Table 2. Multiple regression of int

Table 2. Multiple regression of intergroup perception predictors on attitudes toward majority integration efforts (N = 503)

Standardized coefficients (β) t-value p-value Part correlation
Counterstereotypic intentions (CSI) .60 16.80 .000 .47
Counterstereotypic competence (CSC) .08 2.42 .016 .07
Perceived entitativity (E) –.20 –5.80 .000 –.16
Metawarmth (MW) –.09 –2.78 .006 –.08
Metacultural competence (MCC) .01 0.29 .769 .00
Metageneral competence (MGC) .05 1.83 .067 –.05
Notes. R2 = .60, F(6, 506) = 126.12, p < .001, adjusted R2 = .597.


Table 3. Interaction effects of warmth/competence on attitudes toward majority integration efforts: counter stereotypic information and meta perspectives. Standardized coefficients listed (N = 519)

Independent measure Block Warmth Competence Interaction W × C R2 ΔR2
Counterstereotypic information 1 0.587*** 0.113*** 0.566
2 0.578*** 0.114*** –0.06* 0.570 0.044*
Metaperspectives 1 –0.088* 0.204*** 0.040
2 –0.087* 0.185*** –0.119** 0.054 0.014**
Notes. *p < .05, **p < .01, ***p < .001, all variables are mean centered.

and Counterstereotype Competence (r = .40, p < .001), in- dicating that seeing immigrants as having positive integra- tion intentions (warmth) was associated with also seeing them as competent. In contrast, we found significant neg- ative correlations between perceived entitativity and count- erstereotypic intentions (r = –.55, p < .001) and counterste- reotype competence (r = –.32, p < .001) respectively. This suggests that the more majority members perceived immi- grants as constituting an entitative group, the less they agreed with counterstereotypic information portraying im- migrants as having positive integration intentions or as competent members of the Norwegian society.
A standard multiple regression analysis was performed
to assess the utility of the intergroup perception variables as predictors of majority members’ proactive integration attitudes and, in addition, to identify which variables were most strongly related to MIE attitudes. Analyses were con- ducted to ensure that the assumptions of regression analysis were fulfilled (normality of residuals, linearity, multicollin- earity, and homoscedasticity). All intergroup perception variables were included as independent variables with MIE scores as the dependent variable. Table 2 reports standard- ized regression coefficients, t-values, p-values, and part correlations for all independent variables. 60% of variance in MIE scores was explained by the total set of intergroup perception variables, F(6, 506) = 126.12, p < .001. The per- ceptions of majority members of the integration intentions of immigrants (CSI) was the strongest predictor of MIE attitudes (β = .60, p < .001). This suggests that, if majority members perceived immigrants to have positive intentions to integrate, they were more likely to be in favor of proac- tive integration. The next strongest predictor was perceived

entititavity (β = –.20, p < .001), suggesting that a higher perception of immigrant group cohesiveness predicted an unwillingness to favor MIE attitudes. Other measures that were statistically significant were Counterstereotypic Competence (β = .08, p < .05) and Metawarmth (β = –.09, p < .01).
To better understand key relationships observed in the previous analyses, we examined the potential interaction effects of warmth and competence, entitativity, and meta- perception. All independent variables were first mean-cen- tered to aid in the interpretation of potential effects (see e.g., Aiken & West, 1991). Because the stereotype content model (Fiske et al., 2007) states that stereotypes of different groups may vary due to different combinations of per- ceived warmth and competence, we first examined whether interactions between warmth and competence of counter- stereotypic measures and metaperspectives explained ad- ditional variance in MIE attitudes. Next, we explored whether the predictive effect of entitativity on MIE atti- tudes might also interact with the counterstereotypic mea- sures based upon the observed correlations between the measures (Table 1) and (theoretical) primacy of the univer- sal dimensions of social cognition.
Interaction variables were computed by multiplying to- gether warmth and competence measures of counterstereo- typic and, metaperspectives, respectively,1 and each count- erstereotypic measure and entitativity. We then performed hierarchical regressions in which the original variables for each measure were entered in the first step and then con- trolled for in the second step as the new interaction measure was added. We found a small, but statistically significant, increase in the amount of explained variance when we in-

The independent variable metacultural competence was omitted from these analyses as its content was deemed too specific to be included in the interaction variable and in addition made little contribution to the first multiple regression analysis.




