Content-Type: text/html

This paper was presented at the Association for Education in Journalism and
Mass Communication in San Antonio, Texas August 2005.
         If you have questions about this paper, please contact the author
directly. If you have questions about the archives, email
rakyat [ at ] eparker.org. For an explanation of the subject line, 
send email to
<a href="/cgi-bin/wa?LOGON=A3%3Dind0602A%26L%3DAEJMC%26E%3D8bit%26P%3D7789819%26B%3D--%26T%3Dtext%252Fplain%3B%2520charset%3Diso-8859-1" >[log in to unmask]</a> with just the four words, "get help info aejmc," in the
body (drop the "").

(Feb 2006)
Thank you.
Elliott Parker
====================================================================

The Civic Parent
Awakening the Civic Parent:
The School and Family in Political Socialization



Michael McDevitt,
University of Colorado at Boulder

Spiro Kiousis,
University of Florida



April 1, 2005



Submitted to the Mass Communication & Society Division,
Association for Education in Journalism and Mass Communication



Running head: The Civic Parent

Contact information:

Michael McDevitt
School of Journalism and Mass Communication
1511 University Avenue
478 UCB
Boulder, Colorado 80309-0478

303-735-0460
<a href="/cgi-bin/wa?LOGON=A3%3Dind0602A%26L%3DAEJMC%26E%3D8bit%26P%3D7789819%26B%3D--%26T%3Dtext%252Fplain%3B%2520charset%3Diso-8859-1" >[log in to unmask]</a>
fax: 303-492-0969

This research is supported by the Center for Information & Research 
on Civic Learning
& Engagement and by the John S. and James L. Knight Foundation.


Awakening the Civic Parent:
The School and Family in Political Socialization

Abstract

This paper explores whether schools-through the prompting of 
student-parent conversation-can awaken the civic parent of an adult, 
a role identity that might otherwise remain dormant. Results validate 
a theoretical model in which a school intervention engenders 
political involvement directly, but also indirectly through the 
long-term cultivation of civic parenting. Results are derived from a 
field study of Kids Voting USA as taught to high school students and 
parents in Arizona, Colorado, and Florida.


