Brandeis University Maurice and Marilyn Cohen Center for Modern Jewish Studies

Transcription

1 Brandeis University Maurice and Marilyn Cohen Center for Modern Jewish Studies Young Adults and Jewish Engment: The Impact of Taglit-Birthright Israel Leonard Saxe Shira Fishman Michelle Shain Graham Wright Shahar Hecht November 2013

2 Young Adults and Jewish Engment 2013 Brandeis University Maurice and Marilyn Cohen Center for Modern Jewish Studies Brandeis University Mailstop 014 Waltham, MA The Cohen Center for Modern Jewish Studies, founded in 1980, is dedicated to providing independent, high quality research on issues related to contemporary Jewish life. The Cohen Center is also the home of the Steinhardt Social Research Institute (SSRI). Established in 2005, SSRI uses innovative research methods to collect and analyze socio-demographic data on the Jewish community.

3 Young Adults and Jewish Engment ii Acknowledgments We gratefully acknowledge the support of Taglit-Birthright Israel, an organization that considers independent research a critical element of its program development. We are especially appreciative for the help of Taglit s educational and research consultants, Prof. Barry Chazan and Prof. Gil Troy. We also thank Taglit s executive leadership: Gidi Mark (CEO), Prof. Ada Spitzer (Vice President), and Dr. Zohar Raviv (Director of Education). Appreciation is also expressed to the supporters of the Cohen Center for Modern Jewish Studies at Brandeis University. This project could not have been carried out without the assistance of our research team at Brandeis University. We are appreciative and grateful for their efforts. Working diligently to ensure the highest care going into our data collection was Natanya Cohen. Calling supervisors Rachel Bernstein and Yves Bruno supervised many night and weekend shifts and responded to various questions from respondents. Ben Lefebvre offered technical support for the survey software and the survey manment. Dina Bleckman and Ellie Aitan were responsible for data coding. Theadora Fisher assisted with data cleaning and initial analysis. We acknowledge Eitan Melchior, z"l, for his help in the initial sts of the study. We miss him greatly. We also thank our colleagues Prof. Charles Kadushin and Prof. Ted Sasson, who provided critical review and insights. Deborah Grant and Joshua Davidson turned our words and charts into a physical report. And none of this work would be possible without the support of our colleagues, Masha Lokshin and Gloria Tessler, who man day-to-day operations. Our team of telephone interviewers was critical to project implementation. We acknowledge their work with appreciation. Callers included: Nihan Celiktas, Carmelle Eloi, Nia Fogelman, Lauren Fox, Rachel Gordon, Rebecca Grossman, Yakov Israel, Vicky Negus, Dara Rosenkrantz, Danielle Spencer, Daniel Shpolyansky, Shira Straus, Alie Tawah, Annie Torres, and Kristina Yepez. Our gratitude to reviewers of this report notwithstanding, the authors take full responsibility for the design, conduct, and results of the study.

4 i Young Adults and Jewish Engment

5 Young Adults and Jewish Engment ii Table of Contents List of Figures and Tables... iii Executive Summary... 1 Introduction... 3 Study Population... 9 Taglit s Impact on Attitudes Jewish Engment on the College Campus Post-College Jewish Engment Discussion Notes References Appendix A: Methodology Appendix B: Attitudinal Impact of Taglit for Undergraduates and Non-Undergraduates Appendix C: Tables... 47

6 iii Young Adults and Jewish Engment List of Figures and Tables Figure 1: Taglit applicants who were undergraduates at the time of application by cohort... 9 Figure 2: Age at time of survey by undergraduate status Figure 3: Most intense form of ritual practice during high school by undergraduate status Figure 4: Most intense form of formal Jewish education by undergraduate status Figure 5: Jewish denomination raised by undergraduate status Figure 6: Feelings of Jewish connection by Taglit participation (undergraduate and post-college) Figure 7: Importance of being Jewish and forming a Jewish family by Taglit participation Figure 8: Being invited to activities sponsored by campus-based Jewish organizations by Taglit participation (undergraduates) Figure 9: Participating in activities sponsored by campus-based Jewish organizations by Taglit participation (undergraduates) Figure 10: Attending events sponsored by a Jewish organization (undergraduates) Figure 11: Attending events sponsored by a Jewish organization (post-college) Figure 12: Donating to Jewish causes and volunteering under Jewish sponsorship by Taglit participation (post-college) Figure 13: Engment in Jewish religious life by Taglit participation (post-college) Figure 14: Awareness of NEXT Shabbat by group (Taglit participants) Figure 15: Awareness of NEXT Shabbat by Taglit cohort (Taglit participants) Figure 16: Attending a NEXT Shabbat meal by Taglit cohort (Taglit participants)... 23

7 Young Adults and Jewish Engment 1 Executive Summary The engment of Jewish young adults with their herit and communal institutions has been an ongoing concern of the Jewish community. In North America, a host of initiatives have been developed to eng this population, the largest of which is Taglit-Birthright Israel. Launched in 1999, Taglit s reach is striking. By 2013 more than 350,000 Jewish young adults from around the world had participated in a 10-day educational trip to Israel under the program s auspices. A substantial body of prior evaluative research documents the effects of the program, in particular on participants attitudes toward Israel and Jewish life. When compared to similar others who did not participate, Taglit alumni are more likely to feel a stronger connection to Israel and to the worldwide Jewish community. They are also more likely to consider it very important for them to marry someone Jewish and raise Jewish children. In the short-term, the trip leads to modest behavioral changes, particularly among college-d participants, who are more likely to eng in Hillel activities and take classes focusing on Israel or on Jewish subjects. In the longterm, there is evidence that the program has significant behavioral impact. Recent data from studies of Taglit alumni who are now (on aver) over 30 years old, show that participants, as compared to nonparticipants, are more likely to be married to another Jew, belong to synagogues, celebrate Shabbat, and make charitable donations to Jewish or Israeli causes. The present study examines the short- and moderate-term impact of Taglit. This analysis focuses on the emerging young adult experience the period of time immediately after the trips and the first few years that follow. The study looks at the effects of Taglit on participants, and in particular, on their attitudes toward Jewish life and participation in Jewish activities. The present findings are based on data collected in an online survey (spring 2011) from a random stratified sample of American Taglit applicants in the cohorts (N=7,662). Findings: Six months to four years after the program, Taglit participants felt more connected to Israel and the Jewish community than did nonparticipants. Among all respondents, those with stronger Jewish educational backgrounds reported deeper feelings of connection to Israel and Jewish life, while those who were older reported weaker feelings of connection. Controlling for the impact of Jewish background and, however, participation in Taglit still had a measurable impact on attitudes. Most undergraduate respondents were invited by at least one campus-based Jewish organization to participate in its activities; participants were more likely than nonparticipants to have been invited and to have participated in at least one

8 2 Young Adults and Jewish Engment activity. Most post-college respondents participated in at least one program sponsored by a Jewish organization in the year prior to the survey. Parties and social gatherings were the most popular activities. Taglit participants were generally more likely to participate than nonparticipants. Among post-college respondents who donated to any cause in the past year, Taglit participants were more likely to have donated at least some of their money to a Jewish cause. Similarly, among those who reported volunteering in the past 12 months, Taglit participants were more likely than nonparticipants to have volunteered under Jewish auspices. Despite low levels of engment overall, post-college Taglit participants, as compared to nonparticipants, were more likely to have had a special Shabbat meal on the Friday night prior to the survey, to be synagogue members, and to have attended Jewish religious services in the past month. Forty-three percent of Taglit participants had heard of NEXT Shabbat before being asked about it in the survey. Overall, 13 percent of Taglit participants had attended a NEXT Shabbat meal, either as a guest or a host, including six percent having attended a NEXT Shabbat meal more than once. About half of all those who attended a NEXT Shabbat meal had hosted a meal at least once; the rest had been guests. Consistent with previous research, the present findings indicate that Taglit has a robust impact on attitudes toward Israel and the Jewish community, and a broad, but modest, impact on engment with Jewish life. The level of engment with Jewish life among young adults, including engment with Taglit follow-up programs, however, is relatively low. Given participants expressed strong connection to their Jewish identity, one might expect that Jewish young adults, and Taglit alumni in particular, would be more involved in Jewish activities. But even when attitudes and behaviors are aligned, there are additional barriers to participation. Taglit alumni have a heightened sense of their Jewish identity and it seems likely, as they come into full adult roles, that they will be more highly involved than their peers who did not have the experience. At the same time, the likelihood that they will participate will inevitably be mediated by the ability of Jewish communal institutions to serve their needs. The gap that currently exists between Taglit alumni s positive attitudes and their actual behavior may only be breached by institutions that are able to adapt to the concerns and structural constraints associated with emerging adulthood. Exactly how this can be done is beyond the scope of the present study. It is clear, however, that those organizations that are effective will address a desire for community through the loose connective structures most appropriate for this population s st of life.

9 Young Adults and Jewish Engment 3 Introduction The engment of Jewish young adults with their herit and Jewish institutions has, over the last two decades, been a central issue for the Jewish community (see, e.g., Cohen & Kelman, 2008; Kotler-Berkowitz, et al., 2004). Concerned with assimilation pressures and intermarri, the Jewish community has developed a number of initiatives designed to bolster engment with Jewish life among Jewish young adults and to foster connections with Israel. Taglit- Birthright Israel is the largest of these programs (Saxe & Chazan, 2008) and, since its inception in 1999, has allowed more than 350,000 Jewish young adults from around the world to visit Israel on 10-day educational trips. More than two-thirds of Taglit s participants are from the largest Jewish diaspora community, the United States (Taglit-Birthright Israel, 2013). Taglit, and in particular its North American groups, has been the focus of an extensive program of independent evaluative research. 1 Several dozen studies have now been conducted with multiple cohorts of Taglit applicants and participants to assess the program s impact. Typically, these studies employ quasi-experimental designs that compare participants and equivalent nonparticipant applicants both pre- and posttrip (see, e.g., Saxe, et al., 2004; Saxe, Sasson, & Hecht, 2006). A variety of measures have been used to assess the program s impact three months to nearly 12 years after participation in Taglit. Impact has been assessed by focusing on participants attitudes and feelings about their Jewish identity, Israel, and the Jewish community. In addition, researchers have examined behaviors related to Jewish life such as participation in Jewish organizations on and off campus, observance of Shabbat and holidays, and giving to Jewish causes. Taglit specifically targets young people in the st of life termed emerging adulthood (see Arnett, 2004), a period characterized by questioning, seeking, and developing one s identity. Findings from evaluation research studies regarding Taglit s impact on attitudes have been highly consistent across and within cohorts over time. For example, multiple studies have documented that when participants are compared to similar others who did not participate on the trip, participants are more likely to feel a stronger connection to Israel and to the worldwide Jewish community (Saxe, Kadushin, Kelner, Rosen, & Yereslove, 2002; Saxe, et al., 2008; Saxe, Sasson, Phillips, Hecht, & Wright, 2007). Other studies have provided evidence that these differences persist over time, from one to three years after the trip (Saxe, et al., 2004; Saxe, et al., 2006) and even five to 11 years after the trip (Saxe, et al., 2009; Saxe, Sasson, et al., 2011; Saxe, et al., 2012). Evaluation studies also indicate that participants are more likely to report a stronger commitment to Jewish continuity. In both the short and the long term, participants are more likely to state that it is very important to them to marry someone Jewish and to raise their children Jewish (Saxe, et al., 2004; Saxe, et al., 2002; Saxe, et al., 2008; Saxe, et al., 2009; Saxe, et al., 2006; Saxe, Sasson, et al., 2011; Saxe, et al., 2012).

