The research and null hypothesis

An important element of your IA is stating and operationalising the IV and the DV in the null and research hypothesis.

The research hypothesis is a prediction of the relationship between these variables. It is NOT a conclusion about what has already been found, but a prediction of what will be found when the experiment is conducted. So, a hypothesis should be framed in the future tense, not the present or past.

The best way to understand this is to look at an example with which you are all familiar: the Loftus and Palmer experiment on the effect of post-event information on memory. The conclusion to this study is clearly stated in the Abstract: These results are consistent with the view that the questions asked subsequent to an event can cause a reconstruction in one’s memory of that event. But what was the actual hypothesis being tested in this experiment? What were the IV and DV? The research hypothesis must explicitly mention which variables are actually being tested. So, a statement like: The specific questions asked about the car accident changed the eyewitness’s memory of the event is not a research hypothesis. Neither the IV (the changed verb) nor the DV (estimate of speed) has been identified, so no specific prediction about their relationship has been attempted.

It is important to remember that the purpose of an experiment is to identify a cause-effect relationship between variables. The IV must be manipulated, and the DV must be measurable. In the hypothesis, both the manipulated and the measured variable must be clearly identified and operationalised. In Loftus and Palmer, the 5 conditions of the verb are the IV; the DV is the estimate of speed in mph. So, the research hypothesis could take two forms. In a 2-tailed hypothesis: The changed verb will result in a change in speed estimate; in a 1-tailed hypothesis: The more emphatic the verb, the higher will be the speed estimate. Note that both forms are in the future tense: They are predictions, not conclusions!

A sound understanding of the research hypothesis should make the nature of the null hypothesis obvious. The null hypothesis is a statement that there will be no effect of manipulating the independent variable, and that any observed difference will be the result of chance… It must still be framed in terms of the operationalized variables (like the research hypothesis, the null hypothesis must identify the IV and DV in an operationalized form), and should still be in the future tense.

When the experiment has been conducted, and the inferential statistic applied to the data, the null hypothesis can be either retained (accepted) or refuted (rejected).

This post was inspired by Stuart Cipinko.