Awakening the Civic Parent:
The School and Family in Political Socialization

Proponents of the school as a site for civic development are fond of 
noting that the public school system in the United States was 
established to promote active citizenship (Center for Information & 
Research on Civic Learning & Engagement, 2003). Parents, by contrast, 
were never designed for this purpose. The study of political 
socialization is in many respects an attempt to understand how 
non-political institutions-most notably the family-act on behalf of 
democracy to prepare neophyte citizens. Consequently, the bulk of 
theorizing on family processes does not assume that parent influence 
is deliberate in the sense of parental duty consistently and 
consciously enacted over many years. In response to Hyman's seminal 
book, Political Socialization (1959), scholars applied various 
psychological theories to explain children's attitudes toward 
authority and diffuse support for a political regime (e.g., 
Greenstein, 1965). Freudian and neo-Freudian approaches gave way to 
cognitive modeling, social learning, and reciprocal models of 
parent-child influence (Dudley & Gitelson, 2002; McDevitt & Chaffee, 
2002), but in most cases there was no need for scholars to claim that 
the adoption of political attitudes and perceptions required the 
sustained and purposeful effort of parents.1
Likewise, political behavior research often views active citizenship 
as an indirect consequence of childrearing practices, the social 
standing of parents, and household resources such as newspaper 
subscriptions and online access. In the standard approach to 
predicting adult participation, parental legacy is important, but 
mostly because of socioeconomic status (SES). High-SES youth are more 
likely to attend college and to acquire jobs that provide them with 
resources of time and skills, which are transferable to politics 
(Brady, Verba, & Schlozman, 1995). In studies of developmental 
psychology, child-rearing practices predict learning outcomes in 
social studies and other courses, and parent support for schools is 
associated with student achievement (e.g., Dornbusch, Ritter, 
Leiderman, Roberts, & Fraleigh, 1987), but these benefits are usually 
not traced back to parents' specific attempts at civic instruction.
Parenting in these perspectives is functional for family system 
dynamics that have a logic all their own apart from any implications 
for political socialization. Left mostly unexplored is the 
identification of factors that might encourage adults to consider 
civic nurturing as a parental responsibility. The question arises as 
to whether civic parenting as a behavioral construct is responsive to 
situational influence. If so, what theoretical framework can be used 
to explicate the causes and consequences of childrearing directed at 
civic growth?
	Here we will explore whether schools-through the prompting of 
student-parent conversation-can awaken the civic parent of an adult, 
a role identity that might otherwise remain dormant. We will test the 
validity of a conceptual model in which a school intervention 
engenders political involvement directly, but also indirectly through 
the long-term cultivation of civic parenting. Panel data are drawn 
from a field study of effects of the Kids Voting USA curriculum on 
students and parents in Arizona, Colorado, and Florida. Respondents 
were interviewed following the 2002 election campaign and again one 
year later.
Civic Parenting
By civic parenting, we mean the purposeful activity of parents in 
encouraging the political curiosity, involvement, and expression of 
children. We presume that this effort is most beneficial to an 
adolescent when it is explicitly communicated in the family. Existing 
theoretical perspectives in communication research suggest that 
parents can successfully promote adolescent political involvement 
through family media use and discussion. "Parental mediation" of 
child television viewing (Austin & Pinkleton, 2001), parent news use 
habits and subscriptions (Zerba, 2004), family discussion about the 
news (McDevitt & Chaffee, 2003), and parents' comfort with 
controversial expression (Saphir & Chaffee, 2002) are associated with 
an upward trajectory of child political development. These parenting 
orientations and the availability of political stimulation might be 
inadvertent or haphazard in many families, as when, for example, 
children form impressions of issues based on observing parents 
complaining about news bias (Austin & Pinkleton, 2001). Here we view 
civic parenting as purposeful communication intended to promote 
political awareness and sophistication, but also allowing for 
reciprocal influence in the child-parent dyad. Civic parenting should 
be manifest in encouraging adolescents to pay attention to news and 
to talk about political issues. Furthermore, parents should possess a 
feeling of discussion efficacy vis-…-vis their children-i.e., the 
ability to get an adolescent to talk about politics. This parenting 
influence is, in fact, a major achievement given that many 
individuals, regardless of age, are reticent about speaking up on 
partisan topics (Dutwin, 2003). If parents can nurture such 
expression in the safety of the home, they may go a long way to 
preparing young citizens for confident participation in the more 
intimidating arena of the public sphere. We should note what we do 
not mean by civic parenting. We are not referring to heavy-handed, 
partisan indoctrination, as in a father announcing: "We are Democrats 
in this family so get used to it!" Civic parenting should have 
positive consequences for socialization when the effort is to promote 
awareness and reflection rather than to impose a partisan 
identification (Saphir & Chaffee, 2002). Taken a step further, civic 
parenting should be effective when adolescents are able to influence 
parents. A recent study found that student-initiated discussion and 
the ability to influence parents' news use are associated with 
adolescent political growth (McDevitt, 2004).
Kids Voting USA
Innovations during the Progressive Era offer hints as to how schools 
might cultivate civic parenting. The impulse to assimilate immigrants 
during this period "reached its greatest flowering in the 
kindergarten movement," Putnam writes (2000, p. 395). Reformers set 
out to influence parents' child-rearing techniques, and a nexus of 
involvement sprung forth in forms such as mother's clubs and sewing 
clubs. Progressive reformers understood that they could overcome 
cultural barriers to improve education by creating institutions 
centered on parenting.
Another historical era with implications for civic parenting occurred 
in the 1970s with the eruptions of school controversies about racial 
desegregation, student discipline, and community control and 
decentralization. Jennings (1975) documented how students were quite 
sensitive to pedagogical encroachments into family matters, alerting 
parents when potentially offensive issues arose. Jennings noted, 
"Having a child in school sets up a web of relationships tying 
school, child, and parent together. No other modern institution 
serves as the focal point for such common parent-child interests and 
activities as the school" (p. 49).
	Intricate relationships of school and home suggest a mediation role 
for families in which they are responsive to the exogenous 
stimulation provided by schools. This function is consistent with 
normative theories of democratic education as advanced by Dewey and 
taken up by contemporary scholars (Gutmann, 1987; Angell, 1991). 
Dewey stressed the benefits of constant interaction between an open 
classroom community and the social spheres with which it overlaps. 
The family, then, can be understood in political socialization as a 
social nexus shaped by extra-familial forces such as schools, media, 
elections, and community controversies (Cook, 1985).
Our expectation that schools can stimulate civic parenting is derived 
in part from prior research on the effects of Kids Voting USA. The 
program encompasses a multi-pronged approach based on peer-centered 
learning, information gathering, and hands-on activity. The 
nonpartisan, nonprofit initiative took root on a trial basis in 
several Arizona communities in 1988, and has since spread to 
affiliates in 29 states, encompassing 4.3 million students and 
200,000 teachers in 10,600 schools. The K-12 curriculum is taught 
during the final months of an election campaign to coincide with the 
beginning of a school year. Within the classroom, the Civics Alive! 
curriculum promotes the principle that citizens should study 
candidates and issues. This emphasis is particularly important for 
deliberative dispositions that might carry over into other social 
spheres such as the family and the community. KVUSA also offers 
community service in its Destination Democracy events. This extension 
of the curriculum into the community is important for older students 
as they are offered realistic opportunities to assert themselves in 
activities such as get-out-the vote campaigns. The final aspect of 
the program is voting on Election Day as students cast ballots 
alongside parents.
		The authentic experiences provided by Kids Voting contrast with the 
rote-instruction and passive learning of conventional civic courses 
(Niemi & Junn, 1998). These innovations have attracted a great deal 
of scholarly attention, making Kids Voting perhaps the most studied 
civic curriculum since the mid 1990s. This research tends to follow 
one of two tracks-documentation of increased voting turnout of 
parents and former students at the community level (e.g., Merrill, 
Simon, & Adrian, 1994), and studies of individual and family-level 
processes characterized by stimulated news media use, cognition, and 
opinion formation (e.g., McLeod, Eveland, & Horowitz, 1998; McDevitt 
& Chaffee, 2000, 2002). The latter research is more pertinent to this 
study, and we will highlight some of the key insights from prior 
studies that address how schools and parent interact as socializing 
agents. McDevitt and Chaffee (2000) documented the phenomenon of 
"trickle-up influence" in which KVUSA participation motivated 
children and adolescents to initiate conversations at home. Parents 
responded by paying more attention to news and strengthening their 
opinions in anticipation of future discussion. The authors 
subsequently developed a functional model of family political 
communication to illustrate how the parent's response reflects her 
desire to maintain a leadership role in the family (2002). Family 
role relationships are largely based on norms of communicative 
competence, and when these expectations are violated-as when students 
narrow gaps or surpass parents in civic expertise-tension can arise. 
This strain is resolved when parents regain leadership by paying 
attention to politics. In a subsequent study of curriculum effects, 
McDevitt (2004) illustrated the beneficial consequences of role 
strain. He tested a model of "developmental provocation" in which 
younger adolescents poked and prodded parents in conversations during 
the final weeks of the 2000 election. By generating feedback, 
adolescents benefited by comparing their opinions with those of 
parents. These studies describe how parents adapt to episodes of 
student-initiated conversation, but parent responses are viewed as 
essentially inadvertent and reactive.
A Model of Parent Priming
	We will test a model of parent priming that begins with a school 
intervention prompting students' news attention, discussion, and 
opinion formation as immediate effects. Student-parent
discussion should engender the same outcomes for parents, although to 
a lesser degree, as was shown in prior KVUSA evaluations (McDevitt & 
Chaffee, 2000, 2002). We anticipate that parents should show 
curiosity about civic instruction at school as an early indication 
that they are becoming interested in the political growth of their 
children. This priming of civic parenting will be consequential in 
the long run with parents encouraging adolescent development beyond 
the progress predicted by the school intervention, according to the model.
	Immediate Effects. Once triggered, student-parent discussion bridges 
the classroom with the living room, creating possibilities for 
immediate growth in three areas.
	News relevance. The stark reality is that most American teenagers 
most of the time do not pay much attention to politics (CIRCLE, 
2003). However, episodes of widespread political discussion increase 
the relevance and information utility of news (Atkin, 1981). When 
adolescents anticipate that they will be called upon in class to talk 
about campaign issues, they turn to media (perhaps more than to 
parents) for knowledge and opinions (Kiousis, McDevitt, & Xu, 2004). 
The interpersonal demands of political discussion would require that 
adolescents actively reflect upon the meaning of news events so that 
they comprehend how isolated events and discrete issues are connected 
(Kosicki & McLeod, 1990). Thus, motivated media use and concurrent 
discussion should entail information integration as a processing 
goal. We consequently incorporated measures of news attention, 
knowledge, and information integration to evaluate effects of Kid 
Voting. We will use these same indicators to document parent 
responses to student-initiated discussion and increased news 
attention. The anticipation of future conversations with children, 
coupled with parents' desire to maintain a leadership role in the 
family, should motivate them to fortify their knowledge base.
	Deliberative discussion. An intervention such as KVUSA establishes 
conditions in which school allowances for open discussion and 
partisan expression might conflict with the communication norms in 
families, particularly in homes with authoritarian parents (Saphir & 
Chaffee, 2002). However, political discussion at school tends to 
foster conversations in the home as teenagers become curious about 
how parents will respond to the contentious ideas batted around in 
the classroom (Jennings, 1975; McDevitt, 2004). In keeping with prior 
evaluations, we can expect that Kids Voting influence will translate 
into increased frequency of student-parent discussion about politics, 
along with more frequent conversations with friends and an expansion 
of adolescent networks for political discussion. These interpersonal 
outcomes should be detectable among parents also, although to a 
lesser extent given the inertia of adult dispositions such as an 
aversion to speaking up on touchy topics (Dutwin, 2003).
	For adolescents and parents, opportunities to practice political 
discussion-and to become more comfortable with it inside and outside 
the home-might engender the kinds of dispositions celebrated by 
theorists of deliberative democracy. Evidence for this will be 
assessed in terms of two indicators: willingness to openly disagree 
with others and to listen respectfully to partisan opponents. In the 
contexts of school and family life, we would describe this 
combination as confident and civil deliberation, acted out in 
spontaneous, everyday interaction. And there is good reason to 
believe that these outcome are realistic given recent studies showing 
that news media use is associated with a willingness to argue (Kim, 
Wyatt, & Katz, 1999) and political discussion enhances individuals' 
respect for partisan opponents (Voltmer & Lalljee, 2004).
	Opinion validation. Another benefit of school and family interaction 
is increased opportunities for students to compare, contrast, and 
integrate perspectives from multiple spheres of interaction. Student 
should become more curious about politics, for example, if the views 
of teachers conflict with those of peers, siblings, and parents. The 
increased attention to news and the efforts at information 
integration, described above, should provide a cognitive base for 
opinion formation and an awareness of campaign issues that are most 
salient at the time. Discussion in various settings, meanwhile, 
should allow students to try out tentative opinions and partisan 
identities to see how they play out in social relations. The very act 
of articulating political views forces a discussant to think more 
carefully about the consistency and validity of political opinions, 
and the responses from others provide additional information for 
shaping attitudes (Dutwin, 2003). With parents more active in media 
use and discussion, we can expect some degree of opinion 
crystallization among them as well.
	Finally, we expect that parents will become interested in classroom 
experiences as they observe the civic transformation of children and 
wonder where this inspiration is coming from.
H1: Kids Voting will provide immediate stimulation for students' news 
media use, discussion, and opinion formation.