10 4 Young Adults and Jewish Engment Although evidence of changes in attitudes regarding connection to Israel and toward Jewish life suggests fairly dramatic program impact, the evidence of change in behavior, as a result of participation in Taglit, has been more modest. The findings, however, interact with and the developmental st of the participant. Thus, among participants who are undergraduate college students, there is evidence that Taglit increases engment with Hillel (Saxe, et al., 2008) and leads to increases in the likelihood of taking Israel or Jewish Studies courses (Saxe, et al., 2002). At the same time, for post-college students, there is not much evidence of behavioral change (Saxe, et al., 2008; Saxe, et al., 2007). In a 2009 study, many post-college Taglit alumni describe Jewish life in their communities as inaccessible and unappealing (Chertok, Sasson, & Saxe, 2009). But it is not clear to what extent the findings are a function of the lack of opportunities for this group. Recently, however, as large numbers of participants from the early years of the program assume full adult roles, it has been possible to explore Taglit s impact on a wider variety of potential behavioral measures: marri, childrearing, and adult engment with the Jewish community. Since 2009, four waves of a longitudinal study, with a panel of more than 3,000 individuals who applied to Taglit between 2001 and 2006, have been conducted in order to track Jewish engment as a function of Taglit. The findings indicate substantial program impact. Taglit participants, as compared to nonparticipants, are more likely to be married to another Jew, be synagogue members, celebrate Shabbat, and make charitable donations to Jewish or Israeli organizations or causes (Saxe, 2013; Saxe, et al., 2009; Saxe, Sasson, et al., 2011; Saxe, et al., 2012). Thus, the evidence indicates that over time Taglit has been successful in strengthening Jewish identity, as well as the likelihood of Jewish family formation and participation in Jewish life. The findings indicate that the program is effective for participants with a variety of Jewish backgrounds (i.e., participants with no formal Jewish education, those with several years of Jewish supplementary schooling, and even those with day school experience). Two issues underlie the present research. The first question looks at whether the effects of Taglit observed in previous studies are evident in recent cohorts. Taglit has expanded over time and, since 2008, the number of participants has doubled. In addition, the characteristics of the applicant pool have changed compared to its first eight years. Whereas the majority of Taglit participants in the cohorts were s 21 or younger (Saxe, et al., 2009), by summer 2008, the majority of participants were s 22 or older (Saxe, et al., 2008). There has also been an increase in the proportion of participants with little or no Jewish background. In the cohorts, 37 percent of participants kept kosher at home during high school (Saxe, et al., 2009), whereas only 13 percent of the summer 2008 cohort did so (Saxe, et al., 2008). Thus, the report examines the patterns of attitude change and the levels of Jewish engment among recent Taglit cohorts. The second question addresses Jewish engment among Taglit applicants in their emerging adulthood years a period of time when most of these young adults are not in a secure job or professional role, are highly mobile, and are in the process of developing the social networks that will stabilize only later in life (Arnett, 2004). Specifically, the

11 Young Adults and Jewish Engment 5 analysis focuses on Taglit s role in these young adults Jewish lives and on its ability to impact Jewish engment immediately after the trip and the first few years that follow. Although there is unequivocal evidence of Taglit s impact on the attitudes of participants of diverse Jewish backgrounds, the trajectory of Jewish involvement is somewhat puzzling. Upon immediate return, strong attitudinal changes are not matched by participants levels of engment with the community. Yet, substantial change over the long term (up to a decade or more after participation in Taglit) has been documented, in particular, in terms of marital decision-making. The findings described in the present study derive from data collected in a survey of the cohorts that applied to the program between 2007 and The analyses presented serve to help us better understand what Jewish involvement looks like at this particular developmental st and the ways in which Jewish engment evolves over time. Thus, in addition to examining overall levels of Jewish engment, the report examines how Jewish involvement manifests itself in early young adulthood, both among those in college and those postcollege. Finally, this study reviews the impact of recent efforts to eng Taglit alumni.

12 6 Young Adults and Jewish Engment

13 Young Adults and Jewish Engment 7 Method The present analyses focus on data collected in an online survey of eligible North American applicants to the winter through summer 2010 Taglit trips. A stratified random sample of 67,400 eligible applicants was invited to participate in an online survey that was in the field during spring Survey respondents completed the survey six months to four years after they applied for the trips. A random group (N=3,000) was selected for intensive phone follow-up in order to account for any bias due to non-response. The relatively low response rate in the overall sample did not appear to contribute to significant nonresponse bias. This finding is based on an analysis comparing the overall sample to the special follow-up sample which achieved a much higher response rate. 2 The response rate for the intensive follow-up sample was 48 percent (53 percent for participants, 40 percent for nonparticipants), while the overall response rate was 11 percent (12 percent for participants, 10 percent for nonparticipants. For the purposes of this report, analysis was limited to respondents who resided in the United States at the time of the survey. Full details of sample selection and survey administration are presented in Appendix A. The analytic paradigm of this study compares Taglit participants to similar others who applied to the program but did not go. Comparisons are valid if there are no known pre-existing differences between participants and nonparticipants or, in the event that there are differences that might affect outcomes, the differences can be accounted for statistically. Historically, assignment to the program was practically random (Saxe, et al., 2008), but this changed in recent years. Beginning in winter , Taglit implemented a pre-registration system in order to give preference to applicants who had applied to the trip before. Preference is also given to older applicants who will out of the program. The central findings presented in this report, therefore, use logistic regression (either binary, ordinal, or multinomial, depending on the nature of the dependent variable) to control for factors that are related to participation in Taglit:, Jewish background, 3 and, for non-undergraduates, engment with Jewish organizations in college (e.g., Hillel). By controlling for these factors, the impact of Taglit participation can be isolated from factors associated with being selected to participate. The figures presented are based on predicted probabilities derived from the regression models, holding the control variables at their mean values.

14 8 Young Adults and Jewish Engment

15 Young Adults and Jewish Engment 9 Study Population Undergraduates versus Others Forty-seven percent of Taglit participants in the current sample went on the program while they were undergraduates. The proportion of Taglit participants who were undergraduates declined in both winter and summer trips over the four years of the study period (Figure 1). This was most dramatic for the summer trips, where the proportion of undergraduates declined by one-third from summer 2007 to summer 2010, from 47 percent to only 31 percent. The pattern of decreasing participation by undergraduates means that fewer Taglit participants return to college campuses after their trips. Because post-college alumni face a very different set of opportunities for Jewish engment than undergraduate alumni, this shift in the makeup of participants is critical to understanding the impact of Taglit on post-trip Jewish engment. At the time of the survey, onequarter (24 percent) of respondents were undergraduates, while three-quarters (76 percent) were not. 4 Demographic and Jewish Background Characteristics Age and gender - Most undergraduate respondents were between 20 and 22 years old at the time of the survey, whereas most post-college respondents were in their midtwenties (Figure 2). Women were slightly overrepresented in the population: 54 percent of undergraduates and 57 percent of post-college respondents were female. Figure 1: Taglit applicants who were undergraduates at the time of application by cohort* 100% 80% 60% 40% 20% 60% 47% 57% 40% 48% 44% 55% 31% 0% Winter Summer 2007 Winter Summer 2008 Winter Summer 2009 Winter Summer 2010 * Estimated proportions.

16 10 Young Adults and Jewish Engment Family Jewish ritual practice during high school - Undergraduate and post-college respondents did not differ dramatically on measures of Jewish background. Close to half of the respondents reported that during their high school years their families celebrated both Hanukkah and Passover. About one in five reported that their families also regularly lit Shabbat candles. Undergraduates were more likely to come from families that kept kosher at home (23 percent vs. 16 percent for nonundergraduates) (Figure 3). Figure 2: Age at time of survey by undergraduate status* 100% 80% 60% Non-undergraduates Undergraduates 40% 40% 24% 19% 20% 16% 14% 12% 12% 11% 11% 10% 7% 4% 6% 0% 1% 2% 2% 2% 1% 1% 1% 2% 1% 0% 0% 0% 0% * Estimated proportions. Figure 3: Most intense form of ritual practice during high school by undergraduate status* 100% 80% 6% 3% 8% 12% 60% 47% 40% 19% 20% 16% 0% Non-undergraduates * Estimated proportions. 47% 19% 23% Undergraduates None Hannukkah Seder Shabbat Kosher

17 Young Adults and Jewish Engment 11 Formal Jewish education growing up - Respondents had varying levels of formal Jewish education (Figure 4). About one-fifth had no formal Jewish education, while another fifth attended full-time Jewish day schools. The rest of the respondents had some supplemental Jewish education, such as Hebrew school or Sunday school. While undergraduates and non-graduates were similar in their Jewish education levels, undergraduates were more likely to have attended Jewish day school. Parents religion - Almost three-quarters of respondents had two Jewish parents (73 percent of undergraduates and 71 percent of post-college respondents). Jewish denomination - A plurality of respondents (36 percent of both undergraduates and post-college respondents) were raised Reform or Reconstructionist, and one-quarter were raised Conservative (25 percent of both undergraduates and post-college respondents). Another quarter of respondents were raised secular/culturally Jewish or Just Jewish (27 percent of undergraduates and 24 percent of postcollege respondents). A small number of respondents were raised Orthodox (eight percent of undergraduates and four percent of post-college respondents), and the rest were raised no religion or something else (Figure 5). Jewish background index - The analyses presented in this report control for a single index of childhood Jewish background, which combines the different measures of Jewish background reported above: formal Jewish education, high school ritual practice, parental inmarri, and being raised Orthodox. 5 Figure 4: Most intense form of formal Jewish education by undergraduate status* 100% 22% 17% 80% 60% No formal Jewish education 40% 63% 63% No day school, but some supplementary Jewish education Some day school 20% 15% 20% 0% * Estimated proportions. Non-undergraduates Undergraduates

18 12 Young Adults and Jewish Engment Figure 5: Jewish denomination raised by undergraduate status* 100% 3% 2% 5% 5% 80% 27% 24% Other No religion 60% 40% 36% 36% Secular/culturally Jewish, Just Jewish Reform, Reconstructionist Conservative 20% 25% 25% Orthodox 0% 4% Non-undergraduates 8% Undergraduates * Estimated proportions.

