This is an open access article distributed under the terms of the Creative Commons Attribution License (“You are the first Jewish person I’ve ever met” is a phrase familiar to many Jews residing in Germany. Due to the small number of Jews living in Germany, many non-Jewish people do not have personal contact to Jews. However, they often encounter information on past Jewish–non-Jewish contact through media and public discourse, particularly regarding the massive antisemitic crimes committed by Germans during WWII. Politically, the memory of the Shoah is viewed as a moral compass for the future (Lienen & Cohrs, 2021). At the same time, antisemitism in Germany persists, including memory-defensive forms rejecting ongoing remembrance of the German atrocities (European Union Agency for Fundamental Rights, 2018). Recognizing the need to reduce antisemitic prejudices and addressing the desire of Jews to not be perceived solely through the lens of the Shoah (Branković & Kranz, 2022), the Jewish community in Germany launched two initiatives on intergroup contact in 2006 and 2014, which merged in 2020 to form the project “Meet a Jew”.
“Meet a Jew” is a social project that facilitates encounters between Jewish volunteers and non-Jewish groups throughout Germany. Such contact interventions can be initiated for various forms of groups, for instance by teachers for their classes, or by sports club members for their fellows. We present a field study conducted in cooperation with the project, providing insights into the effectiveness of the intervention in enhancing positive attitudes toward Jews from the angle of the initiators. Drawing on intergroup contact literature, we assessed the initiators’ impressions of the group of attendees along three dimensions relevant in reducing prejudices (knowledge, anxiety, and perspective-taking). Accompanied by an in-depth qualitative analysis of the feedback the initiators received from the attendees, this research is, to our best knowledge, the first to address post-war Jewish–non-Jewish contact in Germany within a real-world intervention program.
Intergroup Contact
Rooted in Allport’s (1954) seminal work, intergroup contact has emerged as a key strategy for reducing prejudices. The Contact Hypothesis posits that interactions between members of different social groups can increase positive attitudes toward one another. These positive attitudes can generalize beyond the individual group members to influence positivity toward the entire group, thus contributing to the reduction of group-based conflicts (Pettigrew, 1998). Indeed, cross-sectional, longitudinal, and experimental studies across various group contexts support the effectiveness of contact in improving intergroup relations (for a comprehensive meta-analysis, see Pettigrew & Tropp, 2006). But what are the effects of contact in societies marked by tremendous historical intergroup violence?
In post-conflict societies, direct intergroup contact is often limited due to continuous segregation or unwillingness to engage in contact (Boin et al., 2021). Particularly, ongoing marginalization and structural segregation may result in the former victim group not being visible within the majority society (Hübscher et al., 2022). Additionally, groups might avoid contact. For instance, perpetrator group members might avoid contact with the victims to evade reminders of their ingroup’s wrongdoing (Kauff et al., 2021). Thus, a first major barrier in post-violent contexts is to establish contact.
When occurring, however, intergroup contact has shown to improve post-conflict relations. For example, direct contact between Catholics and Protestants in Northern Ireland led to more positive intergroup attitudes and increased reconciliation efforts (Tropp et al., 2017). Similar outcomes were observed in South Africa; a context marked by massive violence against Black South Africans by Whites (Tropp et al., 2017). In Bosnia and Herzegovina, intergroup contact further increased willingness to acknowledge ingroup responsibility for historical crimes (Čehajić & Brown, 2010). These findings suggest that among groups involved in substantial historical transgressions, contact may not only have the potential to foster intergroup positivity, but also to reduce reluctance to engage with the ingroup’s past wrongdoings.
These positive contact effects can be based on various affective and cognitive mechanisms (Boin et al., 2021; Pettigrew, 1998). With regard to the specific context of post-conflict contact, Hewstone and colleagues (2014) argued that three of these mechanisms are of particular importance: (1) reduced anxiety arising from the conflicted past and unfamiliarity with the other group, (2) increased perspective-taking, thus gaining new views and empathizing with the other group, and (3) reduced realistic and symbolic threat. The latter appeared to be especially pertinent for minority group members, whose social role or even existence as a group may have been severely threatened during the conflict by the majority society in which they now live. Meta-analytic findings by Pettigrew and Tropp (2008) across diverse contact settings align with the relevance of anxiety and perspective-taking, and further identify knowledge as a central mechanism. Enhancing intergroup knowledge facilitates mutual understanding and identification of similarities with the other group (Dovidio et al., 2003). Especially in post-conflict contexts with limited contact to the former victim group, such understanding may be crucial for reducing prejudices (Bilali & Vollhardt, 2013; Čehajić & Brown, 2010). Against this backdrop, it seems plausible to reason that when examining contact effects in post-violent majority group settings, intergroup knowledge, anxiety, and perspective-taking may be of particular importance.
However, much of the literature on intergroup contact in post-conflict societies is grounded in cross-sectional research rather than interventions studies. Extending these findings, Lemmer and Wagner (2015) conducted a meta-analysis to evaluate the effectiveness of contact interventions, including interventions in conflicted contexts. While the results indeed supported the overall effectiveness of interventions in improving (post-)conflict relations, findings from direct contact intervention programs may not always align with these results. In fact, studies on Jewish-Polish and Palestinian-Jewish contact indicate that the effect of direct contact may crucially depend on the content discussed during the encounter (Bilewicz, 2007) or even be negative (Ditlmann & Samii, 2016). As such, examining the actual potential of contact interventions within specific intergroup contexts seems relevant.
The context of Jewish–non-Jewish contact in Germany has been scarcely examined in empirical studies to date (for one study on the secondary transfer effect of contact with immigrants on antisemitism, see Schmid et al., 2012). The present field study seeks to take a first step toward addressing this gap by approaching the effectiveness of a nationwide contact intervention program aimed at reducing stereotypes and prejudices among non-Jewish attendees. In order to situate the specificity of such a contact intervention, it is important to consider the current state of Jewish–non-Jewish contact in Germany.
The Case of Jewish–non-Jewish Contact in Germany
After WWII, most Jewish displaced persons left Germany. Until the late 1980s, Jewish communities in Germany maintained around 30,000 members (Wenzel, 2022). The Jewish population began to increase only after 1990 due to the immigration of Jewish refugees1 from the Soviet Union, who were a marginalized minority there. Today, about 200,000 Jews live in Germany, constituting 0.2% of the population (Central Welfare Board of Jews in Germany, 2022). The likelihood of direct contact with Jews is thus rather low. Additionally, identifying whether a person is Jewish is often not feasible unless explicitly disclosed or openly shown through signs of Jewish faith and culture, such as wearing a kippa or a necklace with the Star of David.
However, this does not imply a complete absence of contact. Rather, most contact is mass-mediated (Harwood, 2021), including parasocial observation of Jewish portrayals or vicarious observation of intergroup contact in the media. Both can pertain to present-day contact (e.g., documentaries on contemporary Jewish life), or to historical contact, particularly the German crimes during WWII (e.g., news reports on Shoah remembrance). Notably, mass-mediated present-day contact directly involving Jews, such as talk shows or podcasts, represents a rather recent phenomenon of the last decade, primarily found in niche programs or on social media (Wohl von Haselberg, 2021). In contrast, mass-mediated historical contact is largely institutionalized, and has been present in (West-)Germany since the 1980s (Wenzel, 2022).
Apart from mass-mediated contact, Jewish–non-Jewish interactions in Germany often take two other forms, both related to history: interviews with Shoah survivors (usually organized group events), and extended contact across generations. The latter describes transgenerationally transmitted knowledge about one’s ancestors’ contact with Jews, imparted within families, in school lessons, or in memorials (Stasiuk & Bilewicz, 2013). In summary, Jewish–non-Jewish contact in Germany is thus often indirect and tends to revolve around historical crimes, while occasions for direct present-day contact are rare.
Whereas contact is limited, antisemitism in Germany persists. Recent population representative surveys (Decker et al., 2022) reveal that about 28% of Germans hold negative attitudes toward Jews, or believe in old-fashioned stereotypes like the idea that Jews have too much power. Israel-related antisemitism, attributing antisemitic stereotypes to the state of Israel, holding Jews collectively responsible for the actions of the Israeli government, or questioning Israel’s right to exist, and memory-defensive antisemitism, rejecting ongoing engagement with the Shoah, reversing the perpetrator-victim relation, and accusing Jews of perpetuating the memory for selfish reason, were even endorsed by about 31%-62% of the population. Considering the effectiveness of intergroup contact in reducing (post-conflict) prejudices and historical defensiveness, facilitating such contact may help addressing antisemitism in Germany. The “Meet a Jew” intervention serves to cover this need.
The Present Research Context: The Work of “Meet a Jew”
“Meet a Jew” arranges encounters between tandems of Jewish volunteers and non-Jewish groups all over Germany. All Jews in Germany can get involved as volunteers, with over 300 volunteers aged from 14 years being currently active for the project. The intervention represents a one-time event that typically lasts about 90 minutes and occurs in person, except during the Covid-19 pandemic when it shifted to an online format. Specifically, the intervention involves Jewish volunteers visiting the group of attendees, engaging in a conversation about their (Jewish) lives, answering and posing questions, and discussing shared life topics. The encounters do not follow a standardized script but are individually formed by the volunteers and the attendees. For example, some volunteers may choose to start with a brief presentation about themselves, while in other settings, the attendees may guide the meeting along topics they wish to discuss. Importantly, the role of the volunteers is not to act as experts on Jewish religion and culture, but to give authentic insights into their personal experiences and perspectives.
It is particularly relevant for “Meet a Jew” to avoid status asymmetries and enable the identification of commonalities and shared interests, thus supporting relevant boundary conditions of successful intergroup contact (Allport, 1954). Therefore, the project strives to match volunteers and attendees on the basis of a peer-to-peer approach, taking into account age, regional criteria, and common interests. However, the one-time nature of the encounter limits interaction opportunities, arguably impacting the potential for friendship—another crucial condition for the contact effect (Pettigrew, 1998). While lacking extended time together, the encounter may still foster the detection of similarities and enable interpersonal intimacy (an important element of friendship) by allowing individuals to openly ask questions and receive authentic responses. Especially in a context characterized by limited contact, such meaningful interactions may contribute to creating at least hypothetical friendship potential; the previously unknown group member could become a friend.
Arguably, the contact intervention by “Meet a Jew” then has the potential to reduce barriers to positive intergroup relations via at least three routes relevant in (post-conflict) intergroup contact (Hewstone et al., 2014; Pettigrew & Tropp, 2008): increased intergroup knowledge, reduced anxiety, and perspective-taking. By sharing their everyday experiences, volunteers may expand abstract and history-related knowledge about Jews (Branković & Kranz, 2022). In addition, gaining knowledge may facilitate the identification of similarities, fostering a common ingroup identity (Dovidio et al., 2003). Along these lines, intergroup anxiety may be reduced. Anxiety can be particularly pronounced when there is a high level of unfamiliarity with the other group, or if one group fears being rejected by the other group (Stephan & Stephan, 1985). The latter may be prevalent in contexts marked by historical transgressions. In fact, past perpetration can substantially increase the need of perpetrator group members to be (morally) accepted by the victim group (Shnabel & Nadler, 2008), potentially intensifying the fear of (moral) rejection. Finally, the intervention may enhance understanding of Jewish perspectives, which are often not salient or accessible (Wohl von Haselberg, 2021). Perspective-taking, in turn, can lead to a variety of positive intergroup outcomes, such as increased empathy for the other group’s concerns or a greater overlap in the representations of self and others (Aberson & Haag, 2007).
