I am writing an article (and eventually a book) on music listening habits.
For this article I prepared this survey (which should take about 8-15 minutes to complete).
As an aside, I am curious exactly how hard it will be to recruit people to fill out this survey. I plastered the URL on my other social media, but does anybody actually see these things? We will see.
Related: I did this survey on Google Forms (which is pretty slick and easy to use). Google forms works well on mobile devices and lets you break down surveys into multiple pages fairly easily. The hardest part is being able to import (and clean) the data into a statistical analysis app. I left a few optional questions in — and expect some people will start — but not finish the surveys.
I explored various alternatives for importing data. Finally I decided that even if I received 500 responses or more (unlikely), it still would be easy to manually import everything if I needed to.
I think music habits are changing profoundly and well worth studying. Some other remarks about preparing a survey:
I thought about reducing the number of questions by about a third. Ultimately, I decided to leave most of the questions in because I wanted to capture many facets of listening.
I am a novice to survey preparations, but I am an expert at wording questions and have some background in user testing. A lot involves hidden bias but also redundancy. Also, some questions seem to force you into an answer (which is bad). 5 minutes before I published the survey, I took it myself — and noticed certain choices which seem unlikely to be chosen by anyone.
I read several articles suggesting that the best way to create a survey is to ask a series of questions starting with the words “Is there a relationship between A and B?” I tried to do that. At the same time I left a few curiosity questions in because I wanted to expect the unxpected.
About this particular survey, I expect that age more than anything affects how we listen to and discover new music.
One other thing about this particular survey is that I wanted to include several open-ended questions. As tapped in as I am to the music scene, at best I really only know 3-5% of what’s out there, and frankly it helps to hear what resources which other people are using.
I am only guessing, but I imagine that a lot of surveys must uncover a strange correlation –and it exposes a matter which the survey writer never expected. In the best of all worlds, the survey writer would have the opportunity to do follow ups so the survey writer can ask two or three additional questions (which perhaps can be correlated with the respondent’s original answers).