19 Young Adults and Jewish Engment 13 Taglit s Impact on Attitudes Consistent with previous research, participation in Taglit has an impact on attitudes associated with Israel and the Jewish community. Indeed, participants from the Taglit cohorts again demonstrate stronger feelings of connection to Israel and the Jewish community than nonparticipants (Figure 6). Among participants and nonparticipants, Jewish background was naturally associated with stronger feelings of connection, and was associated with weaker feelings of connection. However, controlling for the impact of Jewish background and, Taglit still had a measurable impact on participants. Taglit s greatest impact is observed on connection to Israel, where participants were 2.5 times as likely to feel very much connected to Israel. Higher levels of connection to the worldwide Jewish community, to the local Jewish community, and to Jewish peers were observed among participants, but the effect size was smaller. Taglit s impact on attitudes presented here does not distinguish between college and post-college respondents because Taglit s impact is very similar for both groups. 6 Figure 6: Feelings of Jewish connection by Taglit participation (undergraduate and post-college)* 100% 80% 13% 4% 23% 10% 32% 6% 25% 28% 23% 7% 5% 28% 22% 60% 40% 42% 37% 36% 37% 32% 30% 39% 39% 20% 0% 31% 14% 35% 23% 31% Nonparticipants Participants Nonparticipants Participants Nonparticipants Participants Nonparticipants Participants Connection to Israel Part of a worldwide Jewish community 27% 13% 30% 17% Part of local Jewish community 26% 33% Connected to Jewish peers Very much Somewhat A little Not at all * Predicted probabilities from ordinal logistic regression models controlling for Jewish background and.

20 14 Young Adults and Jewish Engment Figure 7: Importance of being Jewish and forming a Jewish family by Taglit participation* 100% 80% 25% 17% 8% 16% 5% 5% 3% 11% 14% 19% 60% 23% 19% 26% 22% 33% 29% 29% 40% 28% 20% 25% 35% 49% 62% 43% 54% 0% Nonparticipants Participants Nonparticipants Participants Nonparticipants Participants Marry someone Jewish** Raise children Jewish*** Being Jewish Very important Somewhat important A little important Not important * Predicted probabilities from ordinal logistic regression models controlling for Jewish background and. ** Excludes the seven percent respondents who are currently married. *** Excludes the eight percent of respondents who have children. Taglit participants were more likely to rate being Jewish as very much important to their lives, relative to nonparticipants, suggesting the salience of Judaism to their identity. They also demonstrated a stronger desire to form a Jewish family (Figure 7). Participants were 40 percent more likely to say that marrying someone Jewish was very important to them and 27 percent more likely to say that raising their children Jewish was very important to them. These attitudes may eventually translate into behavior. Long-term follow-up research has documented that between six and 11 years after the trip Taglit participants are significantly more likely to be married to a Jew as compared to nonparticipants (Saxe, et al., 2009; Saxe, Phillips, et al., 2011; Saxe, Sasson, et al., 2011; Saxe, et al., 2012).

21 Nonparticipants Participants Nonparticipants Participants Nonparticipants Participants Nonparticipants Participants Young Adults and Jewish Engment 15 Jewish Engment on the College Campus Twenty-four percent of respondents were undergraduates at the time of the survey. The majority (85 percent) went on the trip during their college years. Just as Jewish background and being younger were associated with more positive Jewish attitudes, these two factors were also associated with increased Jewish engment for undergraduates. Controlling for the impact of Jewish background and, Taglit had a small but consistent impact on participants engment with campusbased Jewish groups such as Hillel, as well as on participation in a wide variety of Jewish activities. Campus-Based Jewish Groups Undergraduate respondents were asked whether they had been invited to activities sponsored by Hillel, Chabad-Lubavitch, a Jewish fraternity or sorority (e.g., AEPi), or another campus-based Jewish group in the past 12 months. They were then asked whether they actually engd in any such activities. Overall, most respondents (85 percent) were invited by at least one campus-based organization to participate in its activities. However, Taglit participants were significantly more likely to be invited by these groups (Figure 8). Figure 8: Being invited to activities sponsored by campus-based Jewish organizations by Taglit participation (undergraduates)* 100% 80% 11% 60% 13% Once or twice 11% 3 or more times 40% 20% 0% 52% 73% 9% 34% 43% 13% 25% 14% 32% 4% 11% 5% 18% Hillel Chabad Jewish fraternity or sorority Other campus-based Jewish group * Predicted probabilities from multinomial logistic regression models controlling for Jewish background and.

22 16 Young Adults and Jewish Engment The majority of respondents (67 percent) also participated in at least one activity sponsored by a campus-based Jewish group in the past 12 months. Taglit participants were more likely than nonparticipants to participate in activities sponsored by these organizations (Figure 9). Those who were invited by a campus Jewish group to participate in activities were also much more likely to participate in those activities than those who were not invited. Additional analysis not presented here suggests that Taglit participants greater likelihood of participating in Jewish activities on campus is largely, perhaps even exclusively, due to their greater likelihood of receiving an invitation. Among those who were not invited, Taglit participants were no more likely to participate than nonparticipants. Participation in Activities sponsored by Jewish Organizations To gauge the types of activities offered by Jewish organizations on campus that are most popular and well attended, survey respondents were asked whether they had engd in particular activities with Jewish content or Jewish sponsorship including: (1) a party, happy hour, or social gathering; (2) a lecture, speaker, or class; (3) a cultural event, such as a concert of film screening; Figure 9: Participating in activities sponsored by campus-based Jewish organizations by Taglit participation (undergraduates)* 100% 80% 60% 40% 21% Once or twice 3 or more times 24% 20% 38% 13% 14% 12% 15% 0% 18% 13% 17% Nonparticipants Participants Nonparticipants Participants Nonparticipants Participants Nonparticipants Participants Hillel Chabad Jewish fraternity or sorority Other campus-based Jewish group 13% 17% 3% 4% 5% 8% * Predicted probabilities from multinomial logistic regression models controlling for Jewish background and.

23 Nonparticipants Participants Nonparticipants Participants Nonparticipants Participants Nonparticipants Participants Nonparticipants Participants Young Adults and Jewish Engment 17 Figure 10: Attending events sponsored by a Jewish organization (undergraduates)* 100% 80% 60% 19% 40% 20% 0% 21% 29% 42% 17% 22% 20% 34% 19% 14% 21% 27% 10% 10% 14% 20% 3% 7% Once More than once 4% 12% Party, happy hour, or social gathering Lecture, speaker, or class Cultural event Social justice / activism event or activity Another type of activity * Predicted probabilities from multinomial logistic regression models controlling for Jewish background,, and Jewish engment in college. (4) a social justice/activism event or activity; or (5) another type of event or activity. The majority of undergraduate respondents (69 percent) had participated in at least one activity in the past twelve months. The most popular type of activities were parties and social gatherings sponsored by Jewish organizations, attracting about half of undergraduate respondents at least once in the 12 months prior to the survey. In general, Taglit participants were likely to attend these activities with greater frequency compared to nonparticipants (Figure 10).

24 18 Young Adults and Jewish Engment

25 Young Adults and Jewish Engment 19 Post-College Jewish Engment This section explores the Jewish engment of the 76 percent of respondents who were not undergraduates at the time they were surveyed. 7 Removed from the campus community, Jewish young adults face a different landscape of Jewish opportunities. Many Jewish organizations and congregations cater their services and programming to families with young children (Sheskin & Kotler-Berkowitz, 2007). While there has been recent investment in programming for single Jewish young adults, many barriers to engment remain, including difficulties in finding opportunities and issues of cultural fit (Chertok, et al., 2009; Cohen & Kelman, 2007). Below, Taglit s impact on Jewish involvement is examined, focusing on participation in activities sponsored by Jewish organizations, Jewish congregational membership, attendance at religious services, Shabbat and holiday observances, giving to Jewish causes, and volunteering under Jewish sponsorship. Not surprisingly, Jewish background and engment with campus Jewish life during college were positive predictors of Jewish engment for post-college respondents, while had an inconsistent effect. Controlling for the impact of Jewish background, college Jewish engment, and, Taglit was a predictor of increased Jewish engment across a number of measures. The end of this section examines the role of the largest initiative targeting post-college Taglit alumni NEXT Shabbat in facilitating Jewish engment. Participation in Jewish Activities Whereas almost 70 percent of undergraduates participated in at least one activity sponsored by a Jewish organization in the year prior to the survey, only half of post-college respondents did so. As with undergraduates, among post-college respondents, the most popular of these activities were parties and social events (around 40 percent of respondents reported having participated in such activities in the past 12 months). Taglit participants were generally more likely to participate in Jewish-sponsored activities than nonparticipants (Figure 11). While the differences are not large, they are statistically significant. Giving and Volunteering Respondents were asked whether, in the year prior to the survey, they had made any charitable contributions or engd in any volunteer activities. Among those who donated to any cause in the past year, Taglit participants were more likely to have donated at least some of their money to a Jewish cause (Figure 12). Similarly, among those who reported volunteering in the past 12 months, Taglit participants were more likely than nonparticipants to have volunteered under Jewish auspices (Figure 12). 8 Jewish Religious Engment Post-college Taglit participants were somewhat more likely to eng in Jewish

26 Nonparticipants Participants Nonparticipants Participants Nonparticipants Participants Nonparticipants Participants Nonparticipants Participants 20 Young Adults and Jewish Engment Figure 11: Attending events sponsored by a Jewish organization (post-college)* 100% 80% 60% 40% 20% 0% 16% 17% 18% 24% 11% 13% 12% 17% 14% 15% 9% 13% Once More than once 8% 9% 4% 8% 4% 4% 5% 6% Party, happy hour or social gathering Lecture, speaker or class Cultural event Social justice / activism event or activity Another type of activity * Predicted probabilities from multinomial logistic regression models controlling for Jewish background,, and Jewish engment in college. Figure 12: Donating to Jewish causes and volunteering under Jewish sponsorship by Taglit participation (post-college)* 100% 80% 60% Nonparticipants Participants 40% 20% 0% 21% 27% Donated to Jewish or Israeli organizations or causes** 14% 20% Volunteered under Jewish sponsorship*** * Predicted probabilities from binary logistic regression models controlling for Jewish background,, and Jewish engment in college. ** Limited to respondents who made any charitable contributions. *** Limited to respondents who engd in any volunteer activities.

27 Young Adults and Jewish Engment 21 Figure 13: Engment in Jewish religious life by Taglit participation (post-college)* 100% 80% 60% Nonparticipants Participants 40% 20% 0% religious life than nonparticipants. Participants were significantly more likely to have had a special Shabbat meal on the Friday night before taking the survey, to be synagogue members, and to have attended Jewish religious services in the past month (Figure 13). On the other had, Taglit participants were not significantly more likely to have attended a Passover seder in the previous year. NEXT Shabbat 14% 18% Had special Shabbat meal last Friday night NEXT Shabbat is a national program that endeavors to boost the Jewish engment of post-college Taglit alumni. 9 The program provides monetary reimbursement and educational resources for Taglit participants who invite friends into their homes for a Shabbat meal. Launched in July 2008, NEXT Shabbat has a low threshold for involvement traditional Shabbat rituals and kosher food are not required, and guests 27% 22% 25% Synagogue member do not have to be Jewish. Shabbat meals can happen anywhere in the United States, allowing for broad participation around the country. This section of the report will describe the level and patterns of participation in NEXT Shabbat among Taglit participants. While NEXT is targeted specifically at post-college Taglit alumni, undergraduates appear to be aware of and participate in NEXT at roughly the same rate as non-undergraduates. Consequently, the analyses below will include Taglit participants from both groups. Awareness of NEXT Shabbat 32% Attended Jewish religious services in past month * Predicted probabilities from binary logistic regression models controlling for Jewish background,, and Jewish engment in college Overall, 43 percent of Taglit participants had heard of NEXT Shabbat before being asked about it in the survey. Those participants s 25 and older were somewhat more likely to have heard of the program than younger participants (Figure 14). In addition, those who went on a Taglit trip in

28 22 Young Adults and Jewish Engment Figure 14: Awareness of NEXT Shabbat by group (Taglit participants)* 100% 80% 60% 40% 20% 43% 38% 48% 0% Age Age Age 25+ * Estimated proportions. Figure 15: Awareness of NEXT Shabbat by Taglit cohort (Taglit participants)* 100% 80% 60% 40% 58% 57% 57% 55% 20% 32% 31% 33% 44% 0% Winter Summer 2007 Winter Summer 2008 Winter Summer 2009 Winter Summer 2010 * Estimated proportions.

