Online Program

Meta-Analytic Approaches for Analyzing Survey Data
Jennifer S Gorman, Kutztown University 
*Karen H Larwin, Youngstown State University 

Keywords: meta-analysis, longitudinal, healthcare, employee surveys

The purpose of this paper is to illuminate novel ways to apply effect size and meta-analytic techniques when facing limitations in the evaluation of data from survey research. The presenters will address why significance tests are often inappropriate for small samples, but how small sample size issues can be overcome when used within meta-analysis applications specifically for primary data analysis. This presentation is appropriate for the meta-analysis novice or veteran, and will detail uses of meta-analysis with nested data sets of small sample sizes. This paper will demonstrate the use of the results from the meta-analytic approach relative to the traditional approaches of dealing with survey data from multiple locations of different sizes. While the focus of this presentation is on the analysis method used in a longitudinal evaluation, the data is drawn from a multi-year Health and Human Services, Rural Healthcare study. This research has been conducted across a three year period, and has examined the impact of Wellness Programs on the health of employees working for a number of small businesses in rural communities. Results of a pre and post-training knowledge and behavior assessments will be shared to demonstrate the utility of effect sizes, meta-analysis applications with primary data, random versus fixed effects models, forest plots, and recommendations regarding software. The analyses will demonstrate the power of meta-analytic approaches in understanding the effect of interventions within and across participating organizations.

Recommended Topic Groups: Health Care Study Longitudinal Study Employee Survey Within and Across Establishments