Skip to main content
. 2018 Mar 13;10:260–268. doi: 10.1016/j.dadm.2018.02.004

Table 5.

Correlations of features with socioeconomic variables

Age
Gender
Education
Variable Corr. Variable Corr. Variable Corr.§
MFCC-3 kurtosis −0.200 MFCC-10 kurtosis −0.239 Preposition freq. 0.230
Conjunction freq. 0.182 MFCC-12 variance 0.181 Hapax legomena 0.222
Brunet's W index 0.179 MFCC-13 variance 0.179 Vocabulary size 0.219
General entropy 0.177 MFCC-5 skewness 0.175 Text size 0.207
Auxiliary verb freq. 0.175 MFCC-5 variance 0.174 General entropy 0.201
MFCC-1 variance 0.172 MFCC-8 skewness −0.169 Adjective freq. 0.200
MFCC-6 kurtosis −0.168 MFCC-10 skewness 0.163 Conjunction freq. 0.191
MFCC-5 kurtosis −0.164 Noun freq. 0.190
MFCC-9 kurtosis −0.158 Informativeness t = 60% 0.190
Informativeness score 0.156 Auxiliary verb freq. 0.184
Verb freq. 0.172
Informativeness score 0.169
Brunet's W index 0.167

Abbreviations: Corr, correlation; MFCC, Mel-Frequency Cepstral Coefficient; freq., frequency.

All correlations presented with P value < .001. Variables are shown in descending order with respect to the strength of their correlation.

Controlled for education, gender, and cognitive status.

Controlled for age, education, and cognitive status.

§

Controlled for age, gender, and cognitive status.