An analysis of the relative efficacy of different measures of socioeconomic status in predicting academic achievement in math

Thumbnail Image
Emminger, Jaden
Finch, W. Holmes (William Holmes)
Issue Date
Thesis (M.S.)
Other Identifiers
CardCat URL

The relationship between socioeconomic status SES and academic achievement has long been investigated. The way SES has been operationalized in studies has also often varied. It has been measured through the use of a single variable, multiple variables combined into a composite score, or through multiple variables considered as different aspects of the same model. In this study I attempted to compare the relative efficiency of three models using different methods of measuring SES. By performing this analysis, I hope to identify some strengths and deficiencies in utilizing measures of SES. All of the models I will compare will share in common the same dependent variable of mathematic academic achievement as measured by a standardized math assessment. The differences come in how I am measuring the predictor variable of SES. The models I compare are a model utilizing household income (Yearly Household Income), another using the number of items (Item Composite), including certain rooms within the house, and finally a model that makes use of both household income and the number of items in the house as predictor variables (Hybrid). The results showed that no matter the model it had a positive significant relationship with mathematic academic achievement. I found that the best model based on our data is one that utilizes a combined method of both household income and household items the next best model was one that utilized only household income as a predictor of achievement. I also discuss some possible drawbacks and strengths of the methods used to measure SES.