. 2022 Nov 7;13:927617. doi: 10.3389/fpsyg.2022.927617

Table 3.

Estimation results for the impact of innovation.

Variable	Firm Performance (ROA)
Variable	Random-effect, Model 1	Fixed-effect, Model 2
$R & D_{i, t - 1}$	0.009 (0.006)	0.059^*** (0.010)
$O P I_{i, t - 1}$	0.045^*** (0.004)	0.049^*** (0.005)
$A G R_{i, t - 1}$	0.084^*** (0.004)	0.086^*** (0.004)
$A S T_{i, t - 1}$	−0.191^*** (0.009)	−0.265^*** (0.016)
$N E M_{i, t - 1}$	0.094^*** (0.010)	0.061^*** (0.016)
$A L R_{i, t - 1}$	−0.091^*** (0.012)	0.004 (0.016)
$C F_{i, t - 1}$	−0.035^*** (0.009)	−0.011 (0.011)
$S H E Q_{i, t - 1}$	0.005^*** (0.000)	0.006^*** (0.001)
$D P S_{i, t - 1}$	−0.389^*** (0.036)	−0.523^*** (0.042)
$E P S_{i, t - 1}$	0.661^*** (0.015)	0.697^*** (0.018)
$E N T_{i}$	Categorical variable	_
$Y E A R_{t}$	_	_
Cons	4.825^*** (0.154)	5.411^*** (0.256)
$R^{2}$ : Within	0.786	0.795
$R^{2}$ : Between	0.719	0.668
$R^{2}$ : Overall	0.668	0.635

^***p < 0.01; ^**p < 0.05. According to the Hausman test, $χ^{2}$ (10) = 289.90, p < 0.05. Hence, the fixed-effect model fit the data set better than random-effect model.