A Hierarchical Probabilistic Framework for Recognizing Learners’ Interaction Experience Trends and Emotions

Table 1

Descriptive statistics on intensities of emotions for each interaction trend in the three learning environments. Standard errors given in parentheses. ANOVAs reported in italics (p < 0.001).

Environment	Experience trend	Stress	Confusion	Boredom	Frustration
Trigonometry	Flow	22.72 (1.37)	18.46 (1.36)	07.64 (1.20)	15.28 (1.40)
	Stuck	46.37 (1.94)	57.96 (1.94)	25.03 (1.71)	50.17 (1.99)
	Off-task	49.69 (2.80)	74.30 (2.80)	58.66 (2.46)	67.58 (2.87)
	F(2, 701)	69.47	238.65	180.45	189.163
BDS	Flow	24.91 (1.98)	18.88 (2.07)	09.47 (1.87)	10.62 (1.78)
	Stuck	45.85 (3.53)	57.26 (3.70)	20.47 (3.34)	36.04 (3.17)
	Off-task	50.36 (3.97)	70.36 (4.16)	49.41 (3.75)	58.45 (3.56)
	F(2, 261)	24.51	84.53	45.77	82.36
Logic	Flow	17.13 (1.20)	16.36 (1.23)	06.71 (1.11)	11.25 (1.16)
	Stuck	44.01 (1.72)	57.32 (1.76)	24.73 (1.58)	45.47 (1.66)
	Off-task	56.79 (2.30)	77.19 (2.35)	64.45 (2.11)	66.39 (2.22)
	F(2, 877)	157.47	353.30	297.16	307.07