with deviant silence as the rst factor and the
remaining ve dimensions indicated in Table 3 as the second factor. This conguration was based on the differential
in contribution to total explained variance in Study 2 between deviant silence and the other ve dimensions, and the
conceptual difference between deviant silence and the other ve dimensions. The three-factor model was roughly
based on dividing the second factor in the two-factor model into the two dimensions of quiescence and acquiescence
proposed by (Pinder & Harlos, 2001). The four-factor model was based on separating relational items into a distinct
dimension and was partly guided by prior theorizing on the other-orientation aspect of prosocial silence (Van Dyne
et al., 2003). In the ve-factor model, items representing the difdent dimension in Table 3 were assigned to an
additional factor. This was partly based on the conceptual similarity between items representing this dimension
and individual traits, which have been linked to propensity to express voice (e.g., neuroticism, self-efcacy). The
six-factor model was based on the factor structure that emerged in Study 2.
I show the t statistics for the models (16) in Table 4. Many researchers gauge chi-square relative to its degrees
of freedom, with a ratio of 2 often used as an arbitrary indicator of good t (Arbuckle, 1997). As indicated in Table 4,
according to chi-square calculations, none of the models t particularly well. The CFI and the NNFI have been
shown in Monte Carlo studies to be more resistant to sample size effects (Bentler, 1990). For both the CFI and
NNFI, a value closer to l indicates a better t with a value of 0.90 suggested as an arbitrary indicator of good t
(Wegener & Fabrigar, 2000). Both the CFI (0.96) and NNFI (0.96) suggest that the six-factor model is a good
t to the data. The RMSEA asks the question How well would the model, with unknown but optimally chosen
parameter values, t the population covariance matrix if it were available? (Brown & Cudeck, 1993, pp. 137138).
It then measures that discrepancy relative to the degrees of freedom. Brown and Cudeck (1993) have argued that values
greater than 0.10 indicate poor t, values between 0.08 and 0.10 indicate mediocre t, values between 0.05 and 0.08
indicate reasonable t, and values less than 0.05 indicate good t. As indicated in Table 4, on the basis of the RMSEA
(0.087), the six-factor model falls into the mediocre range. Assessing whether the t of the six-factor model is
signicantly better than that of the other models is traditionally accomplished using a chi-square difference test. For
example, the difference in chi-square between the six- and ve-factor model is 694, which is itself distributed as chi-
square with (395 390 = 5) degrees of freedom. The fact that this value is statistically signicant would suggest that
the six-factor model is signicantly better than the ve-factor model. Similarly, the chi-square difference test indicates
DOI: 10.1002/job
Table 4. Study 3: Silence measuresCFA results.

Model No. RMSEA 90% condence
no. factors N w2 w2 df w2/df CFI NNFI RMSEA interval
1 1 511 7069 5366*** 405 17.45 0.80 0.79 0.22 0.22, 0.23
2 2 511 5177 3474*** 404 12.81 0.86 0.85 0.19 0.18, 0.19
3 3 511 4259 2556*** 402 10.59 0.90 0.88 0.17 0.17, 0.17
4 4 511 3221 1518*** 400 8.05 0.92 0.92 0.14 0.13, 0.14
5 5 511 2397 694*** 395 6.06 0.94 0.94 0.11 0.11, 0.11
6 6 511 1703 0 390 4.36 0.96 0.96 0.087 0.083, 0.091
7 6 255 1089 614*** 390 2.79 0.95 0.95 0.090 0.084, 0.096
8 6 239 1101 602*** 390 2.82 0.95 0.95 0.092 0.086, 0.098
All w2 calculations use Model 6 as a baseline.
Models 16 include all responses (i.e., both scenario types combined).
Model 7 includes only responses for the scenario type of unfair or unethical situation.
Model 8 includes only responses for the scenario type of operational problem or opportunity for improvement.
See Study 3, Analysis and results, for descriptions of the different models.
***p < .001.
the ve-factor model is a better t than the four-factor model, the four-factor model is better than the three-factor model,
and so on. However, the chi-square difference test is only appropriate in comparing nested models. One model is
nested within another if the model is a special case of the other (e.g., a more restricted version of it). It is not clear
whether a six-factor model is a more restricted version of the other models because new latent variables are introduced.
However, model comparisons can also be made using the 90 percent condence interval of the RMSEA (Colquitt, 2001).
This comparison shows that the six-factor model is signicantly better than the ve-factor model because their
condence intervals do not overlap. Moreover, the ve-factor model is signicantly better than the four-factor model,
which is itself signicantly better than the three-factor model, and so on.
To assess generalizability across different types of issues, situations, or concerns to which employees may remain
silent, I also calculated t indices for the six-factor model on the basis of the two subsets of scenario types (unfair or
unethical situation; Model 7) and (operational problem or opportunity for improvement; Model 8). In addition, I
conducted a multigroup CFA to test the invariance of the six-factor model across the two subsets of scenario types,
as well as across the three subsamples of undergraduate students, MBA students, and employees of the participating
organization. Although the t indices reported in Table 4 for Models 7 and 8 indicate a similar level of t as when all
cases were included (i.e., 90 percent condence intervals for RMSEA overlap), this does not necessarily mean that
the models are invariant across these different scenarios. In the multigroup CFA, I compared the six-factor model
with all factor loadings forced to be equal for the different subsamples with one in which all factor loadings were
allowed to vary. Results from the across scenario multigroup analysis revealed that the chi-square difference was
statistically signicant (w2 [36df] = 166), indicating that the factor loadings might not be invariant across these
two different scenarios. However, considering that the t statistics between the two models were very similar, the
factor loadings may be equivalent from a practical standpoint. Similarly, results from the multigroup analysis across
respondent subsamples also revealed that the chi-square difference was statistically signicant (undergraduates
compared with MBAs, w2 [36df] = 77) and (MBAs compared with employee sample, w2 [36df] = 99), indicating
that the factor loadings might not be invariant across the three subsamples. Fit statistics, however, between the three
subsamples revealed notable similarities and differences. The CFI for undergraduate (n = 180), MBA (n = 102), and
employee (n = 229) samples respectively were 0.95, 0.95, and 0.97. The NNFI were respectively 0.94, 0.95, and
0.96. The RMSEA were 0.094, 0.097, and 0.069 respectively across the three subsamples. As indicated, the CFI
and NNFI are similar across these subsamples, but the RMSEA suggests that the six-factor model ts the employee
subsample better than the undergraduate or MBA subsamples. This is further supported by the 90 percent condence
interval of the RMSEA, which is 0.063 to 0.076 for the employee subsample, whereas it is 0.087 to 0.11 for the
undergraduate and MBA subsamples combined.
DOI: 10.1002/job
C. T. BRINSFIELD
To investigate if the six-factor model can be more meaningfully conceptualized at higher orders of abstraction, I
also conducted CFA with different congurations of second-order latent constructs. In the rst of these analyses, I
recongured the model with one second-order latent construct with all six dimensions indicating this higher order
factor (w2 [9df] = 127). The second analysis consisted of two second-order latent factors indicted by defensive,
relational, and difdent, and ineffectual, disengaged, and deviant (w2 [8df] = 122). The third analysis was identical
to the second except that I did not assign deviant to a second-order construct (w2 [8df] = 105). Chi-square
difference tests indicated that the t was signicantly worse for all the second-order factor models compared with
the original rst-order model.
Correlation analyses, means, standard deviations, and scale reliabilities

