Table 5.

Regression analysis results.

Variables	Pro
	Model 1	Model 2	Model 3	95% Confidence Intervals		VIF
Price	0.002 ***	0.002 ***	0.002 ***	0.001	0.003	1.18
	(0.000) t = 4.18	(0.000) t = 4.12	(0.000) t = 3.86
Volume	0.000 ***	−0.000 ***	−0.000 ***	−0.001	−0.000	1.19
	(0.000) t = −7.32	(0.000) t = −7.77	(0.000) t = −7.39
Rating	1.054 ***	1.074 ***	1.005 ***	0.505	1.505	1.00
	(0.258) t = 4.09	(0.257) t = 4.18	(0.255) t = 3.94
Age	−0.003 ***	−0.003 ***	−0.003 ***	−0.003	−0.002	1.01
	(0.000) t = −13.75	(0.000) t = −13.62	(0.000) t = −14.28
Topic 1		1.422 **	1.086 *	0.070	2.102	1.96
		(0.521) t = 2.73	(0.518) t = 2.10
Topic 2		2.571 ***	2.019 ***	0.984	3.054	1.90
		(0.528) t = 4.87	(0.528) t = 3.82
Topic 3		1.735 ***	1.348 **	0.371	2.325	2.10
		(0.500) t = 3.47	(0.498) t = 2.71
Topic 4		2.234 ***	1.677 ***	0.834	2.521	3.05
		(0.429) t = 5.21	(0.430) t = 3.90
Topic 5		2.384 ***	1.943 ***	1.012	2.874	2.35
		(0.476) t = 5.01	(0.475) t = 4.09
Scores			0.110 ***	0.049	0.172	1.15
			(0.031) t = 3.55
Polarity			0.279 ***	0.192	0.366	1.16
			(0.045) t = 6.26
Observations	4.254	4.254	4.254
R 2	0.060	0.068	0.084
Adjusted R2	0.0587	0.0656	0.0814
F	67.35	34.17	35.25

Notes: Significance: *** p < 0.001, ** p < 0.01, * p < 0.05. The first entry in the table is the beta weight. The numbers in brackets are the standard errors. The third entry is the t-value. R² values: 0.25, 0.50, and 0.75 indicate weak, moderate, and substantial predictive power [60]. The 95% confidence intervals for the regression coefficients of model 3 are presented. In order to eliminate multicollinearity, we calculated the VIF value.