Sie sind auf Seite 1von 14

What Did You Do Today?

Children's Use of Time, Family Composition, and the Acquisition of


Social Capital
Author(s): Suzanne M. Bianchi and John Robinson
Source: Journal of Marriage and Family, Vol. 59, No. 2 (May, 1997), pp. 332-344
Published by: National Council on Family Relations
Stable URL: http://www.jstor.org/stable/353474 .
Accessed: 12/09/2013 15:09
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

National Council on Family Relations is collaborating with JSTOR to digitize, preserve and extend access to
Journal of Marriage and Family.

http://www.jstor.org

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:09:26 PM


All use subject to JSTOR Terms and Conditions

SUZANNEM. BIANCHIAND JOHNROBINSON

University of Maryland

WhatDid You Do Today?Children'sUse of Time,


FamilyComposition,andthe Acquisition
of Social Capital

Using time-diary data collected from a statewide


probability sample of California children aged
3-11, we examine the amount of time children
spend on four activities presumed to affect their
or
cognitive and social development-reading
being read to, watching TV, studying, and doing
household chores-and how that time varies by
four family characteristics: parental education,
maternal employment, number of parents in the
household, and family size. As expected, children
of highly educated parents study and read more
and watch TV less. Contrary to expectations, children of mothers who are employed part-time
watch significantly less TV than children of mothers at home full-time. Otherwise, there are few
significant differences by mother's extent of paid
employment, the presence of a father, and the
number of siblings. Thus, the results reinforce the
thesis that parental education is the predominant
predictor of the human and social capital investments that children receive.

Departmentof Sociology, 2112 Art-SociologyBuilding,College Park,MD 20742-1315(bianchi@bssl.umd.edu).


Key Words: children, family composition, social capital, time
use.

332

Changesin the Americanfamily have resultedin


a growing uneasiness about the well-being of
childrenin the United States. Some researchers
note how the increasededucationalattainmentof
parentsand declining family size bode well for
children's economic security (Bianchi, 1990;
Haveman& Wolfe, 1993; Hernandez,1993). At
the sametime, othershave emphasizedtrendslike
the increasein single parentingthatresultin more
povertyandunderminechildren'seconomicwellbeing (McLanahan& Sandefur,1994). Although
the researchon maternalemploymenthas been
mixed and inconclusive (Bronfenbrenner &
Crouter,1982; Presser,1995), there continuesto
be concernaboutthe potentiallynegativeeffects
of working mothers. Debate continues over
whetherthe dramaticincrease in mothers'labor
force participation and more single parenting
have diminishedparentalattentionto childrenand
eroded children's chances of success in school
andin otherarenasof life.
Sociologists have emphasizedthe importance
of parentaltransmissionof cognitive ability and
academicexpectationsfor the educationaland occupational achievement of children (Duncan,
Featherman,& Duncan, 1972; Sewell & Hauser,
1980). The notionof "humancapital,"a termsociologists borrowed from neoclassical
economists, infuses our understanding of the

Journalof Marriageandthe Family59 (May 1997):332-344

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:09:26 PM


All use subject to JSTOR Terms and Conditions

333

Children's Use of Time

goals of many childhood activities, particularly


schooling. In this view, accumulatingskills in
childhoodis of paramountimportance,andunderstandingwhat impedesand whatenhancesdevelopment of competencies that will translateinto
success later in life becomes an importantfocus
of the studyof childrenandchildhoodactivities.
Sociologists, particularly James Coleman
(1988), have broadenedthe economist'snotionof
humancapitalto encompasswhathas been called
"social capital."Social capital,like humancapital, increases children's futureproductivity,but
investmentsdependon the relationshipsin which
childrenare embedded.So, for example, a child
may have a talentedand highly educatedparent
and thus be genetically endowed with great potentialfor cognitivedevelopment,but interactions
with thatparentare neededto convey encouragement and expectations. Social interactionwith
parents thus can create resources that will enhancethe child'srealizationof his or herpotential
for achievement and later success. Parentalresources also place children in certain types of
communitiesandeducationalenvironmentswhere
social interactionscan providefurthersocial capital to facilitate"good"childoutcomes.
McLanahanandSandefur(1994) arguepersuasively thatthe dysfunctionaloutcomesof children
raisedwith one parent-such as the higherlikelihood of droppingout of high school, having an
early teen birth,and having difficultygettingestablished in the labor market-result not only
from the poorereconomiccircumstancesof these
childrenbut also from the diminishedinteraction
these childrenhave with parents(especially fathers).Childrenin one-parentfamiliesalso derive
less social capitalboth from the schools they attend and from the neighborhoodsin which they
live. Indeed,one of Coleman's(1988, 1992) concernsaboutthe increasein single parentingandin
maternalemploymentwas thatboth trendsin the
family underminethe social capital availableto
childrenin far-reaching
ways. Parentswerespending less time with their own childrenand fewer
adultswere at home duringthe day to supervise
the activityof groupsof childrenwhen they came
home to their neighborhoodsafter school each
day. Also, less time was available for volunteerismin schools andotherorganizationsand for
buildingsocial networksaroundthese institutions
thatprovideextensiveresourcesfor children.
The interrelatednotions of humancapitaldevelopment and social capital suggest that how
children spend their time is importantand that

parental(andcommunity)resourcesmay be critical in determiningwhich childrenengage in activities that enhanceintellectualgrowth,encourage responsibility,andgenerallysteerchildrentoward a productive adulthood. The human and
social capitalof childhoodarebuilt over time and
throughthe activities in which children engage
andthe qualityof the resourcesandsocial interactions thatsurroundthem.
However,we know relativelylittle abouthow
children spend their time. To date, there have
been only a handfulof time-diarystudiesof children'stime use, most of limitedscope. Nor do we
know how children'stime in activities varies in
differenttypes of families, althoughfamilies are
presumedto providevaryingdegreesof access to
the social capitalneededfor the successfuldevelopmentof humanpotential.
In this article,we conductan exploratoryanalysis of the amountof daily time childrenspendin
an arrayof activitiesthat might be deemed relevant to their accumulationof skills duringchildhood. We review the literatureon the interrelationship among family characteristicsand parents' and children's time use. With 1989-1990
time diarydatafor childrenin California,we examine the time childrenspend in four important
activities: reading or being read to, studying,
doing housework,and watchingTV. We address
fourquestionsaboutthe interrelationship
between
family characteristicsand time spentin these activities.
(a) Do children of motherswho are at
home full-time spend more time in cognitive enhancingactivities, such as reading,
than children of mothers who are in the
paid labor force? Are there differencesin
the amount of time children spend doing
housework,watching TV, or studying by
the laborforce statusof theirmother?
(b) Do children in two-parent households spend more time reading, doing
household chores, and studying, and less
time watching TV than children in oneparenthouseholds?
(c) Do children in smaller families
spend more time in these four activities
thanchildrenin largerfamilies?
(d) Finally,how does parentaleducation
relate to these activities? Do children of
better educated parents spend more time
studyingand readingand less time watching TV thanchildrenof less well educated
parents?