29 Young Adults and Jewish Engment or 2010 were more likely to have heard of the program than those who went on the trip in 2007 or 2008 (Figure 15). Analyzing multiple factors that impact awareness of the program reveals that in addition to being older and having gone on a trip after 2008, women and those with more intense Jewish backgrounds were also more likely to have heard of the program. Participation in NEXT Shabbat Taglit participants who had heard of NEXT Shabbat were asked how many times in the past 12 months they had hosted or been a guest at a NEXT Shabbat meal. Overall, 13 percent of Taglit participants had attended a NEXT Shabbat meal, either as a guest or a host, with six percent having attended a NEXT Shabbat meal more than once. About half of all those who attended a NEXT Shabbat meal had also hosted a meal at least once; the rest had been guests. The time elapsed since a participant s Taglit trip is associated with the likelihood of attending a NEXT Shabbat meal. Those who went on a Taglit trip in 2009 or 2010 were significantly more likely to have attended a NEXT Shabbat meal in the past 12 months than those who went in 2007 or 2008 (Figure 16). A binary logistic regression model of NEXT participation indicates that several factors are associated with participation in NEXT Shabbat meal. As shown in Figure 16, those who went on the trip in 2009 or 2010 were more likely to have attended a NEXT Shabbat meal. In addition, older Taglit participants and women were somewhat more likely to have attended. Those with more intense Jewish backgrounds were neither more nor less likely to attend a meal. 10 Figure 16: Attending a NEXT Shabbat meal by Taglit cohort (Taglit participants)* 100% 80% 60% 40% Attended or hosted more than once Attended or hosted once 20% 0% 12% 10% 9% 8% 7% 5% 6% 5% 5% 4% 4% 5% 9% 8% 10% 8% Winter Summer 2007 Winter Summer 2008 Winter Summer 2009 Winter Summer 2010 * Estimated proportions

30 24 Young Adults and Jewish Engment

31 Young Adults and Jewish Engment 25 Discussion The present findings document Taglit s impact on participants Jewish identities and Jewish engment post-trip. Consistent with studies of earlier Taglit cohorts, the present study conducted up to three and half years after participation in Taglit demonstrates the program s strong effects on participants attitudes toward their Jewish identity. Participants who took part in a Taglit trip between 2007 and 2010 reported greater feelings of connection to Israel and the Jewish community as compared to nonparticipants and a stronger desire to marry someone Jewish and raise Jewish children. Thus, the program continues to produce positive attitudinal changes among participants even as the program has evolved and matured and the applicant pool has changed. The findings also shed light on the levels of behavioral engment with Jewish life both on and off the college campus. More than two-thirds of undergraduate Taglit applicants were involved in some way in Jewish life on campus in the year prior to the survey. Respondents reported participating in a variety of organizations and activities on campus; Hillel was the most cited campus organization and social gatherings were the most popular activity. Although the overall rates indicate that the majority do not attend these events with much frequency, Taglit has a significant impact on participation in the organizations and activities. Taglit s impact on campus is related to the fact that participation on trips makes alumni more visible to Jewish organizations. These organizations often partner with Taglit, which allows them to more effectively reach out to their target population. Taglit participants were much more likely to be invited to events and were also far more likely to participate. Among post-college respondents, overall levels of engment were significantly lower than those among the undergraduate respondents. Taglit, nevertheless, had a small yet significant impact on participants, with post-college participants being more likely to attend Jewish events, donate to or volunteer with Jewish organizations, have Shabbat meals, join synagogues, and even attend religious services. The analyses used to assess Taglit s impact in these areas also indicated that prior experience with Jewish life on the college campus was a strong predictor of engment post college. Insofar as Taglit increases opportunities for engment in Jewish life on campus, it will indirectly impact Jewish engment later on. To the extent that the majority of current Taglit participants are past their college experience, Taglit may have a lower impact on engment. Overall levels of engment in the flagship program of alumni follow-up, NEXT Shabbat, are also low, but in line with the overall rates of involvement in Jewish life among post-college young adults. Despite the low barriers for participation in the program and the relatively high name recognition of the program (more than a third had heard of it), participation rates are between 16 to 22 percent. Given participants expressed strong connection to their Jewish identity, one might expect that Jewish young adults, and Taglit alumni in particular, would be more

32 26 Young Adults and Jewish Engment involved in Jewish activities. Although Taglit has consistent positive effects, on most measures, the proportion of participants who are engd is less than a majority. Some have argued that engment of Jewish young adults should be measured outside the box of mainstream Jewish institutions (Cohen & Kelman, 2007; Kaunfer, 2010). However, even when considering innovative Jewish initiatives and non-institutionalized Jewish practice, the level of Jewish engment for Taglit alumni, as well as the current generation of which they are a part, is low (Shain, Fishman, Wright, Hecht, & Saxe, in press). To understand the findings, one needs to appreciate the relationship between Jewish identity and behavioral engment. The psychological literature is rich with discussion of the reasons why attitudes do not always predict behavior. From this perspective, a global measure of Jewish identity (e.g., how connected one feels to the Jewish community) needs not be strongly related to actual involvement (i.e., whether one participates in activities or local Jewish institutions). The strength of the attitudebehavior relationship depends on the degree of correspondence (or compatibility) between the attitudes and the behaviors (Ajzen & Fishbein, 1977; Kim & Hunter, 1993; Kraus, 1995). Thus, young adults may have strong feelings about being Jewish, but they may not be in a st of life in which formal synagogue membership or other forms of institutional connection make sense. In addition, real-world situations and opportunities hinder or facilitate individuals ability to actualize their attitudes in the form of specific behaviors. Engment not only requires individual motivation, but also time, resources, and opportunity structures. Some of the issues that lead to a disconnect between individuals Jewish identity and their involvement with the community may be associated with the unsettled nature of emerging adulthood (Arnett, 2004). Most of these young adults do not have a secure job or professional role; as well, they are highly mobile and in the process of developing the social networks that will stabilize only later in life. This suggests that opportunity structures for Jewish engment must be developed that are fluid and meet individuals on their own terms (Chertok, et al., 2009). Taglit alumni have a heightened sense of their Jewish identity and it seems likely, as they come into full adult roles, that they will be more highly involved than their peers who did not have the experience. At the same time, the likelihood that they will participate in Jewish life will inevitably be mediated by the ability of Jewish communal institutions to serve their needs. The gap that currently exists between Taglit alumni s positive attitudes and their actual behavior may only be breached by institutions that are able to adapt to the concerns and structural constraints associated with emerging adulthood. Exactly how this can be done is beyond the scope of the present study. It is clear, however, that the most effective organizations will be the ones which address a desire for community through the loose connective structures most appropriate for this population s st of life.

33 Young Adults and Jewish Engment 27 Notes 1 For a comprehensive list of publications based on findings from the extensive research on Taglit, see: 2 See Appendix A and a detailed analysis in Wright, Fisher, & Saxe, The models used in these reports control for a single index of childhood Jewish background which combines the different measures of Jewish engment reported below, namely: formal Jewish education, high school ritual practice, parental inmarri, and being raised Orthodox. For details of scale construction, see p 12 and Note 5. 4 Among those Taglit participants who were not undergraduates at the time of survey, there was no substantive difference across the various measures of engment between those who went on Taglit during college and those who went after graduating. 5 Because these four variables have different scales of measurement, they could not simply be averd or added together. The Jewish background index therefore added each variable s standard score ( zscore ), which standardizes the mean of each variable at zero and recodes the values of that variable to reflect the standard deviation away from that mean. Loevinger H scalability coefficients were used to determine the optimal scale composition. The Loevinger H coefficient for the adopted scale was For an analysis of college and post-college respondents see Appendix B. 7 The vast majority (93 percent) of non-undergraduates have bachelor s degrees. This section also includes a small number of individuals who never attended or never completed college. 8 Taglit participants and nonparticipants were equally likely to have done any volunteer activities; each had a 67 percent probability of having volunteered (F test, etc.). Taglit is negatively associated with having made any charitable contributions; participants had a 66 percent probability of having donated, compared to a 72 percent probability for nonparticipants (F test, etc.). 9 Canadian alumni are also eligible to participate in NEXT Shabbat. 10 For the regression model, see Appendix C.

34 28 Young Adults and Jewish Engment

35 Young Adults and Jewish Engment 29 References Ajzen, I., & Fishbein, M. (1977). Attitude-behavior relations: A theoretical analysis and review of empirical research. Psychological Bulletin, 84(5), doi: / Arnett, J. J. (2004). Emerging adulthood: The winding road from the late teens through the twenties. Oxford: Oxford University Press. Battaglia, M. P., Izrael, D., Hoaglin, D. C., & Frankel, M. R. (2004). Practical considerations in raking survey data. Paper presented at the American Association for Public Opinion Research, Phoenix, AZ. Chertok, F., Sasson, T., & Saxe, L. (2009). Tourists, travelers, and citizens: Jewish engment of young adults in four centers of North American Jewish life. Waltham, MA: Cohen Center for Modern Jewish Studies, Brandeis University. Cohen, S. M., & Kelman, A. Y. (2007). The continuity of discontinuity: How young Jews are connecting, creating, and organizing their own Jewish lives: Andrea and Charles Bronfman Philanthropies. Cohen, S. M., & Kelman, A. Y. (2008). Uncoupled: How our singles are reshaping Jewish engment. New York: The Jewish Identity Project of Reboot Andrea and Charles Bronfman Philanthropies. Deming, W. E. (1943). Statistical adjustment of data. New York: John Wiley. Deming, W. E., & Stephan, F. F. (1940). On a least squares adjustment of a sampled frequency table when the expected marginals are known. Annals of Mathematical Statistics, 11, Kaunfer, E. (2010). Empowered Judaism: What independent minyanim can teach us about building vibrant Jewish communities. Woodstock, VT: Jewish Lights Publishing. Kim, M.-S., & Hunter, J. E. (1993). Attitude-behavior relations: A meta-analysis of attitudinal relevance and topic. Journal of Communication, 43(1), doi: /j tb01251.x Kotler-Berkowitz, L., Cohen, S. M., Ament, J., Klaff, V., Mott, F., & Peckerman-Neuma, D. (2004). The National Jewish Population Survey : Strength, challenge and diversity in the American Jewish Population. New York: United Jewish Communities. Kraus, S. J. (1995). Attitudes and the prediction of behavior: A meta-analysis of the empirical literature. Psychology Bulletin, 21(1), doi: / Saxe, L. (2013). The Taglit-Birthright Israel generation comes of. Paper presented at the Taglit-Birthright Israel: An Academic Symposium, Jerusalem, Israel. Saxe, L., & Chazan, B. (2008). Ten days of Birthright Israel: A journey in young adult identity. Lebanon, NH: Brandeis University Press/ University Press of New England. Saxe, L., Kadushin, C., Hecht, S., Rosen, M. I., Phillips, B., & Kelner, S. (2004). Evaluating Birthright Israel: Long-term impact and recent findings. Waltham, MA: Cohen Center for Modern Jewish Studies, Brandeis University. Saxe, L., Kadushin, C., Kelner, S., Rosen, M. I., & Yereslove, E. (2002). A mega-experiment in Jewish education: The impact of birthright israel. Waltham, MA: Cohen Center for Modern Jewish Studies, Brandeis University.