H2: Student discussion at home will provide immediate stimulation for 
parents' media use, discussion, opinion formation, and interest in 
civic instruction.

	Long-term Effects. The purpose of KVUSA is to cultivate 
communication and participatory skills that are transferable outside 
the classroom. The hope is that when adolescents develop curiosity 
about election news, and when they achieve civic empowerment by 
acquiring opinions and expressing partisan identities, these 
orientations will reinforce each other as habits long after the 
intervention ends. Episodic attention to election news should allow 
adolescents to develop cognitive schema for more efficient 
comprehension of news, increasing the chances that news exposure will 
become habitual (Kosicki & McLeod, 1990). Likewise, a situational 
increase in discussion could take hold as a long-term habit if 
adolescents experience a sense of empowerment from expressing an 
autonomous civic identity in relation to teachers, peers, and 
parents. That is, teenagers would persist in initiating discussions 
if they begin to appreciate the social utility of possessing 
political views and expressing them.
	The cognitive and attitudinal sophistication that results from media 
use and interpersonal communication should generate motivation, 
efficacy, and skills for civic participation. This follows from 
normative conceptions of deliberative democracy, and recent research 
confirms a close empirical association between these communication 
activities and various participatory behaviors beyond what would be 
predicted from demographics and general political interest (Kim, 
Wyatt, & Katz, 1999). In applying the deliberative democracy 
framework to socialization, we will test whether Kids Voting's 
influence as a communication stimulus translates into campus 
activism, volunteering, and students' intention to vote in 2004.
H3: KVUSA will predict students' media use, discussion, opinion 
formation, and political participation in the long term.

	We have used the metaphor of parent awakening to suggest that 
schools might indirectly influence students by making a civic 
nurturing role more salient to adults in childrearing. If, as we 
expect, parents become more interested in civic instruction at 
school, they might contemplate what their own duties are in preparing 
children for citizenship. For example, this might come about if 
parents feel a sense of competition with teachers or if they simply 
come to realize that they share responsibility with schools in 
preparing their offspring for civic life. This realization might come 
about during the school intervention period, but ultimately we are 
interested in whether interest in school instruction evolves into 
purposeful and sustained efforts at civic parenting. The student 
herself should help to bring this about given our understanding of 
the student-parent dyad as a nexus for reciprocal influence via media 
use and discussion. In light of the upward trajectory of long-range 
growth described above for adolescents, students should  contribute 
political stimulation to the parent via discussion. And if parents 
want to maintain a leadership role in the family in terms of 
politics, they will have to upgrade their own knowledge and 
discussion skills (McDevitt & Chaffee, 2002), providing, in turn, 
motivation and resources for civic parenting. Thus, the phenomenon of 
civic parenting is really a manifestation of reciprocal influence in 
the dyad. Just as the adults in our study should benefit from the 
students' long-term political involvement, parents might retain and 
refine communication dispositions that promote the continued civic 
growth of their children.
	H4a: Parent interest in instruction will predict civic parenting in 
the long term.

	H4b: Civic parenting will predict students' media use, discussion, 
opinion formation, and
participation in the long term.