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:09:26 PM


All use subject to JSTOR Terms and Conditions

Journal of Marriage and the Family

334
STUDIESOF CHILDREN'STIMEUSE

Most previous time-use research on children's activities has focused on the time parents spend interacting with children (primarily preschool-age),
rather than on how these children themselves
spend their time. Although some studies of children's activity patterns have been undertaken
(Beschen, 1972; Medrich, Roisen, Rubin, &
Buckley, 1982; Wallace, 1987), very few of them
have been conducted with large, representative
samples that also provide full coverage of all
daily activities. The one exception to this was a
1981 study of 229 children conducted as a followup of parents interviewed in the 1975 University
of Michigan National Time Use Study (Juster &
Stafford, 1985). However, these data are now
over 15 years old, were obtained from a sample
with less than a 30% response rate, and did not
cover all activities.
Parental Education
In Timmer, Eccles, and O'Brien's (1985) analysis
of the 1981 Michigan data, higher parental education was correlated with more studying, less TV
watching, and more reading on the part of children. They highlighted the longstanding interest
in children's cognitive development and acquisition of aggressive behaviors that characterized the
developmental psychology literature on television
viewing. Underlying the concerns about children's time in front of the TV is the assumption
that, if children viewed less TV, they would engage in other, "more productive" activities like
reading. The 1981 survey showed a large amount
of television viewing on the part of children that
peaked at ages 11 to 12, but it was not clear that
less TV viewing resulted in children spending
more time in activities like reading. Yet the more
television that 9- to 17-year-old children watched,
the lower their scores on a standardized reading
comprehension test (Timmer et al., 1985, p. 370).
Whether less able readers selected to watch more
TV or whether large amounts of TV viewing limited reading comprehension was not clear, given
the cross-sectional nature of the 1981 time-diary
data for children.
The most extensive previous research related to
how children spend time comes indirectly from
educational differences in parental time with children. Early studies of parents of preschoolers suggested that college-educated mothers did devote
more time to child care than less well-educated
mothers. In addition, the quality of interaction was

also seen to be superior because more educated


mothers spent more time teaching children, taking
them on educational outings, and playing and interacting with children (Hill & Stafford, 1974;
Leibowitz, 1974, 1977). Others (e.g., Lindert,
1977) questioned the strength of the relationship
between education and parental time with children
and suggested that the relationship was quite sensitive to model specification and data used.
Subsequent analysis of 1975-1976 nationally
representative U.S. time diary data by Hill and
Stafford (1985) confirmed earlier conclusions that
college-educated mothers devoted significantly
more time to children than mothers with only a
high school education or less. In particular, Hill
and Stafford (1985, p. 427) argued that not only
did college-educated mothers commit more time
to children, but they engaged children in a greater
variety of activities at a higher level of (maternal)
involvement and with more predictability. They
concluded that:
while preschooldifferencesin quantityof care
time are substantial, the differences in predictabilityand varietyof care are also large and
that more educatedparentsinteractmore with
theirchildrenthroughouttheiryouththando less
educatedparents.Hence, it may also be consistency and varietyof interactionover a long time
period that create differencesin child development.(p. 435)
Maternal Employment
The study of the effect of maternal employment
on child outcomes has been fraught with methodological problems. Some research suggests that,
net of socioeconomic status, maternal employment negatively affects student achievement
(Milne, Myers, Rosenthal, & Ginsburg, 1986),
student discipline (Myers, Milne, Baker, & Ginsburg, 1987), and teachers' rating of the cognitive
abilities of elementary school age children
(Stafford, 1987). Other research has found maternal employment to be unrelated to child outcomes
(Dawson, 1991; Leibowitz, 1977) or related only
to outcomes for select subgroups. For example,
the relation is negative for high income boys
(Desai, Chase-Landsdale, & Michael, 1989), negative for boys (but not girls) in single-parent families (Krein & Beller, 1988), negative if employment is full-time and occurs in the first year or
two of life (Belsky & Eggebeen, 1991), but positive if it occurs in subsequent preschool years
(Blau & Grossberg, 1992). In sum, the research
on maternal employment remains inconclusive,

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:09:26 PM


All use subject to JSTOR Terms and Conditions

335

Children's Use of Time

often showing effects that are neitherstrongnor


consistent(Heyns & Catsambis,1986; Parcel &
Menaghan,1994).
Nock andKingston(1988) suggestedfourpossible reasonsfor the failureto find strongnegative effects of maternalemploymenton children.
Employed mothers might substitutequality for
quantityof time with children;fatherswith employed wives might spend more time with children and therebynarrowdifferencesin parental
time between dual-earnerfamilies and motherhomemakerfamilies;relativelysmall amountsof
time may be spent with childrenin any type of
household, hence minimizing the possibility of
differencesby laborforce statusof mother;and,
finally, childrenmay not requiremuch parental
time for good outcomes.
Using the 1975 time-diarydata collected by
JusterandStaffordandthe follow-updatacollected in 1981, Nock and Kingston(1988) foundthat
much more time was spent with children in
households with a mother at home full-timeeven thoughrelativelylittle of thattime was spent
directlyinteractingwith children.That is, during
much of the time mothers spend with children,
they are primarilydoing othertasks, like housework.The care of childrenis an auxiliaryor secondaryactivity. Nonetheless,employed mothers
did not fully compensatefor quantitywith quality. Motherswho were not in the laborforce also
spentmoretime in qualityactivities,suchas playing with and readingto children,althoughdifferences were concentrated among mothers of
preschool rather than school-age children. In
dual-earnerfamilies, fatherscompensatedsomewhatfor less maternaltime but not sufficientlyto
erase differences between two-earnerand oneearnerfamilies. With the data availableto them,
Nock and Kingstonwere not able to addressthe
most interesting, and perhaps most important,
questionaboutmaternalemployment:How much
parentaltime do childrenneed? In general,questions aboutneed are difficultto researchand remainlargelyunanswered(Presser,1995).However, determininghow childrenspendtheirtime and
how this varies acrossfamiliesis a reasonable,if
only a first, step in assessingthe questionof how
childrenshouldbe spendingtheirtime.
A majortheoreticalconcernin the social capital literatureis whetherthe increasein maternal
employmentdepriveschildrenof the supervision
they need and limits connections between the
family and the larger community. Using timediary data collected on various samples of rural
and urbanwomen in the 1920s throughthe late