Table 4. Hierarchical regression of effects of entitativity, counterstereotypic intentions, and competence, and counterster- eotypic interaction of intentions and competence on MIE attitudes. All variables are mean centered (N = 515)

Independent measure Block Standardized regression coefficient R2 ΔR2
Entitativity 1 –0.546*** .309
Entitativity 2 –0.200*** .591 .281***
Counterstereotypic intentions 2 0.604***
Counterstereotypic competence 2 0.077*
Entitativity 3 –0.196*** .594 .003
Counterstereotypic intentions 3 0.597***
Counterstereotypic competence 3 0.078*
Counterstereotypic interaction (I × C) 3 –0.054
Notes. *p ≤ .05, **p ≤ .01, ***p ≤ .001, all variables are standardized.



















Figure 1. The impact of entitativity, counterstereotypic intentions, and competence on MIE attitudes. Standardized coef- ficients listed. *p < .05, **p < .001.


cluded the interaction effects of counterstereotypic mea- sures and metaperspectives in the models, albeit no inter- action effects for entitativity. Table 3 presents the statisti- cally significant interaction effects.
Both significant interaction effects had negative signs; for the counterstereotypic β = –.06 (p < .05) and for meta- perspective β = –.119 (p < .01). To interpret these findings,
the predicted values for the relationship between warmth and MIE attitudes for low, moderate, and high competence values were plotted in a chart for both measures. In both models, the relationship was strongest between warmth and MIE attitudes (i.e., regression line with the steepest slope) when competence was perceived as low.
However, despite achieving statistical significance, the interaction effects did not provide a substantial change to the amount of explained variance in MIE attitudes. In ad- dition, as is apparent in Tables 2 and 3, the metaperspec- tives variables did not make much substantial contribution to predicting MIE attitudes. This seemed to suggest that of the observed main effects, endorsement of counterstereo- typic portrayals of intentions (CSI) and competence (CSC), and perceived immigrant entitativity were the most mean- ingful predictors of MIE attitudes.

To further clarify the relationship between entitativity and measures of warmth (intentions) and competence, we also performed a mediation analysis. As Table 1 shows, perceived entitativity was significantly correlated with both the mediation variables (CSI and CSC) as well as the outcome variable MIE. Both mediation variables were also correlated with the outcome, meaning that steps 1, 2, and
3 in Baron and Kenny’s (1986) mediation analysis were met. Step 4 means that the relationship between perceived entitativity and MIE attitudes should be weakened when controlling for the mediation variables. Therefore, hierar- chical multiple regression analysis was performed to test for potential mediation effects and is presented in Table 4. In the first block, entitativity was entered as the only pre- dictor, counterstereotypic intentions and competence in block 2, and the interaction effect between warmth and competence in block 3. After we entered the mediation variables in the model, the regression coefficient for the relationship between entitativity and MIE attitudes dropped from β = –.546 to β = –.200 (p ≤ .001). This sug- gests that while entitativity still had an independent predic- tive value for MIE attitudes, much of its effect was medi- ated by perceptions of the counterstereotypic intentions




and competence of immigrants. The final model showed that counterstereotypic intentions remained the best indi- vidual predictor of MIE attitudes, followed by entitativity and counterstereotypic competence. The counterstereotyp- ic warmth/competence interaction variable did not make a statistically significant contribution (p = .06) to the overall model and seems to support the claims above regarding the predictive value of individual independent measures as op- posed to interaction variables.
To better illustrate the mediation effects, we present the
relationships between the four variables as a path diagram in Figure 1. As can be seen, the effect of entitativity on MIE attitudes is primarily mediated through counterstereotypic intentions. Sobel’s test for mediation (Sobel, 1982) showed that this mediation effect was statistically significant (Z =
–11.23, p < .001). Although the mediation effect of count- erstereotypic competence was much weaker, Sobel’s test of mediation for this effect was also significant (Z = –2.36, p
= .018).