I investigated the relationships between the various dimensions of employee silence, examined means and standard
deviations, and calculated Cronbachs alphas to assess scale reliabilities. I report these results in Table 5.
Correlations between the various silence dimensions are in all cases signicant, but not so strong as to indicate that
the scales are not measuring distinct dimensions of silence (Law, Wong, & Mobley, 1998). The standard deviations
indicate sufcient discrimination between respondents (DeVellis, 1991), and all of the scales have sufcient
reliability to conclude that each is assessing a singular factor construct (Nunnally & Bernstein, 1994).
Summary of Study 3
The purpose of this study was to quantitatively assess the goodness of t of the measurement model by comparing a
single common factor model, models with different silence dimensions combined, and a model based on the six
dimensions revealed in Study 2. Although the t of the six-factor model as indicated by the RMSEA is not in the
range considered good by many contemporary standards, it is signicantly better than the models with fewer factors.
Moreover, on the basis of the CFI and NNFI, the six-factor model appears to t the data reasonably well. In addition,
the t of the six-factor model is signicantly better with the subsample of actual employees (RMSEA = 0.069),
suggesting that the different forms of silence may be more meaningful to employees in an organizational setting
as opposed to students with typically less work experience. Moreover, although the models are not statistically
invariant across the two scenarios examined in this study (unfair/unethical situation, or operational problem/
opportunity for improvement), the t statistics are similar enough across these scenarios to suggest practical equivalence.
This not only provides some evidence of the generalizability of the measures across different types of incidents to which
employees may remain silent but also suggests that silence is experienced differently in different situations. Next, further
evidence of construct validity is obtained by examining how the scales incrementally predict voice behavior and correlate
to conceptually related factors.
Table 5. Study 3: Descriptive statistics and correlations.

Mean SD 1 2 3 4 5 6
1. Deviant silence 1.62 1.01 (.95)