1970s, Bryantand Zick (1992) reachedthe conclusion that the reduced time for children concomitantwith the rise in marriedmothers'participationin the laborforce may be overstated.Larger families and the relatively heavy burden of
domestic and unpaidagriculturalwork that was
requiredof mothersin earlierdecadeslimitedthe
time mothers had to spend with children. Although the direct time spent with a given child
probablyhas decreasedsomewhatfor very young
children,BryantandZick estimatedthatmaternal
time actually has increasedfor school-age childrenover the courseof this century.
Robinson (1993) found that mothers in his
1985 national time-diary survey spent just as
much time with children,on average,as mothers
in a 1965 nationalsurvey.Moreover,if the higher
maternalemployment levels of 1985 are taken
into account,it couldbe arguedthatmothersgave
childrenslightly more of their available time at
the laterpoint.
Recent evidence by Bryant and Zick (1996)
suggests that the activities parentsdo with children changes with a mother'semployment.Maternalemploymentmay actuallyincreaseleisure
time and houseworktime thatis sharedwith children. On the other hand, it also decreases time
spent in direct family care and supervision of
children.
Single Parenting

The researchon single parenting,as with maternal employment, often has been beset with
methodologicalproblems,particularlythe failure
to controlfor socioeconomicdifferences.However, recent research,which does adequatelymeasure differencesin family income between oneandtwo-parentfamilies,suggeststhatonly half of
the poorer outcomes of children who grow up
with one parentcan be attributedto economicfactors(McLanahan& Sandefur,1994).
McLanahanand Sandefurarguethat much of
the rest of the differencecan be explainedby differences in parentalinvolvement.Maintaininga
single-parentfamilycurtailsthe time thatmothers
(andfathers)areable or willing to invest in monitoring children's activities, supervising homework, and developingrelationshipswith teachers
and parentsof children'sfriends.This argument
suggests there are importantdifferencesin children's activity patternsin one- and two-parent
families.However,althoughTimmeret al. (1985)
found that childrenof single parentsin the 1981
time-diarystudy watchedmore TV on weekends

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:09:26 PM


All use subject to JSTOR Terms and Conditions

336

Journal of Marriage and the Family

and routinely got less sleep, other differences


were minimal.
Robinson (1993) found that, other factors
equal, single mothersspent about30% less time
with their children, on average, than married
mothersin 1985. Not only did childrenin singleparentfamilies miss out on time spent with their
father,buttime with theirmotherwas also lower.

encourageand providethe opportunityfor more


time spent doing housework.And it remains at
least a plausible hypothesis that the greaterthe
numberof children in the household, the more
difficult it will be for parentsto devote time to
reading to children or effectively monitoring
homework.Using the statewideCaliforniatimediarydatadescribedin the next section,we investigateeach of these expectations.

Family Size and Birth Order

Numberof siblingshas consistentlybeen foundto


have a small,negativecorrelationwith educational andoccupationalattainmentin adulthoodin the
statusattainmentliterature(Mare,1995). Using a
number of data sets, Blake (1989a, 1989b)
showed that childrenfrom small families do better on tests of verbal and math ability than childrenfromlargerfamilies.
Social psychologists,most notablyZajoncand
Markus(1975; Zajonc, 1976), found birthposition to be inversely relatedto achievement.The
argumentis that additionalchildrenlower the intellectualstimulationin the family. Only children
and first-born children interact primarily with
adults and, hence, the cognitive stimulation is
higherthan when interactionsinclude more children.A consistenteffect of birthorderis not well
establishedin studies that adequatelycontrolfor
socioeconomic status, however. For example,
Steelmanand Mercy (1980) showed that, net of
socioeconomicstatus,the effect of the numberof
siblings on intelligenceheld in the presenceof a
control for birth order but not vice versa. In a
study of tests of vocabulary and block design
ability of children6-11 years old, the numberof
youngersiblings had a consistentnegativeeffect
on test scores, and the numberof older siblings
tended to reduce verbal but not nonverbal
achievement(Mercy& Steelman,1982).
In sum, past researchon family composition,
children'sandparents'use of time, andchildren's
cognitive achievementsuggests that childrenof
college-educated parents can be expected to
spend more time reading and studying and less
time watching TV than children of less welleducatedparents.Maternalemploymentand single parentingalso may detractfrom time spent
reading and studying and may encourage more
TV watching,althoughhere the supportfor such
an expectationis farmorelimited.
Additionally,because of time constraints,we
might expect childrenin families with either an
employedmotheror a single parentto spendmore
time on housework.Largerfamilies might also

DATAANDMETHODS

In orderto providebasic informationon the daily


activitiesof youngchildrenin California,the Survey ResearchCenterat the Universityof California, Berkeley,conducteda surveyof the daily activitiesandlocationsof a representativesampleof
childrenaged 11 yearsor youngerfor the California Air ResourcesBoardbetweenApril 1989 and
March1990. The principalobjectivesof the survey were to determine the proportion of time
spent in specific indoor and outdoor locations
(e.g., living room, workplace)and in specific activities (e.g., active sports, hobbies) by California's childrenin generalandby theirdemographic andsocioeconomicsubgroups.
The CaliforniaChildren'sActivityPatternSurvey implementeda methodologyof collectingtimediary informationin the context of probabilitybased surveysthat had been employedin several
earlierrepresentativesurveys. (See Wiley et al.,
1991, for backgroundand furtherdetails.)Timediary estimatesfrom these studies thus far have
producedratherreliable results at the aggregate
level (Robinson, 1977, 1985). However, almost
all of the methodologicaland substantivedataon
activity patterns have come from samples of
adults.
Studiesof childrenraise specialissues, particularly because of the unknownabilities of children to reporton their own behavior.Moreover,
because children'sactivities tend to be different
from those of the larger adult population, new
data collection proceduresand coding categories
neededto be developedfor them.To help identify
potentialsourcesof reportingbias and to develop
efficientprobingtechniquesto improveparentrecall, a small pilot study was conductedwith 38
households containing children aged 11 or
younger.The resultsof the pilot study suggested
that interviewerattentivenessand probing were
the most effective ways to enhancethe qualityof
the diaryinformation.Interviewerswere instructed to probe any activity that lasted more than 2
hours to determinewhether the child had done

