1. We predict a main effect of response speed on social desirability scores, with higher social desirability scores in the fast (vs. slow) responding condition. We will perform an ANOVA with the between-subjects factor response speed (fast vs. slow) on the total score of social desireability. BOOTSTRAP /SAMPLING METHOD=SIMPLE /VARIABLES TARGET=socdesire INPUT=speed /CRITERIA CILEVEL=95 CITYPE=PERCENTILE NSAMPLES=10000 /MISSING USERMISSING=EXCLUDE. UNIANOVA socdesire BY speed /METHOD=SSTYPE(3) /INTERCEPT=INCLUDE /EMMEANS=TABLES(speed) COMPARE ADJ(LSD) /PRINT=DESCRIPTIVE /CRITERIA=ALPHA(.05) /DESIGN=speed. 2. We predict a positive correlation between good true self bias scores and social desirability. To calculate the magnitude of the good true self bias, we will average participants' responses to the morally good items and the morally bad items to create two average scores. We will then compute a difference score of the two (morally good average score minus morally bad average score), which will indicate the magnitude of the good true self bias, higher positive scores indicating a bias towards the good true self, and higher negative scores indicating a bias towards a negative true self. (The variable is called Goodminusbadlikert) First, we will test whether there is a significant good true self bias, thus replicating the main finding by Newman et al. (2013). Newman performed an ANOVA with the within-subjects factor vignette valence: (good vs. bad vs. neutral) and the between-subjects factor block (block 1 vs. block 2; presumably vignettes 1-4 good vs. vignettes 1-4 bad), and participants' scores on the Likert items as dependent variable. We will deviate from this analysis in the following way: 1. we only include good vs. bad vignettes, no neutral vignettes. 2. We fully counterbalance which vignettes participants get to see in the morally good or morally bad version. Therefore, we do not include block as a between-subjects factor with 2 levels in our analysis. We will perform a repeated-measures ANOVA with the within-subjects factor vignette valence (good vs. bad), and participants' scores on the Likert items as dependent variable. GLM goodlikert badlikert /WSFACTOR=valence Polynomial /METHOD=SSTYPE(3) /PLOT=PROFILE(valence) /EMMEANS=TABLES(valence) /PRINT=DESCRIPTIVE ETASQ /CRITERIA=ALPHA(.05) /WSDESIGN=valence . We will then test our novel prediction by calculating the Pearson correlation between the magnitude of the good true self bias (i.e., good minus bad items, see above) and social desirability scores. CORRELATIONS /VARIABLES=goodminusbadlikert socdesire /PRINT=TWOTAIL NOSIG /MISSING=PAIRWISE. 3. We predict an interaction of response speed and good true self bias, in that the relationship between good true self bias and social desirability and should be stronger in the fast (vs. slow) response speed condition. To test this prediction, we will perform a regression analysis with the factors response condition (dummy coded), magnitude of good true self bias, and the interaction term. Social desirability scores will be the dependent variable. REGRESSION /DESCRIPTIVES MEAN STDDEV CORR SIG N /MISSING LISTWISE /STATISTICS COEFF OUTS CI(95) R ANOVA /CRITERIA=PIN(.05) POUT(.10) /NOORIGIN /DEPENDENT socdesire /METHOD=ENTER speed goodminusbadlikert speed*goodminusbadlikert. 4. We predict an interaction of response speed and authenticity. Specifically, we predict that there will be no significant relationship between authenticity and social desirability in the slow condition, but there may be a positive relationship in the relationship in the fast condition. To test this prediction, we will perform a regression analysis with the factors response condition (dummy coded), authenticity scores, and the interaction term. Social desirability scores will be the dependent variable. If there is a significant interaction, we will perform separate regression analyses for the two response speed conditions to test whether there is a significant relationship between authenticity and social desirability in the fast but not the slow response condition. REGRESSION /DESCRIPTIVES MEAN STDDEV CORR SIG N /MISSING LISTWISE /STATISTICS COEFF OUTS CI(95) R ANOVA /CRITERIA=PIN(.05) POUT(.10) /NOORIGIN /DEPENDENT socdesire /METHOD=ENTER speed authenticity speed*authenticity. SORT CASES BY speed. SPLIT FILE SEPARATE BY speed. REGRESSION /DESCRIPTIVES MEAN STDDEV CORR SIG N /MISSING LISTWISE /STATISTICS COEFF OUTS CI(95) R ANOVA /CRITERIA=PIN(.05) POUT(.10) /NOORIGIN /DEPENDENT socdesire /METHOD=ENTER authenticity .