What is Reliability?

Reliability is probably one of the most difficult concepts to get a handle on that you have to understand at A-level psychology. So I'll ease you in gently.

Reliability refers to the extent to which something is consistent. It does not imply validity, as something may consistently give incorrect measures, though both can be affected by factors such as participant variables, observer bias, subjectivity and standardisation. 

You can be asked to assess reliability of a number of things, most commonly:

- reliability of a measure
- reliability of research
- reliability of a method

There are two main types of reliability: external reliability and internal reliability. Internal reliability refers to the extent to which a measure is consistent within itself whilst external reliability is consistency over time (i.e. it is externally reliable if the measure taken in similar circumstances would produce similar results).

Generally, the factors that affect reliability are:
- the level of control over confounding variables. The higher the control, the less chance there is of inconsistencies over time and thus the more externally reliable it is likely to be
- standardisation. If research into an area tends to use experiments with standardised procedures, or a manual is used, it's more likely that reliability will be high.
- population and ecological validity (research). If a study has poor generalisability due to an ethnocentric or biased sample, or poor mundane realism (as in the case of many lab experiments) findings when applied to real life cannot be expected to be consistent with that of the research.
-subjectivity. If a measure or study relies on subjectivity to give results, as with all qualitative data, self-reports or observations and much research done in cognitive/psychodynamic psychology, it's unlikely to produce consistent results and will probably lack in internal reliability because interpretation becomes a confounding variable and thus may change from measure to measure.