Table 4.

Nested models showing incremental validity of Warmth and competence representativeness and its interaction with direction, for ChatGPT and SBERT, in predictions of human ratings of valence.

Model#PredictorbtPR2AICχ2P
ChatGPT1Warmth direction0.2625.98<0.001
Competence direction0.3455.18<0.0010.171667.10
2Warmth direction0.2636.05<0.001
Competence direction0.3034.59<0.001
Warmth representativeness−0.415−2.720.007
Competence representativeness0.3983.020.0030.227647.2923.81<0.001
3Warmth direction−0.082−1.290.197
Competence direction0.2422.430.016
Warmth representativeness−0.421−2.900.004
Competence representativeness0.1860.330.333
Warmth interaction1.1987.18<0.001
Competence interaction0.2801.220.2220.322600.7450.55<0.001
SBERT1Warmth direction0.1674.47<0.001
Competence direction0.0581.450.1480.027848.43
2Warmth direction0.1062.870.004
Competence direction0.0431.100.271
Warmth representativeness−1.445−7.08<0.001
Competence representativeness0.0750.280.7810.13804.1248.31<0.001
3Warmth direction0.0060.140.892
Competence direction−0.086−1.750.081
Warmth representativeness−0.766−2.880.004
Competence representativeness−0.756−2.210.028
Warmth interaction1.0073.120.002
Competence interaction1.6843.73<0.0010.167783.3424.78<0.001
Model#PredictorbtPR2AICχ2P
ChatGPT1Warmth direction0.2625.98<0.001
Competence direction0.3455.18<0.0010.171667.10
2Warmth direction0.2636.05<0.001
Competence direction0.3034.59<0.001
Warmth representativeness−0.415−2.720.007
Competence representativeness0.3983.020.0030.227647.2923.81<0.001
3Warmth direction−0.082−1.290.197
Competence direction0.2422.430.016
Warmth representativeness−0.421−2.900.004
Competence representativeness0.1860.330.333
Warmth interaction1.1987.18<0.001
Competence interaction0.2801.220.2220.322600.7450.55<0.001
SBERT1Warmth direction0.1674.47<0.001
Competence direction0.0581.450.1480.027848.43
2Warmth direction0.1062.870.004
Competence direction0.0431.100.271
Warmth representativeness−1.445−7.08<0.001
Competence representativeness0.0750.280.7810.13804.1248.31<0.001
3Warmth direction0.0060.140.892
Competence direction−0.086−1.750.081
Warmth representativeness−0.766−2.880.004
Competence representativeness−0.756−2.210.028
Warmth interaction1.0073.120.002
Competence interaction1.6843.73<0.0010.167783.3424.78<0.001

Model comparison metrics are provided, comparing the three levels of regression for each language model. Higher values for the outcome indicate more positive human ratings of the social categories.

Table 4.

Nested models showing incremental validity of Warmth and competence representativeness and its interaction with direction, for ChatGPT and SBERT, in predictions of human ratings of valence.

Model#PredictorbtPR2AICχ2P
ChatGPT1Warmth direction0.2625.98<0.001
Competence direction0.3455.18<0.0010.171667.10
2Warmth direction0.2636.05<0.001
Competence direction0.3034.59<0.001
Warmth representativeness−0.415−2.720.007
Competence representativeness0.3983.020.0030.227647.2923.81<0.001
3Warmth direction−0.082−1.290.197
Competence direction0.2422.430.016
Warmth representativeness−0.421−2.900.004
Competence representativeness0.1860.330.333
Warmth interaction1.1987.18<0.001
Competence interaction0.2801.220.2220.322600.7450.55<0.001
SBERT1Warmth direction0.1674.47<0.001
Competence direction0.0581.450.1480.027848.43
2Warmth direction0.1062.870.004
Competence direction0.0431.100.271
Warmth representativeness−1.445−7.08<0.001
Competence representativeness0.0750.280.7810.13804.1248.31<0.001
3Warmth direction0.0060.140.892
Competence direction−0.086−1.750.081
Warmth representativeness−0.766−2.880.004
Competence representativeness−0.756−2.210.028
Warmth interaction1.0073.120.002
Competence interaction1.6843.73<0.0010.167783.3424.78<0.001
Model#PredictorbtPR2AICχ2P
ChatGPT1Warmth direction0.2625.98<0.001
Competence direction0.3455.18<0.0010.171667.10
2Warmth direction0.2636.05<0.001
Competence direction0.3034.59<0.001
Warmth representativeness−0.415−2.720.007
Competence representativeness0.3983.020.0030.227647.2923.81<0.001
3Warmth direction−0.082−1.290.197
Competence direction0.2422.430.016
Warmth representativeness−0.421−2.900.004
Competence representativeness0.1860.330.333
Warmth interaction1.1987.18<0.001
Competence interaction0.2801.220.2220.322600.7450.55<0.001
SBERT1Warmth direction0.1674.47<0.001
Competence direction0.0581.450.1480.027848.43
2Warmth direction0.1062.870.004
Competence direction0.0431.100.271
Warmth representativeness−1.445−7.08<0.001
Competence representativeness0.0750.280.7810.13804.1248.31<0.001
3Warmth direction0.0060.140.892
Competence direction−0.086−1.750.081
Warmth representativeness−0.766−2.880.004
Competence representativeness−0.756−2.210.028
Warmth interaction1.0073.120.002
Competence interaction1.6843.73<0.0010.167783.3424.78<0.001

Model comparison metrics are provided, comparing the three levels of regression for each language model. Higher values for the outcome indicate more positive human ratings of the social categories.

Close
This Feature Is Available To Subscribers Only

Sign In or Create an Account

Close

This PDF is available to Subscribers Only

View Article Abstract & Purchase Options

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

Close