Table 2.

Results of the Logistic Regression test.

Gray	Coefficient	Standard Error	z-value	Significance (p-value)
Constant	−6.9967	7.115	−0.983	n.s. (p = 0.325)
Age	0.0009	0.288	0.003	n.s. (p = 0.998)
Education	−3.5295	3.508	−1.006	n.s. (p = 0.314)
Gender	0.7664	1.562	0.491	n.s. (p = 0.624)
Convergence	0.0018	0.003	0.537	n.s. (p = 0.591)
Sessions per day	0.0342	0.023	1.488	n.s. (p = 0.137)
Prediction	11.8794	5.236	2.269	sig. (p < 0.05)

The prediction probability (outputted by the best-performing pipeline), age, education level, number of typing sessions required for probability convergence, average number of typing sessions per day and gender were used as independent variables and the subject group (with/without depressive tendency) as the dependent binary variable. Only prediction probabilities exhibit a statistically significant (p < 0.05) association with subjects’ status. n.s.: not significant; sig.: significant.