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:09:26 PM


All use subject to JSTOR Terms and Conditions

337

Children's Use of Time


anything else during the period. The results of the
pilot study also led to the decision to obtain diary
information directly from children aged 9-11 and
to encourage parents to consult with children
aged 6-8 about how they had spent their time
during the diary day, i.e., the day preceding the
interview. The pilot study did not indicate that the
quality of the information was enhanced when respondents were contacted ahead of time and told
that the researchers would call back the next day
to obtain information on time use.
The target population for the California Children's Activity Survey consisted of children
younger than 12 years old and living in Englishspeaking households with a telephone. The full
sample was drawn using random-digit-dialing
methods of Waksberg (1978) and was stratified to
provide representative estimates for major regions
of the state. Interviews were distributed over the
four seasons (from April 1989 to March 1990) to
take into account seasonal variations in activity
and location patterns. The survey questionnaire
included a complete inventory of children's activities and locations on the diary day and a brief series of items pertaining to the characteristics of
the selected child, the household, and an adult in
the household who served as an informant about
the child's activities. The questionnaire, including
the diary component, was implemented using the
computer-assisted telephone interviewing technology developed as part of the university's Computer Assisted Survey Execution System. A total
of 1,200 eligible respondents were interviewed,
with an overall response rate of 78%.
The questionnaire instrument for the survey
was divided into three main parts: an adult section, a children's section, and a daily time diary.
When the selected child was 9-11 years old, he or
she was the preferred respondent for the children's and diary sections of the questionnaire.
This occurred in about 85% (269 of a total of
316) of the cases. When the randomly selected
child was 8 years old or younger, the preferred respondent for all sections of the questionnaire was
the parent or guardian who spent the most time
with the child on the diary day. Respondents for
children aged 6-8 were encouraged to consult
with the child when filling out the diary portion
of the questionnaire. Approximately 92% of the
adult respondents were the parent or guardian
who spent the most time with the selected child
on the diary day-most had spent at least 8 waking hours with the child on the diary day. (Less
than 1% of the interviews were conducted with
adults who had spent less than 1 waking hour

with the child on the diary day.) Nearly 98% of


the adult respondents were either the father,
mother, stepfather, or stepmother of the selected
child. In fact, over 80% of the adult respondents
were the children's mothers.
Concern about the validity of the diary responses arises quite naturally because of the nature of the task of reconstructing the full range of
activities and locations of the diary day and because the diary information was collected from
children or from proxy interviews with adults
who may not have had full information. In order
to make a subjective assessment of the validity of
diary responses, the interviewers were asked to
record their confidence in the diary responses
after the diary section had been completed. In approximately 80% of the cases, interviewers said
they had "complete confidence" in the diary responses. Eighteen percent of the diaries inspired a
"somewhat confident" rating. Less than 2% received a rating of "not too confident."
One issue of validity that is of particular concern in this study is that it is likely that mothers
with paid jobs will, on average, spend less time
directly observing their children's activities than
mothers at home full-time. Unfortunately, the
data do not contain a question asking who was
with the child when an activity occurred. This
must be kept in mind when interpreting results,
especially those for maternal employment. At the
same time, because older children reported for
themselves or assisted parents, the mother or father was the sole respondent for the diary only for
children under the age of 6.
Diary Coding Procedures
In contrast to most surveys that examine people's
activities in isolation from the natural temporal
context in which they are embedded (e.g., by asking people to compress their actual behavioral experiences by saying whether they "often" or "usually" do something), time-diary accounts report
activities as they naturally and sequentially occur
in daily life. Studies of time use provide the opportunity, then, to examine human activities in
real time-as individuals are actually involved in
the stream of daily behavior.
In the retrospective diary used in the California Children's Activity Survey, respondents reported each activity they engaged in on the diary
day, beginning at midnight of the preceding day.
They reported where they were at the time, if they
were inside, outside, or equally inside and outside, and whether or not a tobacco smoker was

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:09:26 PM


All use subject to JSTOR Terms and Conditions

338

Journal of Marriage and the Family

present. The interviewers recorded the time the


activity began and ended, entered a single-digit
code (one of eight precoded categories) indicating
the major category of the activity, and entered a
brief description (up to 44 characters in length) of
the activity. When the activity was related to
school or to child care, the interviewer entered an
additional code (one of seven precoded categories) that further specified the type of activity
and its location.
If the randomly selected child was younger
than 9 years old and if the diary contained unspecified "day care," "babysitting," or "school"
activities, interviewers asked the adult respondents to estimate the time the child spent in a variety of activities typical of such settings: eating
snacks or meals, napping or resting, watching TV,
doing arts and crafts, actually playing outdoors,
and going on field trips.
Considerable effort was invested in obtaining
a very detailed account of activities during the
day, including all the important changes that
occur during the day. Through prompts and probing, the interviewers attempted to ensure that each
respondent's report was complete and accurate. A
detailed scheme for coding activities was developed as part of an extensive coding operation for
the diary data after the interviews were completed. (See Wiley et al., 1991, Table 2.3 for the coding scheme and Table 4.15 for a comparison of
the California survey with the 1981 University of
Michigan national survey of children. In general,
the California data are similar to the 1981 data
but differ in predictable directions, e.g., children
spend more time in outdoor activities in California than they do in the rest of the nation.)
Comparison of Survey Estimates with
Population Data
A comparison between estimates based on the
California Children's Activity Pattern Survey and
CurrentPopulation Survey data for 1988 provides
a rough basis for evaluating the representativeness of the California Children's Survey sample
and suggests the possibility that groups of lower
socioeconomic status may be underrepresentedin
the California sample (Wiley et al., 1991, Table
4). Both median household income and mean
years of education for adult respondents are higher than the corresponding CurrentPopulation Survey figures. Although these differences may be
due, in part, to differences in the sampling frame
between the Children's Survey and the Current
Population Survey, they are consistent with pat-

