TABLE 3.

Between-cluster comparisons.

	Cluster 1 (n = 1,214)		Cluster 2 (n = 145)		Welch’s ANOVA
	M	SD	M	SD	F	df	p	Cohen’s d
Age	45.238	9.815	45.903	10.396	0.537	1, 176.055	0.465	0.066
Pretest mood	2.530	0.989	3.552	0.969	143.240	1, 181.760	(<0.001	1.044
Task appreciation	1.483	0.772	0.959	0.881	47.006	1, 171.466	<0.001	0.633
	Cluster 1 (n = 1,214)		Cluster 2 (n = 145)		Pearson’s chi-squared test
	% female		% female		X 2	df	p	Phi
Sex	83.526		93.103		9.095	1	0.003	0.082

Cluster construction was based on two continuous classifiers, occupational depression and cognitive performance. Between-cluster differences pertaining to task appreciation remained statistically significant controlling for pretest mood (as well as age and sex) in analyses of covariance. We did not include a Bonferroni correction given the controversies surrounding its use (e.g., Perneger, 1998). ANOVA, analysis of variance; df, degrees of freedom; M, mean; SD, standard deviation.