Quality surveys can be one of the easiest tools to assess people. All it requires is someone who’s willing to be honest and answer some questions. Although surveys can simultaneously be infinitely more complex. Attempting to ask fair questions that really get at what one is trying to study is very challenging. Then the process of analyzing and interpreting that data can be even worse.
Take, for example, the question “How much do you enjoy using this product?”. It’s a simple question and seems acceptable enough. Although to analyze it a bit deeper, the question does lead the participant a bit. Specifically, it’s assuming and leading the consumer to think that they did enjoy the product. A better question would instead be “How do you rate your experience with the product?”.
In the same theme of something being surprisingly complex, exporting data from a survey isn’t always easy. Not only is data collected on multiple sites at times, but it’s collected in multiple different formats as well. Exporting these into a uniform and usable dataset isn’t as easy as hitting the export button.
Large datasets in particular tend to use specific software and tools for analysis to avoid this. Even then, if the file is exported improperly, the desired software may simply not work. Even worse, it may work but give out improper analyses and conclusions. This means that even on this relatively mundane step of the process caution must be taken.
These are just two simple examples of how surveys can be surprisingly complex. It bears repeating that this won’t matter for many surveys. Although it shows what goes into the most rigorous and massive surveys. Mistakes can be miniscule and make massive differences in the conclusions drawn from the data. This is what makes survey collection complex. Not the process itself, but how easy it is to make mistakes along the way.
