Easier Exploratory Analysis for Epidemiology: A Grad Student 'How-To' Paper

Free registration required

Executive Summary

In epidemiology, exploratory analyses of existing data can screen new hypotheses and provide rationale for future studies. Graduate students often perform an exploratory analysis for a dissertation because it is cost- and time-efficient. Exploratory analyses may examine multiple variables and outcomes across strata of data. Thousands of pages of statistical output may need to be examined for meaningful patterns. The challenges of an exploratory analysis are Organization: Which variables does one want to examine?; Execution: How does one code hundreds of statistical models?; and Summarization: How does one combine hundreds of results across models? This paper gives graduate students the tools to perform an efficient exploratory analysis.

  • Format: PDF
  • Size: 124.7 KB