Location: Children's Nutrition Research CenterTitle: Psychometric assessment of scales for a Model of Goal Directed Vegetable Parenting Practices (MGDVPP) Author
|Baranowski, Tom - Children'S Nutrition Research Center (CNRC)|
|Beltran, Alicia - Children'S Nutrition Research Center (CNRC)|
|Chen, Tzu - Children'S Nutrition Research Center (CNRC)|
|Thompson, Deborah - Debbe|
|O'connor, Teresia - Children'S Nutrition Research Center (CNRC)|
|Hughes, Sheryl - Children'S Nutrition Research Center (CNRC)|
|Diep, Cassandra - Children'S Nutrition Research Center (CNRC)|
|Baranowski, Janice - Children'S Nutrition Research Center (CNRC)|
Submitted to: International Journal of Behavioral Nutrition and Physical Activity
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/19/2013
Publication Date: 9/22/2013
Citation: Baranowski, T., Beltran, A., Chen, T.A., Thompson, D.J., O'Connor, T., Hughes, S., Diep, C., Baranowski, J. 2013. Psychometric assessment of scales for a Model of Goal Directed Vegetable Parenting Practices (MGDVPP). International Journal of Behavioral Nutrition and Physical Activity. 10:110.
Interpretive Summary: Little is known about why parents of preschoolers might want to use techniques that are known to get a young child to eat more vegetables, or the techniques known to be ineffective at doing so. A Model of Goal Directed Vegetable Parenting Practices was proposed to provide this understanding. This Model includes 11 general variables (e.g. intentions, desire, attitudes, norms, habit, barriers, anticipated emotional responses, behavioral control, autonomy, relatedness, self-efficacy) to predict parenting behavior. We generated and tested 29 different sets of items which measure different aspects of these 11 variables. We demonstrated that most of these sets of items were related to parent use of effective or ineffective parenting practices. These sets of items can now be used in models predicting Effective and Ineffective Vegetable Parenting Practices to identify important predictors, and guide the development of interventions to enable preschoolers to eat more vegetables.
Technical Abstract: Vegetable intake has been related to lower risk of chronic illnesses in the adult years. The habit of vegetable intake should be established early in life, but many parents of preschoolers report not being able to get their child to eat vegetables. The Model of Goal Directed Behavior (MGDB) has been employed to understand vegetable parenting practices (VPP) to encourage a preschool child's vegetable intake. The Model of Goal Directed Vegetable Parenting Practices (MGDVPP) provides possible determinants and may help explain why parents use effective or ineffective VPP. Scales to measure effective and ineffective vegetable parenting practices have previously been validated. This manuscript presents the psychometric characteristics and factor structures of new scales to measure the constructs in MGDVPP. Participants were 307 parents of preschool (i.e. 3 to 5 year old) children, used for both exploratory (EFA) and confirmatory factor analyses (CFA). Data were collected via an internet survey. First, EFA were conducted using the scree plot criterion for factor extraction. Next, CFA assessed the fit of the exploratory derived factors. Then, classical test theory procedures were employed with all scales. Finally, Pearson correlations were calculated between each scale and composite effective and ineffective VPP as a test of scale predictive validity. Twenty-nine subscales (164 items) within 11 scales were extracted. The number of items per subscale ranged from 2 to 13, with three subscales having 10 or more items and 12 subscales having 4 items or less. Cronbach's alphas varied from 0.13 to 0.92, with 17 being 0.70 or higher. Most alphas <0.70 had only three or four items. Twenty-five of the 29 subscales significantly bivariately correlated with the composite effective or ineffective VPP scales. This was the initial examination of the factor structure and psychometric assessment of MGDVPP scales. Most of the scales displayed acceptable to desirable psychometric characteristics. Research is warranted to add items to those subscales with small numbers of items, test their validity and reliability, and characterize the model's influence on child vegetable consumption.