Pedestrian, you make an excellent point, and it's related to external validity of the research. I am gathering data from many places and many types of music fans... from Robert Johnson to Justin Beiber; from people who discover music through recommendations of trusted friends to people who discover music through prime time television.
By gathering data from many places through many channels I strengthen the external validity of my study - meaning my sample will be representative of the music-listening population. I have no doubt that people who reach my survey through Music Banter will answer questions much differently than people who reach it from other music-related websites, or social media sites such as Facebook. This is why I don't look at any one answer individually. Instead, I will analyze the full data set that is made up of everyone's individual perceptions.
I am getting excellent response from Music Banter... yet like other places where people click in from it represents a small percentage of the complete data set. This is the point of doing a widespread survey. Each of us perceives music differently. The goal of a statistical analysis isn't just to see basic statistics, it is to measure how perceptions of various aspects of how we enjoy music interact with other related aspects of how we enjoy music. For you fellow stats nerds out there, I'm interested in the covariation that lives in my data set.
Also, keep in mind that my survey is one part of my dissertation research which focuses on music consumers, but keep in mind "consumer" doesn't mean "buyer" - I am interested in how people use/consume/listen to music in contemporary society. The survey is a result of qualitative research, theoretical framework, philosophical foundation, and a pilot survey used to clarify the items on the survey that is posted.
I hope that was more helpful than it was confusing. The bottom line is this... there are no right or wrong answers - we're all individuals who add equal benefit to the study.
Thank you.
|