36 30 Young Adults and Jewish Engment Saxe, L., Phillips, B., Boxer, M., Hecht, S., Wright, G., & Sasson, T. (2008). Taglit-Birthright Israel: Evaluation of the North American cohorts. Waltham, MA: Cohen Center for Modern Jewish Studies, Brandeis University. Saxe, L., Phillips, B., Sasson, T., Hecht, S., Shain, M., Wright, G., et al. (2009). Generation Birthright Israel: The impact of an Israel experience on Jewish identity and choices. Waltham, MA: Cohen Center for Modern Jewish Studies, Brandeis University. Saxe, L., Phillips, B., Sasson, T., Hecht, S., Shain, M., Wright, G., et al. (2011). Intermarri: The impact and lessons of Taglit-Birthright Israel. Contemporary Jewry, 31(2), doi: /s z Saxe, L., Sasson, T., & Hecht, S. (2006). Taglit-Birthright Israel: Impact on Jewish identity, peoplehood, and connection to Israel. Waltham, MA: Cohen Center for Modern Jewish Studies, Brandeis University. Saxe, L., Sasson, T., Hecht, S., Phillips, B., Shain, M., Wright, G., et al. (2011). Jewish Futures Project: The impact of Taglit-Birthright Israel, 2010 update. Waltham, MA: Cohen Center for Modern Jewish Studies, Brandeis University. Saxe, L., Sasson, T., Phillips, B., Hecht, S., & Wright, G. (2007). Taglit-Birthright Israel evaluation: 2007 North American cohorts. Waltham, MA: Cohen Center for Modern Jewish Studies, Brandeis University. Saxe, L., Shain, M., Wright, G., Hecht, S., Fishman, S., & Sasson, T. (2012). Jewish Futures Project: The impact of Taglit-Birthright Israel: 2012 update. Waltham, MA: Cohen Center for Modern Jewish Studies. Schmitz, C. (2009). LimeSurvey (Version 1.80RC3) [computer program]. Shain, M., Fishman, S., Wright, G., Hecht, S., & Saxe, L. (in press). DIY Judaism: How contemporary Jewish young adults express their Jewish identity. Jewish Journal of Sociology. Sheskin, I. M., & Kotler-Berkowitz, L. (2007). Synagogues, Jewish Community Centers, and other Jewish organizations: Who Joins, who doesn't? Journal of Jewish Communal Service, 82(3). StataCorp. (2009). Stata (Version 10.1) [computer program]. College Station, TX: StataCorp. Taglit-Birthright Israel. (2013). CEO report 2013: Celebrating our bar/bat mitzvah year. Jerusalem: Taglit-Birthright Israel. Werner, J. (2003). QBAL (Version 1.52M) [computer program]. Pittsfield, MA: Jan Werner Data Processing. Wright, G., Fisher, T., & Saxe, L. (2012). Using dual sample surveys to examine the relationship between response rate and bias. Paper presented at the American Association of Public Opinion Research 67th Annual Conference, Orlando, FL.

37 Young Adults and Jewish Engment 31 Appendix A: Methodology Study Design Target population - This report discusses results from a survey of eligible applicants to the Taglit- Birthright Israel program. The program offers free ten-day educational peer trips to Israel. Trips are offered twice a year in the winter and in the summer (each trip season is hereafter referred to as round ). The target population of the survey discussed here is the pool of applicants for the rounds between winter of 2006/2007 and the summer of There were a total of 131,804 eligible applicants to these rounds. 1 Sampling plan- The objective of this study was to survey the entire target population (N=131,804), essentially conducting a failed census. However, it was expected that without extensive follow-up efforts (which would not be feasible for such a large group) the achieved response rate would not be high enough to ensure an unbiased respondent pool. Thus, in order to assess (and potentially correct for) the extent of any bias due to low response rate in the surveyed population, a representative sample (N=3,000) was selected for intensive follow-up with the aim of achieving a sufficiently high response rate for this sub-group. This group is referred to here as the sample and the remainder of the target population is referred to as the frame. Stratification - The target population was stratified by round, participant status, (over/under 25), and gender. To cut down on the number of strata, round was collapsed into year of application so, for example winter 2006/7 and summer 2007 were both treated as 2007 for purposes of stratification. This led to the creation of 32 mutually exclusive strata. The selected sample (N=3,000) corresponds to approximately 2.276% of the total target population and was designed to be perfectly representative of it. Thus, 2.276% of the applicants in each stratum were randomly selected to be included in the sample. The remainder of each stratum was assigned to the frame. For logistical reasons both the sample and the frame were divided into four equally sized and equally representative replicates. Due to the use of random selection and the stratification scheme each of the eight replicates are representative of the entire population with respect to the stratification variables. 1 For applicants who applied to multiple rounds, the latest eligible round of application was chosen. Ineligible and duplicate records were dropped.

38 32 Young Adults and Jewish Engment Table 1: Characteristics of Sampling Strata Year Participant Status Age Sex total N % of total Sample N Frame N 2007 Nonparticipant Under Male 1, , Female 1, , Male 1, ,562 Female 1, ,606 Participant Under Male 4, , Female 6, , Male 6, ,916 Female 6, , Nonparticipant Under 25 Participant Male 3, ,099 Female 3, , Male 2, ,346 Female 2, ,460 Under Male 7, , Female 9, , Male 6, ,711 Female 7, , Nonparticipant Under Male 5, , Female 6, , Male 2, ,417 Female 2, ,626 Participant Under Male 4, , Female 5, , Male 3, ,425 Female 3, , Nonparticipant Under 25 Male 3, ,032 Female 4, , Male 1, ,193 Female 1, ,398 Participant Under Male 4, , Female 5, , Male 2, ,113 Female 2, ,429 Total 131,804 3, ,804

39 Young Adults and Jewish Engment 33 Final survey population The survey was administered to the replicates in a staggered manner first being released to two frame replicates. During field operations it became evident that response rates for the initial two frame replicates were even lower than expected. It was decided to forgo the release of the final two frame replicates and simply use the first two replicates as a large representative sample, instead of the failed census that would have resulted from release of all four replicates. Consequently the frame group mentioned below refers only to these two released replicates (N=64,454) and not the entire remainder of the target population. Since each replicate (for either the sample or frame ) is a stratified random sample of the entire population, the frame and sample can both be considered stratified random samples of the underlying population, as can their combination. See below for a description of the response rates for the two groups. Field Operations This study utilized an online web based survey. The Web survey was designed using an online instrument, created in LimeSurvey (Schmitz, 2009). 2 The survey was administered between February and March of Individuals in the frame were sent s inviting them to take the survey, and offered entry into a lottery for one of a number of $100 or $200 Amazon.com gift cards if they completed the survey. Three additional reminders were also sent to nonrespondents encouraging them to complete the survey. Members of the sample were likewise sent invitations (and two additional reminders) but promised a guaranteed Amazon.com gift card for survey completion. As a methodological experiment, members of the third sample replicate were offered $25 Amazon.com gift cards, while members of three other sample replicates were offered $15 Amazon.com gift cards. Approximately two weeks after the initial invitation members of the sample who had not yet responded were called and encourd to complete the survey online. The callers did not actually administer the survey to the respondents, but simply encourd the respondents to complete it on their own, and, in many cases, re- ed the unique URL to an address of the respondent s choosing. Because the survey was, in all cases, administered online, there are no mode effects across the different groups. 2 Cohen Center staff made some modifications to the source code of Lime Survey before using it for this study (LimeSurvey is open-source software released under the terms of the GNU General Public License v. 2). These modifications were mainly to allow greater compatibility between Lime Survey and the in-house CATI and bulk sending systems.

40 34 Young Adults and Jewish Engment Dispositions and Response rate Table 2 shows the final dispositions and response rates for the sample and frame. As expected, the response rate for the sample was significantly higher than that for the frame. Table 2: Final Dispositions and Response Rates Taglit Participants Taglit Non-Participants Total Frame Sample Overall Frame Sample Overall Frame Sample Overall Complete 1 4,184 1,047 5,231 1, ,184 5,964 1,451 7,415 Partial Break-Off Non-respondent 38, ,975 20, ,619 58,075 1,519 59,594 Total Sample 42,493 1,993 44,486 21,961 1,003 22,964 64,454 2,996 67,450 AAPOR Response Rate % 52.53% 11.76% 8.11% 40.28% 9.51% 9.25% 48.43% 10.99% AAPOR Response Rate % 53.19% 12.10% 8.40% 40.68% 9.92% 9.60% 49.0% 11.36% 1 Completed at least 80% of the questions asked. 2 Completed 50-80% of the questions asked. 3 Completed less than 50% of the questions asked. Weighting Design Weights The design weights for a stratified survey are simply the inverse of the probability of selection. For a given strata h, the design weight is calculated as the total population of that strata over the number of respondents in that strata: Thus each case is assigned a weight equal to the number of elements in the population of the frame it represents. In this case individuals in a given strata were assigned a weight equal to the frame population of that strata. Cases from both the sample and the main are treated identically in the calculation of design weights. Comparison of sample and frame It was expected that the sample would have a significantly higher response rate than the frame, due to the addition of guaranteed incentives and phone follow up. By comparing the two groups the bias associated with the lower response rate could be assessed, and, in principal, the lower response rate frame cases could be adjusted to the marginals of the higher response rate sample. However, when the two groups were compared on a number of key variables, there was virtually no difference.

