View Single Post
Old 07-05-2007, 08:04 AM   #19
Heather's Avatar
Join Date: Sep 2005
Posts: 10,704

S/C/G: 295/225/back to Onederland

Height: 5'5"


I'll see if I can take a closer look at that link, but one of the problems with interpreting this research is the question: Who were the participants?

If I remember right, this is one of the questions that was addressed in Kolata's book. With work like this, we need to know if the results can apply to everyone -- or to whom they can be applied.

For example, the very famous "nurses study" has followed a large sample of nurses over many years as the women provided a lot of information about their lifestyles. But can the results of this study apply to people who are NOT like these nurses? Is there something about the nurses that make them different from other people? Probably, though knowing WHAT that is -- who knows? They may be smarter, or more detail oriented, or be different in any number of ways...

The only way to have results that are generalizable to a larger population is to draw a sample of people from a group in such a way that EVERYONE IN THE POPULATION HAS AN EQUAL CHANCE OF BEING IN THE STUDY. This is called random sampling or representative sampling.

I don't know if that sounds easy, but it's not. Think about a population such as all the students in a school. You could get a random sample of that population by taking a list of ALL students (this is important), and then sampling a certain number of them. Again, it has to be in a way such that everyone has an equal chance to be picked. Like dumping all names in a hat and picking so many of them.

But you can't get a random sample of a population if you don't have a way to get a list of everyone...

Without a random sample, it's hard to say how the results generalize beyond that sample, because there may be something that makes them different from others. For example, if you collect data at a mall, the people there are all capable of leaving their homes and willing to spend money and at the mall at a certain time. They may be different from the population as a whole in a number of ways: wealthier, mobile, unemployed? Who knows.

So, look at the studies discussed in Kolata's book and her discussion of the participants -- who was taken OUT of the studies and for what reasons? That's an important issue.

I'm sorry to go on and on... I have no idea if this is of interest to anyone and don't want to blather on if not!!

My 5 C's of healthy living: Commitment to conscious control, with the understanding that choices have consequences
Heather is offline