Table 2.

Cumulative SHAP values for different sources and types of linguistic cues in the perceived helpfulness predictive model.

	Counselors source			Counselees source			Counselor-counselee synchrony
	Negative effect	Positive effect	Total effect	Negative effect	Positive effect	Total effects	Negative effect	Positive effect	Total effects	Total effects
Affective processes	−0.1090	0.1082	0.2172	0	0	0	\	\	\	0.2172
Biological processes	−0.0775	0.0713	0.1488	0	0	0	\	\	\	0.1488
Cognitive processes	−0.5318	0.2821	0.8139	0	0	0	\	\	\	0.8139
Drives	−0.0592	0.0060	0.0653	0	0	0	\	\	\	0.0653
Informal language	−0.0942	0.1055	0.1998	−0.0047	0.0579	0.0625	\	\	\	0.2623
Numbers	−0.0576	0.0115	0.0691	0	0	0	\	\	\	0.0691
Perceptual processes	−0.0709	0.1291	0.2000	0	0	0	\	\	\	0.2000
Personal concerns	−0.1341	0.1781	0.3123	0	0	0	\	\	\	0.3123
Personal pronouns	−0.4995	0.7505	1.2500	0	0	0	\	\	\	1.2500
Prepositions	−0.0685	0.0118	0.0803	0	0	0	\	\	\	0.0803
Relativity	−0.0458	0	0.0458	0	0	0	\	\	\	0.0458
Social processes	−0.0409	0	0.0409	0	0	0	\	\	\	0.0409
Stylistic	−0.6109	1.2431	1.8539	0	0.1074	0.1074	\	\	\	1.9613
Time orientations	−0.1785	0.2092	0.3877	0	0	0	\	\	\	0.3877
AffectSIM	\	\	\	\	\	\	−0.0300	0.0636	0.0936	0.0936
LSM (mean)	\	\	\	\	\	\	−0.0255	0.0288	0.0544	0.0544
SymptomsSIM	\	\	\	\	\	−0.0279	0.0387	0.0665	0.0665	\

See Table A1 for definitions and descriptions of the types of language cues in the above table.