41 Young Adults and Jewish Engment 35 Poststratification Weights Since there were little or no substantive differences between the sample and frame groups poststratification weights were calculated for the entire achieved sample in order to adjust for any differences between the distribution of known characteristics of the sample and known characteristics of the target population (known characteristics were derived from the Taglit registration database). In addition to the characteristics used in initial stratification (participant status, year of birth, and sex), information on Jewish denomination at time of application to the trip was available. Poststratification weights ( ) were created by raking within weighting stratum, where the sum of the weights was set to remain constant. 3 (See below for a description of raking.) The subscript j (where ) is used to distinguish poststratification weights, which could vary across cases within weighting stratum, compared to the design weights,, which remained constant within weighting stratum. Table 3. Characteristics of Weights Weight n Mean Std. Max:Min Min Max dev. ratio Unweighted 7, Design weights 7, Final raked weights 7, Calculation of Confidence Intervals Confidence intervals in tables and figures in this report were calculated at the 95% level using Stata s (2009) survey commands set up for a stratified survey (where the strata are defined as the weighting strata) with simple random sampling within strata. 3 Raking was carried out using QBAL (Werner, 2003).

42 36 Young Adults and Jewish Engment Raking Raking, also known as sample balancing and iterative proportional fitting (Deming, 1943), is a procedure that adjusts the marginal frequencies of a survey to the known marginal frequencies of a population. For example, one might have a population divided on sex and handedness (left and right) as follows: Population Survey Sex Sex Handed M F Total Handed M F Total R R L L Total Total Compared to the population, right-handers are somewhat underrepresented in the survey while lefthanders are somewhat overrepresented. Initially, all right-handers would receive weights of 0.90/0.85 (c. 1.06), while left-handers would receive weights of (c. 0.67). The resulting adjusted table would then be: Sex Handed M F Total R L Total Subsequently, sex would be adjusted to match the desired marginal totals, with males receiving an additional weight of 0.50/0.298 (c ) and females receiving a weight of 0.5/0.702 (c ). After this transformation, the weighted frequencies would be: Sex Handed M F Total R L Total Further raking would yield additional weights of c for men and for women and a marginal frequency of.4998 for men and.5002 for women. Additional iterations could take place until a desired level of precision was reached. (Precision is defined in raking in terms of the sum of the weighted squares of the residuals, the difference between the expected and observed frequency in a cell; Battaglia, Izrael, Hoaglin, & Frankel, 2004; Deming & Stephan, 1940.) The final weights for each cell are approximately for male right-handers, for male left-handers, for female righthanders and for female left-handers.

43 Young Adults and Jewish Engment 37 Appendix B: Attitudinal Impact of Taglit for Undergraduates and Non- Undergraduates Figure 1: Feelings of Jewish connection by Taglit participation (undergraduates)* * Predicted probabilities from ordinal logistic regression models controlling for Jewish background and.

44 38 Young Adults and Jewish Engment Figure 2: Feelings of Jewish connection by Taglit participation (post-college)* * Predicted probabilities from ordinal logistic regression models controlling for Jewish background,, and Jewish engment in college

45 Young Adults and Jewish Engment 39 Figure 1: Feelings of Jewish connection by Taglit participation (undergraduates) Connection to Israel Survey: Ordered logistic regression Number of strata = 31 Number of obs = 7338 Number of PSUs = 7338 Population size = Subpop. no. of obs = 1720 Subpop. size = Design df = 7307 F( 3, 7305) = Prob > F = Linearized conisr Odds Ratio Std. Err. t P> t [] participant jbackground /cut /cut /cut Note: 1 stratum omitted because it contains no subpopulation members.. local j=meanjbackground. local a=mean. prvalue, x(participant=0 jbackground=`j' =`a') ologit: Predictions for conisr Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') ologit: Predictions for conisr Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] x=

46 40 Young Adults and Jewish Engment Part of a worldwide Jewish community Survey: Ordered logistic regression Number of strata = 31 Number of obs = 7327 Number of PSUs = 7327 Population size = Subpop. no. of obs = 1709 Subpop. size = Design df = 7296 F( 3, 7294) = Prob > F = Linearized conwrldjcomm Odds Ratio Std. Err. t P> t [] participant jbackground /cut /cut /cut Note: 1 stratum omitted because it contains no subpopulation members.. local j=meanjbackground. local a=mean. prvalue, x(participant=0 jbackground=`j' =`a') ologit: Predictions for conwrldjcomm Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') ologit: Predictions for conwrldjcomm Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] x=

47 Young Adults and Jewish Engment 41 Connection to local Jewish Community Survey: Ordered logistic regression Number of strata = 31 Number of obs = 7327 Number of PSUs = 7327 Population size = Subpop. no. of obs = 1709 Subpop. size = Design df = 7296 F( 3, 7294) = Prob > F = Linearized conlocaljcomm Odds Ratio Std. Err. t P> t [] participant jbackground /cut /cut /cut Note: 1 stratum omitted because it contains no subpopulation members.. local j=meanjbackground. local a=mean. prvalue, x(participant=0 jbackground=`j' =`a') ologit: Predictions for conlocaljcomm Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') ologit: Predictions for conlocaljcomm Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] x=

48 42 Young Adults and Jewish Engment Connection to Jewish peers Survey: Ordered logistic regression Number of strata = 31 Number of obs = 7327 Number of PSUs = 7327 Population size = Subpop. no. of obs = 1709 Subpop. size = Design df = 7296 F( 3, 7294) = Prob > F = Linearized conjpeers Odds Ratio Std. Err. t P> t [] participant jbackground /cut /cut /cut Note: 1 stratum omitted because it contains no subpopulation members.. local j=meanjbackground. local a=mean. prvalue, x(participant=0 jbackground=`j' =`a') ologit: Predictions for conjpeers Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') ologit: Predictions for conjpeers Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] x=

49 Young Adults and Jewish Engment 43 Figure 2: Feelings of Jewish connection by Taglit participation (postcollege). Controlling for campus engment to make comparable with other non-undergrad models. Connection to Israel Survey: Ordered logistic regression Number of strata = 32 Number of obs = 6863 Number of PSUs = 6863 Population size = Subpop. no. of obs = 5034 Subpop. size = Design df = 6831 F( 4, 6828) = Prob > F = Linearized conisr Odds Ratio Std. Err. t P> t [] participant jbackground campus /cut /cut /cut local j=meanjbackground. local a=mean. prvalue, x(participant=0 jbackground=`j' =`a') ologit: Predictions for conisr Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] campus x= prvalue, x(participant=1 jbackground=`j' =`a') ologit: Predictions for conisr Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] campus x=

50 44 Young Adults and Jewish Engment Connection to worldwide Jewish community Survey: Ordered logistic regression Number of strata = 32 Number of obs = 6837 Number of PSUs = 6837 Population size = Subpop. no. of obs = 5008 Subpop. size = Design df = 6805 F( 5, 6801) = Prob > F = Linearized conwrldjcomm Coef. Std. Err. t P> t [] participant jbackground campus orthodox /cut /cut /cut local j=meanjbackground. local a=mean. prvalue, x(participant=0 jbackground=`j' =`a') ologit: Predictions for conwrldjcomm Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] campus orthodox x= prvalue, x(participant=1 jbackground=`j' =`a') ologit: Predictions for conwrldjcomm Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] campus orthodox x=

51 Young Adults and Jewish Engment 45 Connection to local Jewish community Survey: Ordered logistic regression Number of strata = 32 Number of obs = 6838 Number of PSUs = 6838 Population size = Subpop. no. of obs = 5009 Subpop. size = Design df = 6806 F( 4, 6803) = Prob > F = Linearized conlocaljcomm Odds Ratio Std. Err. t P> t [] participant jbackground campus /cut /cut /cut local j=meanjbackground. local a=mean. prvalue, x(participant=0 jbackground=`j' =`a') ologit: Predictions for conlocaljcomm Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] campus x= prvalue, x(participant=1 jbackground=`j' =`a') ologit: Predictions for conlocaljcomm Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] campus x=

52 46 Young Adults and Jewish Engment Connection to Jewish peers Survey: Ordered logistic regression Number of strata = 32 Number of obs = 6836 Number of PSUs = 6836 Population size = Subpop. no. of obs = 5007 Subpop. size = Design df = 6804 F( 4, 6801) = Prob > F = Linearized conjpeers Odds Ratio Std. Err. t P> t [] participant jbackground campus /cut /cut /cut local j=meanjbackground. local a=mean. prvalue, x(participant=0 jbackground=`j' =`a') ologit: Predictions for conjpeers Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] campus x= prvalue, x(participant=1 jbackground=`j' =`a') ologit: Predictions for conjpeers Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] campus x=

53 Young Adults and Jewish Engment 47 Appendix C: Tables Figure 6: Feelings of Jewish connection by Taglit participation (undergraduate and post-college) Connection to Israel Survey: Ordered logistic regression Number of strata = 32 Number of obs = 7193 Number of PSUs = 7193 Population size = Subpop. no. of obs = 7193 Subpop. size = Design df = 7161 F( 3, 7159) = Prob > F = Linearized conisr Odds Ratio Std. Err. t P> t [] participant jbackground /cut /cut /cut local j=meanjbackground. local a=mean. prvalue, x(participant=0 jbackground=`j' =`a') ologit: Predictions for conisr Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') ologit: Predictions for conisr Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] x=

54 48 Young Adults and Jewish Engment Connection to worldwide Jewish community Survey: Ordered logistic regression Number of strata = 32 Number of obs = 7154 Number of PSUs = 7154 Population size = Subpop. no. of obs = 7154 Subpop. size = Design df = 7122 F( 3, 7120) = Prob > F = Linearized conwrldjcomm Odds Ratio Std. Err. t P> t [] participant jbackground /cut /cut /cut local j=meanjbackground. local a=mean. prvalue, x(participant=0 jbackground=`j' =`a') ologit: Predictions for conwrldjcomm Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') ologit: Predictions for conwrldjcomm Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] x=

55 Young Adults and Jewish Engment 49 Importance of marrying a Jew Survey: Ordered logistic regression Number of strata = 32 Number of obs = 7155 Number of PSUs = 7155 Population size = Subpop. no. of obs = 7155 Subpop. size = Design df = 7123 F( 3, 7121) = Prob > F = Linearized conlocaljcomm Odds Ratio Std. Err. t P> t [] participant jbackground /cut /cut /cut local j=meanjbackground. local a=mean. prvalue, x(participant=0 jbackground=`j' =`a') ologit: Predictions for conlocaljcomm Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') ologit: Predictions for conlocaljcomm Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] x=

56 50 Young Adults and Jewish Engment Importance of raising kids Jewish Survey: Ordered logistic regression Number of strata = 32 Number of obs = 7153 Number of PSUs = 7153 Population size = Subpop. no. of obs = 7153 Subpop. size = Design df = 7121 F( 3, 7119) = Prob > F = Linearized conjpeers Odds Ratio Std. Err. t P> t [] participant jbackground /cut /cut /cut local j=meanjbackground. local a=mean. prvalue, x(participant=0 jbackground=`j' =`a') ologit: Predictions for conjpeers Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') ologit: Predictions for conjpeers Pr(y=Not_at_a x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_muc x): [ , ] x=

