Date of Award

Summer 6-9-2014

Degree Type

Dissertation

Degree Name

Executive Ed.D. in Education Leadership Management and Policy

Department

Education Leadership, Management and Policy

Advisor

Elaine Walker, Ph.D.

Committee Member

Daniel Gutmore, Ph.D.

Committee Member

Bruce L. Dennis, Ed.D.

Keywords

SGP, Longitudinal Study, Large Scale Assessments, Standardized Testing, NELS 88, Long Term Outcomes

Abstract

This study examined the predictive power of student growth for large-scale assessments on meaningful life outcomes, focusing on the three categories of health, career, and societal involvement. Analysis was conducted using the NELS:88/00 dataset–a longitudinal study that followed a nationally-representative sample of over 12,000 eighth grade students from 1988 to 2000, until the students were 26 years old and entered into the work force. The large-scale assessment variables included math and reading performance in the 1988 cognitive batteries administered by NELS. To gauge growth levels, I generated Student Growth Percentiles (SGP) from tests administered by NELS from 1988 to 1992. Measurable outcomes related to health included binge drinking and cigarette use. Career outcomes included yearly income and job satisfaction. Outcomes related to societal involvement included voting habits, social integration, and the frequency of obtaining information from the outside world.

This quantitative study revealed that student growth on large-scale assessments is meaningfully predictive for three of seven outcome variables: binge drinking, cigarette smoking, and social involvement. Interestingly, I found that students’ performance growth on large-scale exams did not yield more desirable outcomes linearly. For occurrences of binge drinking at age 26, only low reading growth increased the likelihood of binge drinking. Typical and high growths in reading performance were statistically identical in reducing binge-drinking occurrences. The use of cigarettes at age 26 saw similar results for both reading and math growth: only low growth on the large-scale assessments increased the likelihood of the respondent smoking as a young adult. Finally, only respondents who had exhibited typical growth in math performance were more likely to be highly socially involved as young adults.

From the methods and conclusions of this study, I support two major recommendations. First, I recommend that policymakers and school leaders make a habit of collecting longitudinal data along with large-scale assessment results in order to allow researchers and school personnel to investigate long-term program effectiveness. Second, I recommend that a philosophical shift occur among educational researchers in the interest of tracking long-term outcomes that benefit the adult lives of students and society instead of short-lived gains in performance scores and signals.

Download

Share

COinS