ABSTRACT
<target id="page_61" target-type="page">61</target>Chapter summary
An experimental hypothesis is a prediction of what will happen in a given experiment. More specifically, it refers to the predicted effect of an independent variable on a dependent variable.
Confounding variables should be avoided when using the experimental method. This can be done by controlling as many variables as possible.
Participants should form a representative sample from some population. This can be done by random or quota sampling.
Within the experimental method, there are independent designs, repeated measures designs, and matched participants designs.
Repeated measures designs produce unwanted order effects, which can often be handled by counterbalancing.
When using the experimental method, we should use standardised procedures, avoid experimenter effects, and try to achieve ecological validity.
Field experiments have the advantage that behaviour can be assessed without participants’ knowledge. However, such experiments can pose ethical issues, and they often have poor experimental control without random assignment of participants to conditions.
Observational studies can be very informative. However, there is a lack of experimental control and there are issues about measurement reliability.
Cross-sectional studies involve various groups studied at a single point in time, whereas longitudinal studies involve one group studied several times. Longitudinal studies are generally preferable because all findings are based on the same individuals. However, they can be misleading because participants tested repeatedly often show practice effects.
Correlational studies involve comparing two measures of behaviour from the same participants. The findings are hard to interpret because correlations do not indicate what has caused any given behaviour.
Case studies involve investigating one or two individuals in detail and typically involve the collection of qualitative data. It is often hard to generalise from such studies.
Socially sensitive research raises ethical issues because of its potentially damaging effects on many people not directly tested.
Ethical research is based on respect, competence, responsibility, and integrity.
When deciding whether a proposed piece of research is ethically acceptable, a cost–benefit analysis should be carried out.