terns of underrepresentationof groups with lower


income and less education that have been observed in many telephone and personal interview
surveys of adult respondents (Goyder, 1987). This
should be kept in mind when attempting to generalize from the Children's Survey estimates.
Compared with statewide public school enrollment figures for children in grades kindergarten
through six, the grade distribution for the Children's Survey is not significantly different from
the enrollment figures, but the distribution of children by race and ethnicity is significantly different from the enrollment data (Wiley et al., 1991,
Table 5). This suggests some underrepresentation
of non-White and Hispanic children in the California Children's Survey sample, due, in part, to
the exclusion of households that did not speak
English. In addition, because race and ethnicity
are correlated with socioeconomic status, underrepresentation of certain racial and ethnic groups
also may be due to selection bias with respect to
socioeconomic factors.
Despite its restriction to English-speaking California households with children and its possible
underrepresentationof households with lower socioeconomic status, the 1989-1990 California
Children's Time Use Study stands as the most
comprehensive recent addition to the study of
children's activity patterns. In the next section,
we report descriptive findings on the amount of
time children in the study spent reading or being
read to, watching TV, studying, and doing housework during the diary day. Given the nature of the
activities we examine, we restrict the sample to
the 887 children who were ages 3-11. That is, although we have diaries on infants and toddlers,
we exclude children under the age of 3 from the
analysis.
We use tobit regression models to assess
whether family background, particularly parental
education, maternal employment, single parenting, and family size are predictive of the number
of minutes a child engages in these activities. A
tobit estimator is used because it corrects for the
censored distribution of children's time in each
activity. That is, a sizable number of children
spend no minutes in a given activity on the diary
day, and the tobit estimation takes into account
this censoring at zero minutes.
FINDINGS
Implicit in the literature on family background
and child outcomes is the notion that children experience a more varied, cognitively stimulating

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:09:26 PM


All use subject to JSTOR Terms and Conditions

339

Children's Use of Time


childhood in some family environments than in
others. The child outcomes that result are, therefore, better in some situations than in others. The
challenge is to determine which activities and investments matter most and what environments
foster achievement. We cannot, with the data
here, address the question of what activities are
best for children, but we can add to what is
known about how children spend their time and
how this varies by family composition.
In selecting the activities to focus on, we are
drawn to activities that seem to generate strong
sentiments on the part of parents, teachers, or researchers about how children should spend time.
For example, we argue that many teachers (and
parents) believe that the more children are read to
in their preschool years, the more they read themselves as they get older, the better they do in
school, the better they perform on standardized
tests, and the more likely they are to attain higher
levels of education. The presumption is also that
good study habits improve academic performance.
Too much TV viewing is often considered bad or
at least of questionable value for children. Housework is more neutral, but at least one scholarly
work (Goldscheider & Waite, 1991) argues that
doing housework may be good for children because it prepares them for running their own
households later in life and teaches responsibility.
However, other research, more focused on paid
employment than on household chores, suggests
that work (of any kind) may detract from time
children devote to study or to developing cognitive skills (Zill, Nord, & Loomis, 1995).
Table 1 shows the percentage of children
(aged 3-11) in the 1989-1990 California time-use
study who spent some amount of time during the
diary day in each of four activities: reading or
being read to, watching TV, doing housework,
and studying. In addition, for those children who
engaged in a particular activity, the average number of minutes spent doing that activity is shown.
About one quarter of the sampled children
spent some time reading or being read to, whereas
almost nine in ten watched TV. Those who engaged in reading spent about 3/4 of an hour per

day in this activity. TV watchers spent an average


of 23/4hours in front of the TV, 3 hours when we
include watching TV and doing some other activity such as homework (data not shown). Only
22% spent any time studying, and those who reported studying spent an hour in this activity. It
should be noted that studying is not an activity in
which many preschool-aged children engage. Because the sample combines preschoolers and
school-aged children, this lowers the percentage
who report studying. In addition, interviews are
spread over the calendar year, with some children
on summer vacation when the interview occurred,
and this also lowers the percentage studying. The
regression models include controls for summer
and weekend interviews in order to improve the
modeling of time spent studying.
About 40% of children reported doing some
housework. Children who did this activity during
the diary day averaged close to an hour in household chores. These estimates seem relatively high
and may reflect the fact that time spent doing
household chores is reported by children themselves at ages 9-11, rather than by parents.
Among younger children, housework time may
include time helping an adult who is engaged in
housework such as dinner preparation.
In Table 2, we show means (percentages for
categorical measures) and standard deviations for
variables used in the multivariate analysis of children's time use. Control variables include
whether a child is male, a member of a minority
group, the child's age, whether the diary day was
a weekend day, and whether the interview took
place in the summer.
A family's socioeconomic status is measured
by two variables: the educational attainment of
the parent and family income. Parental education
is the education of the adult identified as knowing
most about the child's activities, usually the
mother of the child. About one quarterof the children live with a parent who has a college education (or more), and an additional 29% have a parent who attended college but did not graduate.
One third of the children have a parent who is a
high school graduate with no further schooling,

TABLE 1. TIME THAT CHILDREN SPEND READING, WATCHING TV, STUDYING, AND DOING HOUSEHOLD CHORES

Percentagewho did the activity on the previous day


Average minutesper day in activity for those who did the activity
Average minutesper day for entire sample

Reading

WatchingTV

Studying

Doing
Housework

27%
45
12

89%
168
150

22%
63
14

40%
54
22

Note: n = 887 Californiachildren,aged 3-11.

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:09:26 PM


All use subject to JSTOR Terms and Conditions

340

Journal of Marriage and the Family

TABLE 2. MEANS AND STANDARD DEVIATIONS FOR


VARIABLES PREDICTING CHILDREN'S TIME USE

Mean/
Percentage
Controlvariables
Child is male
Child is minority
Child's age (in years)
Weekend diary day
Summerinterview
Parentaleducation
Less thanhigh school
High school graduate
Some college
College graduate
(Postgraduate)
1988 family income (in $10,000)
One-parentfamily
Mother's labor force status
Employed full-time
Employed part-time
Student
Other-looking, laid off, retired
(Keeping house)
Numberof children

51%
30%
7.09
37%
30%
100%
11%
33%
29%
12%
15%
3.44
20%
100%
42%
19%
5%
7%
27%
2.16
100%
Sibling position
28%
Only child
Oldest child-1 (or more) siblings
21%
(Youngerchild-1 (or more) siblings) 51%

SD
50%
46%
2.70
48%
46%
32%
47%
46%
33%
35%
8.54
40%
49%
39%
22%
25%
44%
1.04
45%
40%
50%

Note: Omittedcategories in multivariateanalysis are in


parentheses;n = 887.

and about 11% have a parent who dropped out


before completing high school.
Family type is indicated by whether a child
lives with only one parent. Almost 20% of the
sample lives in one-parent families, almost always a family headed by the mother. The maternal employment variable indicates whether a
child's mother is employed full- or part-time, is a
student, is unemployed, or is a full-time homemaker. About one quarterof the sample lives with
a mother at home full-time, whereas 42% have a
mother who holds a full-time job. "Maternal employment" actually measures the employment status of the father for the 16 (out of 887) children in
the sample who do not live with a mother (or
stepmother) but reside only with their father. The
average number of children in the sample households is just over two children. About 28% of the
children for whom the diary is collected are the
only child in their household. The remaining 72%
of the children in the sample live in households
with two or more children. About 21% are the
oldest child in the household, and 51% have at
least one older sibling.

