A structural equation modeling approach for examining position effects in large-scale assessments

Large-scale Assessments in Education

An IERI – International Educational Research Institute Journal

Table 6 Summary of the model fit information from the SEM models

Model	Number of parameters	Loglikelihood value	Scaling factor	AIC	BIC
M₁: Rasch model	45	−257,591	0.992	515,272	515,604
M₂: Form effect	48	−257,590	1.008	515,277	515,631
M₃: Passage position effect	52	−257,571**	1.007	515,246	515,629
M₄: Item position effect	90	−257,420**	0.991	515,019	515,683
M₄: Item position effect*	68	−257,437**	0.992	515,010	515,511

* This model only included the significant item position effects in M₄
** The model fits the data significantly better than M₁ at the alpha level of α = .05