Table 5.

Proportion of variance corresponding to each predictor.

	Criterion variable: decision reaction times
	Firsta	Lastb
Severity of ataxia	.30
BDI-II		.22

	First	Last	LMGc	PMVDd
Severity of ataxia	.40
Cognitive-affective factore	.15	.07	.11	.14
Somatic-Motivational factore	.08	.01	.04	.01

^aFirst measure informs on the explained variance value when the variable is the first to be entered in the model.

^bLast measure informs on the explained variance value when the variable is the last to be entered in the model.

^cLMG measure decomposes R² into nonnegative contributions based on semipartial correlation coefficients that automatically sum to the total R².

^dProportional Marginal Variance Decomposition decomposes R² analogously to LMG keeping in mind data-depended weight. This eliminates the possibility of cooperative suppression

^eFactors always entered in the model after severity of ataxia.