# regMixR2

*regMixR2*** R function description:**

This function reads in regression mixture parameter estimates and outputs all relevant R^{2} measures and decompositions. That is, when Equation (1) or (2) (see Rights & Sterba, in press) is fit, R^{2}’s in Table 1a and decompositions in Table 3 are outputted. When Equation (7) is fit, R^{2}’s in Table 1b and decompositions in Table 4 and 5 are outputted. Additionally, barcharts for total R^{2} (and decompositions) and level-2 R^{2} (and decompositions) are produced, as in Figures 4, 7, and 8.

Any number of level-1 and/or level-2 classes is supported (including the special case of *K*=1, a multilevel mixture with classes only at level-2, or* H*=1, a single-level mixture). When fitting a multilevel mixture with classes at levels 1 and 2, if the number of level-1 classes differs across *h *(as described in manuscript Footnote 15), a researcher can constrain all parameters equal across class-combinations within any *h *(e.g., if there are *K*=3 level-1 classes, constraining two of these equal within *h* effectively yields *K _{h}* =2). Parameter estimates should still be entered as described below, which would involve entering all constrained parameters as though they were belonging to separate classes.

Rights, J.D., & Sterba, S.K. (in press). A framework of R-squared measures for single-level and multilevel regression mixture models. *Psychological Methods.*

*regMixR2*** input description:**

*data *– Data set with rows denoting observations and columns denoting variables.

*covariate.cols* – List of numbers corresponding to the columns in the data set that represent the predictors used in the regression model

*H* – Number of level-2 classes.

*K* – Number of level-1 classes.

*intslopes* – Vector of coefficient estimates for all class-combination-specific (or class-specific) intercepts and all class-combination-specific slopes, to be entered in the following order: (1) all intercepts going in order of increasing *k *(level-1 class) then increasing *h *(level-2 class) (e.g., *k*=1,*h*=1; *k*=2,*h*=1; *k*=1,*h*=2; *k*=2,*h*=2); (2) all slopes for each class-combination (classes increasing as in (1)), e.g., xslope1_k1h1, xslope2_k1h1, xslope1_k2h1, xslope2_k2h1, etc.) If coefficients are constrained equal across certain classes, then the same estimates would be entered for those classes.

*resvar* – Vector of class-combination-specific (or class-specific) residual variance estimates (entered in the order of classes described in *intslopes* above)

*mcwi* – Vector of level-1 class multinomial intercept estimates, entered in order of increasing *k*, with 0 entered for *K*

*mcws* – Vector of multinomial slopes of *k *on *h estimates *(entered in the order of classes described in *intslopes* above, with 0 entered for every *k=K *and *h=H*)

*mcbi* – Vector of level-2 class multinomial intercept estimates, entered in order of increasing *h*, with 0 entered for *H*

*regMixR2*** R function code and example input:**

## Connect with Vanderbilt

©2019 Vanderbilt University ·

Site Development: University Web Communications