(Re-)Designing a measure of student’s attitudes toward science: a longitudinal psychometric approach

International Journal of STEM Education

Table 4 Longitudinal measurement invariance: results of the measure on attitudes toward science within the two-factor model (mATSI:3)

Measurement invariance on the science attitude
Model tested	Loglik	Number of parameter	SCF¹ (from H0)	CD² (correction)	TRd³	TRd (p)	RMSEA	CFI	Diff.CFI	Pass?	SRMR
Conf	− 37,303.417	185	1.389				0.033 [0.030, 0.036]	0.951			0.057
Weak	− 37,315.271	164	1.420	1.140	20.794	0.472	0.033 [0.030, 0.036]	0.951	0.000	Yes	0.058
Strong	− 37,330.379	143	1.485	0.983	30.735	0.078	0.032 [0.030, 0.035]	0.950	0.001	Yes	0.058
Strict	− 37,418.044	116	1.398	1.858	94.373	0.000	0.034 [0.031, 0.036]	0.942	0.008	Yes	0.060

SCF¹ indicates scaling correction factor in the Mplus output; CD² indicate corrected scale difference; and TRd³ indicates the Chi-square difference test described in Satorra and Bentler (2010)