Friday, May 16, 2014

How to humanize a survey

I recently wrote an article about humanizing surveys which suggested that more casual language may create a better survey experience for responders while not comprising data quality nor research results. Here are a few tips on how to achieve those results for yourself.
  1. Don’t compromise on grammar. Even though we’re trying to loosen up and use a more casual writing style, we don’t need or want to compromise on grammar. This is not the place to forget how to use a comma, switch around your verb tenses, or generally be sloppy.

  2. Shorter is better. Charles Dickens is well known for his ability to write perfectly crafted sentences of 100 words or more. Surveys are not the place for that. Once sentences creep over the 15 word mark, figure out they can be broken down into more readable lengths. This long question can easily be shortened: “For each of the following descriptions of shopping behaviors, please indicate whether the description is highly characteristic, somewhat characteristic, slightly characteristic, or not characteristic at all of you when you visit a membership-only warehouse club store.”  Instead try, “How descriptive are these characteristics when you visit a membership-only warehouse club store.”

  3. Don’t over apply grammar. We’ve all heard the adage of not ending a sentence in a preposition. Well, as part of natural language, we do it all the time. Don’t be scared to do it in a survey if the language sounds natural. Instead of awkwardly yet correctly saying “Into which of the following groups do you fall?” why not simply say “Which group do you fall into?”

Avoiding ending a sentence in a preposition is not something you need to strive for.

  1. Keep grid headers short. Researchers like to be as descriptive as they possibly can when writing surveys, perhaps to the point of being over-descriptive. Do we really need to ask what someone “currently owns” instead of just asking what they “own.” Do you we really need to ask what someone has “used in the past week” instead of just asking what they “use a lot.”

  2. Loosen up your wording. Try using some different scales. Instead of using a scale from “Strongly Agree” to “Strongly Disagree,” what about a scale of “Love, Like, Neutral, Dislike, Hate” or “Awesome” to “Terrible.” Yes, the words are much more casual but they will create differentiation among your responders and that is your true goal, and could even generate more meaningful results.

  3. Add a little humor. There’s no denying that humor is tricky. Jokes about politics, religion, and the usual iffy suspects remain off the table but that’s no reason to avoid all humour. Mention a currently popular meme (“This survey may not be as fun as your favorite cat playing the piano video but we hope you like it anyways!”), a generally popular movie (“May the survey force be with you”), or spice up your answer options with some fun descriptors (“Zero, Zip, Zilch!”).

This survey may not be as fun as your favorite cat playing the piano video but we hope you like it anyways!

  1. Say please and thank you. Whether it’s minding your manners or treating others as you’d like to be treated, don’t forget to be polite throughout the survey experience. There’s no need to plaster it onto every question, but a little reminder now and then is much appreciated. Research participants like to know that there’s a human being on the other side of the research. And of course, use your Ps and Qs in a more casual way. Instead of “Thank you for your participation,” why not try “Thanks a bunch for all your help” or “You’ve been a great help. Thanks so much!

If you apply these techniques carefully and don’t overdue it, you too could benefit from happier responders. May the survey force be with you! 

No comments: