Parameter estimates of mixed-effects models predicting contribution change.

In model 1 INTRA subjects and intra-groups were treated as random effects, whereas in all other models subjects, intra-groups, and matched groups were treated as random effects. The presented models are superior regarding AIC/BIC to other model specifications (e.g. including interaction terms). REML = restricted maximum likelihood. AIC = Akaike information criterion. BIC = Bayesian information criterion. * p<.01, ** p<.001, *** p<.0001.

在模型1中，受试者内效应与组内效应被设定为随机效应；而在其余所有模型中，受试者、组内效应以及匹配组均被设定为随机效应。本文所呈现的模型在赤池信息准则（Akaike information criterion, AIC）与贝叶斯信息准则（Bayesian information criterion, BIC）上的表现均优于其他模型设定形式（例如包含交互项的模型）。其中REML指代限制性极大似然估计（restricted maximum likelihood）。* p<0.01，** p<0.001，*** p<0.0001。