2. Relational silence 3.36 1.66 .26** (.92)
3. Defensive silence 3.12 1.48 .33** .53** (.92)
4. Difdent silence 2.77 1.30 .39** .49** .58** (.89)
5. Ineffectual silence 3.14 1.43 .31** .25** .51** .54** (.90)
6. Disengaged silence 2.65 1.44 .51** .52** .41** .55** .39** (.83)
Alpha reliabilities are shown on the diagonal in parentheses.
N = 511.
*p < .05; **p < .01.
DOI: 10.1002/job
Study 4
To begin explication of the nomological network (Cronbach & Meehl, 1955), and to investigate the incremental
value in examining employee silence motives, I administered the scales developed in the previous studies to a
different group of respondents. In addition, I administered measures of four related constructs (employee voice
behavior, psychological safety, neuroticism, and extraversion), as well as several demographic variables commonly
used in prior voice-related research (e.g., LePine & Van Dyne, 1998).
Employee voice behavior
One of the most prevalent approaches to this broad eld of inquiry (see Brinseld et al., 2009), which the study of
silence motives seeks to complement, is the study of employee voice behavior. One of the most widespread
conceptualizations of employee voice behavior is that of Van Dyne and LePine (1998), which they dened as
promotive behavior that emphasizes expression of constructive challenge intended to improve rather than merely
criticize (p. 109). Investigations of this conceptualization of voice often rely on a supervisors or co-workers
ratings of the focal employees voice behavior (e.g., Morrison et al., 2011). Although a central tenet on which this
research is based is that silence motives are not simply equivalent to an absence of voice behavior, it does not mean
that they are unrelated. Quite the contrary, because motives are an integral part of behavioral intention and
subsequent behavior (Ajzen & Fishbein, 1980), I expected the silence motives to be negatively related to voice
behavior that is discernible by ones supervisor. Moreover, considering that the study of voice behavior is arguably
the most prevalent approach to this area of inquiry (Morrison, 2011), a logical step in establishing overall construct
validity of silence motives is to examine convergent, divergent, and incremental predictive validity relative to voice
behavior.
Hypothesis 1: Voice behavior will be negatively related to each of the six employee silence motives.
Psychological safety
Edmondson (1999) dened team psychological safety as a shared belief that the team is safe for interpersonal risk
taking (p. 354). This construct has also been successfully adapted to measure individual-level perception of
psychological safety with the organization as the foci (e.g., Baer & Frese, 2003). Previous research has shown a
signicant relationship between psychological safety and employee voice behaviors (e.g., Edmondson, 2003).
Considering that defensive silence is based on an underlying motive to protect oneself from negative consequences
associated with speaking up, I expected psychological safety to be signicantly negatively related to the defensive
dimension of employee silence. Considering that difdent silence is based on personal insecurities and uncertainties,
I expected psychological safety to be signicantly negatively related to the difdent dimension of employee silence.
This is because in a psychologically safe environment, employees personal insecurities and uncertainties may be
less salient or personal vulnerabilities perceived as less likely to be exploited. Considering that relational silence
is based on not wanting to harm a relationship or general relational concerns, I expected this dimension of employee
silence to be signicantly negatively related to psychological safety. This is because in a psychologically safe
environment, speaking up may not be expected to have as negative of an impact on a relationship, or damage to
a relationship may not be expected to have as negative of consequences.
Hypotheses 2a2c: Psychological safety will be negatively related to (a) defensive silence, (b) difdent silence,
and (c) relational silence.
DOI: 10.1002/job
C. T. BRINSFIELD
Neuroticism
Neuroticism is a personality trait described as anxious, depressed, angry, embarrassed, emotional, worried, or insecure
(Barrick & Mount, 1991). Previous research has shown a negative relationship between ones willingness to express
themselves and neuroticism (e.g., LePine & Van Dyne, 2001). Previous research also has shown a positive relationship
between neuroticism and deviant workplace behavior (Colbert, Mount, Harter, Witt, & Barrick, 2004; Cullen & Ones,
2001). Therefore, on the basis of both the typically negative relationship between neuroticism and self-expression, and
evidence indicating a positive relationship between neuroticism and deviant workplace behavior, I expected deviant
silence to be positively related to neuroticism. Difdent silence is based on ones insecurities, self-doubt, or uncertainties
regarding the situation or what to say. Previous research also has shown a positive relationship between neuroticism- and
difdent-related constructs (e.g., self-efcacy; Judge, Bono, Erez, & Locke, 2005). On the basis of both the typically
negative relationship between neuroticism and self-expression, and evidence indicating a positive relationship between
neuroticism- and difdent-related constructs, I expected difdent silence to be positively related to neuroticism.
Hypotheses 3a3b: Neuroticism will be positively related to (a) deviant silence and (b) difdent silence.
Incremental validity
Part of the value in examining silence motives rests not only on their own reliability and internal validity but on whether
they are distinct from and demonstrate incremental validity over other predictors of voice behavior. I therefore
conducted analyses in which I controlled for demographic factors commonly used in prior research on voice (gender,
age, and organizational tenure). To further test the incremental effects of silence motives on voice behavior, I also
controlled for psychological safety and neuroticism (see preceding discussions). In addition, I also controlled for
extraversion, a personality trait described as sociable, gregarious, assertive, talkative, and active (Barrick & Mount,
1991), and which has been shown to predict voice behavior (LePine & Van Dyne, 2001).
Hypothesis 4: Each of the six silence motives predict voice behavior in analyses controlling for demographic
variables, psychological safety, neuroticism, and extraversion.
Sample
I recruited a total of 318 employees, and 221 (69 percent) completed the survey. Their supervisor also rated 190
(86 percent) from this sample for voice behavior. Respondents consisted of full-time front-line and administrative
personnel of a midwestern manufacturing company, and the human resource department recruited them by sending
an e-mail on my behalf. We assured the respondents that their responses would be kept condential; however, we
asked them for identifying information to enable matching to their supervisors ratings of voice behavior. The
average age of respondents was 36 years, the average tenure with this organization was ve years, 42 percent were
female, and all were distinct from the respondents in the previous studies.

