Validity and Reliability

Validity is often the most important issue in any psychological research. There are two broad categories of validity, internal validity and external validity. The best research will be high in every type of validity, although this is often difficult to achieve in practice, leading some to suggest that psychology is not scientific. 

Internal Validity  – The extent to which we can be sure that changes in the DV are caused by changes in the IV and nothing else.

External Validity – The extent to which findings of research can be generalised to other situations, people or times in history.

Concurrent Validity  – The extent to which scores recorded on a scale are matched by scores on an equivalent, well respected scale.

Content Validity  – The extent to which a scale measures all aspects of the characteristic it is designed to measure.

Construct Validity  – The extent to which a scale measures only the characteristic it is designed to measure.

Ecological Validity – The extent to which findings of research can be generalised to real life.

Population Validity  – The extent to which findings of research can be generalised to other people.

Cultural Validity  – The extent to which findings of research can be generalised to other cultures.

Temperal Validity – The extent to which findings of research can be generalised to other eras in history.

Experimenter Bias

- Double blind design – neither pps nor researcher know which level of the IV pps are in.

- Use standardised instructions

Demand Characteristics

- Double blind design

Individual Differences

- Randomisation - Ps are randomly allocated to levels of the IV

Social Desirability

- Single blind design - purpose of research is withheld from ps / Reasurance of privacy

Lack of Control

- Reduce extraneous and confounding variables

Cultural Validity

- Carry out study in variety of cultural settings

Temperal Validity

- Replicate the study in more recent times

Population Validity

- Use a larger/more representative sample

Ecological Validity

- Mundane realism; Change the task to be more realistic.

- Do the study in a more realistic setting appropriate to the behaviour being measured

Internal Reliability  – The extent to which a method consistently measures a dependent variable.

External Reliability  – The extent to which a procedure consistently delivers the same results.

Study only completed once (can't know consistancy)

Occurs in; Experiments and Interveiws

Test/re-test – Replicating the procedure as closely as possible in a follow up study.

Observer Reliability (consistancy of cheaklist)

Occurs in; Naturalistic observations, Interveiws

Inter-rater reliability – A technique used to assess how consistently different observers are using a behavioural checklist.

Reliability of Measuring Tool (is every item measuring the same thing?)

Occurs in; Questionnaires, Scales, Interveiws

Split Half Method  – A technique used to assess the internal reliability of a scale. Each ‘half’ of a participants responses are compared for consistency.