Blind and Double-blind Experiments (page 2)

The experimental design called blind experiment is used most often when the subjects in the experiment are people. Sometimes the subjects know too much about the experiment or are biased. Therefore, the experimenter must work out a way to keep them "blind" about the independent variable, that is, without fore-knowing what has changed. Here is an example that illustrates the place that blind experiments hold in science investigations.

Blind Experiments

Jack enjoys experimenting with cookery, changing recipes to learn how to make some of the foods he prepares taste better. On the grocery store shelf he finds a box of pure cocoa with no sweeteners or other additives. He asks himself, How would chocolate cookies made with that pure cocoa compare with cookies made with the baking chocolate I always use? He finds in a cookbook a formula for substituting cocoa for baking chocolate, and he bakes two batches of cookies. He refers to those with cocoa as "A" and those with baking chocolate as "B"

Jack invites a group of friends over and offers the cookies, which he has separated into two cookie jars, one labeled ''A" and the other labeled "B" He explains to them that, without knowing the difference, or while "blind" to the difference, they are to try each kind of cookie and tell him which they prefer.

The blind experiment is much used by scientists who are investigating painkillers, dietary additives, and other variables. Blind experiments are used when it is important that the subjects do not know or are "blind" to which treatment they are getting, if any. That way the subjects can be completely honest about the effects of their treatment. In such experiments, the scientists often will give one group of subjects a "dummy" or placebo pill, one that does not contain any of the medicine being tested. The subjects, of course, are not told which they are getting-the active medicine or the placebo. Even then there is often a problem because the subjects in a blind experiment sometimes convince themselves that they are getting—the active medicine, and their belief tends to affect them as if they were actually getting the medicine.

An investigator can avoid this problem by using the following approach: the investigator makes up two different painkillers or analgesics in identical forms such as tablets or capsules, and issues them to subjects with pain problems without telling them which kind of pill they are getting. Then, as the subjects report back about the results of the two painkillers (or analgesics), the investigator can make a decision about their effectiveness.

For your own investigations, you might consider blind experiments in the area of cookery, as with Jack's experiment with cocoa. Or, you might want to try comparing the effects of coffee, with or without caffeine, on performance in mathematics, such as memorizing digits, on athletic performance, and so on. You might find other ways for using blind experiments. The prejudice problem on the part of experimental subjects tends to demand it.