Discussion

The present research advances the point of vie

Hasil (Bahasa Indonesia) 1:

Table 2. Multiple regression of intergroup perception predictors on attitudes toward majority integration efforts (N = 503) Standardized coefficients (β) t-value p-value Part correlationCounterstereotypic intentions (CSI) .60 16.80 .000 .47Counterstereotypic competence (CSC) .08 2.42 .016 .07Perceived entitativity (E) –.20 –5.80 .000 –.16Metawarmth (MW) –.09 –2.78 .006 –.08Metacultural competence (MCC) .01 0.29 .769 .00 Metageneral competence (MGC) .05 1.83 .067 –.05 Notes. R2 = .60, F(6, 506) = 126.12, p < .001, adjusted R2 = .597.Table 3. Interaction effects of warmth/competence on attitudes toward majority integration efforts: counter stereotypic information and meta perspectives. Standardized coefficients listed (N = 519)Independent measure Block Warmth Competence Interaction W × C R2 ΔR2Counterstereotypic information 1 0.587*** 0.113*** 0.566 2 0.578*** 0.114*** –0.06* 0.570 0.044*Metaperspectives 1 –0.088* 0.204*** 0.040 2 –0.087* 0.185*** –0.119** 0.054 0.014** Notes. *p < .05, **p < .01, ***p < .001, all variables are mean centered. dan kompetensi Counterstereotype (r =.40, p <.001), melihat imigran memiliki niat integra-tion positif (kehangatan) itu terkait dengan juga melihat mereka sebagai kompeten di dicating. Sebaliknya, kami menemukan korelasi signifikan neg-mengaktifkan antara dianggap entitativity dan niat hitung-erstereotypic (r = –. 55, p <.001) dan kompetensi counterste-reotype (r = –. 32, p <.001) masing-masing. Hal ini menunjukkan bahwa mayoritas anggota lain dianggap immi-hibah sebagai merupakan sebuah kelompok entitative, yang kurang mereka setuju dengan counterstereotypic informasi yang menggambarkan im-migran memiliki niat integrasi positif atau sebagai anggota masyarakat Norwegia yang kompeten.Standar dilakukan analisis regresito assess the utility of the intergroup perception variables as predictors of majority members’ proactive integration attitudes and, in addition, to identify which variables were most strongly related to MIE attitudes. Analyses were con- ducted to ensure that the assumptions of regression analysis were fulfilled (normality of residuals, linearity, multicollin- earity, and homoscedasticity). All intergroup perception variables were included as independent variables with MIE scores as the dependent variable. Table 2 reports standard- ized regression coefficients, t-values, p-values, and part correlations for all independent variables. 60% of variance in MIE scores was explained by the total set of intergroup perception variables, F(6, 506) = 126.12, p < .001. The per- ceptions of majority members of the integration intentions of immigrants (CSI) was the strongest predictor of MIE attitudes (β = .60, p < .001). This suggests that, if majority members perceived immigrants to have positive intentions to integrate, they were more likely to be in favor of proac- tive integration. The next strongest predictor was perceived entititavity (β = –.20, p < .001), suggesting that a higher perception of immigrant group cohesiveness predicted an unwillingness to favor MIE attitudes. Other measures that were statistically significant were Counterstereotypic Competence (β = .08, p < .05) and Metawarmth (β = –.09, p < .01).To better understand key relationships observed in the previous analyses, we examined the potential interaction effects of warmth and competence, entitativity, and meta- perception. All independent variables were first mean-cen- tered to aid in the interpretation of potential effects (see e.g., Aiken & West, 1991). Because the stereotype content model (Fiske et al., 2007) states that stereotypes of different groups may vary due to different combinations of per- ceived warmth and competence, we first examined whether interactions between warmth and competence of counter- stereotypic measures and metaperspectives explained ad- ditional variance in MIE attitudes. Next, we explored whether the predictive effect of entitativity on MIE atti- tudes might also interact with the counterstereotypic mea- sures based upon the observed correlations between the measures (Table 1) and (theoretical) primacy of the univer- sal dimensions of social cognition.Interaction variables were computed by multiplying to- gether warmth and competence measures of counterstereo- typic and, metaperspectives, respectively,1 and each count- erstereotypic measure and entitativity. We then performed hierarchical regressions in which the original variables for each measure were entered in the first step and then con- trolled for in the second step as the new interaction measure was added. We found a small, but statistically significant, increase in the amount of explained variance