I administered a web-based survey to respondents with the following instructions:
Below are different reasons employees may have for wanting to remain silent to their supervisor in response to
encountering an operational problem or opportunity for improvement at work. Please indicate the extent to which
you have experienced each of the following reasons for wanting to remain silent to your supervisor in response to
encountering an operational problem or opportunity for improvement at work.
DOI: 10.1002/job
Supervisor was specied as the target of silence because of the relatively high frequency that supervisors were
reported as targets of silence in Study 1 and because the dependent variable was supervisors rating of employee
voice. The type of incident to which one has experienced the various motives for silence was constrained to an
operational problem or opportunity for improvement because of (i) the ample frequency that this type of incident
was reported in Study 1, (ii) prior research on employee voice and silence has often focused on these types of
incidents (e.g., LePine & Van Dyne, 1998; Van Dyne et al., 2003), and (iii) management at the participating
organization was particularly interested in silence in this context.
Immediately following these instructions, I presented the silence scales to respondents. I presented measures of
psychological safety, neuroticism, and extraversion following the silence scales. I measured psychological safety
by adapting seven items from Edmondsons (1999) measure of team psychological safety to the organization as
the foci (Baer & Frese, 2003). I assessed both neuroticism and extraversion with three-item bi-polar instruments
developed by Langford (2003). Next, I asked respondents several demographic questions. I coded respondents
gender (0/1, 1 = female). I assessed respondents age using six categories ranging from 18 to 24 to 56 or older.
I assessed respondents tenure with the organization with six categories ranging from less than six months to
16 years or longer. To mitigate common-method variance (cf. Campbell & Fiske, 1959), I administered a web-
based survey to supervisors asking them to rate all their direct reports voice behavior with six items adapted from
Van Dyne and LePine (1998), which I then matched to the employee responses.
Table 6 presents the means, standard deviations, correlations, and estimated reliabilities (on the diagonal).
Cronbachs alphas for ve of the six silence scales are above the .80 threshold commonly interpreted as good
internal consistency (Nunnally & Bernstein, 1994). The modest alpha for disengaged silence may be due to the scale
being composed of only three items, whereas the other silence scales contain ve or six items. However, in Study 3,
disengaged silence had an alpha of .83, suggesting that further research is necessary to determine if more items are
needed. In general, correlations are as expected (e.g., voice behavior is positively correlated with both psychological
safety and extraversion). Also as expected, the six silence motives are positively correlated, but not so strongly as to
suggest that they are not each reecting a distinct phenomenon.
The results illustrate that ve of the six dimensions of employee silence are signicantly negatively related to
voice behavior. This provides partial support for Hypothesis 1. In addition, the low to moderate strength of these
relationships indicate that the dimensions of employee silence motives are not simply equivalent to an absence of
voice behavior. Deviant silence is not related to voice behavior that is discernible by ones supervisor. This may
be due to harmful behavior typically being more covert than innocuous behavior (Robinson & Bennett, 1995),
and therefore less recognizable, but more research is needed to determine this.
The results also illustrate that defensive silence, difdent silence, and relational silence are all signicantly
negatively related to psychological safety, providing support for Hypotheses 2a, 2b, and 2c. This provides further
evidence of construct validity as the underlying nature of these three dimensions of silence are to varying extents
conceptually related to negative consequences associated with speaking up, albeit the nature of the negative
consequences associated with each of these dimensions are different (Study 2). As expected, the results also indicate
that deviant silence, ineffectual silence, and disengaged silence are not signicantly related to psychological safety.
This appears to make sense as the nature of these dimensions is not based on perceptions of negative consequences
associated with speaking up. The results furthermore indicate that neuroticism is positively related to deviant silence,
providing support for Hypothesis 3a. The results also indicate that neuroticism is positively related to difdent
silence, providing support for Hypothesis 3b.
To test Hypothesis 4, I performed a series of hierarchical multiple regression analyses (Table 7). Model 1 is the
baseline model wherein I entered the demographic variables, psychological safety, neuroticism, and extraversion as
control variables. Together, these variables accounted for 13 percent of the variance in voice behavior reported by
DOI: 10.1002/job
C. T. BRINSFIELD
Table 6. Study 4: Descriptive statistics and correlations.

Mean SD 1 2 3 4 5 6 7 8 9 10 11 12 13
1. Gender 0.42 0.49
Copyright 2012 John Wiley & Sons, Ltd.

2. Age 35.58 12.96 .06
3. Organizational tenure 4.73 4.83 .12 .88**
4. Deviant silence 2.00 1.20 .11 .06 .02 (.94)
5. Relational silence 3.38 1.48 .14 .04 .02 .20** (.93)
6. Defensive silence 3.71 1.41 .03 .03 .01 .24** .47** (.91)
7. Difdent silence 3.00 1.09 .07 .04 .03 .34** .48** .45** (.86)
8. Ineffectual silence 3.64 1.30 .05 .06 .07 .13 .22** .33** .38** (.87)
9. Disengaged silence 3.43 1.23 .05 .02 .03 .37** .30** .36** .38** .29** (.68)
10. Voice behavior 4.10 0.95 .04 .23** .26** .02 .28** .29** .21** .20** .20** (.63)
11. Psychological safety 4.07 0.94 .02 .01 .01 .05 .28** .38** .25** .02 .02 .17* (.68)
12. Neuroticism 3.50 1.35 .08 .01 .01 .16* .20** .21** .22** .12 .10 .14 .13 (.86)
13. Extraversion 4.55 1.32 .02 .03 .03 .01 .10 .07 .16* .01 .04 .26** .02 .38** (.84)
Alpha reliabilities are shown on the diagonal in parentheses. Gender coded 0/1, 1 = female.
N = 190.
*p < .05; **p < .01.
DOI: 10.1002/job
J. Organiz. Behav. (2012)
Table 7. Study 4: Silence motives as predictors of voice behavior.