Methodological Constraints of the Present Field Study
Several scholars have highlighted the importance of complementing experimental approaches with field studies in validating social psychological theories (e.g., Brislin, 1983; Mortensen & Cialdini, 2010). Field research can keep “a hypothesis honest” (Brislin, 1983, p. 372), testing the robustness of phenomena in ecologically more valid settings. While they can thus compensate for some weaknesses of laboratory experimentation, field studies are subject to other relevant weaknesses and real-world constraints. Our research was particularly confronted with three challenges that require careful consideration in contextualizing the present results.
First, in cooperation with a social project, we had to ensure that our research does not impair the project’s work. The intervention offered by “Meet a Jew” represents for many people in Germany a rare opportunity to consciously meet Jewish people. As such, the project strives to create positive conditions for the encounter and minimize additional burdens, such as repeated requests to evaluate it. Moreover, “Meet a Jew” expressed concern that attendees’ awareness of participating in a research study may influence their experience with the intervention (i.e., Hawthorne effect; Sommer, 1968), for instance by creating barriers to asking questions. Thus, we decided to assess data only from those attendees who initiated the intervention, such as teachers reporting on their student group, or initiating sports club members reporting on their fellow group before and after the intervention, as well as at the end of the intervention year.
The second challenge in field research involves the difficulty of eliminating all possible confounders that may influence the effect of interest (Mortensen & Cialdini, 2010). In our research, differences in intergroup attitudes between the before and after or year-end measurements could result from actually enhanced knowledge, reduced anxiety, and perspective-taking. Yet, they may also stem from a confirmation bias (Oeberst & Imhoff, 2023), influencing the initiator’s post-intervention responses to align with their expectations, or strategic considerations. For instance, initiators may actively avoid portraying the attendees as reserved toward Jews to not jeopardize the intervention.
A third challenge concerned participants’ cooperation. Selective non-participation or attrition is not unique to field studies (Zhou & Fishbach, 2016). However, in field research focusing on limited populations and unable to freely recruit additional participants, non-participation can become particularly evident. In our study, over 85% of initiators responded only to one measurement point. The number of responses before and after the intervention was comparable, whereas the number at year-end was half as high. The choice of the measurement time could have been not at random. For instance, initiators perceiving the intervention as highly successful or unsuccessful might have been more motivated to express their opinion afterwards than those with moderate views (Anderson, 1998). Alternatively, practical reasons like time constraints could have led to one-time participations.
Given these limitations, we can only approach an understanding of the effectiveness of the contact intervention. Despite its limitations, fieldwork can corroborate the broader applicability of theories about human behavior. Considering furthermore the need to test existing theories in previously unexplored contexts, we sought to take a first step in that direction within the context of Jewish–non-Jewish contact in Germany.
Present Research
The present fieldwork aims at approaching the impact of a contact intervention on its attendees through the lens of initiators of the intervention, who themselves also attended the intervention. To that end, we applied a mixed-methods approach using a convergent design (i.e., qualitative and quantitative data collection and analysis were conducted in parallel) within the framework of an intervention (Fetters et al., 2013). The quantitative analysis compared the initiator’s impressions of the group of attendees along the dimensions of knowledge, anxiety, and perspective-taking. These findings were complemented by an in-depth qualitative analysis of the positive and critical comments that the initiators received from the attendees about the intervention. This allowed us to gain a more nuanced understanding of the attendees’ perception of the intervention, thus enabling a more profound comprehension of the outcomes. The present work is exploratory (i.e., non-preregistered). We provide Supplementary Online Materials (SOM) comprising additional analyses, materials and sample quotations (see Kazarovytska & Ionescu, 2024).
Method
Dataset
We collected data over the course of one year. The initiators were asked to report their impressions of the intervention and the group of attendees in an online survey at three time points: up to one month before the intervention, up to one month after the intervention, and at the end of the year in which the intervention took place (i.e., between October 2021 and March 2022).2 The period of data collection was thus January 1st 2021-March 31st 2022.
In total, 482 responses of initiators were recorded, with 444 responses containing information beyond the first few demographic and contextual details. As only n = 6 initiators participated at all three time points, and n = 47 at two of the three time points,3 we did not have sufficient data to conduct longitudinal analyses as intended. Therefore, we decided to explore differences between the three time points, using the before time point as a baseline measure. To conduct independent group comparisons, we kept only the response at one time point from each initiator in the dataset. The other responses were removed randomly, but proportionally to the number of responses per time point by an independent coder blind to the research question and all responses except demographic and contextual details (see SOM-A for details on the exclusion procedure). This resulted in a final sample of N = 385 initiator responses, each relating to another intervention (nbefore = 160, nafter = 144, nyear-end = 81). As in 2021 a total of 587 interventions took place, these responses concerned 66% of all interventions.4 A sensitivity analysis in G*Power (Faul et al., 2007) indicated that we had over 80% power to detect differences of Cohen’s d = 0.32 between before- and after-groups, and d = 0.38-0.39 between before- or after- and year-end-groups in independent t-tests (α = .05).
In the following, we report all measures relevant to the present research question aimed at approaching the effectiveness of the intervention for attendees across three time points. These measures were assessed as part of a larger research project, which also included additional questions concerning contextual details of the intervention (e.g., group size), supplementary questions asked only at one time point (e.g., topics associated with Judaism at year-end), as well as questions focused on the initiators’ experiences rather than the attendees’ (e.g., feedback from colleagues or private contacts regarding the initiated encounter), which were not included in the following analyses. For a descriptive report including also these measures, see Ionescu and Kazarovytska (2022).
Quantitative Analysis
Responding to all measures was voluntarily. Unless indicated otherwise, we measured the variables at all three time points, with items ranging from 1 = very little/fully disagree to 7 = very much/fully agree. An overview of all items used to measure the following variables is given in SOM-B.
Mediators
In a first step, we measured variables relevant to the effectiveness of contact interventions, informed by prior literature (Hewstone et al., 2014; Pettigrew & Tropp, 2008). Two items assessed the perspective of initiators on two different facets of attendees’ knowledge: “How much do attendees know about Judaism in general (religion, history, culture...)?”, and “…about today’s Jewish life in Germany?” Perceived anxiety was measured by four items adapted from Stephan and Stephan (1985), e.g., “Attendees [are nervous with regard to/were nervous during] the encounter”; anchored at 1 = fully disagree to 5 = fully agree. Since the anxiety items referred to contact during the intervention, they were assessed only before and after the intervention. Three items adapted from Aberson and Haag (2007) further captured how initiators evaluated the ability of attendees to take the perspective of Jews (e.g., “I think, attendees can understand the perspective of Jews”).
Variable of Interest
Two items measured perceptions of intergroup attitudes (“I think, the attendees’ attitude toward Jews is…”; ranging from 1 = very negative to 7 = very positive, and “I think, the attendees are…toward Jews”; ranging from 1 = very reserved to 7 = very open).
Contextual Information
Furthermore, we assessed information about the context of the intervention, including the institutions initiating the intervention, reasons for initiating the intervention, and topics discussed during the intervention. We also inquired about of the frequency of conflicts on topics related to the intervention (1 = no conflicts to 7 = several conflicts per day), and the age range of attendees. Seven of the ranges formed an ordinal scale (1 = 5-10 years to 7 = 50-60 years or older), whereas the last was a residual category (8 = intergenerational) and thus excluded from all analyses using ordinal age information.
Qualitative Analysis
Adding to the above measures, initiators were asked at all three time points to describe the feedback they received from attendees in two open-ended questions addressing separately positive and negative feedback: “Did attendees make [positive/negative] comments about the [upcoming] encounter? If so, what did they report?” In addition, after the intervention, the initiators had the opportunity to further comment on the attendees’ knowledge about Judaism and Jewish life in Germany.
To analyze the responses to these open-ended questions, we conducted a qualitative content analysis. This approach is situated between the methodological traditions of standardized and reconstructive social research (Kuckartz, 2009). Specifically, we applied the method of content-structuring content analysis (Kuckartz, 2014; Schreier, 2012, 2014). As an interpretative method, it focusses on the reconstruction of characteristic patterns of meaning along pre-defined categories. In our analysis, we deductively selected the two categories, positive and negative feedback (coding frame), as an overarching system to organize the responses (Kuckartz, 2009; Schreier, 2014). These categories were chosen—in correspondence to the survey questions—to comprehensively capture attendees’ experiences and structure them along their contribution to a (non-)successful intervention. In a second step, we applied an inductive, material-driven approach to identify relevant subcategories, representing core themes.
The second author read all open-ended responses and coded them in terms of the two overarching categories. We classified feedback as positive/negative based on its content rather than the question it was responding to, as initiators sometimes provided positive comments when asked for negative feedback, and vice versa. Subsequently, the second author identified initial core themes within these categories. Core themes were qualified by text segments of similar meaning occurring multiple times and thus constituting recurring, characteristic motifs. The first author served as an analytic auditor (Elliott et al., 1999) and critically reviewed the coding frame. Both authors then discussed the initial core themes, evaluating them in terms of their consistency and validity (i.e., whether they adequately reflect the meaning of the material; Schreier, 2014). This led to the modification of the coding frame, and the refinement of themes that lacked coherence. Finally, the second author coded all materials using the revised coding frame.
The approach of a content-structuring content analysis seemed particularly suitable for combining the quantitative and qualitative perspectives, as it can provide a more nuanced understanding of attendees’ feelings and perceptions associated with the intervention. Thus, the results can inform and situate the findings from the quantitative analyses. The data corpus used for the analysis was the entirety of the open comments provided by the initiators included in the quantitative analyses. Aimed at discovering similarities and differences between the groups, we analyzed the responses separately for all three time points.
Results
Results of Quantitative Analysis
Figure 1 outlines the intervention context, displaying the initiating institutions, reasons for initiation as reported before the intervention, and the topics discussed during the intervention as reported after the intervention. The vast majority of interventions took place at schools, and were initiated due to a lack of contact with Jews and curiosity about Jewish life. Commonly discussed topics included Jewish life, holidays, identity, and antisemitism.
Figure 1
Institutions Initiating the Intervention (Upper), Reasons for Initiating the Intervention (Middle), and Topics Discussed (Lower)
Note. In the latter two questions, multiple answers were possible. Percentages indicate the proportion of initiators who chose each answer.
Descriptive statistics, reliabilities and correlations are displayed in Table 1. Consistent with contact literature (Hewstone et al., 2014; Pettigrew & Tropp, 2008), knowledge and perspective-taking were positively, but anxiety was negatively related to positive attitudes toward Jews. Attitudes moreover correlated negatively with the frequency of conflicts on topics related to the intervention, but—unexpectedly—positively with age.
Table 1
Descriptive Statistics, Reliabilities, and Correlations
| Variable | Before | After | Year-End | α | 1 | 2 | 3 | 4 | 5 | 6 | |||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M | SD | M | SD | M | SD | ||||||||
| 1. Knowledge Judaism | 3.26 | 1.34 | 5.15 | 1.17 | 4.69 | 1.25 | – | – | – | – | – | – | – |
| 2. Knowledge Jewish Life Today | 2.30 | 1.19 | 4.96 | 1.31 | 3.86 | 1.34 | – | .77*** | – | – | – | – | – |
| 3. Anxiety | 1.69 | 0.62 | 1.75 | 0.67 | – | – | .66 | -.06 | -.03 | – | – | – | – |
| 4. Perspective-Taking | 4.66 | 1.22 | 5.72 | 1.01 | 5.32 | 1.16 | .88 | .47*** | .56*** | -.22* | – | – | – |
| 5. Attitudes | 5.18 | 0.96 | 5.65 | 0.93 | 5.83 | 0.91 | .81 | .35*** | .35*** | -.25** | .36*** | – | – |
| 6. Age | 2.76 | 1.08 | 2.82 | 1.21 | 3.51 | 1.77 | – | .05 | .07 | -.17 | .20 | .27** | – |
| 7. Conflicts | 1.12 | 0.32 | 1.13 | 0.41 | 1.15 | 0.45 | – | .01 | -.01 | .07 | .01 | -.31*** | -.05 |
Note. We report Pearson’s r, except for correlations with age (ordinal scale) for which we report Spearman’s ρ. p-values were adjusted for multiple comparisons according to Holm’s methods.
*p < .05. **p < .01. ***p < .001.
Further compatible with literature, analyses of variance (Table 2) indicated significant differences between groups in both types of intergroup knowledge, perspective-taking, and attitudes, but—deviating from intergroup contact theory—not in anxiety. Post-hoc tests revealed that after the intervention and at year-end, the levels of knowledge about Judaism and Jewish life today, as well as perspective-taking, were significantly higher than before the intervention. Both types of knowledge and perspective-taking were also higher immediately after the intervention than at year-end. Attitudes, in turn, were more positive in the after- and year-end-groups compared to the before-group, without significant differences between after- and year-end-groups. Eventually, age of attendees was higher at year-end than in the other two groups. Given that higher age correlated with positivity toward Jews, differences to the year-end-group need to be interpreted with particular caution as they may represent an artefact of this correlation.
Table 2
Differences Between Time Points
| Variable | F-Test (η2) | Bonferroni-Holm corrected Post-Hoc Tests (Cohen’s d) | ||
|---|---|---|---|---|
| Before–After | Before–Year-End | After–Year-End | ||
| Knowledge Judaism | F(2, 339) = 78.62*** (.32) | t(250) = 12.16*** (1.51) | t(172) = 7.72*** (1.09) | t(151) = 2.65** (0.38) |
| Knowledge Jewish Life Today | F(2, 337) = 143.70*** (.46) | t(260) = 17.25*** (2.12) | t(145) = 8.37*** (1.24) | t(155) = 5.82*** (0.85) |
| Anxiety | F(2, 237) = 0.21 (.01) | t(237) = 0.46 (0.06) | – | – |
| Perspective-Taking | F(2, 323) = 27.29*** (.14) | t(217) = 7.28*** (0.94) | t(171) = 3.76*** (0.55) | t(141) = 2.53* (0.37) |
| Attitudes | F(2, 346) = 14.18*** (.08) | t(265) = 4.09*** (0.50) | t(171) = 4.88*** (0.69) | t(164) = 1.36 (0.19) |
| Age | F(2, 351) = 8.11*** (.05) | t(270) = 0.46 (0.05) | t(92) = 3.24** (0.56) | t(101) = 2.89** (0.48) |
| Conflicts | F(2, 333) = 0.37 (.00) | t(248) = 0.04 (0.04) | t(197) = 0.54 (0.07) | t(193) = 0.22 (0.03) |
**p < .01. ***p < .001.
In order to delignate which factors might be particularly relevant in understanding differences in attitudes between the before- and after-groups, we performed a mediation analysis using the R package lavaan (Rosseel, 2012). We mean-centered mediator variables and dummy-coded the condition (before as reference category 0) prior to analysis. Confidence intervals for indirect and total effects were calculated using 5000 bootstrap resamples. It is important to note that mediation analyses cannot inform about causality and thus can only be considered as compatible with—but not as proofs of—causal assumptions (Fiedler et al., 2011). The noted limitations of the present fieldwork further complicate the interpretation of causality. Therefore, the following mediation analysis should be considered as a more detailed description of the data, not as proof of causality.
While literature suggests that all three examined mediators are relevant in reducing prejudices (Pettigrew & Tropp, 2008), our data revealed indirect effects (IE) only via perspective-taking, IE = 0.12, 95% CIboot [0.05, 0.20], p = .002, but not knowledge about Judaism, IE = 0.06, 95% CIboot [-0.05, 0.17], p = .279, knowledge about Jewish life today, IE = 0.07, 95% CIboot [-0.08, 0.23], p = .359, or anxiety, IE = -0.00, 95% CIboot [-0.03, 0.02], p = .766. Accompanied by a significant total effect in absence of a direct effect (Figure 2), these results were compatible with perspective-taking fully mediating the difference in attitudes between groups.
Figure 2
Mediation Results (Completely Standardized Solution)
Note. Standard errors are given behind the path estimates in parentheses. Non-significant paths are shaded.
**p < .01. ***p < .001.
Due to the high intercorrelation between the two knowledge dimensions, which may cause problems with multicollinearity, we conducted additional mediations with only one knowledge dimension per model. The pattern of results remained identical, with the only significant indirect effect found for perspective-taking (SOM-C). Additional mediation analyses of before- and year-end-groups supported the key role of perspective-taking, but also of knowledge about Jewish life today (SOM-D). However, as noted, particular caution is warranted in interpretating results including the year-end-group. Considering that perspective-taking may mediate not only the link between intervention and attitudes but also the connection between contact-based knowledge and attitudes (Dovidio et al., 2003), we tested a corresponding sequential mediation model, which was supported (SOM-E).
Finally, we tested whether age of attendees, frequency of conflicts, or macro-level factors such as the electoral success of right-wing populists (on the role of political orientation in the effectiveness of intergroup contact, see Graf & Sczesny, 2019), and the percentage of Jews living in the federal state where the intervention took place (on the relevance of context-level contact effects, see Christ et al., 2014) could moderate the differences between before- and after-groups in positive attitudes, which was not the case (SOM-F).
Results of Qualitative Analysis
The quantitative analyses were accompanied by a content-structuring content analysis, aimed at exploring which aspects of the intervention were particularly noteworthy to attendees, and thus could be relevant in situating the intervention outcomes. We present supporting sample quotations we consider illustrative for relevant patterns of meaning and mark them by adding the first letter of the group they refer to (B for before, A for after, AK for additional knowledge-related comments after the intervention, Y for year-end) and the category they belong to (p for positive, n for negative). An overview of categories, themes, and additional example quotations is given in Figure 3. All quotations excerpted are given in full in SOM-G. The quotations were translated into English by the authors.
Figure 3
Overview of Categories, Themes, and Additional Sample Quotations
Before: “Great Excitement Because Never Had Contact With Jewish Life Before”
Before the intervention, n = 55 initiators reported that the majority of attendees expressed positive views about the upcoming encounter. According to the initiators’ comments, the attendees expressed interest, curiosity, and expectancy regarding the intervention (B_p1; B_p3-B_p6; B_p8). One initiator described, “great excitement” because the attendees “never had contact with Jewish life before” (B_p7). Another comment explained, “the group likes the idea of experiencing Jewish culture without—as is usual in school—focusing on catastrophes (persecution of Jews in the Middle Ages, Shoah)” (B_p2). This highlights the desire to learn about contemporary Jewish life.
In contrast, four initiators described negative statements from attendees about the upcoming intervention: Two initiators reported disinterest regarding to the encounter (B_n1; B_n2). One initiator highlighted the “critical view of Israeli politics by Muslim students” (B_n3). A fourth initiator revealed that some attendees expressed demand for historical closure (Kazarovytska & Imhoff, 2024), “they ‘do not want to be constantly confronted with Germany’s Nazi past.’” (B_n4). This perspective indicates that Jewish life in Germany is perceived as clearly connected to the historical crimes committed, with the confrontation of this past apparently being linked to Jewish individuals.
After: “The Encounter Has ‘Opened Doors’”
A total of n = 112 initiators reported (very) positive comments made by attendees after the intervention. Responses within the category of positive feedback could be structured along three themes. One theme pertained to the value of a real-life encounter (Theme A_p1), which broadened the attendees’ understanding of Jewish life and “has ‘opened doors’” (A_p1). For instance, the responses highlighted, “all participants, without exception, expressed positive opinions. The encounter had been very lively and informative, and had made them aware of aspects that many had not previously thought of” (A_p2), or, “they were enthusiastic to have met Jewish people ‘in real life’” (A_p3), which they never had before (A_p4).
Initiators also drew attention on the relevance of perspective-taking (Theme A_p2). For example, one initiator emphasized, “after the encounter, the attendees were able to put themselves much better into the perspective of Jews” (A_p6; see also A_p7; A_p9). Another response explicated, “exciting input, new insights, completely new perspectives; one became more aware of one’s own clichés [and gained a] new view on [the] media” (A_p8). At the same time, initiators underscored that the volunteers and their lives were perceived as “normal” and “like everyone else” (A_p16), or, “virtually no different from themselves” (A_p17). Finally, the openness of the volunteers (Theme A_p3) to talk about themselves and answering questions was described repeatedly as particularly noteworthy (A_p11-A_p14).
Compared to the positive category, the number of responses within the category of negative feedback was limited (n = 16). Responses concerned several different aspects of the encounter. Some initiators reported that the attendees expressed their regret regarding the online format due to the Covid-19 pandemic. A “real” offline meeting would have been more intense (A_n2; A_n5; A_n9). Furthermore, the limited duration of the encounter was critically reflected. There was not enough time for all questions and topics (A_n3; A_n6; A_n7). Another critical comment pointed out, “the speeches of the volunteers were too long, so that there was not enough room for all questions” (A_n1). Initiators also reported that attendees found the meeting boring (A_n11; A_n14), that the answers were perceived as “not very diverse and therefore a little exhausting” (A_n10), that the responses of the Jewish volunteers were difficult to follow (A_n16), or that not enough pictures and objects were shown (A_n15). Moreover, one group of attendees expressed a preference for engaging with younger Jewish volunteers (A_n8). Finally, one initiator reported that some Muslim attendees criticized the Jewish volunteers for “avoid[ing] the topic of Israel-Palestine” (A_n12).
In addition, the initiators were given the opportunity to comment on the attendees’ level of knowledge. Responses (n = 44) could be organized in two themes. First, the initiators emphasized the great learning gain (Theme AK_1), reflected in a (large) increase in knowledge (AK_ p2-AK_p4). For instance, one initiator explained, “we learned a lot and the meeting was highly interesting” (AK_p1). Particularly noteworthy, however, were the responses referring to the primacy of (first) direct experiences and current perspectives (Theme AK_2) as compared to gaining knowledge. This is even more remarkable as the initiators were asked to explicitly comment on the knowledge of the attendees. Specifically, initiators emphasized that experiencing contact with Jews and understanding current Jewish perspectives was even more important for the attendees, than gaining knowledge about Judaism (AK_p6; AK_p7; AK_p9). The “very concrete insight into the current life of Jewish people in Germany” that could be gathered was “the greatest profit of the encounter” (AK_p8). This was backed up by responses highlighting that the interaction provided perspectives that went beyond information from the literature and museums (AK_p11; AK_p12).