when we in- The independent variable metacultural competence was omitted from these analyses as its content was deemed too specific to be included in the interaction variable and in addition made little contribution to the first multiple regression analysis. Table 4. Hierarchical regression of effects of entitativity, counterstereotypic intentions, and competence, and counterster- eotypic interaction of intentions and competence on MIE attitudes. All variables are mean centered (N = 515)Independent measure Block Standardized regression coefficient R2 ΔR2Entitativity 1 –0.546*** .309 Entitativity 2 –0.200*** .591 .281***Counterstereotypic intentions 2 0.604*** Counterstereotypic competence 2 0.077* Entitativity 3 –0.196*** .594 .003Counterstereotypic intentions 3 0.597*** Counterstereotypic competence 3 0.078* Counterstereotypic interaction (I × C) 3 –0.054 Notes. *p ≤ .05, **p ≤ .01, ***p ≤ .001, all variables are standardized.Figure 1. The impact of entitativity, counterstereotypic intentions, and competence on MIE attitudes. Standardized coef- ficients listed. *p < .05, **p < .001. cluded the interaction effects of counterstereotypic mea- sures and metaperspectives in the models, albeit no inter- action effects for entitativity. Table 3 presents the statisti- cally significant interaction effects.Both significant interaction effects had negative signs; for the counterstereotypic β = –.06 (p < .05) and for meta- perspective β = –.119 (p < .01). To interpret these findings,nilai diperkirakan untuk hubungan antara kehangatan dan MIE sikap untuk nilai-nilai kompetensi rendah, sedang, dan tinggi sedang diplot di bagan untuk langkah-langkah kedua. Dalam kedua model, hubungan itu terkuat antara kehangatan dan sikap MIE (yaitu, regresi baris dengan lereng rel) ketika kompetensi dianggap sebagai rendah.Namun, meskipun mencapai signifikansi statistik, efek interaksi tidak memberikan perubahan substansial jumlah varians dijelaskan dalam sikap MIE. Dalam iklan-dition, seperti yang terlihat di tabel 2 dan 3, Variabel yang metaperspec-tives tidak membuat banyak kontribusi substansial untuk memprediksi MIE sikap. Ini tampaknya menyarankan bahwa efek utama diamati, dukungan dari counterstereo - typic penggambaran niat (CSI) dan kompetensi (CSC), dan dianggap entitativity imigran adalah prediktor berarti-ingful kebanyakan MIE sikap. Untuk lebih memperjelas hubungan antara entitativity dan langkah-langkah kehangatan (niat) dan kompetensi, kami juga melakukan analisis mediasi. Sebagai tabel 1 menunjukkan, dirasakan entitativity adalah signifikan berkorelasi dengan kedua mediasi variabel (CSI dan CSC) serta hasil variabel MIE. Kedua variabel mediasi yang juga berkorelasi dengan hasil, berarti bahwa langkah 1, 2, dan3 in Baron and Kenny’s (1986) mediation analysis were met. Step 4 means that the relationship between perceived entitativity and MIE attitudes should be weakened when controlling for the mediation variables. Therefore, hierar- chical multiple regression analysis was performed to test for potential mediation effects and is presented in Table 4. In the first block, entitativity was entered as the only pre- dictor, counterstereotypic intentions and competence in block 2, and the interaction effect between warmth and competence in block 3. After we entered the mediation variables in the model, the regression coefficient for the relationship between entitativity and MIE attitudes dropped from β = –.546 to β = –.200 (p ≤ .001). This sug- gests that while entitativity still had an independent predic- tive value for MIE attitudes, much of its effect was medi- ated by perceptions of the counterstereotypic intentions and competence of immigrants. The final model showed that counterstereotypic intentions remained the best indi- vidual predictor of MIE attitudes, followed by entitativity and counterstereotypic competence. The counterstereotyp- ic warmth/competence interaction variable did not make a statistically significant contribution (p = .06) to the overall model and seems to support the claims above regarding the predictive value of individual independent measures as op- posed to interaction variables.To better illustrate the mediation effects, we present therelationships between the four variables as a path diagram in Figure 1. As can be seen, the effect of entitativity on MIE attitudes is primarily mediated through counterstereotypic intentions. Sobel’s test for mediation (Sobel, 1982) showed that this mediation effect was statistically significant (Z =–11.23, p < .001). Although the mediation effect of count- erstereotypic competence was much weaker, Sobel’s test of mediation for this effect was also significant (Z = –2.36, p= .018).DiscussionThe present research advances the point of vie