Lightspeed Blog

Leveraging Lightspeed GMI Honesty Detector for Research Quality

Posted by Tiama HD Fowler on Nov 26, 2014

Depending on who you ask, the idea of a ‘lie detector’ might spark visions of Orwellian instruments of torture or Tom Cruise sweating it out in Minority Report (or Tom Cruise sweating it out under E-Meter for that matter!). At Lightspeed GMI, we think about lie detectors as less sinister and more scientific. As part of our commitment to helping clients ensure they’re drawing business conclusions from accurate data, we spend a lot of time thinking about measuring and ensuring quality. We even have a suite of services that we offer as standard on our business to make sure we’re truly delivering on our promise of making good research easy to conduct. Lightspeed GMI Honesty Detector is one of services we use. ‘HD’, as we affectionately call it, helps evaluate the extent to which survey respondents are truthful in answering questions. It’s these respondents who provide a portion of the foundational data for a client’s business decision.

As our clients have begun leveraging HD as part of their research process, we’ve encountered a wealth of anecdotes about the impact it’s had on resulting research quality. We’ve compiled many of those and worked with clients to create a series of Case Studies to help others learn about and get inspired by how HD might help improve their research. This week, we’re featuring a recent example where a consumer electronics corporation client used HD to gain a clearer understanding of what drives decision makers.

What’s important to know about this particular set of research is that it was going to drive strategy around a broad array of products. If the sample didn’t accurately represent their target professional ecosystem, the costs would be high. Focus on the wrong drivers and their recommendations might misrepresent the importance of a technology change vs. speed or cost savings vs. use habits/practices, etc. as they relate to driving adoption. As you’ll learn in this short read when we asked about adoption habits and whether respondents self-categorized as ‘visionary adopter’, ‘early adopter’, ‘early majority’, or ‘late majority’, people who passed our HD service reported early visionary at 27% vs. 73% for dishonest respondents. Among those who self-identified as late majority adopters (i.e., a ‘less desirable’ category), 98% passed the HD test and only 2% were dishonest. That pattern of dishonest respondents tending to over claim in desirable categories can have damaging effects on your research -- as they tend to continue to over claim in areas like purchase intent, ownership and likelihood to recommend, etc.

The bottom line is that for all research, ensuring your sample is of the highest quality is critical. Until HD though, there wasn’t a strong industry service to measure and control for a respondent’s tendency to be dishonest or untruthful. We’re happy to report though that now that we’ve built HD into our main research platform, all of our clients can benefit from this technique. It’s applied to clean up our own panel, but all respondent sources exhibit these same behaviors and so our system also ensures that we are able to apply HD to any respondent source prior to entering a client’s survey. The truth about that is it’s good for everyone’s business.

Topics: Honesty Detector

Subscribe to Email Updates

Posts by Topic

see all