Table 2 Correlations between clinical variables and ML-predicted/clinically rated scores.

From: Using machine learning of computerized vocal expression to measure blunted vocal affect and alogia

 

Blunted vocal affect

Alogia

 

Clinical ratings

Predicted scores

Clinical ratings

Predicted scores

Global psychiatric symptoms

BPRS: Agitation

−0.09

−0.22

0.02

0.15

BPRS: Positive

0.28*

0.06

0.14

−0.02

BPRS: Negative

0.82*

0.63*

0.46*

0.20

BPRS: Affect

−0.11

−0.13

−0.06

0.06

Schizophrenia-spectrum symptoms

SAPS: Hallucinations

0.37*

0.19

0.13

−0.05

SAPS: Delusions

0.27

0.26

0.27

0.05

SAPS Bizarre behavior

0.07

−0.02

0.18

0.03

SAPS: Thought disorder

−0.13

−0.12

0.17

0.19

SANS: Blunted affect

0.85*

0.62*

0.43*

0.20

SANS: Alogia

0.46*

0.34*

1.00

0.57*

SANS: Apathy

−0.17

0.10

0.07

0.00

SANS Anhedonia

0.14

0.18

0.19

0.08

Functioning

Cognition

−0.30*

−0.29*

−0.14

0.03

Social functioning

−0.27+

−0.28+

−0.28+

−0.31*

  1. Bivariate correlations between ML “Predicted” scores (from Machine Learning) and Clinically Rated Blunted Vocal Affect and Alogia scores and clinical symptom and functioning variables. ML scores for each audio recording were averaged across participants (total K samples = 1745, n = 55).
  2. ML machine learning.
  3. *p < 0.05; +p < 0.10.
Back to article page