The experimental method is the main technique used by the classical sciences: the researcher attempts to eliminate all environmental factors in order to be able to study the effect of an independent variable (IV). Obviously when studying behaviour, it is very difficult to eliminate all extraneous factors - for example no two people will have had the same upbringing and life experiences - but that is the aim.
The IV corresponds to the cause of what is being measured. The thing that is measured is called the dependent variable (DV).
For example, an experiment might be devised to determine how long water takes to evaporate in different temperatures. In this case, the IV would be the temperature of the water, and the DV the time taken for it to evaporate. By keeping all other factors (the volume of water, the humidity, etc.) constant, measuring the time of evaporation (DV) can show what effect a change in temperature (IV) has.
In designing an experiment, a hypothesis is stated which describes what is to be tested. A hypothesis is a statement which can be proved or not proved by a subsequent experiment. For example, "The time taken for water to evaporate varies with temperature". Or, for a pyschological test of learning, "There will be a difference between the end of year results of children who have received daily praise from their teacher and those who receive no praise."
Such a hypothesis is called a research hypothesis and is typically abbreviated to H1. A null hypothesis is one that states that no difference will be found, e.g. "Water evaporates at the same speed regardless of temperature", or "There will be no difference between the end of year results of children who have received daily praise from their teacher and those who receive no praise." The null hypothesis may be abbreviated as H0
The result of an experiment will be that one of H1 or H0 will have been accepted, and the other rejected.
Note that a hypothesis such as "There will be a difference between the end of year results of children who have received daily praise from their teacher and those who receive no praise." does not predict what the difference will be, i.e. whether praise will have a benefit or not. Such a hypothesis is two tailed or non-directional. A two-tailed hypothesis is used when there is no clear-cut reason for expecting one thing or another. In contrast, "Children who are praised will score higher end of year results than children who are not praised" is a directional or one-tailed hypothesis.
For both one- and two-tailed hypotheses, the experimental settings will be the same, but the results for a one-tailed hypothesis are tested more stringently.
Back to index page
Read about the 3 experimental designs on the handout.
Answer questions below:
For each of the following hypotheses or studies state
a. whether the hypothesis should be one-tailed or two-tailed
b. what the I.V. and D.V. are for each
c. whether it would be preferable to use a repeated measures or an
independent groups design: