The negative secondary transfer effect: underlying processes and facilitating conditions

Abstract

Background The Secondary Transfer Effect (hereafter: STE) describes how contact with members of a (primary) outgroup alters also evaluations of uninvolved, secondary outgroups. Positive contact leads to better evaluations (positive STE), negative contact to worse evaluations (negative STE). Literature proposes three different kinds of underlying mechanisms (in methodological terms mediators). For mediators involving the outgroup (like ‘threat’ and ‘attitude generalization’), contact affects secondary outgroup evaluations by shifting primary outgroup evaluations. For mediators involving the ingroup (like ‘ingroup pride’), contact affects secondary outgroup evaluations by shifting ingroup evaluations. For mediators involving the self (like ‘multiculturalism’), contact affects secondary outgroup evaluations by shifting the ideological worldview. Literature also suggests facilitating conditions for STE processes. Similarity in stereotype content associated with involved outgroups would facilitate threat or attitude generalization, and salience of recalled cultural elements STE via multiculturalism. Objective This dissertation tackles several gaps in STE research. Research gap 1 concerns the lack of research on negative STE. Research gap 2 concerns the recurring practice to study STE mediators cross-sectionally (which cannot disentangle implied causal sequences), focus on few theorized processes beyond attitude generalization, and study them one-at-a-time (which makes it difficult to disentangle or compare them). Research gap 3 concerns the fact that above mentioned facilitating conditions for STE processes were rarely empirically investigated in comparative study designs. Lastly, emergent results may help tackle the open question of whether positive and negative STE occur oppositely via the same processes. Methods Using secondary survey data, three studies are conducted. They draw upon a sample of native Germans who took part in the German ALLBUS survey and subsequently three waves of the GESIS panel. This enabled a multi-context investigation by three different primary outgroups: Foreigners living in Germany (Study 1; N= 1553 & Study 2; N= 390), Muslims living in Germany (Study 3, scenario A; N= 385) or Sinti & Roma living in Germany (Study 3, scenario B; N= 396). Data collection happened between 2015 and 2017 during the so-called 'Refugee crisis'. Refugees are thus always the secondary outgroup. Study 1 investigates positive and negative STE cross-sectionally via attitude generalization, multiculturalism, and ingroup pride within a parallel mediation path model using manifest indicators. Open answers to a question on conception of the group lable ‘Foreigners’ aided identifying respondents that perceived primary and secondary outgroup as overlapping. Study 2 longitudinally investigates STE via attitude generalization, multiculturalism, and threat within a parallel mediation CLPM with manifest indicators, accompanied by power analyses. Study 3, replicates study 2 within two further intergroup contexts, varying by primary outgroup. External classification frameworks are employed to assess similarity in stereotype content and salience of recalled cultural elements, and then hypothesize about the emergence of STE mediators. All studies include a control measure of secondary outgroup contact. Results Study 1 finds cross-sectional evidence for positive and negative STE, directly or indirectly via attitude generalization and multiculturalism (but not via ingroup pride). Study 2 longitudinally replicates finding the STE mechanisms from study 1, but finds no evidence for STE via primary outgroup threat. Scenario A of study 3 (primary outgroup 'Muslims') finds longitudinal evidence for positive and negative STE via multiculturalism (but not via attitude generalization or threat) and direct positive (but not direct negative) STE. Scenario B of study 3 (primary outgroup 'Sinti & Roma') finds longitudinal evidence for direct positive and negative STE, but no evidence for STE via attitude generalization, threat or multiculturalism. Findings from study 3 did not always reflect the hypothesized similarity-based emergence of mediators. Statistically significant longitudinal STE paths had satisfactory statistical power. Discussion & conclusion In summary this dissertation helps tackle the identified research gaps within the conducted studies: more research on negative STE is conducted (c.f. Research Gap 1), whereby longitudinal study designs are utilized (c.f. Research Gap 2) and a multi-context investigation that uses similarity classification systems to empirically investigate theorized facilitating conditions (c.f. Research Gap 3). All studies found empirical evidence for negative STE within the context of the ‘Refugee Crisis’ in Germany, where individuals likely generalized from past contact experiences with other migrant outgroups to form an opinion on the newly arrived refugees. Negative STE thus seems to be a phenomenon that may co-occur with positive STE and (as the results indicate) via the same processes but in the opposite direction. These processes may include direct STE, attitude generalization or multiculturalism (which could newly be established as a mediator of negative STE). No indication for STE via threat or ingroup pride emerged in the investigated scenarios. However, the results indicate a contextually instable nature of STE. It remains yet unclear whether this contextuality can indeed be systematically linked to similarity in stereotype content as theory proposed. Methodologically refined further comparative experimental research seems necessary.