15 - Strategies for Cleaning Your Evaluation Data Using Excel, R, and SPSS

Abstract Information: Evaluation data, like a lot of research data, can be messy. Rarely are evaluators given data that is ready to be analyzed. Missing data, coding mistakes, and outliers are just some of the problems that evaluators should address prior to conducting analyses for their evaluation report. Even though data cleaning is an important first step to data analysis the topic has received little attention in the literature and the resources that are available tend to be complex and not always user friendly. The goal of this workshop is to go step-by-step through the quantitative data cleaning process, discuss the decisions that need to be made at each step, explore the various solutions an evaluator can choose at each step, and provide hands-on examples on how to clean your evaluation data at each step using Excel, R, and SPSS.

Relevance Statement: Over the years we have encountered many emerging evaluators state to us that the process of data cleaning is confusing, time consuming, and at times seems completely arbitrary. Even though data cleaning is an important first step to data analysis the topic has received little attention in the literature (Van den Broweck, Cunningham, Eeckels, & Herbst, 2005) and available resources tend to be complex and not always user-friendly. Also, as evaluators working in a variety of settings (e.g., universities, government, non-profit) we don’t always have access to expensive proprietary software (e.g., SPSS, SAS) so methods we learning in graduate school to clean/analyze our data offer little guidance when we only have relatively common (e.g., Excel) or freeware (e.g., R) software options available to us. Our facilitators have combined over 35 years of experience training emerging evaluators and applied researchers on how to organize the process of data cleaning and clean evaluation data using a variety of software packages (e.g., Excel, R, SPSS). This workshop has two primary goals: 1) to introduce participants to a step-by-step approach to organizing and cleaning their evaluation data and 2) engage participants in hands-on experiences cleaning evaluation data using three different software packages (e.g., Excel, R, SPSS). The workshop will be a mix of discussion, demonstration, small group hands-on activities, and large group reflection. Participants will be provided detailed guides on how to clean evaluation data (one for Excel, R, and SPSS), a step-by-step guide on how to organize and clean evaluation data (Morrow’s 12-step guide for data cleaning), real datasets that they can practice their data cleaning on (both in the workshop and after the conference), an extensive list of resources on this topic. Our goal is for participants to walk away from the workshop more confident in their ability to organize, make decisions about data cleaning steps, and conduct data cleaning prior to analyzing data to address their evaluation questions.