Method
This study explores the influence of Kids Voting in three 
communities: Maricopa County, Arizona; El Paso County, Colorado; and 
Broward/Palm Beach counties, Florida. The panel design encompasses 
interviews of student-parent dyads following the 2002 election and 
interviews of the same respondents one year later. Each of the three 
sites included the presence of both Kids Voting schools and 
non-participating schools. Similar demographics between the two 
groups of schools would help us to eliminate extraneous factors as 
explanations for Kids Voting effects. To test this assumption, a 
multiple regression model was created with the following demographic 
variables serving as predictors of Kids Voting exposure: gender, 
ethnicity, grades earned in school, family SES, and parental vote 
turnout in prior elections. The demographic predictors made no 
significant contribution to the variance explained. Nevertheless, the 
design does not fit entirely the requirements for a fully controlled 
experiment in that we did not randomly assign students to 
independent-variable conditions. We consequently characterize this 
study as a quasi-experiment, in which the selection to contrasting 
conditions appears to be unbiased but is not literally randomized 
(Campbell & Stanley, 1963). A particular student's participation in 
the curriculum was determined by decisions made by school 
administrators and teachers, eliminating self-selection (of students) 
as a threat to internal validity.
Kids Voting programs vary from community to community, depending upon 
the amount of volunteer support and the discretion of district 
administrators and individual teachers. Any given teacher might 
decide to use all, some, or none of the Kids Voting lesson plans, and 
we accordingly assessed Kids Voting exposure as a continuous 
variable. It is also worth noting that many of the curriculum 
activities we measured are not necessarily unique to Kids Voting, as 
any enterprising teacher might decide to implement similar lesson 
plans. Thus, we are not engaged in a conventional curriculum 
evaluation of clearly contrasting conditions. However, the presence 
of Kids Voting in the three study sites did serve our purposes by 
increasing variance in the types and intensity of civic learning 
experiences. Ultimately we are interested in observing what happens 
to family dynamics in response to any sufficiently intensive school 
intervention.
Data Collection & Sampling. The sampling frame is defined as all 
families in the three regions with at least one student in the 11th 
or 12th grade during the fall of 2002. The total sample includes 
students representing more than 150 schools.2 We obtained lists of 
students and parents from a major vendor for survey research sample 
frames. To maximize the response rate for a self-administered mail 
back, Dillman's (2000) Total/Tailored Design Method was used, which 
includes follow-up contacts to non-respondents. In addition, we 
included small incentives ($5 phone cards) in the initial mailing and 
a 1-800 number in case students or parents had questions. We also 
provided a Web-based survey, anticipating that this option would be 
especially attractive for adolescents. Finally, we conducted 
telephone interviews to reach students and parents who failed to 
respond initially.3
For the first wave of interviews (T1), the initial questionnaire 
mailing took place on November 19, 2002 (after schools had 
implemented the curriculum and the election finished). Reminder post 
cards were mailed to non-respondents on December 9. Telephone 
follow-up began December 17. During this phase, at least 10 attempts 
were made before coding a number as unreachable. Web surveys were 
completed throughout the field period. Data collection ended on 
February 25, 2003. The N for the final sample is 497 student-parent 
dyads. The basic demographics for the student sample are as follows: 
53% juniors and 47% seniors; 57% female and 43% male; 64% Anglo, 12% 
Hispanic, 7% African-American, 3% Asian, 1% Native American, and 13% 
"other." In terms of SES, 50% of the parents indicated that they 
graduated from college, and 75% said they earn at least $41,000 annually.
The cooperation rate for student-parent dyads represents the ratio of 
completed questionnaires/interviews to eligible respondents 
contacted. The rates are 58% for Arizona, 62% for Colorado, and 55% 
for Florida. These rates are consistent with a recent effort to reach 
young adults on matters of civic engagement without the benefit of 
school-site administration (National Survey of Student Engagement, 2002).
	We refined interview procedures for the second wave (T2) by dropping 
the online questionnaire option but increasing substantially our 
efforts in the telephone and mail-back modes. We provided the same 
response incentive. Interviews were initially attempted by telephone 
with all dyads in the sample, beginning on November 6, 2003. Many of 
the students had graduated from high school and had to be reached at 
new addresses. Telephone calls were first made to the homes of the 
parents. If students no longer lived at home, parents were asked for 
the new phone numbers. The range of attempts per telephone number was 
1 to 35. Often, the student was now in college and reaching them was 
difficult. Questionnaires were mailed at two different times to 
parents and students not reached by phone. In total 308 parents and 
313 students completed the questionnaire at T2, either by phone or by 
mail, comprising a total of 288 dyads. This represents a completion 
rate of 58 percent from the baseline number of dyads.
	The basic demographics for the student sample at T2 are as follows: 
59% female and 41% male; 71% Anglo, 8% Hispanic, 7% African-American, 
2% Asian, and 12% "other." In terms of SES, 54% of the parents 
indicated that they graduated from college, and 76% said they earn at 
least $41,000 annually. These breakdowns are similar to those at T1, 
particularly with respect to gender and parent income level, although 
we do see drop off among the Hispanic respondents, from 12% at T1 to 8% at T2.
Kids Voting Participation. The T1 questionnaire included 10 items 
used to prompt recall of Kids Voting experiences. Each item was 
standardized and summed to form an index. No single item is 
definitive evidence of participation, but responses to the items 
collectively provide a probabilistic approach. Reliability of the 
scale (_ = .62) was similar to the exposure scale used by Chaffee et 
al. (1995) in their study of the program in San Jose, California (_ = .67).
For the first two questions, students used a 1-to-5 scale with 1 
meaning "never" and 5 meaning "very often." Respondents were asked, 
"In school this fall, how often has the election been discussed in 
your classes?" and "How often have your teachers encouraged you to 
say what you think about politics, even if the topic is 
controversial?" Students then answered "yes" or "no" to questions 
about their school experiences from that year. These items asked 
whether they took sides in a debate, analyzed political cartoons, 
analyzed political ads, participated in a "service learning" program, 
worked at a polling site on Election Day, encouraged people to 
register to vote, did any homework assignments on the election that 
involved family participation, or voted with a parent on Election Day.
Demographics. The following demographic controls, assessed at T1, 
were incorporated for both students and parents: gender, ethnicity, 
religious group membership, state of origin, and family SES. We also 
measured student grade level and grades earned in school. We asked 
parents to report voting or non-voting in previous elections, which 
represents a variable not controlled for in prior evaluations of Kids 
Voting. Item wording and coding are provided in the Appendix.
Civic Involvement. Kids Voting effects for students are evaluated 
using measures of media use and cognition, discussion, and opinion 
formation at T1 and at T2. Dimensions of civic participation are 
measured at T2. For parents, curriculum influence at T1 is measured 
in the areas of media use and cognition, discussion, opinion 
formation, and interest in civic instruction. The civic parenting 
variable is measured at T2. The corresponding measures for student 
and parent political involvement are identical or nearly identical. 
In most cases the corresponding T1 and T2 indicators are identical. 
The Appendix provides item wording, coding, and reliability scores.
Results
We created a stringent test of Kids Voting effects given the many 
demographic controls. As shown in Table 1, a hierarchical regression 
model controls first for the demographics. The second equation 
incorporates the measure of KVUSA exposure. H1 predicted that Kids 
Voting would stimulate student media use, discussion, and opinion 
formation. The curriculum provided a strong impetus for increased 
attention to election news (_ = .33, p < .001) and information 
integration (_ = .28, p < .001). To a lesser extent, the curriculum 
also promoted knowledge gain and issue salience. We suspect that the 
knowledge outcome is mostly an indirect effect of the curriculum via 
news processing.
-Table 1 here-
	Situational increases in news attention are often accompanied by an 
upsurge in political discussion (Atkin, 1972), and this is the case 
here as KVUSA also stimulated student-parent conversation (_ = .30, p 
< .001). A similar magnitude of effect was evident with size of 
discussion network, and KVUSA influence was even stronger for 
discussion with friends, accounting for 14 percent of incremental 
variance. The dimensions of interpersonal communication represent 
three of the strongest effects. While demographics accounted for an 
average of 4.3% of variance explained across the three measures, 
KVUSA accounted for 12.7%. The curriculum also promoted deliberative 
habits of willingness to disagree and willingness to listen to 
opponents, although to a lesser degree. KVUSA provided a modest 
impetus to partisanship (_ = .11, p < .05). While the result is not 
spectacular, opinion formation is accompanied by evidence of news 
media use and cognitive processing. We can reasonably presume that 
these partisan views are backed up with some contemplation.
The promotion of student-parent conversation is a particularly 
important finding as student discussion acts as a lynchpin in our 
model of family political activation. According to H2, student 
discussion will stimulate parents' media use, discussion, opinions, 
and interest in civic instruction. In the regression model used for 
Table 2, the demographic block is entered in the first equation, 