Table 3 displays estimates from the tobit equations predicting time spent reading, watching TV,
studying, and doing housework. Before turning to
the family composition indicators, a comment on
the control variables is in order. The control for
whether the diary charts activities for a weekend
or a weekday shows that children watch more
TV, study less, and do more housework on the
weekends than during the week. Not surprisingly,
the estimate of time spent studying is greatly reduced in summer interviews.
Three demographic characteristics of the child
are included in the model, largely as control variables: the child's age, minority status, and gender.
The most important is age. Not surprisingly, as
age increases, the likelihood of studying, doing
household chores, and TV viewing increases.
Among 3- to 11-year-olds, the likelihood of reading (or being read to) does not differ significantly
by age. There are no major differences between
minority and nonminority children, except that
White, non-Hispanic children (the omitted category in the regression) spend slightly more time
doing household chores. Finally, there are two
gender differences of note. Boys may spend
slightly more time in front of the TV, and boys
spend significantly less time doing household
chores than girls. This finding of a significant
gender difference in housework among young
children parallels the findings for teenagers (and
young adults living at home) reported by Goldscheider and Waite (1991).
Our main research questions revolve around
whether family environments-particularly the
level of parental education, maternal employment, single parenting, and family size-influence the likelihood of a child participating in various activities. Our multivariate findings suggest
the following.
Parental Education and Income
Children's time spent reading or being read to is
significantly higher in households in which the
parents are college educated. Minutes spent reading per day do not differ between children who
live with a college graduate and children who live
with a parent who has attended graduate school or
completed postgraduate education (the omitted
category). However, children of parents who have
attended (but not completed) college, who are
high school graduates, or who did not complete
high school read fewer minutes per day than children of college graduates. In addition, the higher

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:09:26 PM


All use subject to JSTOR Terms and Conditions

341

Children's Use of Time


TABLE 3. PARAMETER ESTIMATES OF TOBIT EQUATIONS PREDICTING MINUTES PER DAY SPENT READING,
WATCHING TV, STUDYING, AND DOING HOUSEWORK

Reading
Intercept
Controlvariables
Child is male
Child is minority
Child's age
Weekend diaryday
Summerinterview
Parentaleducation
Less thanhigh school
High school graduate
Some college
College graduate
Family income (in $10,000)
One-parentfamily
Mother's labor force status
Employed full-time
Employedpart-time
Student
Other-looking, laid off, retired
Numberof children
Sibling position
Only child
Oldest child-i (or more) siblings

-14.86

Studying

Doing
Housework

8.66

-104.28***

-94.65***

-9.27
-7.62
-1.05
-13.58*
4.01

18.11*
16.66
10.26***
41.46***
2.00

-11.84
7.56
17.77***
-96.28***
-86.50***

-16.37**
-15.43*
7.38***
17.01**
5.94

-30.98**
-37.27***
-34.87***
-12.17

82.29***
42.76***
28.94**
0.35
-0.85
0.47

-6.47
-25.78**
-34.03***
-17.47

-21.29
-8.62
-12.24
-19.98
-0.20
-14.12

0.98***
4.64
-5.65
12.75
11.08
-1.41
1.28
9.47
1.45
75.28**
-1619

Sigma
Log likelihood

WatchingTV

-0.60
0.71
-16.02
17.27
0.77
-29.80
3.35

1.11
-30.62**
-40.62*
0.27
5.11
5.39
2.22
132.27***
-5042

11.94
11.75
85.63***
-1486

5.42
-3.84
11.68
-23.38
13.67***
14.25
-11.59
92.1***
-2467

Note: n = 887 Californiachildren,aged 3-11.


*p<.10. **p<.05. ***p<.01.

the family income, the more time children spend


reading-perhaps reflecting the ability or willingness of families with higher incomes to purchase
reading materials.
The coefficients for TV viewing suggest an inverse relationship between TV viewing and
parental education. Significantly more time is
spent watching TV by children whose parents are
not college graduates. Children whose parents attended but didn't complete college (or who only
competed high school) also study less than children of highly educated parents. The results for
studying are not as consistent as for reading or
TV viewing, however, because the coefficient for
children whose parent is a high school dropout is
not statistically different from that for children
whose parent obtained a postgraduate education.
Finally, parental education and income are not
predictive of the amount of time children spend
doing household chores. The lack of statistical
significance suggests little association between
parental education and the amount of housework
done by children.

Marital Status of Parent


Do children's activities in two-parent homes differ significantly from those of children who live
with only one parent? We find no indication that
a child who lives with only one parent reads less,
watches TV more, or studies less than a child who
resides with two parents. Nor do we find differences in the amount of time spent doing housework between children in the two types of households, even though we expected there might be
greater need for children's help in families with
only one parent.
Maternal Employment
We noted at the outset that there continues to be
concern about whether children receive the supervision they require when their mothers are employed. Our investigation suggests little difference in the time use of children by the employment status of their mother. The one significant
finding is that children of a mother employed
part-time watch less TV than children with a

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:09:26 PM


All use subject to JSTOR Terms and Conditions

342

Journal of Marriage and the Family

mother who is a full-time homemaker. It appears


that latch-key children do not take advantage of
parental absence to increase their TV viewing.
Perhaps more surprisingly, there were no differences in the likelihood of children doing
household chores by mother's labor force status.
We expected mothers employed full-time might
require more help with household chores, but
there was no evidence of this in the data.
It is possible that reporting on children's activities is less accurate for employed mothers, who
may not spend as much time with their children
as mothers who are not employed. For example,
an alternative interpretation of the finding of no
difference in TV viewing on the part of children
with mothers employed full-time and those whose
mothers are homemakers is that children of mothers employed full-time watch more TV than other
children, but employed mothers do not realize this
and underestimate the time their children spend in
this activity. Mothers who are at home full-time
may more accurately report how long their children are engaged in this activity. Although we
cannot rule out such an interpretation of the findings, the use of older children's self-reports of activity (regardless of maternal employment status)
helps mitigate against potential bias. Also the
finding of significant differences between mothers employed part-time (who, it might be conjectured, know more about their children's activities
than mothers who work full-time do) and homemakers makes it harder to dismiss findings for
maternal employment as merely an artifact of
data quality.
Family Size
Although we expected that family size might detract from study time and the likelihood of reading or being read to, we find no evidence of this.
The one activity that seems to occupy more of a
child's time in a large family is housework. As
the number of children in the household increases, the average number of minutes a child spends
in housework also increases significantly. Birth
order (i.e., being an only child or a first-born
child) is not predictive of differences in the
amount of time spent in any of the activities we
examine.
SUMMARY AND CONCLUSION

