SURVEYS AND FOCUS GROUPS
If you want to obtain a quick sense for the range of views or knowledge in any large group, say residents in a polluted community or plaintiffs in a class action, I can conduct focus groups for you with a few of those people.
Focus groups often surprise you, revealing hot issues and hidden knowledge you never could have guessed at otherwise. Almost always these one-time discussion groups of 6 to 12 "representative" people will give you a rough, qualitative gut feeling about the range of the larger group's views.
Unless you are positive you have already heard most of the facts or opinions people might offer up in a focus group, surveying a random sample of the larger group without first holding a couple of focus groups can be a major mistake.
Surveys are essential when you need to know not just the range of opinions, but the precise percentages of people who hold each opinion. Those numbers tell you exactly what you're up against.
Bear in mind that all surveys are not created equal. Many market research outfits don't select respondents randomly from their larger group. They wind up not really knowing if the larger group believes what their small, biased sample of people believes.
Dewey "defeated" Truman on the front page of the Chicago Tribune in the 1948 Presidential election because the pollster only quizzed subscribers to a literary magazine that attracted well-educated people with money. Back then that meant Republicans. So the sampling was biased, and the conclusion was flat wrong.
Once they choose a random household, even more survey organizations don't randomly select which adult member of the family to question. They wind up with way too many retired people and housewives, people who stay at home. If those kinds of folks disagree with employed people on your issue, you're sunk, and you'll never know it.
In contrast, I learned surveying at arguably the most prestigious survey organization in the United States, the Institute for Social Research at the University of Michigan. I sat through (and greatly enjoyed) the two semesters of turgid math that prove random samples give accurate results. I worked on the federal government's well-known annual survey of teenagers' drug use. My Ph.D. dissertation was based on an Institute survey. I also worked on a variety of academic and marketing studies at the well-known Detroit polling firm, Market Opinion Research.
If you need to be certain your research results will be accurate, give me a call. 216 321-5018.
To see examples of Dr. John Miller's original social-science research, click here.