IMPACT EVALUATION SERIES: Collect Detailed Impact Data from Successfully Adopted Participants
Prior to sending the survey, it is a good idea to build the data analysis model that used to summarize the data. It usually takes the form of a spreadsheet or database. In a spreadsheet, the data structure in each row is one participant response with the columns being the data.
When to collect data is a factor of the number of course deliveries and volume of participants trained. Other guidelines on the timing or frequency of data collection include the following:
- Intention Evaluation : collect at the conclusion of each course. For classroom courses, collect a goal sheet from each participant. For e-learning courses, embed the goal sheet into the end-of-course programming and submit it to the data analysis model.
- Adoption Evaluation : collect data when a sufficient number of participants have graduated and have had the opportunity to transfer their skills to the workplace. If you have a sufficient population size, you do not have to collect adoption data from all participants. Collect enough data to ensure a reasonable return rate and sample to obtain a high degree of confidence.
- Impact Evaluation : collect data immediately after the Adoption data have been analyzed and reported.
Survey Administration
Draft an email from a company executive soliciting survey completion and watch the response rate (the ratio of number of people who answered the survey divided by the number of people in the sample, usually expressed in the form of a percentage).
A reminder email is sent requesting completion of the survey. You want to get the highest return rate as possible, because obtaining a high response rate can bolster statistical power, reduce sampling error, and enhance the generalizability of the results to the population (participants).
Data Scrubbing
When collection has ceased, you need to scrub the Impact data. Data scrubbing, sometimes, called data cleansing, is the process of detecting and removing or correcting any information in a database (or spreadsheet) that has some sort of error. This error can be because the data are wrong, incomplete, formatted incorrectly, or are a duplicate copy of another entry (the participant responded multiple times). Simply review the data and make the necessary corrections.
As always, please send me your thoughts on this method. Next blog: Analyze Impact Survey Data.
Previous Blogs in the Impact Evaluation Series
