Table 3.

Overview of the analytic process using a Rasch model approach

Step	Psychometric property	Statistical approach and criteria	Results original 10-item GSE	Results reduced 7-item GSE (omits items with poor fit) a
1	Rating scale functioning: Does the rating scale function consistently across items? (substantive validity)	• Average measures for each step category and threshold on each item should advance monotonically	• Rating scale met criteria for all items but item 2. Scale steps 1 and 2 reversed. Recoded into 1(2)34scale	• Rating scale met criteria
		• z-values < 2.0 in outfit mean square (MnSq) values for step category calibrationsb
2	Internal scale validity: How well do the actual item responses match the expected responses from the Rasch model? (content validity)	Item goodness-of-fit statistics • MnSq values < 1.3c	• 3 items failed to meet criteriond: • Item 2: MnSq=1.64 (1) • Item 3: MnSq=1.39 (2) • Item 1: MnSq=1.38 (3)	• All items met criterion
3	Internal scale validity: Is the scale unidimensional (i.e., does it measure a single construct)? (structural validity)	Principal component analysis • ≥ 50% of total variance explained by first component (general self-efficacy)e • Any additional component explains < 5% (or eigenvalue<2.0) of the remaining variance after removing first componente No more than 5% (or 1 out of 20) of the residual correlations >.30	• First component explained 61.3% of total variance • Second component • explained 6.9% of total variance, but eigenvalue <2.0 (1.8) • One out of 45 (2.2%) residual correlations >.30 (#3 - #8: r = -.31)	• First component explained 64.5% of total variance • Second component explained 8.9% of total variance, but eigenvalue <2.0 (1.7) • One out of 21 (4.8%) residual correlations >.30 (#4 - #9: r = -.33)
4	Person-response validity: How well do the individual responses match expected responses from the Rasch model? (substantive validity)	Person goodness-of-fit statistics • Infit MnSq values < 1.5 and z-value ≤ 2.0f • ≤ 5% of sample fails to demonstrate acceptable goodness-of-fit valuesf	• 13/14 respondents (9.2/9.9% of sample) failed to demonstrate acceptable goodness-of-fit values	• 8/9 respondents ( 5.7/6.4% of sample) failed to demonstrate acceptable goodness-of-fit values
		• Infit MnSq values < 1.5 and z-value ≤ 2.0f
		• ≤ 5% of sample fails to demonstrate acceptable goodness-of-fit valuesf
5	Person-separation reliability: Can the scale distinguish ≥3 distinct groups of self-efficacy in the sample tested? (reliability)	Person-separation index • ≥ 2.0g	• 2.75	• 2.67
6	Internal consistency: Are item responses consistent with each other? (reliability)	Cronbach’s alpha coefficient • > 0.8g	• 0.93	• 0.93

Note. After initial evaluation of the original 10-item GSE, a stepwise process was used whereby items failing to meet criteria were removed one at a time, and only those meeting criteria in earlier steps advanced to subsequent steps. If more than one item failed to meet a criterion, the item with the worst fit was removed and the step was repeated with the remaining items. The last column includes ^a 7-item version omitting misfitting items 1, 2, and 3.

^b[38].

^c[52].

^dItems are listed in the order of removal and the MnSq values shown reflect the iteration prior to the item’s removal.

^e[47].

^f[53].

^g[48].