Poster #137 - Factor Analysis of a new Parenting Self-Efficacy Measure

• In Event: 1-115 - Poster Session 03
  In Poster Session: PS 03 Section - Parenting & Parent-Child Relationships

Integrative Statement

Parenting self-efficacy is defined as the belief that a parent has regarding their ability to successfully perform parenting tasks such that they can positively impact their child’s development (Guimond, Wilcox, & Lamorey, 2008). Parenting self-efficacy is related to positive parenting skills with Latino mothers, including family involvement in learning activities (Machida, Taylor, & Kim, 2002) and acceptance of the child (Dumka, Stoerzinger, Jackson, & Roosa, 1996). However, there is limited research on the factor structure of parenting self-efficacy measures in diverse populations, and most measures lack theoretical grounding (Wittkowski, Garrett, Calam, & Weisberg, 2017).
This study involved the validation of a parenting self-efficacy measure with a sample of low-income, urban mothers in the Southwest. Data were collected from 310 mothers as part of an evaluation of a home visiting program. In this population, 83% of the mothers were Latino and 12% of the mothers were African American. Surveys were collected in group meeting areas at elementary schools that included parenting self-efficacy items. Multiple imputation was used to fill in missing data.
Bandura (1997)’s model of general self-efficacy, specific self-efficacy, and outcome expectancy guided creation of the scales. Two measures, the Parent Involvement and Efficacy Scale and a measure of questions related to Bandura’s three types were analyzed. Items were grouped that applied Bandura’s concepts to parenting to make three scales: Competence, or a general sense of confidence in parenting; Skills, or confidence in specific parenting skills; and Outcome Expectancy, or the influence that parents believe they can have in parenting. The 38 items were sorted by scales and were deleted during an iterative process of examination of coefficient alphas until a satisfactory alpha was reached. Exploratory factor analysis was conducted using Maximum Likelihood with promax rotation that set the number of factors equal to three factors prior to rotation. Findings supported a three-factor solution, with items loading on expected factors. Four items were dropped because they did not have loadings of at least .3 on any factor.
A confirmatory factor analysis was conducted to validate the three-factor solution. The model showed a good fit. The chi-square value was χ²(132) = 251.06, p < .001. Because of the large sample size, a significant chi-square was expected. The RMSEA value was .05, less than the cutoff of .08 for a good fit. The CFI value was .93, higher than the .90 cutoff. The SRMR value was .06, less than the cutoff of .08. Standardized loadings ranged from .28 to .78 and were all statistically significant at the p < .001 level, reflecting significant relationships between the items and the latent factor. The R² values ranged from .08 to .60 and were all statistically significant at the p < .05 level, showing good reliability of items as measuring the latent factors.
In conclusion, this parenting self-efficacy measure was reliable and valid in a largely Latino sample. It suggests that Bandura’s model can be applied to parenting with Latino mothers. Generalizability of the factor structure to other ethnicities should be explored.

Authors

• Laura Nathans, Penn State Scranton
  Presenting Author

• M. Angela Nievar, University of North Texas
  Non-Presenting Author