Year-End: “They Wish They Could Have More Encounters Like This!”
At year-end, we identified four themes within the category of positive responses (n = 66). First, the opportunity to learn about present Jewish life (Theme Y_p1) was evaluated positively (Y_p2; Y_p3). One initiator wrote, “it was the only way to get a glimpse of Jewish life in Germany today” (Y_p1). This parallels the desire to learn more about contemporary Jewish life observed in the responses before the intervention. A further aspect repeatedly highlighted was the young age of the Jewish volunteers, along with the match in age (Theme Y_p2) between volunteers and attendees (Y_p7; Y_p8), which allowed “meeting at eye level” (Y_p5) and reduced the “barrier to ask[ing] questions” (Y_p6).
Replicating Theme A_p3, initiators reported that the attendees were impressed by the openness of the volunteers (Theme Y_p3), which allowed for interesting responses that were “true-to-life” (Y_p11). One initiator elaborated, “for the more rural, small-town youth, it was their first encounter, and they were impressed by the unforced atmosphere and friendly, open interaction” (Y_p10). Further, the encounter enabled understanding of Jewish perspectives on current political events, such as the emergence of Covid-related antisemitic conspiracy theories on social media: “No one was aware before what that means for Jewish people” (Y_p12). Supplementing Theme AK_1, responses highlighted the learning gain (Theme Y_p4; Y_p17), particularly regarding the diversity of Jewish life in Germany (Y_p14-Y_p16).
There were seven negative comments. Besides four comments underscoring that the meeting format, or the volunteers, did not meet the attendees wishes or expectations (Y_n4; Y_n5; compare also Figure 3), critical responses related to the time frame. Two attendees thought the intervention was too long (Y_n2), while one group of attendees would have preferred to have more time (Y_n3). Finally, one negative response commented on the effect one volunteer had on the attendees: “One of the two Jewish people did not seem listening and understanding, but arrogant and judgmental” (Y_n7). It is unclear to what extent this negative perception made the encounter more difficult or obstructed it. Yet, the comment reflects that the way Jewish volunteers are perceived and evaluated seems to matter.
Discussion
Contact between Jewish and non-Jewish individuals in Germany received limited empirical attention, and there is a lack of data and systematic research on respective intergroup encounters (Hübscher et al., 2022). The present work contributes toward filling this gap. Combining quantitative and qualitative analyses, we examined field data from a real-world, cross-generational intervention program. Results underscore the potential of Jewish–non-Jewish encounters in fostering positive attitudes toward Jews. Our results further suggest that particularly the opportunity to get into direct contact and gain insights into Jewish perspectives are a crucial component of successful intergroup contact.
Implications
Our findings offer important theoretical and empirical contributions to the literature on intergroup contact. While compatible with the general assumption that intergroup contact can effectively tackle prejudices, our results provide a more nuanced view on potential factors conducive to this effectiveness, partly deviating from summative findings (Pettigrew & Tropp, 2008) and theorizing on intergroup contact in conflicted contexts (Hewstone et al., 2014). Diverging from the literature, we found no significant difference in intergroup anxiety before and after the intervention. Instead, the content analysis revealed that already before the intervention, the attendees expressed various positive emotions with regard to the upcoming contact, including curiosity and excitement. Considering that curiosity may increase desire for intergroup contact, thereby alleviating anxious concerns (Migacheva & Tropp, 2013), intergroup anxiety may have been rather low from the outset. Low anxiety levels could also result from a self-selection bias. As noted by Kauff et al. (2021), anxiety can cause individuals to avoid intergroup group contact. Consequently, particularly anxious groups may not have been represented in our sample.
Knowledge about Judaism and Jewish life today as well as perspective-taking, in turn, differed significantly between before and after groups. The content analysis elaborated on these findings. Mirrored in the main themes after the intervention, for many attendees the intervention represented their first conscious encounter with Jews, bringing a substantial increase in intergroup knowledge.
However, a further key theme concerned the primacy of direct experiences and insights into current perspectives: Even more than acquiring knowledge, the attendees seemed to value learning about Jewish perspectives and gaining firsthand insights into the lived experiences of Jews in Germany. Notably, mediation results also reflected a primacy of perspective-taking. Despite the initial positive correlation between both knowledge types and positive intergroup attitudes, perspective-taking remained the only significant intermediate factor linking the intervention to intergroup attitudes. Dovidio and colleagues (2003) proposed an interrelation between knowledge and perspective-taking, suggesting that knowledge can enhance understanding and empathy, which then foster positive relations. Indeed, compatible with this reasoning, we found a positive relationship between knowledge and favorable attitudes, mediated through perspective-taking.
The particular relevance of perspective-taking in the effectiveness of intergroup contact aligns with results from other (post-)violent contexts: Perspective-taking played an important mediating role in the contact effects among Greek and Turkish Cypriots (Husnu & Crisp, 2015), in the context of Bosnia and Herzegovina (Čehajić & Brown, 2010), and in the Polish–Ukrainian borderland region (Stasiuk & Bilewicz, 2013). Especially in post-violent contexts, awareness of the other group’s view on the past may promote a shared understanding of history and thus reduce negative sentiments. However, adding on the research in post-genocidal contexts that has demonstrated the relevance of promoting historical perspective-taking (Bilali & Vollhardt, 2013), our results illustrate the relevance of taking present prospects. In contexts like Germany, characterized by strong presence of rather established historical narratives (Lienen & Cohrs, 2021), but lack of access to Jewish views on current issues, it may be particularly the possibility to get to know present-day perspectives that constitutes a key factor for effective intergroup contact.