followed by Kids Voting, and finally student-parent discussion in a 
third equation. The effects of Kids Voting on parents in aggregate 
are less pronounced than those on the adolescents. This is not 
surprising given that parents were not directly exposed to the 
curriculum, as KVUSA influence is mostly mediated by the prompting of 
student discussion in the home (McDevitt & Chaffee, 2000). Even so, 
Kids Voting generated significant effects on parents' frequency of 
political conversations with friends, size of discussion network, 
partisanship, and interest in civic instruction.
-Table 2 here-
Student discussion at home generated significant effects with respect 
to all four measures of parent media use and cognition, conversations 
with friends, and willingness to disagree. Adolescent-parent 
conversations probably involved a fair amount of disagreement, given 
the electoral backdrop, and thus the students appear to be 
socializing their parents to be more assertive in expressing partisan 
views inside and outside the home. Finally, student-parent discussion 
accounted for 11% of incremental variance in parents' interest in 
civic instruction. This last finding might help to explain why KVUSA 
was so effective in the immediate stimulation of students' political 
involvement (Table 1). As evident in the operationalization of this 
variable, interest in instruction is quite evident to the students as 
parents ask children about homework assignments, talk to them about 
classes, and encourage them to express opinions at school. Parent 
interest should heighten the relevance of Kids Voting lesson plans 
for students as they come to understand that parents believe these 
topics to be worthy of discussion at home.
We now consider the possibility of long-term effects of Kids Voting 
on students, as well as indirect effects via the stimulation of civic 
parenting. The hierarchical regression model used in Table 3 reflects 
our theoretical expectation of sequential processes that culminate in 
civic parenting. Demographics are entered in the first equation, 
followed by KVUSA exposure at T1. We then added the T1 student 
outcome measure as an additional control to help us evaluate the 
strength of civic parenting as a predictor of gains beyond the T1 
levels. Civic parenting as measured at T2 is entered in the final equation.
-Table 3 here-
	KVUSA influence endured as shown in every indicator. All told this 
pattern provides impressive evidence for the assertion that the 
intervention would promote habits of news media use, cognition, 
discussion, and participation (H3). Both the curriculum and the 
election campaign had ended in the fall of 2002, and yet students 
continued to pay attention to news and to participate in political 
conversations with parents and friends. Convincing evidence for the 
induction of news attention and discussion as long-term habits is 
found in the knowledge effect. The most intriguing results probably 
involve knowledge, the two deliberative dispositions, and 
partisanship. In all four cases, KVUSA influence is stronger at T2 
than at T1.
	Implied but not shown directly in Table 3 are influences of Kids 
Voting and student-parent discussion on civic parenting. Parent 
interest in instruction, once stimulated by student-parent 
conversation, should provide motivation and cognitive resources for 
civic parenting in the long term, according to our theoretical model. 
We traced this process in path modeling in which we controlled for 
the demographics and assumed a causal ordering in which KVUSA would 
prompt student-parent discussion at T1, leading to parent interest in 
instruction at T1, with civic parenting as the final outcome at T2. 
The results confirmed that student exposure to KVUSA did not exert 
any direct influence on civic parenting, but student discussion in 
the home predicted civic parenting at T2. This influence was mostly 
indirect, mediated by parent interest in civic instruction. These 
findings support the assertion that parent interest in civic 
instruction is consequential as a mechanism for cultivating parent 
commitment to civic nurturing (H4a).
With these mediating processes confirmed, we can consider 
consequences of civic parenting for student growth in the long run 
(H4b). Civic parenting effects are most consistent in the area of 
media use and cognition, with increments to variance of 2% to 4%. 
Civic parenting also predicted adolescent-parent discussion at T2. 
Our previous analyses suggest that students contributed to initiating 
conversations at home at T1, but here we are controlling for the T1 
measure of student-parent discussion. As shown in Table 3, civic 
parenting helped to sustain this activity. Civic parenting also 
accounted for incremental variance in the T2 indicators of student 
discussion with friends, willingness to disagree, listening to 
opponents, and partisanship, but failed to predict participation.4
Discussion
The results in aggregate support our model of parent priming in which 
a school intervention engenders students' news attention, discussion, 
and opinion formation as immediate effects. Student-parent discussion 
fosters the same outcomes for parents, although to a smaller extent. 
Parents then become curious about instruction at school as an early 
indication that they are developing a commitment to the civic 
nurturing of children. This priming of civic parenting is meaningful 
in the long run as parents encourage adolescent development beyond 
the progress originally set in motion by the school intervention. We 
take from the findings that effects of civic education must be 
understood beyond patterns of direct influence, as systematic as they 
may be in the case of an intervention such as Kids Voting. 
Student-parent discussion, once initiated, seems to transform the 
family environment, making the home more hospitable for habitual news 
use and political conversation.
	The results provide insights as to how schools, students, and 
parents each contribute to the activation of civic parenting as a 
mechanism of political socialization. While the three sources of 
influence appear to be synergistic in relationship to each other, it 
is worth noting the unique contributions of each. Judging by the 
incremental variances reported in Table 3, KVUSA is particularly 
adept at fostering discussion, and this makes intuitive sense given 
that the home cannot compete in terms of opportunities for peer 
interaction. The public, secular environment of schools might also 
make a civic curriculum more effective than parents in preparing 
adolescents for political participation. This was shown clearly in 
Table 3. Parents did contribute to media use and cognition, and this 
also fits with what we know of the home as the primary site for 
spontaneous and habitual media access. But parents also helped to 
promote the deliberative dispositions of willingness to disagree and 
listening to opponents. The home can provide a relatively safe domain 
for practicing opinion expression, and thus parents can play an 
important role in preparing children for participation in discursive 
venues outside the home. As for the adolescent's role, she is the one 
who must make the effort to grow, of course, but she also plays a 
crucial role by initiating discussions that allow for the integration 
of school and parent influence.
Limitations & Future Research. Our documentation of curriculum 
effects is based on a careful accounting of student, parent, and 
family demographic traits that would normally predict levels of civic 
involvement. Numerous demographic controls, along with the panel 
design, strengthen internal validity with respect to curriculum 
effects on students. While prior evaluations of KVUSA effects 
incorporated standard demographics, this study included an additional 
control in the form of parents' prior voting. We can also address 
concerns about social desirability in interview responses with our 
tests for student and parent knowledge, which were not subject to 
fake answers or exaggerated claims. Response set is another concern 
of questionnaire data, but data were obtained from different 
respondents; we were careful to interview students and parents 
separately. Our analyses involved use of student measures to predict 
parent outcomes, and parent measures to predict student outcomes.
Future research is needed, however, to shore up assumptions of 
causality in the model of parent priming. For example, the results 
seem to suggest a symbiotic relationship between KVUSA and parents as 
socializing agents for students, with schools and parents reinforcing 
each other. However, we do not possess direct evidence that parents' 
interest in instruction at T1 heightens the impact of civic 
instruction by making the curriculum more relevant to students. The 
causal or reciprocal relationship between student curriculum 
participation and parent interest in instruction is difficult to 
assess in that both variables were measured at T1.
Implications for Civic Education. Debate on the role of public 
schools in civic education is often framed in dichotomous terms: Do 
schools or parents have responsibility for certain areas of 
instruction, as in the teaching of "values" that might intersect 
religion and politics? This conflict schema is potentially 
misleading. Parents and schools are potentially linked via student 
discussion, as we have shown here. The resulting flow of idea should 
reduce tension as the parent's traditional role of asserting family 
values is activated while the adolescent is exposed to a larger 
universe of secular views in school. The adolescent benefits by 
comparing and integrating perspectives from multiple sources: 
parents, teachers, friends, and news media. Unfortunately, social 
studies instructors are often leery about allowing for partisan 
discussion in class, particularly when topics might ruffle parents 
(CIRCLE, 2003).
 From our perspective, teachers should actively encourage students to 
talk about how their parents' views inform their beliefs, rather than 
skirting this topic as taboo. But parent organizations, political 
parties, and local journalists must be tolerant of this kind of 
instruction rather than contributing to inflammatory rhetoric about 
teachers upstaging parents. In a worst-case scenario, the goal of 
political identity exists as a kind of secret project of the 
adolescent, who must figure out on her own how to integrate 
perspectives. By contrast, when schools promote civic parenting, they 
create opportunities for students to express political identities 
that bridge family and school life.