57 Young Adults and Jewish Engment 51 Figure 7: Importance of being Jewish and forming a Jewish family by Taglit participation Importance of marrying a Jew Survey: Ordered logistic regression Number of strata = 32 Number of obs = 6025 Number of PSUs = 6025 Population size = Subpop. no. of obs = 6025 Subpop. size = Design df = 5993 F( 3, 5991) = Prob > F = Linearized recimpmrryjew Odds Ratio Std. Err. t P> t [] participant jbackground /cut /cut /cut local j=meanjbackground. local a=mean. prvalue, x(participant=0 jbackground=`j' =`a') ologit: Predictions for recimpmrryjew Pr(y=1 x): [ , ] Pr(y=2 x): [ , ] Pr(y=3 x): [ , ] Pr(y=4 x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') ologit: Predictions for recimpmrryjew Pr(y=1 x): [ , ] Pr(y=2 x): [ , ] Pr(y=3 x): [ , ] Pr(y=4 x): [ , ] x=

58 52 Young Adults and Jewish Engment Importance of raising kids jewish Survey: Ordered logistic regression Number of strata = 32 Number of obs = 6459 Number of PSUs = 6459 Population size = Subpop. no. of obs = 6459 Subpop. size = Design df = 6427 F( 3, 6425) = Prob > F = Linearized recfutchild Odds Ratio Std. Err. t P> t [] participant jbackground /cut /cut /cut local j=meanjbackground. local a=mean. prvalue, x(participant=0 jbackground=`j' =`a') ologit: Predictions for recfutchild Pr(y=1 x): [ , ] Pr(y=2 x): [ , ] Pr(y=3 x): [ , ] Pr(y=4 x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') ologit: Predictions for recfutchild Pr(y=1 x): [ , ] Pr(y=2 x): [ , ] Pr(y=3 x): [ , ] Pr(y=4 x): [ , ] x=

59 Young Adults and Jewish Engment 53 Importance of being Jewish Survey: Ordered logistic regression Number of strata = 32 Number of obs = 7152 Number of PSUs = 7152 Population size = Subpop. no. of obs = 7152 Subpop. size = Design df = 7120 F( 3, 7118) = Prob > F = Linearized impbejew Odds Ratio Std. Err. t P> t [] participant jbackground /cut /cut /cut local j=meanjbackground. local a=mean. prvalue, x(participant=0 jbackground=`j' =`a') ologit: Predictions for impbejew Pr(y=Not_impo x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_imp x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') ologit: Predictions for impbejew Pr(y=Not_impo x): [ , ] Pr(y=A_little x): [ , ] Pr(y=Somewhat x): [ , ] Pr(y=Very_imp x): [ , ] x=

60 54 Young Adults and Jewish Engment Figure 8: Being invited to activities sponsored by campus-based Jewish organizations by Taglit participation Being invited to Hillel Survey: Multinomial logistic regression Number of strata = 31 Number of obs = 7281 Number of PSUs = 7281 Population size = Subpop. no. of obs = 1663 Subpop. size = Design df = 7250 F( 6, 7245) = Prob > F = Linearized rechillelin~e Coef. Std. Err. t P> t [] (base outcome) Once_or_twice participant jbackground _cons times participant jbackground _cons Note: 1 stratum omitted because it contains no subpopulation members.. local j=meanjbackgroundu. local a=meanu. prvalue, x(participant=0 jbackground=`j' =`a') mlogit: Predictions for rechillelinvite Pr(y=Once_or_ x): [ , ] Pr(y=3+_times x): [ , ] Pr(y=0 x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') mlogit: Predictions for rechillelinvite Pr(y=Once_or_ x): [ , ] Pr(y=3+_times x): [ , ] Pr(y=0 x): [ , ] x=

61 Young Adults and Jewish Engment 55 Being invited to Chabad. svy, subpop(subpop if undergradsvy==1): mlogit recchabadinvite, base(0) (running mlogit on estimation sample) Survey: Multinomial logistic regression Number of strata = 31 Number of obs = 7266 Number of PSUs = 7266 Population size = Subpop. no. of obs = 1648 Subpop. size = Design df = 7235 F( 6, 7230) = Prob > F = Linearized recchabadin~e Coef. Std. Err. t P> t [] (base outcome) Once_or_twice participant jbackground _cons times participant jbackground _cons Note: 1 stratum omitted because it contains no subpopulation members.. local j=meanjbackgroundu. local a=meanu. prvalue, x(participant=0 jbackground=`j' =`a') mlogit: Predictions for recchabadinvite Pr(y=Once_or_ x): [ , ] Pr(y=3+_times x): [ , ] Pr(y=0 x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') mlogit: Predictions for recchabadinvite Pr(y=Once_or_ x): [ , ] Pr(y=3+_times x): [ , ] Pr(y=0 x): [ , ] x=

62 56 Young Adults and Jewish Engment Being invited to J frat Survey: Multinomial logistic regression Number of strata = 31 Number of obs = 7279 Number of PSUs = 7279 Population size = Subpop. no. of obs = 1661 Subpop. size = Design df = 7248 F( 6, 7243) = 8.98 Prob > F = Linearized recjfratinv~e Coef. Std. Err. t P> t [] (base outcome) Once_or_twice participant jbackground _cons times participant jbackground _cons Note: 1 stratum omitted because it contains no subpopulation members.. local j=meanjbackgroundu. local a=meanu. prvalue, x(participant=0 jbackground=`j' =`a') mlogit: Predictions for recjfratinvite Pr(y=Once_or_ x): [ , ] Pr(y=3+_times x): [ , ] Pr(y=0 x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') mlogit: Predictions for recjfratinvite Pr(y=Once_or_ x): [ , ] Pr(y=3+_times x): [ , ] Pr(y=0 x): [ , ] x=

63 Young Adults and Jewish Engment 57 Being invited to other Survey: Multinomial logistic regression Number of strata = 31 Number of obs = 7270 Number of PSUs = 7270 Population size = Subpop. no. of obs = 1652 Subpop. size = Design df = 7239 F( 6, 7234) = 5.91 Prob > F = Linearized recothgrpin~e Coef. Std. Err. t P> t [] (base outcome) Once_or_twice participant jbackground _cons times participant jbackground _cons Note: 1 stratum omitted because it contains no subpopulation members.. local j=meanjbackgroundu. local a=meanu. prvalue, x(participant=0 jbackground=`j' =`a') mlogit: Predictions for recothgrpinvite Pr(y=Once_or_ x): [ , ] Pr(y=3+_times x): [ , ] Pr(y=0 x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') mlogit: Predictions for recothgrpinvite Pr(y=Once_or_ x): [ , ] Pr(y=3+_times x): [ , ] Pr(y=0 x): [ , ] x=

64 58 Young Adults and Jewish Engment Figure 9 : Participating in activities sponsored by campus-based Jewish organizations by Taglit participation Engd Hillel Survey: Multinomial logistic regression Number of strata = 31 Number of obs = 7276 Number of PSUs = 7276 Population size = Subpop. no. of obs = 1658 Subpop. size = Design df = 7245 F( 6, 7240) = Prob > F = Linearized rechillelen~e Coef. Std. Err. t P> t [] (base outcome) Once_or_twice participant jbackground _cons times participant jbackground _cons Note: 1 stratum omitted because it contains no subpopulation members.. local j=meanjbackgroundu. local a=meanu. prvalue, x(participant=0 jbackground=`j' =`a') mlogit: Predictions for rechilleleng Pr(y=Once_or_ x): [ , ] Pr(y=3+_times x): [ , ] Pr(y=0 x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') mlogit: Predictions for rechilleleng Pr(y=Once_or_ x): [ , ] Pr(y=3+_times x): [ , ] Pr(y=0 x): [ , ] x=

65 Young Adults and Jewish Engment 59 Engd Chabad Survey: Multinomial logistic regression Number of strata = 31 Number of obs = 7262 Number of PSUs = 7262 Population size = Subpop. no. of obs = 1644 Subpop. size = Design df = 7231 F( 6, 7226) = Prob > F = Linearized recchabaden~e Coef. Std. Err. t P> t [] (base outcome) Once_or_twice participant jbackground _cons times participant jbackground _cons Note: 1 stratum omitted because it contains no subpopulation members.. local j=meanjbackgroundu. local a=meanu. prvalue, x(participant=0 jbackground=`j' =`a') mlogit: Predictions for recchabadeng Pr(y=Once_or_ x): [ , ] Pr(y=3+_times x): [ , ] Pr(y=0 x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') mlogit: Predictions for recchabadeng Pr(y=Once_or_ x): [ , ] Pr(y=3+_times x): [ , ] Pr(y=0 x): [ , ] x=

66 60 Young Adults and Jewish Engment Engd Jewish fraternity Survey: Multinomial logistic regression Number of strata = 31 Number of obs = 7269 Number of PSUs = 7269 Population size = Subpop. no. of obs = 1651 Subpop. size = Design df = 7238 F( 6, 7233) = 9.16 Prob > F = Linearized recjfrateng~e Coef. Std. Err. t P> t [] (base outcome) Once_or_twice participant jbackground _cons times participant jbackground _cons Note: 1 stratum omitted because it contains no subpopulation members.. local j=meanjbackgroundu. local a=meanu. prvalue, x(participant=0 jbackground=`j' =`a') mlogit: Predictions for recjfrateng Pr(y=Once_or_ x): [ , ] Pr(y=3+_times x): [ , ] Pr(y=0 x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') mlogit: Predictions for recjfrateng Pr(y=Once_or_ x): [ , ] Pr(y=3+_times x): [ , ] Pr(y=0 x): [ , ] x=

67 Young Adults and Jewish Engment 61 Engd other Survey: Multinomial logistic regression Number of strata = 31 Number of obs = 7249 Number of PSUs = 7249 Population size = Subpop. no. of obs = 1631 Subpop. size = Design df = 7218 F( 6, 7213) = 7.78 Prob > F = Linearized recothgrpen~e Coef. Std. Err. t P> t [] (base outcome) Once_or_twice participant jbackground _cons times participant jbackground _cons Note: 1 stratum omitted because it contains no subpopulation members.. local j=meanjbackgroundu. local a=meanu. prvalue, x(participant=0 jbackground=`j' =`a') mlogit: Predictions for recothgrpeng Pr(y=Once_or_ x): [ , ] Pr(y=3+_times x): [ , ] Pr(y=0 x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') mlogit: Predictions for recothgrpeng Pr(y=Once_or_ x): [ , ] Pr(y=3+_times x): [ , ] Pr(y=0 x): [ , ] x=

68 62 Young Adults and Jewish Engment Figure 10 : Attending events sponsored by a Jewish organization (undergraduates) Parties Survey: Multinomial logistic regression Number of strata = 30 Number of obs = 7145 Number of PSUs = 7145 Population size = Subpop. no. of obs = 1671 Subpop. size = Design df = 7115 F( 6, 7110) = Prob > F = Linearized acttypeparty Coef. Std. Err. t P> t [] Never (base outcome) Once participant jbackground _cons More_than_once participant jbackground _cons Note: 2 strata omitted because they contain no subpopulation members.. local j=meanjbackgroundu. local a=meanu. prvalue, x(participant=0 jbackground=`j' =`a') mlogit: Predictions for acttypeparty Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') mlogit: Predictions for acttypeparty Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] x=

