Description Material for the Next Quantitative Brown Bag Sessions
Comparisons of the use Structural Equation Models for Current Research
Nita Guzman's Description
My research is in the proposal phase.
I’m looking at two constructs from motivation literature: academic self- efficacy and attributions for success and failure to see how they predict academic achievement in undergrad business school students. The academic self- efficacy items were designed by me. I piloted them 1 year ago. They break themselves out into two factors. The attribution theory items come from another researcher. I have permission to use his items. His items break down into four factors.
I’m interested in studying how the two constructs of academic self-efficacy and attributions for success and failure impact academic achievement separately as well as how those constructs interact to impact academic achievement of undergraduate business school students.
Here are my research questions:
Is there a relationship between academic self-efficacy for business coursework and academic achievement, as measured by GPA?
Is there a relationship between attributions for success and failure and academic achievement, as measured by GPA?
Does race/ethnicity act as a moderating variable in the relationship between academic self-efficacy for business coursework and attributions for success and failure in their relationship to academic achievement?
Does gender act as a moderating variable in the relationship between academic self-efficacy for business coursework and attributions for success and failure in their relationship to academic achievement?
Does first generation college student status act as a moderating variable in the relationship between academic self-efficacy for business coursework and attributions for success and failure in their relationship to academic achievement?
Here’s how I see myself doing this and what I need input from others on:
I see myself running several models. A model with the self- efficacy items and GPA. A model with the attribution items and GPA. A model with the strongest self-efficacy latent variables and strongest attribution latent variables and GPA (trying to combine the most predictive items from each of the constructs, does that make sense?) and then separately look at the moderating variables.
Based on these variables and these research questions, are their other ways that I should be thinking about my model building process?
What kind of “pitfalls” should I be looking out for using this SEM technique (like checking for normality of data, multicolinearity, etc).
Practical Application:
A product of this research will be a model that includes a subset of items from the two constructs and demographic variables that are highly predictive. Use this model to develop and instrument that can be used as a screening tool for students who are at risk for academic difficulty based on an assessment of these motivational constructs.
Professor Isabelle Chang's Description
(Click Here for Microsoft Word version)
Children’s reading ability develops from a variety of cognitive skills and ecological factors. The underlying assumption is that learning opportunities are embodied in family interaction, which generates sufficient auditory input to enhance children’s language development. The extant research has been unable to examine the complex causal sequence pertaining to children’s reading achievement in a holistic way and utilize the existing statistical techniques such as structure equation modeling to explain the pathway among variables. This study will attempt to explore auditory input generated during the early parent-child shared book reading activities at home and its contribution to children’s reading achievement. The hypothesized pathway in reading development on the causal sequence among variables will be explored according to the following models.
The research question proposed to be explored in this study is:
What are the relationships among variables inside and outside of the classroom, which account for children’s reading achievement?