Notes

1.  Scholars have recognized that many parents do express their 
partisan preferences to children. For example, Tedin (1974, p. 1581) 
concluded that some parents "convey to their children quite early 
that `we' are Democrats or `we' are Republicans." But party 
identification stands out from other attitudes in its salience to 
parents and visibility to children.

2.  Cooperation of school officials is often sought because of the 
reduced costs of administering questionnaires to students. However, 
the logistics of working with school districts typically results in 
researchers gathering data from just one or several schools, bringing 
up concerns about external validity. The design of this study, 
however, called for interviews of parents as well, and we did not 
want to restrict data collection to just a few schools. But there are 
certainly limitations to the sample obtained in terms of an upward 
tilt in family SES. We can point out that the narrowed SES range 
works against us in statistical tests given that prior studies showed 
that KVUSA is most effective for low-income families (McDevitt & 
Chaffee, 2000).

3.  The distribution of respondents for each data collection mode was 
54% for telephone, 36% for mail, and 10% for Web. A preliminary 
analysis showed that collection mode is not significantly correlated 
with KVUSA exposure or with the outcome variables.

4.  We explored possible interaction effects of Kids Voting X civic 
parenting to predict long-term outcomes but the results were not 
statistically significant.


References

Angell, A. V. (1991). Democratic climates in elementary classrooms: A 
review of theory and
	research. Theory and Research in Social Education, 19 (3), 241-266.

Atkin, C. (1972). Anticipated communication and mass media 
information seeking. Public
	Opinion Quarterly, 36, 188-199.

Atkin, C. (1981). Communication and political socialization. In D. 
Nimmo & K. Sanders (Eds.),
	Handbook of political communication (pp. 299-328). Beverly Hills, CA: Sage.

Austin, E. W., & Pinkleton, B. (2001). The role of parental mediation 
in the political
socialization process. Journal of Broadcasting & Electronic Media, 45 
(2), 221-240.

Brady, H. E., Verba, S., & Schlozman, K. L. (1995). Beyond SES: A 
resource model of
participation. American Political Science Review, 89, 271-294.

Campbell, D. T., & Stanley, J. C. (1963). Experimental and 
quasi-experimental designs for
research. Boston: Houghton Mifflin.	

Center for Information & Research on Civic Learning & Engagement 
(2003). The civic mission of schools. CIRCLE and Carnegie Corporation 
of New York.

Chaffee, S. H., Moon, Y., McDevitt, M., Pan, Z., McLeod, J. M, 
Eveland, W. P., & Horowitz,
	E. (1995). Stimulation of communication: Reconceptualizing the study 
of political
Socialization. Paper presented at the annual meeting of the 
Association for Education in Journalism and Mass Communication, 
Washington, D.C.

Cook, T. E. (1985). The bear market in political socialization and 
the costs of misunderstood
psychological theories. American Political Science Review, 79, 1079-1093.

Dillman, D. A. (2000). Mail and internet surveys: The tailored design 
method. New York: John
Wiley.

Dornbusch, S. M., Ritter, P. L., Leiderman, P. H., Roberts, D. F., & 
Fraleigh, M. (1987). The
relation of parenting style to adolescent school performance. Child 
development, 58, 1244-1257.

Dudley, R. L., & Gitelson, A. R. (2002). Political literacy, civic 
education, and civic
  engagement: A return to political socialization? Applied 
Developmental Science, 6 (4), 175-182.

Dutwin, D. (2003). The character of deliberation: Equality, argument, 
and the formation of
public opinion. International Journal of Public Opinion Research, 15, 239-264.

Gutmann, A. (1987). Democratic education. Princeton, NJ: Princeton 
University Press.

Greenstein, F. (1965). Children and politics. New Haven: Yale University Press.

Hyman, H. H. (1959). Political socialization. Glencoe, IL: Free Press.

Jennings, K. M. (1975). Discontent with schools: Some implications 
for political socialization
and behavior. Youth & Society, 7 (1), 49-68.

Kiousis, S., McDevitt, M., & Wu, X. (2004). The genesis of civic 
awareness: Agenda-setting
	in political socialization. Paper presented at the annual meeting of 
the International
	Communication Association, New Orleans, LA.

Kim, J., Wyatt, R. O., & Katz, E. (1999). News, talk, opinion, 
participation: The part played by
	conversation in deliberative democracy. Political Communication, 16, 359-360.

Kosicki, G., & McLeod, J. (1990). Learning from political news: 
Effects of mass media images
and information-processing strategies. In S. Kraus (Ed.), Mass 
communication and political information processing (pp. 69-83). 
Hillsdale, NJ: Erlbaum.

McDevitt, M. (2004). Poking and prodding parents: The child's 
assertion of political identity.
Paper presented at the annual meeting of the International 
Communication Association, New Orleans, LA.

McDevitt, M., & Chaffee, S. H. (2000). Closings gaps in political 
communication and
	knowledge: Effects of a school intervention. Communication Research, 
27 (3), 259-292.

McDevitt, M., & Chaffee, S. H. (2002). From top-down to trickle-up 
influence: Revisiting
assumptions about the family in political socialization. Political 
Communication, 19, 281-301,

McLeod, J. M., Eveland, W. P. Jr., & Horowitz, E. M. (1998). Going 
beyond adults and voter
turnout: Evaluating a socialization program involving schools, family 
and media. In T. J. Johnson, C. E. Hays, & S. P. Hays (Eds.), 
Engaging the public: How government and the media can reinvigorate 
American democracy (pp. 217-234). Lanham, MD: Rowman and Littlefield.		

	Merrill, B. D., Simon, J., & Adrian, E. (1994). Boosting turnout: 
The Kids Voting program,
		1990. Journal of Social Studies Research, 18, 3-7.	

