During a presentation last week from Vance, our Senior Analyst and co-author on the Bunker Blog, a new term was brought to my attention. While I’ve always understood the concept of homophily in market research, I was never aware of the term that was used to describe it. Homophily means literally “love of the same.” Social scientists use the term to refer to the phenomenon that we tend to associate and bond with those who are similar to us. Homophily has been observed in a wide number of studies regarding social networks. Research has confirmed that similarity in terms of demographic and lifestyle traits will facilitate social connections. Or, as the old adage says more simply, “Birds of a feather flock together.”
Why is Homophily Bad for Market Research?
Homophily is the reason why we won’t allow your friend or spouse to attend the focus group with you, the reason why we won’t complete more than one telephone survey with a household, and the reason why surveys through social media networks cannot be counted on for statistically reliable information.
The homophily of our personal networks (friends, spouses, social networks) ensures that any market research that allows personal referral will inhibit diversity and as a result, fail to be statistically representative. Data will skew toward potential demographics such as gender, education, age, income and arguably most important, behavioral traits. People who network with people they like tend to share similar behavioral traits. So when you are recruiting for a focus group and you allow a recently recruited Alex to invite a friend, you will most likely be getting two very similar Alex profiles. This profile duplication can be compounded through telephone surveys and even worse through social media. So when results come back to the client for your 350 complete online survey saying “80% of respondents in the market area are satisfied with your product”, it should be read like “80% of respondents who fit the profile of Alex are satisfied with your product”.
Homophily points to the importance of random sampling for truly reliable information. By calling random households and cell phone numbers, using an online survey with a well-diversified panel of members, and recruiting fresh focus group participants, it will all avoid a fair share of bias that is incurred through networking homophily.
Research & Marketing Strategies (RMS) is a market research firm located in Upstate New York. If you are interested in using market research or conducting a survey or focus group contact our Business Development Director, Sandy Baker, at SandyB@RMSresults.com or by calling (866) 567-5422.
[…] before making any decisions based off of the results (e.g., if you’re using a referral sample, homophily is going to be a major concern and you can expect the respondents to have similar traits and […]