**3. Results**

The results demonstrated that before the intervention, the level of kindness was comparable in the kindness condition ( *M* = 3.83, *SD* = 0.60) and the control ( *M* = 3.85, *SD* = 0.66), *t*(120) = −0.14, *p* > 0.05. There were also no significant differences in the pretest measurements of other study variables, all *p* ≥ 0.10.

We conducted correlation analysis of the relationships of kindness traits measured at the pretest with the other variables measured at the post-test in order to explore the predictive value of kindness (see Table 1). For simplicity of the data presentation and due to the lack of hypotheses regarding the effect of the experimental manipulation on the relationships between kindness and the other variables, we present the combined data for both conditions in Table 1. Kindness had significant positive correlations with life satisfaction (*r* = 0.30, *p* < 0.001), internal aspirations (*r* = 0.30, *p* < 0.001), and affiliative aspirations (*r* = 0.37, *p* < 0.001) and significant negative correlations with materialism (*r* = −0.21, *p* < 0.05), external aspirations (*r* = −0.21, *p* < 0.05), and intention to shop impulsively (*r* = −0.23, *p* < 0.01). Tables presenting correlations between all study variables separately for each condition can be found in the supplementary materials (see Tables S1 and S2).

**Table 1.** Correlations between the variables used in the study.


Note: T1—pretest; T2—post-test; \* *p* ≤ 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001; *n* = 122.

We used repeated measures ANOVA to examine hypotheses 1 and 3 for dependent variables (life satisfaction, materialism, external aspirations, affiliative aspirations) with one between-subject factor (practicing acts of kindness group vs. active placebo control group), and one within-subject factor (before and after the intervention). Furthermore, for exploratory purposes, we included components of internal (self-acceptance, community feeling) and external aspirations (social recognition, appearance, financial success) as dependent variables. The results of these analyses are presented in Table 2. We examined hypothesis 1d using the Student's t-test due to the lack of a pretest for the intention to buy impulsively.

**Table 2.** Testing the kindness intervention on life satisfaction, materialism and aspirations.


Note: \* *p* ≤ 0.05, \*\*\* *p* < 0.001.

We found a significant main effect of time on life satisfaction, *p* < 0.001, *ηp* 2 = 0.27 (large effect); materialism, *p* < 0.001, *ηp* 2 = 0.10 (medium effect); community feeling, *p* < 0.05, *<sup>η</sup>p*<sup>2</sup> = 0.05 (small effect); and financial success, *p* < 0.05, *<sup>η</sup>p*<sup>2</sup> = 0.03 (small effect). Specifically, the intervention recipients in both groups showed increased life satisfaction and community feeling and decreased materialism and financial success over time. All other main effects of time were nonsignificant, *p* > 0.05.

There was a significant interaction between time and kindness intervention on (1) affiliation, *p* < 0.05, *<sup>η</sup>p*<sup>2</sup> = 0.05 (small effect); and (2) appealing appearance, *p* = 0.05, *<sup>η</sup>p*<sup>2</sup> = 0.03 (small effect). Participants who performed the kindness intervention had higher affiliation on the post-test than on the pretest (supporting H3). The level of appealing appearance did not significantly change over time in the experimental group. All other interactions of time and kindness intervention were nonsignificant, *p* > 0.05 (contrary to H1a,b,c). There were no significant differences in the declared level of intention of impulse shopping, *t*(120) = 0.14; *p* > 0.05, between participants in the experimental group (*M* = 3.68, *SD* = 1.46) and participants in the control group (*M* = 3.64, *SD* = 1.45) (contrary to H1d).

To examine hypotheses 2 and 4, three mediation analyses (Model 4) were performed using the PROCESS macro v.4.1 [49]. We chose mediation analysis, as it allows testing of how a causal antecedent directly affects a variable; when the mediating variable is causally located in-between them, the indirect effects are tested [49]. The indirect effects were tested with bias-corrected bootstrapping (*n* = 5000) and 95% confidence intervals (CIs). The partially standardized indirect effects estimated the effect size of the mediated relationship. We introduced type of intervention (practicing acts of kindness group vs. active placebo control group) as the independent variable, two mediators (changes between the pretest and post-test in life satisfaction and affiliative aspirations) and one covariate (age) into each of the three models for the dependent variables (changes between the pretest and post-test in materialism and external aspirations, intention to buy impulsively). The results of the analyses showed that models exploring changes between the pretest and post-test in materialism, *F*(4, 117) = 0.94, *p* > 0.05, *R*<sup>2</sup> = 0.03, and intention to buy impulsively, *F*(4, 117) = 1.15, *p* > 0.05, *R*<sup>2</sup> = 0.04, were nonsignificant (contrary to H2a,c, H4a,c). The model for changes between the pretest and post-test in external aspirations was significant, *F*(4, 117) = 31.06, *p* < 0.001, *R*<sup>2</sup> = 0.52. The partially standardized indirect effects of kindness intervention on changes in external aspirations via changes in affiliative aspirations (*β* = 0.031, *SE* = 0.132, 95% *CI* = [−0.226, 0.291]) and changes in life satisfaction were nonsignificant (*β* = 0.005, *SE* = 0.014, 95% *CI* = [−0.025, 0.035]) (contrary to H2b, H4b).
