Data.xlsx

This data includes for the first experiment- empathy ratings by participants in response to pictures of losing and winning athletes. Before analyzing our data, we readjusted our scale so that the zero would be in the middle.We ran a paired samples t-test using the program IBM SPSS statistics v23 (2015). As for the second experiment- we used a statistical analysis to investigate which face and body features predicted the highest and lowest scores of empathy. On a five-point scale (from 1, very sad, to 5, very happy), the participants rated the level of empathy they felt with regard to the images on the screen. Since the dependent variable was ordinal (empathy level 1 to 5), we ran an ordinal cumulative mixed model with fixed effects, by participants and by items (i.e, by pictures). Each predictor had different numbers of levels. To standardize the contrast, we chose a simple coding scheme. The coding scheme was designed to compare the mean of the dependent variable for a given level to the overall mean of that level (e.g., for mouth, neutral/relaxed mouth was the reference level, for body, standing was the reference level). Finally, we ran a post-hoc analysis using the pairwise ordinal paired test using package lsmeans in R. This test calculates the p-values for each independent variable level from the paired cumulative link model.<br>本数据集包含两项实验的数据：第一项实验为参与者对胜负运动员图像所产生的共情评分。在开展数据分析前，我们对量表进行了调整，将零点置于量表中间位置。我们使用IBM SPSS Statistics v23（2015）软件完成了配对样本t检验。 第二项实验中，我们通过统计分析探究了哪些面部与身体特征能够预测最高与最低共情得分。参与者借助1至5分的五点量表（1对应“非常悲伤”，5对应“非常快乐”），对屏幕中图像引发的共情程度进行评分。由于因变量为有序变量（共情等级1至5），我们采用了包含固定效应的有序累积混合模型，分别按参与者与项目（即图像）进行建模。每个预测变量包含不同数量的水平。为标准化对比设置，我们选择了简单编码方案：该方案旨在将某一特定水平下的因变量均值与该水平的总均值进行比较（例如，对于面部特征，以中性/放松的嘴部作为参照水平；对于身体姿态，以站立姿态作为参照水平）。最后，我们使用R语言lsmeans软件包中的成对有序配对检验开展事后分析，该检验可从配对累积链接模型中计算每个自变量水平对应的p值。