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7
Control variables
Gender 0.01 0.01 0.03 0.01 0.01 0.01 0.01
Age 0.01 0.01 0.03 0.04 0.01 0.01 0.03
Organizational tenure 0.26 0.26 0.23 0.22 0.26 0.27 0.23
Psychological safety 0.18* 0.18* 0.10 0.07 0.14 0.16* 0.17*
Neuroticism 0.07 0.07 0.01 0.01 0.41 0.04 0.05
Extraversion 0.23** 0.23** 0.22** 0.24** 0.22** 0.23** 0.23**
Silence motives
Deviant silence 0.01
Relational silence 0.24**
Defensive silence 0.25**
Difdent silence 0.13
Ineffectual silence 0.19**
Disengaged silence 0.18**
R2 0.13 0.13 0.18 0.18 0.15 0.17 0.16
R2 0.00 0.05 0.05 0.02 0.04 0.03
F 5.99** 5.11** 6.95** 6.98** 5.68** 6.50** 6.25**
F 0.02 10.77** 10.93** 3.33 8.15** 6.69**
Standardized regression coefcients are reported.
Model 1 is the baseline model for Models 2 through 8.
N = 190.
*p < .05;
**p < .01.
respondents supervisors. In Models 2 through 7, I entered each of silence motives separately after all of the control
variables had been entered. As indicated in Table 7, relational, defensive, ineffectual, and disengaged silence
motives each accounted for a signicant incremental increase in voice behavior over the control variables. Neither
deviant nor difdent silence motives accounted for an incremental increase in voice behavior over the control
variables. Therefore, these results offer only partial support for Hypothesis 4.
Summary of Study 4
This study was designed to further test the construct validity of the employee silence scales developed in the previous
studies and to show the value of empirically examining the dimensions of silence as distinct constructs. Clearly, more
work is needed in both regards, but this study does provide preliminary evidence that the scales generally perform as
expected in relation to voice behavior, psychological safety, neuroticism, and extraversion. The ndings that ve of
the six of silence motives are related to voice suggests that they have predictive validity with regard to speaking up
behaviors. The ndings that four of the six silence motives predict voice behavior when controlling for variables
previously linked to voice further demonstrates the incremental value in examining silence motives.
Although this study shows the predictive validity of the different silence motives with regard to supervisors
assessments of voice, part of the value in examining silence motives is that employees may experience different
motives for silence, which are not necessarily reected in assessments of voice behavior. In this regard, the ndings
of this research do support that employees experience distinct motives for silence as indicated by the means,
standard deviations, and alpha reliabilities in Table 6. Moreover, the moderate strength of the relationships between
the silence motives and voice behavior suggests that employees may experiences motives for remaining silent,
which are not manifest in observations of voice.
DOI: 10.1002/job
C. T. BRINSFIELD
General Discussion
There is considerable evidence that employees willingness to speak up about substantial matters at work is related
to important individual, team, and organizational outcomes (Morrison, 2011). However, it has been argued that there
is still a great deal that we do not know about situations where employees have something meaningful to say, but are
compelled to remain silent (Detert & Edmondson, 2011; Edwards & Greenberg, 2009). It has been suggested that
reasons for this include (i) the assumption that silence is essentially equivalent to an absence of voice behavior,
(ii) research that has focused on silence has mainly been conceptual or qualitative, and (iii) prior research has largely
focused on employees wanting to avoid the risks associated with speaking up to the exclusion of other possible
motives for silence. The four studies presented here, to varying degrees, begin to address all of these concerns.
Construct proliferation and construct redundancy represent signicant problems in management sciences research
(Le, Schmidt, Harter, & Lauver, 2010). Considering the wide variety of voice- and silence-related constructs
previously developed (for reviews, see Brinseld et al., 2009; Morrison, 2011), it is important to establish the value
in studying employee silence motives as distinct constructs. Although more conceptual and empirical work is
needed to afrm this, the research presented here provides evidence that the empirical investigation of silence
motives is warranted. Previous research on employee voice- and silence-related constructs have largely focused
on, and operationalized, the behavioral aspects of the phenomena (e.g., Tangirala & Ramanujam, 2008). However,
given the potentially covert and ambiguous nature of silence, measuring voice or silence behavior and inferring the
nature and extent of intentional silence are insufcient. In addition, many other constructs have been developed and
studied, which relate to employee voice behaviors (e.g., psychological safety, trust, organizational justice); however,
silence motives appear conceptually distinct from these existing constructs.
Another important contribution of this research is the illumination of the variety of different underlying motives
for silence. Whereas prior research on employee silence has mainly focused on silence in response to perceived risks
associated with speaking up, this research shows that motives for silence other than fear of consequences also are
common. In fact, items loading on defensive silence only accounted for 12.65 percent of the total response in Study
1. It is clear from this research that if we want to more fully understand the scope of silence in organizations, we
have to focus on other motives in addition to risk avoidance.
Practitioners also stand to benet from the development of measures of distinct silence motives. Most managers
say that they want employees to be willing to speak up about important matters (Bennis, Goleman, & OToole,
2008), but the tools available for managers to assess and understand intentional silence in the workplace have been
limited. The ability to reliably measure the strength of the various motives for silence will provide managers with
important insights regarding the nature and extent of workers reluctance to speak up. This will in turn enable more
targeted and effective strategies for managing silence in the workplace. This is especially important considering the
inconspicuous nature of silence and that because of initiatives such as open door policies, managers may believe
that they are hearing more than they actually are.
Limitations
As with any new construct, extensive research is required to rene the conceptual domain and improve validity, as
construct and scale development are iterative processes. As such, the research presented here does not constitute
comprehensive validation. Although these studies make numerous contributions, there are several limitations that
may impair the validity and generalizability of the results. This research utilized a mixed-methods design, employing
both qualitative and quantitative techniques. The qualitative aspects of this study likely provided a broader
understanding of employee silence than would have been possible from a limited number of deductively predetermined
questions. Notwithstanding this important benet, this technique does have its drawbacks. As the data gathered in this
DOI: 10.1002/job
study could be interpreted and categorized in a variety of different ways, there is inherent subjectivity in the process,
which could jeopardize the validity of the results. To mitigate this subjective element, I took several important steps.
First, I provided detailed documentation of the data gathering and analyses. Second, I used independent response
analysts during the coding and categorization phases. Third, I calculated ratings of agreement and found them to be
satisfactory for this type of research. In addition, to attenuate social desirability bias, I provided respondents assurance
of anonymity. And nally, the diversity of the respondents and ample number of silence reasons reported in Study 1
(n = 2277) appears to have yielded a saturated list of silence reason exemplars.
Pinder and Harlos (2001) noted that one of the difculties associated with studying employee silence was coming
to grips with what constitutes the boundaries of the construct itself (p. 332). Although they made this assertion over a
decade ago, this continues to be the case (Edwards & Greenberg, 2009). Studies 1 and 2 were intentionally broad
with respect to the boundary constraints imposed as the objective was to inductively investigate the content domain
of employee silence and foster a greater understanding of the implications of two important boundary constraints,
(i) the type of issue, situation, or concern to which one is remaining silent, and (ii) the target of silence. Because
the silence reason exemplars used to create the scale items were derived within this broad conceptual space, it is
