Table 4 Kruskal–Wallis test results comparing linguistic features across generation methods (N = 100 messages, 20 per method)

From: Fine-tuning LLMs in behavioral psychology for scalable health coaching

Feature

H Statistic

p-value

Character length

57

<0.001

Word count

60.65

<0.001

Sentiment

11.71

0.069

Action verb count

36.55

<0.001

Temporal reference

16.58

0.011

Type–Token Ratio

7.44

0.282

Exclamation Count

54.65

<0.001

Readability (FRE)

10.7

0.098