How else might we run our crosswords experiment?

We have already seen that things are complicated by how human behaviour varies. And there are further complications. As far as we know iron filings don’t have feelings or emotions. People do. So how will a person’s feelings and emotions influence their performance in our experiments? It is quite possible that we have no way of knowing. In acknowledging this, psychologists distinguish between things that they have direct control over – we call these controlled variables. And things that they have no control over – we call these uncontrolled variables. In our crosswords experiment we do have control over the place where the testing will take place and the puzzles we use. We also have control over and vary the time of testing.

What we have no control over is how our participants are feeling. For instance, they may be overcoming a hangover in the morning! If we test enough people however we expect the influence of the uncontrolled variables to wash out. Hopefully, most of our participants wont have hangovers. As a matter of fact, there are two ways we could run our experiment. First we can run it as described where everyone is tested twice. Once in the morning and once in the afternoon. This is known as repeated measures experiment or alternatively a within-participants experiment Second we can run the testing between-participants. This is where different people are testing in the morning and in the afternoon. Each sort of experiment has its strengths and weaknesses.

If we repeatedly test people we need to be wary of two things. People may get tired or bored as testing unfolds hence they will get worse on the second testing. In contrast, they may become practised at the task and get better on the second testing. These are apparent weaknesses of a within- participant experiment but with some further thought we can take steps to address these. What would you do? A strength of a within-participants experiment is that all the other uncontrolled variables that differ between people from another can, essentially, be ignored. This is because we are gauging each persons afternoon performance relative to their own morning performance.

In contrast, with the between-participants case we might worry that the results of our experiment reflect nothing other than the inherent differences between morning people and afternoon people. Maybe the really good crossword puzzlers only showed up in the morning ? Think how we might guard against this? A strength of the between-participants experiment is that we need worry much less about practice and fatigue effects because everyone is only tested once Clearly therefore are there are strengths and weaknesses to both kinds of experiments and you need to bear these in mind when thinking about which Is best suited to testing your ideas.

In addition there are some experimental questions that can only be answered with between-participant designs and others that perhaps best suited for within-participant designs. You should now be in a position to consider what sorts of questions these are.