Our analysis consists of multiple steps: **1. We did both a MANOVA and ANOVA to compare groups on both the separated communality items (attractiveness and quality separated) and the communality_total scores.** There were no differences for communality and communality_total (both Fs < 1) for the warm and cold cup conditions. This effect was not different for people that were not searching for a house (Fs < 1), or when controlling for gender (Fs < 1). **2. We also ran a t-test an ANOVA to compare groups on need for affiliation.** There was no significant effect (at N=113) for NFA, F(1, 112) = 1.965, p = .164, although one could expect this effect to appear at the full sample, given that the effects were in the predicted direction (cold>warm; the effect was somewhat more visible when omitting people that were looking for a house; F(1, 92) = 2.862, p = .094). **3. To discover if the proposed mediating effect between the relationship between temperature condition and the evaluation of the house via Need for Affiliation was present, we leveraged a bootstrap procedure using Preacher and Hayes' (2008) INDIRECT macro, with 1000 resamples.** We did not run this analysis, because there was no direct effect. **4. Finally, beyond our temperature (cup) condition, we also explored the role of daily temperature on our effects.** When including daily temperature in a multiple regression model, comparable effects started to emerge as in Study 1. That is, with lower daily temperatures, NFA started to become higher (t(92) = -1.52, p = .131) (interactions with the cup condtions also emerged). Communality of one's own house could also increase when temperature is lower (t(92) = -1.12, p = .268), while communality of one's own house is correlated with how long people live in the house (t(92) = 1.685, p = .096; these were all for people who were not actively searching for a house). Given our previous (pilot) study, it is possible that the cup condition is moderated by several factors (how warm is it outside, and what is the nature of the house - advertisement versus one's own house). For this reason, we chose to discontinue the present study for now, and adapted the study to *only* investigate daily temperature and one's relation to one's own house. Despite that we did not detect an interaction, we do expect an interaction in our replication study (and we will analyze this interaction with and without people that experience outside temperature M-1SD).