Most research on the interrelationship of family


environments and child outcomes assumes that

the opportunities and activities of childhood differ


across family settings. Although the most accurate way to assess how children spend time is to
collect time-diary data from children or their parents, there has been only limited information
about children's activities gathered in this format.
We examine recently collected time-diary information from a relatively large statewide sample
of preschool and elementary school-aged children
to explore whether children's activity patterns
vary in predictable ways by family characteristics. These diary figures provide striking documentation of the dominance of TV watching as an
activity among preschool and elementary schoolaged children. Almost 90% of children watch TV
in any given day, whereas only about 25% read a
book or have someone read to them. The amount
of time spent viewing TV far exceeds the amount
of time spent reading. Although it is possible that
a sizable share of TV watching among preschoolers is of educational programs like Sesame Street,
the extensive amount of viewing time reinforces
concerns about TV's influence on children's
physical fitness, cognitive functioning, and aggressive behavior.
Of the four family characteristics we examine,
the largest differences are found for parental educational attainment. College-educated parents appear to limit their children's time in front of the
TV more than less-educated parents do. Highly
educated parents also appear to encourage their
children to read and study more than do less educated parents.
Recent empirical and theoretical work has
raised new questions about the negative effect of
single parenting and maternal employment on
children, but our findings suggest rather minimal
differences on our four time-use outcomes (time
reading, studying, watching TV, and doing housework) between children in two-parent families
and those who live only with their mother. Furthermore, our findings on maternal employment
do not support the popular hypothesis that young
children fare best when they have a mother at
home full-time. Children of mothers who are employed part-time (and mothers who are students)
watch less TV than children with mothers who
are at home full-time. Our findings suggest that it
may be mothers who achieve a balance between
working outside the home and spending time with
their children who are most successful at steering
their children toward productive use of their time.
Finally, we find relatively little difference in
children's activities by family size. In larger fami-

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:09:26 PM


All use subject to JSTOR Terms and Conditions

Children's Use of Time


lies, children do more housework, perhaps because there is more housework to do. However,
oldest (or only) children's time in activities such
as reading or studying or housework do not appear to be much different than that of younger
siblings in households.
Although the diary data used in this research
do not directly assess the relative value of various
activities, they do provide estimates of how much
time children are spending in activities that may
be related to cognitive and behavioral outcomes.
With diary data, we begin to illuminate the connections between parental inputs and child outcomes. Well-educated parents appear to structure
their children's time differently than less educated
parents, and this may provide the children of
better-educated parents with the social capital
necessary for cognitive enhancement. At the same
time, our results do not support the assumption
that full-time mothering is critical to good child
outcomes. And until we can show systematic differences between how children (and parents)
spend time in single-parent families and how they
spend time in two-parent families, caution is warranted in attributing negative outcomes for children in single-parent families to a dearth of social
capital or inadequate supervision.
In his seminal work on social capital, Coleman
argued that a child's access to adults in the family
(as well as the attention given by adults to children) was crucial to their acquisition of human
capital. In this analysis, we have related differences
in children's time use to several factors that Coleman argued were key to the development of social
capital in the family-the number of parents present in the household, materal employment, family size, and parental education. Although the diary
measures provide a provocative first look at family
variation in children's activities, to be true to Coleman's notion of social capital, we must move beyond measures of simple access and the availability of parents to measures that also capture the degree to which parents are involved with children.
For example, how much of a child's time in various activities is time that is shared with other family members? How much affect and positive reinforcement is transmitted? How much and what
types of time that children allocate to reading or
television viewing actually affect their cognitive
development? California representsmore than 10%
of the country, but the present results need to be
replicated and extended at the national level to assess more fully the role that families play in determining positive child outcomes.

343
NOTE
We wouldlike to acknowledgethe helpfulcommentsof
Melissa Milkie, HarrietPresser,and three anonymous
reviewers.We also benefitedfrom the researchassistance of LekhaSubaiyaand RongjunSun, thanksto a
training grant from the William and Flora Hewlett
Foundationto the Centeron Population,Gender,and
Social Inequality,Departmentof Sociology, University
of Maryland.
REFERENCES
Belsky, J., & Eggebeen,D. (1991). Maternalemployment and young children'ssocioemotionaldevelopment. Journal of Marriage and the Family, 53, 1083-

1098.

Beschen, D. A. (1972). Nationwide personal transportation study: Transportation characteristics of school

children.WashingtonDC: U.S. Departmentof Transportation,FederalHighwayAdministration.


Bianchi, S. M. (1990). America's children: Mixed
prospects. Population Bulletin, 45, 1-39.
Blake, J. (1989a). Family size and achievement. Berke-

ley: Universityof CaliforniaPress.


Blake, J. (1989b). Numberof siblings and educational
attainment.Science, 245, 32-35.

Blau, F. D., & Grossberg,A. J. (1992). Maternallabor


supplyandchildren'scognitivedevelopment.Review
of Economics and Statistics, 74, 474-481.

U., & Crouter,A. C. (1982). Workand


Bronfenbrenner,
familythroughtime and space.In S. B. Kammerman
& C. D. Hayes (Eds.), Families that work: Children
in a changing world (pp. 39-83). Washington, DC:

NationalAcademyPress.
Bryant,W. K., & Zick, C. D. (1992). Are we investing
less in the next generation? Historical trends in time
spent caring for children. Consumer Economics and

HousingResearchPapers(92-9). Ithaca,NY: Cornell


University.
Bryant,W. K., & Zick, C. D. (1996). An examinationof
parent-child shared time. Journal of Marriage and the
Family, 58, 227-238.

Coleman,J. S. (1988). Social capitalin the creationof


human capital. American Journal of Sociology, 94,

S95-S120.
of soColeman,J. S. (1992). The rationalreconstruction
ciety. American Sociological Review, 58, 1-15.

