Research Magazine mentions today in its newsletter an interesting paper from Nader Tavassoli, Associate Professor of Marketing at London Business School, and Gavan Fitzsimons, Associate Professor of Marketing at the Fuqua School of Business, Duke University funded by the Centre for Marketing at London Business School.
The article shows that people respond differently to the same question when typing an answer compared to speaking it.
This is a good academic demonstration of what I believe the industry is already familiar with, the notion that data collection methodologies can deeply impact the data collected. In the short term, it means that researchers who consider switching their trackers to online research should be careful if they don’t want to affect their norms. In practice, most of the research agencies that I have worked with (as an employee or partner) have accompanied the transition of their products to online research with parallel tests, in order to identify/understand the impact on data.
The findings of the study also highlight the challenges behind the adoption of multi-mode research (conducted online, on the phone or face-to-face depending on demographics and preferences).
More importantly, I think that this study raises two important questions:
- first of all, if the two data collections methods provide different results, is one method more accurate, better than the other? A few years ago, in my previous work at Ciao, I was one of the first people in France to “sell” access to online respondents for market research. France is a very conservative country when it comes business practices, and instead of embrassing the technology, I used to hear all the time from market researchers that they didn’t want to do online research because the data was different from CATI or face-t-face. So, despite my arguments why data was different in a good way, people who are averse to change tend to focus on the word ‘different’ than on the word ‘better’
- more importantly, if the interviewing mode has such an impact on the asnwers collected, doesn’t it mean that any interviewing process is problematic by definition? If so, it is high time for market researchers to find new ways to listen to conversations in a more natural way, instead of an artificial laboratory environment.
0 Responses to “Academic study shows consistency in response mode is critical”