69 Young Adults and Jewish Engment 63 Lectures Survey: Multinomial logistic regression Number of strata = 31 Number of obs = 7284 Number of PSUs = 7284 Population size = Subpop. no. of obs = 1666 Subpop. size = Design df = 7253 F( 6, 7248) = Prob > F = Linearized acttypelecture Coef. Std. Err. t P> t [] Never (base outcome) Once participant jbackground _cons More_than_once participant jbackground _cons Note: 1 stratum omitted because it contains no subpopulation members.. local j=meanjbackgroundu. local a=meanu. prvalue, x(participant=0 jbackground=`j' =`a') mlogit: Predictions for acttypelecture Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') mlogit: Predictions for acttypelecture Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] x=

70 64 Young Adults and Jewish Engment Cultural events Survey: Multinomial logistic regression Number of strata = 30 Number of obs = 7130 Number of PSUs = 7130 Population size = Subpop. no. of obs = 1656 Subpop. size = Design df = 7100 F( 6, 7095) = Prob > F = Linearized acttypecult Coef. Std. Err. t P> t [] Never (base outcome) Once participant jbackground _cons More_than_once participant jbackground _cons Note: 2 strata omitted because they contain no subpopulation members.. local j=meanjbackgroundu. local a=meanu. prvalue, x(participant=0 jbackground=`j' =`a') mlogit: Predictions for acttypecult Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') mlogit: Predictions for acttypecult Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] x=

71 Young Adults and Jewish Engment 65 Activism Survey: Multinomial logistic regression Number of strata = 30 Number of obs = 7139 Number of PSUs = 7139 Population size = Subpop. no. of obs = 1665 Subpop. size = Design df = 7109 F( 6, 7104) = Prob > F = Linearized acttypecomm Coef. Std. Err. t P> t [] Never (base outcome) Once participant jbackground _cons More_than_once participant jbackground _cons Note: 2 strata omitted because they contain no subpopulation members.. local j=meanjbackgroundu. local a=meanu. prvalue, x(participant=0 jbackground=`j' =`a') mlogit: Predictions for acttypecomm Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') mlogit: Predictions for acttypecomm Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] x=

72 66 Young Adults and Jewish Engment Other Survey: Multinomial logistic regression Number of strata = 31 Number of obs = 7270 Number of PSUs = 7270 Population size = Subpop. no. of obs = 1652 Subpop. size = Design df = 7239 F( 6, 7234) = 3.47 Prob > F = Linearized acttypeoth Coef. Std. Err. t P> t [] Never (base outcome) Once participant jbackground _cons More_than_once participant jbackground _cons Note: 1 stratum omitted because it contains no subpopulation members.. local j=meanjbackgroundu. local a=meanu. prvalue, x(participant=0 jbackground=`j' =`a') mlogit: Predictions for acttypeoth Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] x= prvalue, x(participant=1 jbackground=`j' =`a') mlogit: Predictions for acttypeoth Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] x=

73 Young Adults and Jewish Engment 67 Figure 11: Attending events sponsored by a Jewish organization (post-college) Parties Survey: Multinomial logistic regression Number of strata = 32 Number of obs = 6769 Number of PSUs = 6769 Population size = Subpop. no. of obs = 4940 Subpop. size = Design df = 6737 F( 8, 6730) = Prob > F = Linearized acttypeparty Coef. Std. Err. t P> t [] Never (base outcome) Once participant jbackground campus _cons More_than_once participant jbackground campus _cons local j=meanjbackgroundnu. local a=meannu. local c=meancampusnu. prvalue, x(participant=0 jbackground=`j' =`a' campus=`c') mlogit: Predictions for acttypeparty Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] campus x= prvalue, x(participant=1 jbackground=`j' =`a' campus=`c') mlogit: Predictions for acttypeparty Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] campus x=

74 68 Young Adults and Jewish Engment Lectures Survey: Multinomial logistic regression Number of strata = 32 Number of obs = 6754 Number of PSUs = 6754 Population size = Subpop. no. of obs = 4925 Subpop. size = Design df = 6722 F( 8, 6715) = Prob > F = Linearized acttypelecture Coef. Std. Err. t P> t [] Never (base outcome) Once participant jbackground campus _cons More_than_once participant jbackground campus _cons local j=meanjbackgroundnu. local a=meannu. local c=meancampusnu. prvalue, x(participant=0 jbackground=`j' =`a' campus=`c') mlogit: Predictions for acttypelecture Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] campus x= prvalue, x(participant=1 jbackground=`j' =`a' campus=`c') mlogit: Predictions for acttypelecture Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] campus x=

75 Young Adults and Jewish Engment 69 Cultural events Survey: Multinomial logistic regression Number of strata = 32 Number of obs = 6744 Number of PSUs = 6744 Population size = Subpop. no. of obs = 4915 Subpop. size = Design df = 6712 F( 8, 6705) = Prob > F = Linearized acttypecult Coef. Std. Err. t P> t [] Never (base outcome) Once participant jbackground campus _cons More_than_once participant jbackground campus _cons local j=meanjbackgroundnu. local a=meannu. local c=meancampusnu. prvalue, x(participant=0 jbackground=`j' =`a' campus=`c') mlogit: Predictions for acttypecult Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] campus x= prvalue, x(participant=1 jbackground=`j' =`a' campus=`c') mlogit: Predictions for acttypecult Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] campus x=

76 70 Young Adults and Jewish Engment Activism Survey: Multinomial logistic regression Number of strata = 32 Number of obs = 6747 Number of PSUs = 6747 Population size = Subpop. no. of obs = 4918 Subpop. size = Design df = 6715 F( 8, 6708) = Prob > F = Linearized acttypecomm Coef. Std. Err. t P> t [] Never (base outcome) Once participant jbackground campus _cons More_than_once participant jbackground campus _cons local j=meanjbackgroundnu. local a=meannu. local c=meancampusnu. prvalue, x(participant=0 jbackground=`j' =`a' campus=`c') mlogit: Predictions for acttypecomm Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] campus x= prvalue, x(participant=1 jbackground=`j' =`a' campus=`c') mlogit: Predictions for acttypecomm Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] campus x=

77 Young Adults and Jewish Engment 71 Other Survey: Multinomial logistic regression Number of strata = 32 Number of obs = 6712 Number of PSUs = 6712 Population size = Subpop. no. of obs = 4883 Subpop. size = Design df = 6680 F( 8, 6673) = 7.32 Prob > F = Linearized acttypeoth Coef. Std. Err. t P> t [] Never (base outcome) Once participant jbackground campus _cons More_than_once participant jbackground campus _cons local j=meanjbackgroundnu. local a=meannu. local c=meancampusnu. prvalue, x(participant=0 jbackground=`j' =`a' campus=`c') mlogit: Predictions for acttypeoth Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] campus x= prvalue, x(participant=1 jbackground=`j' =`a' campus=`c') mlogit: Predictions for acttypeoth Pr(y=Once x): [ , ] Pr(y=More_tha x): [ , ] Pr(y=Never x): [ , ] campus x=

78 72 Young Adults and Jewish Engment Figure 12: Donating to Jewish causes and volunteering under Jewish sponsorship by Taglit participation (post-college) Donating Survey: Logistic regression Number of strata = 32 Number of obs = 6755 Number of PSUs = 6755 Population size = Subpop. no. of obs = 4926 Subpop. size = Design df = 6723 F( 4, 6720) = Prob > F = Linearized bindonatejew Odds Ratio Std. Err. t P> t [] participant jbackground campus _cons local j=meanjbackgroundnu. local a=meannu. local c=meancampusnu. prvalue, x(participant=0 jbackground=`j' =`a' campus=`c') logit: Predictions for bindonatejew Pr(y=1 x): [ , ] Pr(y=0 x): [ , ] campus x= prvalue, x(participant=1 jbackground=`j' =`a' campus=`c') logit: Predictions for bindonatejew Pr(y=1 x): [ , ] Pr(y=0 x): [ , ] campus x=

79 Young Adults and Jewish Engment 73 Volunteering Survey: Logistic regression Number of strata = 32 Number of obs = 6767 Number of PSUs = 6767 Population size = Subpop. no. of obs = 4938 Subpop. size = Design df = 6735 F( 4, 6732) = Prob > F = Linearized binvoljew Odds Ratio Std. Err. t P> t [] participant jbackground campus _cons local j=meanjbackgroundnu. local a=meannu. local c=meancampusnu. prvalue, x(participant=0 jbackground=`j' =`a' campus=`c') logit: Predictions for binvoljew Pr(y=1 x): [ , ] Pr(y=0 x): [ , ] campus x= prvalue, x(participant=1 jbackground=`j' =`a' campus=`c') logit: Predictions for binvoljew Pr(y=1 x): [ , ] Pr(y=0 x): [ , ] campus x=

80 74 Young Adults and Jewish Engment Figure 13: Engment in Jewish religious life by Taglit participation (post-college) Shabbat Meal Survey: Logistic regression Number of strata = 32 Number of obs = 6765 Number of PSUs = 6765 Population size = Subpop. no. of obs = 4936 Subpop. size = Design df = 6733 F( 4, 6730) = Prob > F = Linearized lstfri Odds Ratio Std. Err. t P> t [] participant jbackground campus _cons logit: Predictions for lstfri Pr(y=1 x): [ , ] Pr(y=0 x): [ , ] campus x= logit: Predictions for lstfri Pr(y=1 x): [ , ] Pr(y=0 x): [ , ] campus x=

81 Young Adults and Jewish Engment 75 Synagogue membership (running logit on estimation sample) Survey: Logistic regression Number of strata = 32 Number of obs = 6768 Number of PSUs = 6768 Population size = Subpop. no. of obs = 4939 Subpop. size = Design df = 6736 F( 4, 6733) = Prob > F = Linearized syn Odds Ratio Std. Err. t P> t [] participant jbackground campus _cons logit: Predictions for syn Pr(y=Yes x): [ , ] Pr(y=No x): [ , ] campus x= logit: Predictions for syn Pr(y=Yes x): [ , ] Pr(y=No x): [ , ] campus x=

82 76 Young Adults and Jewish Engment Religious service attendance (running logit on estimation sample) Survey: Logistic regression Number of strata = 32 Number of obs = 6767 Number of PSUs = 6767 Population size = Subpop. no. of obs = 4938 Subpop. size = Design df = 6735 F( 4, 6732) = Prob > F = Linearized relservbin Odds Ratio Std. Err. t P> t [] participant jbackground campus _cons logit: Predictions for relservbin Pr(y=1 x): [ , ] Pr(y=0 x): [ , ] campus x= logit: Predictions for relservbin Pr(y=1 x): [ , ] Pr(y=0 x): [ , ] campus x= NEXT participation (p. 23) Survey: Logistic regression Number of strata = 32 Number of obs = 7188 Number of PSUs = 7188 Population size = Subpop. no. of obs = 4924 Subpop. size = Design df = 7156 F( 10, 7147) = 9.42 Prob > F = Linearized binnext Odds Ratio Std. Err. t P> t [] reground female jbackground _cons

83 The Maurice and Marilyn Cohen Center for Modern Jewish Studies at Brandeis University is a multi-disciplinary research institute dedicated to the study of American Jewry and religious and cultural identity. The Steinhardt Social Research Institute, hosted at CMJS, is committed to the development and application of innovative approaches to socio-demographic research for the study of Jewish, religious, and cultural identity. Brandeis University