The finding that attendees particularly valued learning about current Jewish life—which contrasts with predominantly history-related mass-mediated contact (Wohl von Haselberg, 2021)—aligns with research indicating that the content of intergroup contact can decisively impact its outcomes. Bilewicz (2007) found that Jewish-Polish relations improved only after encounters discussing present-day issues, not historical ones. The author argues that present-day discussions, as compared to historical discussions, had greater potential to foster the identification of commonalities and a sense of shared identity. Alternatively, discussing history may have highlighted the ingroup’s past moral misdeeds (i.e., the Polish passivity or complicity during the Shoah), eliciting defensive reactions against confronting the past and negative attitudes toward the former victim group.
In the context of Jewish–non-Jewish contact in Germany, both explanations seem applicable. On the one hand, several comments after the intervention emphasized the attendees’ enthusiasm about finding similarities with Jewish people and their lives today. These responses seem compatible with the potential of present-day contact to foster a shared identity. On the other hand, comments from the before-group indicated that some attendees looked forward to not discussing Jewish persecution, or even were less excited about the upcoming encounter, precisely because they associated contact with Jews with a confrontation of Germany’s perpetrator past. As such, the attendees’ notable emphasis on the present-day focus of the intervention may, in part, be indicative of defensiveness against engagement with the past (Kazarovytska & Imhoff, 2024).
Our results also bear practical implications. To counter antisemitism, political institutions have proposed improving knowledge about Judaism. For instance, the Organization for Security and Cooperation in Europe published in 2019 the teaching aid “Increasing Knowledge about Jews and Judaism.” Likewise, the “National Strategies against Antisemitism and for Jewish Life” developed by German Government in 2022 propose to enhance knowledge about Judaism and Jewish history as a central strategy in preventing antisemitism. Our results suggest that such attempts may benefit from enriching this knowledge by current Jewish perspectives. Although our data clearly indicated a preference for direct contact over receiving information from public sources (on the specific potential of direct contact in reducing prejudices, as compared to mass-mediated contact, see also Graf & Sczesny, 2019), direct encounters cannot always be realized. When encounters are not feasible, utilizing materials presenting current Jewish perspectives and contemporary Jewish life could, potentially, serve as an indirect alternative (Christ et al., 2014).
However, intergroup contact should not be considered a panacea against negative attitudes toward Jews. As evidenced by the negative responses in our data, the contact may be perceived as boring, exhausting, or uncomfortable, and the Jewish interaction partners may be viewed as arrogant or unsympathetic. It is inherent to the nature of intergroup contact that it is not universally positive. Negative contact, in turn, has the potential to exacerbate negative attitudes (Paluck et al., 2019). As such, contact can be characterized “as a double-edged sword” (Graf & Sczesny, 2019, p. 106) capable of yielding positive intergroup outcomes, but not guaranteed to do so.
Limitations and Directions for Future Research
Our results rely on the perspectives of the intervention initiators regarding the attendees. Hence, our data cannot provide direct access into the attendees’ actual perspectives. This additional interpretative layer requires careful consideration when drawing conclusions from these results. Critically, initiators may not accurately report on attendees’ views but may reproduce their own impressions. As contact is typically initiated by individuals with rather positive intergroup views (Kauff et al., 2021), these impressions may be positively biased, which further limits the generalizability of our results. If possible, future studies should collect data directly from attendees to gain more precise insights into the effectiveness of the intervention and explore its generalizability across different ages, levels of prior knowledge, and political orientations.
Our analysis is also limited to only three mediators. Although the chosen mediators are highly prominent in the literature (Pettigrew & Tropp, 2008) and are among the mechanisms considered particularly relevant in post-conflict societies (Hewstone et al., 2014), a variety of further factors may explain the effectiveness of contact interventions (for an overview, see e.g., Boin et al., 2021). To expand the understanding of the mechanisms underlying the present contact intervention, future studies could investigate, for instance, the role of tolerant norms (Lienen & Cohrs, 2021) and meta-stereotypes (i.e., beliefs about negative stereotypes held by Jews about non-Jewish people; Boin et al., 2021), which both may be salient against the backdrop of Germany’s past perpetration.
Future research may also devote more attention to subtyping effects. Interactions with group members considered atypical of their group may trigger cognitive separation between the atypical exemplar and the broader group. This may hinder generalization of contact-based positive intergroup attitudes (Pettigrew, 1998). In the current context, subtyping effects may have occurred since “Meet a Jew” explicitly aims at introducing individual Jewish experiences. These individual experiences may be contrasted from a stereotypical representation of ‘the entire Jewish group.’ As a result, potential positive effects of the intervention may not extend to Jews who did not participate in the intervention.
Eventually, experiences from a contact intervention may not readily transfer to everyday experiences. As argued by O’Donnell and colleagues (2021), the nature of contact set up in an intervention often differs from everyday contact situations. Routine contact may be brief or trivial, with contextual factors like an individual’s professional role as a colleague outweighing their religious or ethnic identity—identity aspects that are highly present in an intervention. Therefore, findings from contact intervention research may inform only to a limited extent our understanding of intergroup dynamics in everyday contexts.
Conclusion
Germany represents a context where the history of past atrocities looms large, but contact with contemporary Jewish life is limited. In this regard, Germany shares similarities with other (post-)conflict contexts. Our results suggest that in such contexts, attempts to promote positive intergroup attitudes may particularly benefit from incorporating not only historical but also present-day perspectives from the respective other group. However, perspective-taking, just as intergroup contact itself, does not inevitably lead to positive intergroup relations in all settings (Bilali & Vollhardt, 2013; Graf & Sczesny, 2019). Therefore, analyzing the effectiveness of contact interventions within specific intergroup contexts becomes particularly relevant for advancing our theoretical and empirical knowledge in this area.