I think most businesses have changed their menus around, relocated products in the store, and in other ways changed aspects of the experience, specifically to see how customers would respond so that they could improve their operations.
Facebook got in a lot of hot water for changing its algorithm around to see how its users would react. Now, the users aren't exactly customers, as they're not paying for anything and are volunteering to use the service. But that just means that if FB is going to get into hot water for such studies, other businesses that are analyzing their customers' behavior might get into hot water too.
But there's really no standard or discussion, that I know of (and maybe I'm quite wrong here, which would be fine) as to when IRB approval would be necessary for businesses to change their way of doing business to test customer activity. Should there be, and what should that protocol be?
Gathering data and information for purposes of organizational assessment, quality assurance, or quality improvement (e.g., surveys about customer satisfaction with organizational services, analyses of the effectiveness of certain programs, market surveys) does not generally require IRB review or approval, because such activities usually serve to assess and document matters specific to the organization, rather than contribute to generalizable knowledge. "Generalizable knowledge" is the key factor here. Businesses are usually not trying to attain that.
Great response, Gordon L Warren. Would you consider whether the experiment was active or passive another factor? For instance, if Facebook passively analyzed data that it was obtaining through normal user interaction, how much of a difference would that make from say if they altered an independent variable to see how it affected results?
The Human Subjects Institutional Review Board is a function of educational institutions. Any faculty member or staff members who is doing research using data collected from human subjects must obtain IRB permission first.
I am not aware of any such requirement for businesses performing research.
If the business is contracting with the university to do research, then IRB research WOULD be required, because it would be university employees doing the work. Often this would be done as part of the pre-coordination for the grant or contract.
Michael W. Marek, in the United States at least, I'm pretty sure that whether it's an educational institution or not, studies on human subjects require IRB approval. http://www.eandireview.com/index.php?option=com_content&view=article&id=173&Itemid=121
Daniel Goldman Yes, this resources gives examples where the requirement applies. But this is not all human subjects research, i.e. research using data collected from people.
When I walk through a mall and am invited to fill out a marketing questionnaire, I don't believe any IRB review is necessary. When Facebook analyses user data, I don't believe any IRB review is necessary.
It very quickly becomes a grey area. That was the real point of this discussion. In some cases, businesses really are manipulating people in ways that constitute scientific experiments on humans.
- Badges
- Science topic
- Similar topics
- Philosophy
- Ethics