not clear if different exemplars would have emerged if specic boundary constraints had been imposed. It is likely
that taking a broad approach led to a more comprehensive list of exemplars, but the relative frequencies of
occurrence would vary if specic boundary constraints were imposed (e.g., defensive-related exemplars may be
more prevalent when the target of silence is a supervisor than when it is a subordinate). Moreover, as indicated in
Study 3, when the measures derived in this broad context were adapted to more specic boundary constraints
concerning the nature of the silence incident (unfair or unethical situation) and operational problem or opportunity
for improvement, the t statistics were similar enough across these scenarios to suggest practical equivalence.
Another important limitation involves the self-report nature of all of the variables used in this study, with the
exception of supervisors ratings of voice behavior. This was unavoidable with silence motives as they are only
knowable by the person who is experiencing them. This is also the case with extraversion, neuroticism, and individual
perception of psychological safety. Therefore, it is not clear to what extent the resultant relationships were impacted by
common-method bias (Campbell & Fiske, 1959).
Suggestions for Future Research
Although this research has advanced our understanding of the nature of employee silence motives, we are just
beginning to build a foundation of knowledge on the topic. This research, in part, not only conrmed silence motives
previously proposed in the literature but also found evidence for motives not previously investigated in the employee
silence research. It is clear that further research is necessary, perhaps additional qualitative research in a variety of
contexts to corroborate and/or augment the ndings. Moreover, this research initially adopted a broad domain
relative to what is considered employee silence and then narrowed it relative to the type of incident an employee
is remaining silent in response to (i.e., an idea for improvement), and the target of silence (i.e., top management,
supervisor). However, more work is needed, which examines the validity of these silence motives across different
types of incidents (e.g., ambiguous threats, policy issues) and other types of targets (e.g., team members,
subordinates, outside authorities).
In addition, more research that builds on the idea of a climate of silence (Morrison & Milliken, 2000; Vakola &
Bouradas, 2005) promises to be informative. At this point, it is not clear if the dimensions found in this study would
be applicable to the team, unit, or organizational level of analysis. Moreover, this research assumes a face-to-face
context, but as prior research has found, peoples communicative behaviors may be affected by the type of
communication medium (i.e., telephone, e-mail, video-conferencing, etc.; see Kiesler & Sproull, 1992). The items
developed in this study may behave differently if the communication medium is specied.
DOI: 10.1002/job
C. T. BRINSFIELD
In Study 4, I examined the relationships between employee voice behavior, psychological safety, and the
dimensions of employee silence. However, because of methodological limitations, this study was unable to
establish causality. Laboratory studies may be useful for further investigating the causal nature of these and
other fundamental relationships. Past research by Moss and Martinko (1998) suggested that at least some
of the factors associated with employee silence can be manipulated in a laboratory setting. More expansive
experimental designs should build upon this past research to test the effects of contextual (e.g., organizational
policies and norms), target (e.g., power of the target relative to the actor), and individual-level factors
(i.e., self-esteem, self-efcacy, personality, etc.) that have been linked to withholding information. In addi-
tion, and as Detert and Edmondson (2011) pointed out, psychological safety has often been positioned as a
mediator between various antecedents and voice behavior. Considering the important role that psychological
safety plays in employee voice and silence, more work is needed, which examines how different aspects
of psychological safety (e.g., participative psychological safety; West, 1990) inuence silence motives and
related outcomes.
Also in Study 4, the dependent variable of interest was supervisors assessment of employee voice behavior. As
previously discussed, however, the different silence motives may be differentially related to other individual and
organizational outcomes. For example, in highly political organizational cultures, defensive or relational silence
may be more strongly related to performance outcomes, whereas in less threatening cultures, other forms of silence
may be more relevant. Moreover, future research that examines how the different motives interact to shape other
performance and psychological outcomes may be illuminating. For example, how are performance outcomes such
as team innovation, or psychological outcomes such as stress, affected when employees experience multiple motives
for silence?
Finally, it is recognized, of course, that employees should not speak up about all issues at all times. As
Milliken and Morrison (2003) observed, such organizations would be chaotic and disordered. Indeed, a
certain amount of silence is perhaps quite functional. Similarly, Van Dyne et al. (2003) proposed the idea that
silence may be based on prosocial motives, but there has been no empirical examination of this form of
silence. As indicated in this study, silence based specically on prosocial motives did not emerge from the
inductive investigation in Study 1. This may be due to pejorative assumptions associated with silence in
the workplace, which precluded respondents from thinking about this type of silence. This also may be due
to the expectation of benets sometimes associated with performing prosocial behaviors (Zahavi, 2000) and
that because of the concealed nature of silence, people would not expect this type of prosocial behavior to
result in the associated benets. However, these ideas are only speculation, and more research focusing on
the prosocial aspects of silence is needed.
Conclusion
Studying the motives for employee silence builds upon and extends previous theoretical and empirical work on
employee silence and voice by bringing a new perspective to this existing body of research. Whereas previous
research has generally examined the reasons for, and effects of, voice-related behaviors, the direct examination
of silence motives promises to further our understanding of situations where employees have something mean-
ingful to say, but remain silent. Perhaps most importantly, practitioners stand to benet greatly from this line of
inquiry because existing measures of voice do not assess the nature or extent of deliberate silence. Considering
the abundant evidence regarding the role of failures to communicate in such diverse events as the Bernie
Madoff scandal (Arvedlund, 2009) and the abuses at Abu Ghraib prison (Bennis et al., 2008), there is little
doubt that organizations of all types could benet from better understanding the nature and extent of intentional
silence in the workplace.
DOI: 10.1002/job
Author biography
Chad T. Brinseld is an assistant professor of management at the Opus College of Business, University of St.
Thomas. He earned his Ph.D. from The Ohio State Universitys Fisher College of Business. His research focuses
on employee silence and voice, workplace commitment, and trust.
References
Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ: Prentice Hall.
Arbuckle, J. L. (1997). Amos users guide (Version 3.6). Chicago, IL: Small Waters Corporation.
Argyris, C., & Schn, D. (1978). Organisational learning: A theory of action perspective. Reading, MA: Addison-Wesley.
Arvedlund, E. (2009). Too good to be true: The rise and fall of Bernie Madoff. New York, NY: Penguin Publishing.
Baer, M., & Frese, M. (2003). Innovation is not enough: Climates for initiative and psychological safety, process innovations, and
rm performance. Journal of Organizational Behavior, 24, 4568.
Barrick, M. R., & Mount, M. K. (1991). The Big Five personality dimensions and job performance: A meta-analysis. Personnel
Psychology, 44, 126.
Bennis, W., Goleman, D., & OToole, J. (2008). Transparency: How leaders create a culture of candor. San Francisco, CA:
Jossey-Bass.
Bentler, P. M. (1990). Comparative t indices in structural models. Psychological Bulletin, 107, 238246.
Brinseld, C. T., Edwards, M. S., & Greenberg, J. (2009). Voice and silence in organizations: Historical review and current
conceptualizations. In J. Greenberg, & M. S. Edwards (Eds.), Voice and silence in organizations (pp. 333). Bingley:
Emerald Group.
Brown, M. W., & Cudeck, R. (1993). Alternative ways of assessing model t. In K. A. Bollen, & J. S. Long (Eds.), Testing
structural equation models (pp. 136162). Thousand Oaks, CA: Sage.
Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait-multimethod matrix.
Psychological Bulletin, 56, 81105.
Clapham, S., & Cooper, R. (2005). Factors of employees effective voice in corporate goverance. Journal of Management and
Governance, 9, 287313. doi: 10.1007/s10997-005-7421-3
Colbert, A. E., Mount, M. K., Harter, J. K., Witt, L. A., & Barrick, M. R. (2004). Interactive effects of personality and perceptions
of the work situation on workplace deviance. Journal of Applied Psychology, 89, 599609.
Colquitt, J. A. (2001). On the dimensionality of organizational justice: A construct validation of a measure. Journal of Applied
Psychology, 86, 386400.
Cortina, L. M., & Magley, V. J. (2003). Raising voice, risking retaliation: Events following interpersonal mistreatment in the
workplace. Journal of Occupational Health Psychology, 8, 247265.
Cronbach, L. J., & Meehl, P. E. (1955). Construct validity in psychological tests. Psychological Bulletin, 52, 281302.
Cullen, M. J., & Ones, D. S. (2001). The role of conscientiousness and neuroticism in predicting police corruption. Paper
presented at the meeting of the Society for Industrial and Organizational Psychology, San Diego, CA.
Detert, J. R., & Edmondson, A. C. (2011). Implicit voice theories: Taken-for-granted rules of self-censorship at work. Academy of
Management Journal, 54, 461488.
DeVellis, R. F. (1991). Scale development: Theory and applications. London: Sage.
Edmondson, A. C. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44, 350383.
Edmondson, A. C. (2003). Speaking up in the operating room: How team leaders promote learning in interdisciplinary action
teams. Journal of Management Studies, 40, 14191452.
Edwards, M. S., & Greenberg, J. (2009). Sounding off on voice and silence. In J. Greenberg, & M. S. Edwards (Eds.), Voice and
silence in organizations (pp. 275291). Bingley: Emerald Group.
Fleiss, J. L. (1971). Measuring nominal scale agreement among many raters. Psychological Bulletin, 76, 378382.
Fleiss, J. L., Levin, B., & Paik, M. C. (2004). Statistical methods for rates and proportions (3rd ed.). New York, NY: Wiley.
Graham, J. W. (1986). Principled organizational dissent: A theoretical essay. In B. M. Staw, & L. L. Cummings (Eds.), Research
in organizational behavior (Vol. 8, pp. 152). Greenwich, CT: JAI Press.
Graham, J. W., & Van Dyne, L. (2006). Gathering information and exercising inuence: Two forms of civic virtue organizational
citizenship behavior. Employee Responsibilities and Rights Journal, 18, 89109.
Gruys, M. L., & Sackett, P. R. (2003). Investigating the dimensionality of counterproductive work behavior. International
Journal of Selection and Assessment, 11, 3042.
DOI: 10.1002/job
C. T. BRINSFIELD
Hinkin, T. T. (1998). A brief tutorial on the development of measures for use in survey questions. Organizational Research
Methods, 1, 104121.
Jreskog, K. G., & Srbom, D. (1993). LISREL 8: Users reference guide. Chicago, IL: SPSS.
Judge, T. A., Bono, J. E., Erez, A., & Locke, E. A. (2005). Core self-evaluations and job and life satisfaction: The role of self-
concordance and goal attainment. Journal of Applied Psychology, 90, 257268.
Kahn, W. A. (1990). Psychological conditions of personal engagement and disengagement at work. Academy of Management
Journal, 33, 692724.
Kiesler, S., & Sproull, L. (1992). Group decision making and communication technology. Organizational Behavior and Human
Decision Processes, 52, 96123.
Kohler, W. (1992). Gestalt psychology. New York, NY: Liveright.
Langford, P. H. (2003). A one-minute measure of the Big Five? Evaluating and abridging Shafers (1999a) Big Five markers.
Personality and Individual Differences, 35, 11271140.
Law, K. S., Wong, C.-S., & Mobley, W. H. (1998). Toward a taxonomy of multidimensional constructs. Academy of
Management Review, 23, 741.
Le, H., Schmidt, F. L., Harter, J. K., & Lauver, K. J. (2010). The problem of empirical redundancy of constructs in organizational
research: An empirical investigation. Organizational Behavior and Human Decision Processes, 112, 112125.
LePine, J. A., & Van Dyne, L. (1998). Predicting voice behavior in work groups. Journal of Applied Psychology, 83, 853868.
LePine, J. A., & Van Dyne, L. (2001). Voice and cooperative behavior as contrasting forms of contextual performance: Evidence
of differential relationships with Big Five personality characteristics and cognitive ability. Journal of Applied Psychology, 86,
326336.
Martin, R., & Woollard, S. (2009). Recognising and dealing with workplace stress in challenging times. Retrieved from http://
www.speechlys.com/recognising_and_dealing_with_workplace_stress
Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis: An expanded sourcebook (2nd ed.). Thousand Oaks, CA:
Sage.
Milliken, F. J., & Morrison, E. W. (2003). Shades of silence: Emerging themes and future directions for research on silence in
organizations. Journal of Management Studies, 40, 15631568.
Milliken, F. J., Morrison, E. W., & Hewlin, P. F. (2003). An exploratory study of employee silence: Issues that employees dont
communicate upward and why. Journal of Management Studies, 40, 14531476.
Morrison, E. W. (2011). Employee voice behavior: Integration and directions for future research. The Academy of Management
Annals, 5, 373412. doi: 10.1080/19416520.2011.574506
Morrison, E. W., & Milliken, F. J. (2000). Organizational silence: A barrier to change and development in a pluralistic world.
Academy of Management Review, 25, 706725.
Morrison, E. W., Wheeler-Smith, S. L., & Kamdar, D. (2011). Speaking up in groups: A cross-level study of group voice climate
and voice. Journal of Applied Psychology, 96, 183191.
Moss, S. E., & Martinko, M. J. (1998). The effects of performance attributions and outcome dependence on leader feedback
behavior following poor subordinate performance. Journal of Organizational Behavior, 19, 259274.
Noelle-Neumann, E. (1974). The spiral of silence: A theory of public opinion. Journal of Communication, 24, 4351.
Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York, NY: McGraw-Hill.
ODriscoll, M. P., & Coghlan, A. M. (2006). The psychology of ownership: Work environment structure, organizational
commitment, and citizenship behaviors. Group & Organization Management, 31, 388416.
Pinder, C. C. (1998). Motivation in work organizations. Upper Saddle River, NJ: Prentice Hall.
Pinder, C. C., & Harlos, K. P. (2001). Employee silence: Quiescence and acquiescence as response to perceived injustice. In
G. R. Ferris (Ed.), Research in personnel and human resources management (Vol. 20, pp. 331369). Greenwich, CT: JAI
Press.
Preacher, K. J., & MacCallum, R. C. (2002). Exploratory factor analysis in behavior genetics research: Factor recovery with small
sample sizes. Behavior Genetics, 32, 153161.
Robinson, S. L., & Bennett, R. J. (1995). A typology of deviant workplace behaviors: A multidimensional scaling study.
Academy of Management Journal, 38, 555572.
Schwab, D. P. (1980). Construct validity in organizational behavior. In B. M. Staw, & L. L. Cummings (Eds.), Research in
organizational behavior (Vol. 2, pp. 343). Greenwich, CT: JAI Press.
Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics. Boston, MA: Allyn and Bacon.
Tangirala, S., & Ramanujam, R. (2008). Employee silence on critical work issues: The cross level effects of procedural justice
climate. Personnel Psychology, 61, 3768.
Thorngate, W. (1976). In general vs. it depends: Some comments on the Gergen-Schlenker debate. Personality and Social
Psychology Bulletin, 2, 404410.
Tyler, T. R., & Blader, S. L. (2003). The group engagement model: Procedural justice, social identity, and cooperative behavior.
Personality and Social Psychology Review, 7, 349361.
DOI: 10.1002/job
Vakola, M., & Bouradas, D. (2005). Antecedents and consequences of organisational silence: An empirical investigation.
Employee Relations, 27, 441458.
Van Dyne, L., & LePine, J. A. (1998). Helping and voice extra-role behaviors: Evidence of construct and predictive validity.
Academy of Management Journal, 41, 108119.
Van Dyne, L., Ang, S., & Botero, I. C. (2003). Conceptualizing employee silence and employee voice as multidimensional
constructs. Journal of Management Studies, 40, 13591392.
Wegener, D. T., & Fabrigar, L. R. (2000). Analysis and design for nonexperimental data: Addressing causal and noncausal
hypotheses. In H. T. Reis, & C. M. Judd (Eds.), Handbook of research methods in social and personality psychology
(pp. 412450). Cambridge: Cambridge University Press.
West, M. A. (1990). The social psychology of innovation in groups. In M. A. West, & J. L. Farr (Eds.), Innovation and creativity
at work: Psychological and organizational strategies (pp. 309333). Chichester: John Wiley & Sons.
Whetten, D. A. (1989). What constitutes a theoretical contribution? Academy of Management Review, 14, 490495.
Zahavi, A. (2000). Altruism: The unrecognized selsh traits. Journal of Consciousness Studies, 7, 253256.
DOI: 10.1002/job

