Survey Software Data Management

Posted on : March 16, 2017 - by :

Most survey programs, and nearly all cloud offerings, collect answers, but have very few tools to manipulate the data once collected. Instead they rely on letting you use other tools, if you want to clean or manipulate the data in any way. The most sophisticated survey software includes extensive data management, checking and manipulating abilities.

Many people who work with research data don’t need to clean or manipulate the data after it is collected, but people who do the most sophisticated analysis often do need to have those abilities. People who need to use those kinds of tools find it much more convenient to be able to do those kinds of manipulations in the same program they used to collect the data, rather than having to export the data to a more capable program.

One kind of manipulation is to recode answers. For example, you may be doing research in a metropolitan area and asking people in which town or city they live. You want to take their location into consideration when performing your analysis, but might find that some towns or cities are not represented by enough people in your sample to make analyzing them individually meaningful. You could combine the answers from people in several nearby locations into a single group that is large enough to analyze. One way to do this combining is to have your survey software change the answer codes for these individual locations into a code that will be used for the group.

Another common manipulation is to combine ranges of answers into groups for analysis. For example, you might ask people what age they are in a fill-in-the-blank format. Having the individual years of age gives you the most flexibility in your analysis, but usually you need to group ages into ranges to have enough people in each group to analyze. You could for example tell your program that ages 18-24 should be code 1, while 25-34 is code 2 and so on. While you could ask people to pick a range initially, having them give you individual ages that you can combine as desired gives you more flexibility. After using the groups I just mentioned, you might decide what happens if instead you group people from 18-30 in a group, 31-40 in another group and so on.

To be continued…