Psychometric assessment of scales for a Model of Goal Directed Vegetable Parenting Practices (MGDVPP)

Table 3 Items and factor loadings from an exploratory two factor solution of norms toward use of vegetable parenting practices and confirmatory factor analysis model fit criteria

Norm items	Factor 1 loadings	Factor 2 loadings
Norm items	Descriptive norms	Normative expectations
*Descriptive norms*
Most parents have their child eat enough vegetables&	.865	.140
Most children eat enough vegetables	.840	.058
Most parents try to get their child to eat more vegetables	.406	.033
*Normative expectations*&&^, &&&
It is important to the [Most Important Person] that my child eats more vegetables. x It is important to me to please the [Most Important Person] when it comes to getting my child to eat more vegetables.	.052	.885
It is important to the [Second Most Important Person] that my child eats more vegetables. x It is important to me to please the [Second Most Important Person] when it comes to getting my child to eat more vegetables.	.128	.875
Eigenvalue	1.945	1.259
Variance explained	38.9%	25.2%
	Pearson correlation
*Descriptive norms*		0.087
*Normative expectations*
Model Fit Indices from a Two Factor Confirmatory Factor Analysis
X ²	0.728
df	1
p	0.394
RMSEA	0.000
SRMR	0.018
CFI	1.000
TLI	1.007

&This item was included in the exploratory factor analysis, but was excluded from the confirmatory factor analysis to enable the analysis to converge.
&&Due to concerns for the complications from missing data, we included responses in regard to the second most important person, but not the third. See Table 2.
&&&Exploratory factor analyses of the norms items were conducted in two ways. First, we included the three descriptive norm statements and two normative expectations statements about the important person expecting the child to eat vegetables (data not shown). Second, we included the three descriptive norm statements and the two normative expectation statements, but the values for the latter two were multiplied by the extent to which the respondent wanted to please the important person (possible range of scores: 1 to 9). The factor structure with the importance items multiplied by the extent of desire to please yielded the most interpretable structure (Table 3).

ISSN: 1479-5868