Table 4 Proportion of variance explained by the acoustic measures for emotion recognition and confidence ratings by stimulus type.

From: Emotion recognition and confidence ratings predicted by vocal stimulus type and prosodic parameters

Stimulus types

Acoustic parameters

Full modela

Frequency

Energy/amplitude

Temporal

MF0

MinF0

MaxF0

SDF0

Jitter

Amp. (dB)

Peak amp.

Shimmer

MHNR

MaxHNR

SDHNR

Duration

Peak time

 

Emotion recognition

AB

0.0027**

0.0008

–

0.0024**

–

0.0018*

0.0660***

0.0019*

0.1229***

0.0110***

0.0335***

0.0006

0.0170***

0.2380***

Anna

–

0.0463***

0.0187***

0.0024***

0.0012

0.0291***

0.0025***

–

0.0011

0.0138***

0.0527***

0.0276***

0.0010

0.1814***

SNeN

0.0023***

0.0005

–

0.0011*

0.0187***

0.0269***

0.0327***

0.0139***

0.0126***

0.0008

0.0018***

0.0300***

0.0131***

0.1447***

PW

0.0176***

0.0006

0.0051***

0.0014**

–

–

0.0105***

0.0082***

0.0107***

0.0358***

0.0056***

0.0005

0.0003

0.0927***

LS

0.0026***

0.0011*

0.0007

0.0330***

0.0005

0.0050***

0.0035***

0.0039***

–

0.0011*

–

0.0403***

0.0008

0.0897***

PS

0.0039***

0.0205***

–

0.0032***

0.0007

0.0053***

0.0017***

0.0008*

0.0003

0.0025***

0.0139***

0.0213***

0.0070***

0.0786***

SPN

0.0040***

0.0009*

0.0004

–

0.0237***

0.0037***

0.0003

0.0100***

0.0005

–

0.0100***

0.0174***

–

0.0691***

SNN

–

–

0.0066***

0.0018***

0.0016**

0.0090***

0.0019***

0.0013**

0.0033***

0.0009

0.0049***

0.0216***

0.0139***

0.0650***

NS

–

0.0013**

–

0.0085***

0.0022***

0.0250***

0.0014**

0.0065***

0.0052***

0.0031***

0.0009

0.0018***

0.0035***

0.0580***

Confidence

AB

0.0017***

0.0040***

0.0005

0.0012**

0.0013**

0.0024***

0.0064***

0.0028***

0.0038***

0.0011**

0.0063***

0.0428***

0.0009*

0.0737***

LS

0.0208***

0.0070***

0.0003

0.0038***

0.0047***

0.0017***

0.0013***

0.0019***

–

0.0005**

0.0032***

0.0073***

0.0012***

0.0531***

PW

0.0169***

0.0074**

0.0001

0.0014***

–

0.0008**

0.0002

0.0122***

0.0077***

0.0022***

0.0009***

0.0041***

0.0025***

0.0490***

SNeN

–

–

–

–

0.0045***

–

0.0100***

0.0040***

0.0030***

0.0016***

0.0004*

0.0074***

0.0035***

0.0379***

NS

0.0067***

0.0067***

0.0005*

–

0.0003

0.0174***

0.0002

0.0005*

0.0016***

0.0024***

0.0016***

0.0002

–

0.0374***

PS

0.0134***

0.0116***

0.0005**

0.0022***

0.0016***

0.0013***

0.0027***

0.0003

–

0.0004

0.0008***

0.0017***

0.0013***

0.0372***

Anna

0.0046***

0.0003

0.0015***

0.0007*

0.0020***

0.0097***

–

0.0006*

0.0001

0.0011***

0.0143***

–

–

0.0342***

SNN

0.0003

0.0018***

0.0002

0.0014***

0.0007***

0.0045***

0.0037***

0.0002

0.0004

0.0029***

0.0003

0.0053***

0.0051***

0.0261***

SPN

0.0008**

0.0011***

0.0017***

0.0002

0.0038***

0.0004

0.0007**

0.0020***

0.0003

–

0.0045***

0.0057***

0.0012***

0.0217***

  1. Note: Emotion recognition values represent pseudo-R2 for logistic regression (see Cox and Snell, 1989 for details) and were calculated as follow: 1–e2*(dev(0) – dev (model))/df, where dev(0) is the deviance of the null model and dev (model) is the deviance of the full model. Confidence ratings values represent the adjusted R2 for linear regression. All p-values were Bonferroni corrected by the number of acoustic parameters (***p < 0.001; **p < 0.01; *p < 0.05). AB = affect bursts, SNeN = semantic neutral nouns; PW = pseudo-words.
  2. LS = lexical sentences, PS = pseudo-sentences, SPN = sematic positive nouns, SNN = semantic negative nouns, NS = neutral sentences, M = mean, Min. = minimum, Max. = maximum, SD = standard deviation, Amp. (dB) = amplitude (decibel), HNR = harmonics-to-noise ratio.
  3. aThe proportion of variance, in the full model, explained by the acoustic measures in each stimulus type is displayed in decreasing order. No values were calculated for the parameters that were dropped by the backward variable selection.
Back to article page