McLeod, J. M. (2000). Media and civic socialization of youth. Journal 
of Adolescent Health,
	27 (2), 45-51.

National Survey of Student Engagement. (2002). The NSSE 2002 report: 
 From promise to
	progress: How colleges and universities are using student engagement 
results to
	improve collegiate quality. Bloomington, IN: Indiana University Center for
	Postsecondary Research and Planning.
Niemi, R. G., & Junn, J. (1998). Civic education: What makes students 
learn. New Haven:
	Yale University Press

Putnam, R. D. (2000). Bowling alone: The collapse and revival of 
American community. New
York: Simon & Schuster.

Saphir, M. N. & Chaffee, S. H. (2002). Adolescents' contribution to 
family communication
		patterns. Human Communication Research, 28, 86-108.

Tedin, K. L. (1974). The influence of parents on the political 
attitudes of adolescents. The
American Political Science Review, 68 (4), 1579-1592.

Voltmer, K. & Lalljee, M. (2004). Communication and political 
disagreement: The effects of
	deliberation on attitudes toward others. Paper presented at the 
annual meeting of the
	International Communication Association, New Orleans, LA.

Zerba, A. (2004). Growing up with parents who read and watch the 
news: What is the effect on
	college students? Paper presented at the annual meeting of the 
Association for Education
	in Journalism and Mass Communication.

Appendix

Student demographics
	
These measures were assessed during the first year of data collection (T1).

Grade level. A single item determined year in school:
What grade are you in at school? Coded: 11th=1, 12th=2.

  Grades earned. A single item measured grades received in school.
Would you say your grades are mostly A's, B's, C's or D's? mostly 
A's=4, mostly B's=3, mostly C's=2, mostly D's=1.

Gender. A single item determined gender.
What is your gender? female=1, male=2.

Ethnicity. An item asked about ethnic background.
Of what ethnic group do you consider yourself? Hispanic (including 
Chicano and Spanish), Native American, African American, Asian, and 
other= dummy 1; white=dummy 2.

Religious group membership. One item assessed membership.
Are you a member of a religious group or club?" no=0, yes=1.
	
Parent demographics
	
Gender, ethnicity, and religious group membership were identical to 
the student measures. Data for these measures were also assessed at T1.

SES. A two-item scale measured family socioeconomic status based on 
the parent's report of income and education. We standardized the 
coded values for each item and summed the scores. The correlation is 
.36 (p <. 001).
For statistical purposes, we need to estimate household income before 
tax. Indicate the category that fits you. less than $15,000=1, 
$16,000 to $25,000=2, $26,000 to $40,000=3, $41,000 to $60,000=4.
Indicate your level of formal education completed. some high 
school=1, graduated from high school=2, some college=3, graduated 
from college=4, attended graduate school=5.
	
Prior voting. A summed, three-item scale assessed frequency of prior voting.
Did you vote in this year's election (2002)? Coded no=0, yes=1. The 
alpha is .79.
Did you vote in the 2000 presidential election between Al Gore and 
George W. Bush? no, don't recall=0, yes=1.	
Did you vote in the 1996 presidential election between Bill Clinton 
and Bob Dole?

Political involvement at T1 & T2

News attention. Students and parents were asked how much attention 
they paid to the election campaign at T1 and to politics in general 
at T2. They used a 1-to-5 response scale with 1 meaning "none" and 5 
"a great deal."

Information integration. Two items comprised a summed scale at T1. 
Students and parents  were asked to assess how well the following 
statements described them. The response options and coding were as 
follows: not at all like me/not sure=1; somewhat like me=2; a lot 
like me=3. The correlation is .43 (p < .001) for students and .47 for 
parents (p < .001).
When I came across election stories, I found myself tying the stories 
to ideas I had before.
When I join in political conversations, I find myself tying the 
arguments to ideas I had before.
Four items comprised the student measure at T2. The alpha is .67.
When I join in political conversations, I find myself tying the 
arguments to ideas I had before.
When I see or read a news story about an issue, I try to figure out 
if it is biased.
News about people running for office makes me wonder how they might 
change things.
When I hear news about politics, I try to figure out what is REALLY going on.
	For parents, the items about people running for office and about 
conversations were dropped to improve reliability. The remaining 
items were correlated at .42 (p <. 001).

Political knowledge. For students at T1, four questions were used to 
create a summed scale. Answers were coded 0 for incorrect, 1 for 
don't know (DK), and 2 for correct. The alpha is .60.
Which party controls the U.S. House of Representatives?
Which party controls the U.S. Senate?
What is the party affiliation of Matt Salmon/Bill Owens/Jeb Bush?
What is the party affiliation of Janet Napolitano/Rollie Heath/Bill McBride?
	For parents, the questions above and two additional items were used. 
The alpha is .61.
Which party would you say is more in favor of school vouchers?
Which party has been more supportive of privatizing Social Security 
investments?
	For students at T2, seven questions were used to create a summed 
scale. Answers were coded 0 for incorrect, 1 for don't know (DK), and 
2 for correct. The alpha is .60.
Which party do you consider more liberal?
Which party is more in favor of tax cuts to help stimulate the economy?
Which party controls the U.S. House of Representatives?
Which party controls the U.S. Senate?
What is the party affiliation of General Wesley Clark?
What is the party affiliation of Richard Cheney?
What is the party affiliation of Howard Dean?
	For parents at T2, the questions above were used along with two 
additional items. The alpha is .72.
Which party would you say is more in favor of school vouchers?
Which party is more in favor of reducing government regulations to 
help stimulate the economy?
What is the party affiliation of Tom Daschle?
	
Frequency of student-parent discussion. A summed two-item measure was 
used. Students and parents answered with a 1-to-5 scale with 1 
meaning "never" and 5 meaning "frequently." The correlation is .47 
(p  < .001) at T1 and .38 (p < .001) at T2.
T1: How often did you talk about the election campaign with your child/parents?
T2: How often do you talk about politics with your child/parents?
	
Frequency of discussion with friends. Students and parents answered 
using a 1-to-5 scale with 1 meaning "never" and 5 meaning "very often."
T1: How often did you talk about the election campaign with your friends?
T2: How often do you talk about politics with your friends?
	
	Size of discussion network. We used the original number provided by 
students and parents for this measure.
How many friends do you have who like to talk about politics?
	
	Willingness to disagree. A single item was used. Students and 
parents answered with a 1-to-5 scale with 1 meaning "never" and 5 
meaning "frequently."
In conversations, how often do you openly disagree with people about politics?
	
	Listening to opposing views. A single item was used; students and 
parents answered with the same scale.
How often do you listen to people talk about politics when you know 
that you already disagree with them?

