Tips for Best Results

The best way to know if data collection will work as planned is to try it out and get feedback from participants. So, one of the top tips for best results is to pilot your evaluation.

Here is list of more data collection tips from Sun Associates, a leading provider of evaluation tools and services for educational purposes.

1. Challenge your assumptions -- It is tempting to make assumptions about why things are the way they are.
   
   
2. Collect data from a wide audience -- Remember that you have a lot of participants so, collect data from all of these participant populations. Just as important as sample size is sample scope - think about all of the individuals who will be affected by the issue at hand.
   
   
3. Collect data that can't be directly observed -- The best data is often that which provides insights into attitudes and opinions. You tend to get this data as notes from discussions/interviews or written comments to open-ended survey questions. Sure, this is harder to tabulate, but it provides rich insight into what people are thinking versus just what they are doing (or saying that they are doing). Besides, you can get the "doing" data without actually surveying.what you want to collect is that information you do not otherwise see. Data is more than quantitative information and "yes/no" answers to survey questions.



4. Always have more than one data source -- Surveys alone cannot provide all of your relevant data. Therefore, always have more than one data source (e.g., surveys AND interviews AND observations) and then create your analysis from a comparison between the sources. Self-reported surveys are always subjective to some degree, so having lots of other data sources for comparison will help to paint the clearest possible picture of the people behind the survey responses.



5. Have multiple survey-writers -- When creating surveys, try to have more than one person involved in writing questions. Have people read each other's work. This helps address the fact that most questions can be perceived of in more than one way. You need to test your survey language among the survey-writers before springing this on your survey population.



6. Set reasonable expectations -- You should try to "encourage" the reluctant to participate (as they add a very interesting and worth while aspect to your overall data picture), but you really cannot force people to participate.
   
   
7. Mailbox surveys do not work -- Participants will not respond in large numbers to surveys that are simply left in their mailboxes. When your survey becomes just one more piece of junkmail, it will most often be ignored. One way to increase your returns is to let participants complete your survey electronically, through a web-based form.
   
   
8. Choose your survey questions carefully -- You probably will have no more than one chance to hear from participants through a survey. Make it count! Be as specific as possible with your survey questions and make sure you get to the heart of the issues that are important to your team.
   
   
9. Be prepared to report back results and take action -- People will always be more willing to participate in a survey (and participate again later) if they know that the results will be communicated to them and that the data they provide will make a difference in future plans.



10. Survey, and then survey again -- Surveys often play a role at different stages throughout an evaluation process. For example, you may take a survey to collect baseline data that informs a planning process, then take another survey to measure how things are going the following year.

Top

--------------------
Source: Sun Associates. (n.d.). 10 tips for data collection. Retrieved April 14, 2004 from the Sun Associates web site:
http://www.sun-associates.com/news/surveytips.html