10 1002@job 1829

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

10 1002@job 1829

Hochgeladen von

Copyright:

Verfügbare Formate

Journal of Organizational Behavior, J. Organiz. Behav.

Received 6 January 2012

Design and procedure

Analysis and results

Table 1. Study 1: Silence incident types.

Experienced unfair treatment 260 21.0

Table 2. Study 1: Silence targets.

Upper management 504 40.6

Motives for remaining silent

Silence factor/dimension (Study 2)

23 I believed that speaking up 90 0.06 0.06 0.84 0.01 0.08 0.12

Silence factor/dimension (Study 2)

45 Management did not 88 0.05 0.01 0.09 0.15 0.95 0.09

Design and procedure

Analysis and results

Design and procedure

Analysis and results

Table 4. Study 3: Silence measuresCFA results.

Correlation analyses, means, standard deviations, and scale reliabilities

Table 5. Study 3: Descriptive statistics and correlations.

1. Deviant silence 1.62 1.01 (.95)

Employee voice behavior

Design and procedure

Analysis and results

Table 6. Study 4: Descriptive statistics and correlations.

1. Gender 0.42 0.49

Copyright 2012 John Wiley & Sons, Ltd.

Table 7. Study 4: Silence motives as predictors of voice behavior.

Suggestions for Future Research

Das könnte Ihnen auch gefallen