Issue salience. In each study site, voters considered a state 
amendment. Colorado voters contemplated whether to restrict bilingual 
education, Florida voters decided whether to limit class-size, and 
Arizona voters chose whether to expand casino gambling. At T1, a 
single item was used to measure the salience of the ballot issue, and 
at T2 a question referred to the economy. Students and parents 
answered with a 1-to-5 scale with 1 meaning "not important" and 5 
meaning "very important."
T1: How important is the issue of expanding gambling/restricting 
bilingual education/limiting class size in Arizona/Colorado/Florida?
At T2: How important is the issue of the economy?

Partisanship. We created this measure by summing responses to a party 
identification item and a political ideology question. Party ID was 
coded so that "Republican," "Democrat" = 2; "independent," "not a 
member of another party," and "don't know" = 1. Ideology was coded so 
that "conservative," "liberal" = 2; "no, I'm moderate or middle of 
the road," "no, I don't think of myself that way," "don't know" = 1. 
For students, the correlation is .18 (p < .001) at T1 and .26 (p < 
.001) at T2.  For parents, the correlation is .24 (p < .001) at T1 
and .18 (p < .01) at T2.
Do you think of yourself as a Republican, a Democrat, or something like that?
When it comes to politics, do you consider yourself as liberal or 
conservative?
	
Parent interest in civic instruction. Six items were used for this T1 
measure. For the first four items, parents answered with a 1-to-5 
scale with 1 meaning "never" and 5 meaning "frequently." For the 
final two items, respondents answered with a 1-to-5 scale with 1 
meaning "not like me" and 5 meaning "a lot like me." The alpha is .72.
During this school year, how often have you asked your child about 
homework assignments?
How often have you visited the school to volunteer for activities?
How often have you asked your child about a civic or government class?
How often have you told your child that he or she should express an 
opinion during a civic class?
I am interested in what my child is learning in a civic class.
I want my child to express his or her political views in a civic class.
  	
Political involvement at T2

Campus politics. We used a branching question to first identify 
whether a respondent was still a student in high school or in 
college. If so, the student was asked:
At your campus this year, have you participated in any political 
activities such as protests or demonstrations? Coded: yes=1, no=0.
	
Volunteering. A single item measured students' political volunteering 
(same coding).
Have you volunteered this year for any political organizations or causes?

Intention to vote. Students were asked how well the following 
statement described them:
I DEFINITELY plan to vote in the 2004 presidential election. Coded: 
not at all like me/not sure, DK=1; somewhat like me=2; a lot like me=3.

Civic parenting. Two items were combined to form this index. Parents 
were asked to assess how well two statements described them. The 
response options and coding was as follows: not at all like me=1; 
somewhat like me=2; a lot like me=3. The correlation is .37 (p < .001).
I frequently encourage a child to pay attention to news events.
It's easy for me to get a child to talk about politics.
Table 1: Immediate Effects of Kids Voting on Students (Hierarchical 
Regression)


Demographics
R2

Kids Voting
  _R2


beta

Total R2
T1 Outcomes
Media use & cognition
    News attention
.03
.10***
.33***
.13***
    Information integration
.06**
.07***
.28***
.13***
    Knowledge
.13***
.02**
.14**
.15***
    Issue salience
.01
.03***
.18***
.04***
Discussion
    With parents
.10***
.09***
.30***
.19***
    With friends
.05**
.14***
.39***
.19***
    Size of discussion network
.02
.12***
.36***
.14***
Deliberative habits
    Willingness to disagree
.05**
.07***
.27***
.12**
    Listening to opponents
.03
.02**
.14**
.05**
Partisanship
.07***
.01*
.11*
.08***
* p < .05; ** p <.01; *** p < .001. N = 497 dyads.

Note. The first column reports the amount of variance accounted for 
by ethnicity, gender, grade in school at T1, grades earned, religious 
group membership, SES, and the parents' voting history, which were 
entered simultaneously in the first equation. The second column 
reports the incremental variance attributed to Kids Voting exposure, 
which was entered in the second equation. The third column reports 
the beta produced by Kids Voting in the second equation. The fourth 
column reports the total variance explained.



Table 2: Immediate Effects of Kids Voting and Student Discussion on Parents
(Hierarchical Regression)



Demographics
R2


Kids Voting
_R2

Student-parent
discussion
_R2



Total R2
T1 Outcomes
Media use & cognition
    News attention
.07**
.00
.10***
.17***
    Information integration
.12***
.00
.06***
.18***
    Knowledge
.22***
.00
.08***
.30***
    Issue salience
.02
.00
.02**
.04**
Discussion
    With friends
.04*
.03***
.14***
.21***
    Size of discussion network
.04#
.02*
.00
.06*
Deliberative habits
    Willingness to disagree
.03#
.00
.04***
.07***
    Listening to opponents
.03
.01#
.01
.05#
Partisanship
.08***
.01**
.01#
.10***
Interest in instruction
.05*
.02**
.11***
.18***
# p < .10; * p < .05; ** p <.01; *** p < .001. N = 497 dyads.

Note. The first column reports the amount of variance accounted for 
by parent ethnicity, gender, SES, religious group membership, voting 
history, student grade in school at T1, and student grades earned, 
which were entered simultaneously in the first equation. The second 
column reports the incremental variance attributed to student 
exposure to Kids Voting, which was entered in the second equation. 
The third column reports the incremental variance attributed to 
student-parent discussion, which was entered in the third equation. 
The fourth column reports the total variance explained.

Table 3: Long-term Effects of Civic Parenting on Students 
(Hierarchical Regression)


T1 Measures


Demographics
R2




Kids Voting
_R2



Student  outcome
_R2

T2 Measure

Civic parenting _R2





Total R2
T2 Outcomes
Media use & cognition
    News attention
.03
.06***
.12***
.03**
.24***
    Information integration
.06#
.07***
.13***
.04**
.30***
    Knowledge
.10**
.04**
.07***
.02*
.23***
    Issue salience
.01
.05***
.02*
.03*
.11***
Discussion
    With parents
.06#
.04***
.17***
.04***
.31***
    With friends
.09**
.06***
.14***
.03**
.32***
    Size of network
.05
.10***
.07***
.00
.22***
Deliberative habits
    Willingness to disagree
.09**
.10***
.19***
.03**
.41***
    Listening to opponents
.09**
.04**
.05***
.03**
.21***
Partisanship
.09**
.03*
.19***
.01*
.32***
Civic participation
    Campus politics
.03
.02*
--1
.00
.05*
    Volunteering
.07*
.02*
--
.00
.09*
    Intention to vote
.12***
.02*
--
.00
.14***
# p < .10; * p < .05; ** p <.01; *** p < .001. N = 288 dyads.

Note. The first column reports the amount of variance accounted for 
by student ethnicity, gender, grade in school at T1, grades earned at 
T1, religious group membership, SES, and the parents' voting history, 
which were entered simultaneously in the first equation. The second 
column reports the incremental variance attributed to Kids Voting 
exposure at T1, which was entered in the second equation. The third 
column reports the incremental variance attributed to the T1 
dependent variable. The fourth column reports the incremental 
variance attributed to civic parenting at T2. The fifth column 
reports the total variance explained.

1 Not measured at T1.