First, we have to take into consideration where the sample comes from. If you are getting a sample from a local shopping area or mall then the only people that will be participants in the research are those that have the time to spare. These people are obviously at the mall – therefore there are TONS of confounding variables that can influence the results, such as the idea they might be the type of people that are more likely to buy at point-of-purchase.

In most market research you cannot control your sample as you would in an academic environment. There is no randomization, just people who are wanting to do it for whatever reward they are offered. Once again this poses another confounding variable – the psychographics and behavioristics of these people may lean toward these people being the types that search out coupons and look for good deals, etc. All simply because of the how they were approached.

When I was in undergrad I worked as a market research recruiter. In other words, it was my job to find the people to do the surveys. I can tell you firsthand that a significant number of the results that people get are faked or flawed. It was a common practice for people who worked there to call their friends and have them do the surveys or make them up altogether. I quit this job when the owner of the company told me to lie to Kraft Foods, a very major food company in the US.

The design on the studies was very poor as well. You could tell that the research these companies were paying thousands of dollars for were poorly designed. Rarely did I ever see any reverse polarization on the scales, thus often I would see the halo effect come into play when people just wanted to hurry up and finish it so they could get their three dollars.