External Validity

• External validity is the ability to take your findings from your sample and generalize them to the population.

• The question that you must ask (and even test) is whether the people in your sample were representative of the greater population.

• This is not only obtained through a good sample, but it also must be ensured that the research setting that you set up did not alter people’s responses.

• External validity can be determined in a few ways:
  • The first is a re-test. This tests to see if your sample was like other samples that you could have gotten. Rules of probability state that the more samples you take, the closer to the actual population you will get.
  • Another method is to conduct interviews with the people involved about your results this method of obtaining concurrent validity means that you see if the people who you are studying agree with what you said.
  • A third way is to compare the results that you got with similar studies. This is probably the easiest way, however other studies can still have the problem that they are set-up experiments which may not exactly replicate the real world.

• As stated earlier, the best way to have high external validity is to draw a good random sample and have questions which operationalize the concepts well.

• Every member of the group you wish to generalize to must have an equal (or at least known) chance to be drawn in this sample.

• Formulating good questions avoids confusion and replicates reality the best.