Dawson,D. A. (1991). Familystructureand children's


health and well-being. Journal of Marriage and the
Family, 50, 619-648.

Desai, S. P., Chase-Lansdale,R. L., & Michael,R. T.


(1989). Motheror market?Effects of maternalemploymenton the intellectualabilityof 4-year-oldchildren. Demography, 26, 545-562.

Duncan, O. D., Featherman, D. L., & Duncan, B.


(1972). Socioeconomic background and achievement.

New York:AcademicPress.
Goldscheider,F. K., & Waite,L. J. (1991). Newfamilies, no families? The transformation of the American

home.Berkeley:Universityof CaliforniaPress.

Goyder, J. (1987). The silent minority: Nonrespondents

on samplesurveys.Cambridge,England:PolityPress.
Haveman,R., & Wolfe, B. (1993). Children'sprospects
and children's policy. Journal of Economic Perspectives, 7, 153-174.

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:09:26 PM


All use subject to JSTOR Terms and Conditions

Journal of Marriage and the Family

344
Hernandez, D. (1993). America's children: Resources
from family, the government, and the economy. New

York:RussellSage.
Heyns, B., & Catsambis,S. (1986). Mother'semploymentandchildren'sachievement:A critique.Sociology of Education, 59, 140-151.

Hill, C. R., & Stafford,F. P. (1974). Allocationof time


to preschool childrenand educationalopportunity.
Journal of Human Resources, 9, 323-341.

Hill, C. R., & Stafford,F. P. (1985). Parentalcare of


children: Time diary estimates of quantity, predictability, and variety. In F. T. Juster & F. P.
Stafford (Eds), Time, goods, and well-being (pp. 415-

437). AnnArbor,MI:Institutefor SocialResearch.


Juster, F. T., & Stafford, F. P. (Eds). (1985). Time,
goods, and well-being. Ann Arbor, MI: Institute for

SocialResearch.
Krein,S. F., & Beller,A. A. (1988). Educationalattainment of childrenfrom single-parentfamilies:Differences by exposure,gender,and race. Demography,
25, 221-234.
Leibowitz, A. (1974). Home investmentsin children.
Journal of Political Economy, 82, 111-131.

Leibowitz, A. (1977). Parentalinputs and children's


achievement. Journal of Human Resources, 12, 247-

267.
Lindert,P. (1977). Sibling position and achievement.

Presser,H. B. (1995). Are the interestsof womeninherently at oddswiththe interestsof childrenor the family? A viewpoint.In K. O. Mason & A. M. Jensen
(Eds.), Gender and family change in industrialized

countries(pp. 297-319). London:OxfordUniversity


Press.
Robinson, J. P. (1977). How Americans use time: A social psychological analysis of everyday behavior.

New York:Praeger.
Robinson,J. P. (1985). The validityandreliabilityof diaries versus alternativetime use measures.In F. T.
Juster & F. P. Stafford (Eds.), Time, goods, and well-

being(pp.33-62). AnnArbor,MI:Institutefor Social


Research.

Robinson, J. P. (1993). The demographics of time. Itha-

ca, NY: AmericanDemographics.


Sewell, W. H., & Hauser,R. M. (1980). The Wisconsin
longitudinalstudyof social andpsychologicalfactors
in aspirationsandachievements.Researchin Sociology of Education and Socialization, 1, 59-99.

Stafford,F. P. (1987). Women'swork,siblingcompetition, and children's school performance.American


Economic Review, 77, 972-980.

Steelman,L. C., & Mercy,J. A. (1980). Unconfounding


the confluencemodel:A test of sibshipsize andbirthorder effects on intelligence. American Sociological
Review, 45, 571-582.

Journal of Human Resources, 12, 198-219.

Timmer,S. G., Eccles, J., & O'Brien,K. (1985). How


childrenuse time. In F. T. Juster& F. P. Stafford

union: America in the 1990s. Vol. 1. Economic trends

AnnArbor,MI:Institutefor SocialResearch.
Waksberg,J. (1978). Sampling methods for random-

Mare,R. D. (1995). Changesin educationalattainment


andschoolenrollment.In R. Farley(Ed.),Stateof the
(pp. 155-213).New York:RussellSage.
McLanahan,S., & Sandefur,G. (1994). Growing up
with a single parent. Cambridge, MA: Harvard Uni-

versityPress.
Medrich,E. A., Roizen, J., Rubin,V., & Buckley, S.

(1982). The serious business of growing up: A study


of children's lives outside school. Berkeley: Universi-

ty of CaliforniaPress.
Mercy,J. A., & Steelman,L. C. (1982). Familialinfluence on the intellectualattainmentof children.American Sociological Review, 47, 532-542.

Milne,A. M., Myers,D. E., Rosenthal,A., & Ginsburg,


A. (1986). Single parents,workingmothers,and the
educationalachievementof school children.Sociology of Education, 59, 125-139.

Myers,D. E., Milne,A. M., Baker,K., & Ginsburg,A.


(1987). Studentdiscipline and high school performance. Sociology of Education, 60, 18-33.

Nock, S. L., & Kingston,P. W. (1988). Time with children: The impact of couples' work-time commitments. Social Forces, 67, 59-85.

Parcel,T. L., & Menaghan,E. G. (1994). Parents'jobs


and children's lives. New York: Aldine de Gruyter.

(Eds.), Time, goods, and well-being (pp. 353-382).

digit dialing. Journal of the American Statistical Association, 73, 40-46.


Wallace, L. (1987). The total exposure assessment
methodology (TEAM) study. Summary and analysis,

(Vol. 1). WashingtonDC: EnvironmentalProtection


Agency.
Wiley, J. A., Robinson,J. P., Cheng,Y. T., Piazza,T.,
Stork,L., & Pladsen,R. (1991). Studyof children's
activitypatterns(Final Report,ContractNo. A733149). Sacramento:CaliforniaEnvironmentalProtectionAgency,Air ResourcesBoard.
Zajonc,R. B. (1976). Familyconfigurationand intelligence. Science, 192, 227-236.

Zajonc.R. B., & Markus,G. B. (1975). Birthorderand


intellectual development. Psychological Review, 82,

74-88.
Zill, N., Nord,C. W., & Loomis,L. S. (1995). Adolescent time use, risky behavior, and outcomes: An analysis of national data (Report prepared for the Office

of HumanServicesPolicy,U.S. Departmentof Health


andHumanServices).WashingtonDC:Westat.

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:09:26 PM


All use subject to JSTOR Terms and Conditions

Das könnte Ihnen auch gefallen