Validity of Qualitative & Quantitative Methods (AQA GCSE Psychology): Revision Note

Validity refers to whether a piece of research is a genuinely accurate measurement or explanation of behaviour

• The whole point of good research design is to improve the validity of the results

Validity isn't just one thing; two types matter most for GCSE:

• Internal validity: whether we can be confident the IV genuinely caused the change in the DV, rather than some other, uncontrolled factor

• Ecological validity: how well the findings generalise beyond the study itself, to real-life behaviour outside the research setting

In experimental designs, each of the three experimental designs can create problems for internal validity if not carefully controlled:

• Repeated measures: results can be affected by order effects, which can be addressed using counterbalancing

  • To improve validity, half the participants complete condition A then B, the other half B then A

• Independent groups: participant variables may act as an extraneous variable, since different people are in each condition

  • To improve validity, participants should be randomly allocated to conditions to reduce this risk

• Matched pairs attempts to solve both problems by pairing participants on key variables across conditions, though matching is never perfect

Lab experiments tend to use artificial tasks in a controlled environment. This can lower ecological validity, since:

• the setting and task may not reflect real-life behaviour

• participants are aware their behaviour is being measured, which can reduce how naturally they behave

Field experiments take place in more natural settings, which can improve ecological validity

• However, the increased number of uncontrolled extraneous variables makes it harder to be certain that any change in the DV was caused by the IV, weakening internal validity

Quantitative methods more broadly can also have validity issues:

• reducing complex human behaviour to a single numerical score (e.g. a digit-span score in a memory experiment) may give little insight into the wider behaviour being studied

The main issue with samples in psychological research is that they often fail to be representative of the target population they're drawn from, e.g.:

• Researchers frequently use the easiest sample to obtain which limits generalisability to that specific group

  • e.g. students at their own university

• Opportunity/convenience sampling only includes people who happen to be available at the time, automatically excluding everyone who wasn't

Stratified sampling is one of the most representative sampling methods, as it matches the target population proportionally

• e.g. if the target population is 58% female, the sample will also be 58% female

Qualitative methods such as interviews and case studies tend to be higher in validity than quantitative/experimental methods, because:

• they deal in descriptions rather than numbers

  • Words add expression, explanation and meaning, whereas numbers just show what happened, how many, or what percentage

• they typically use small samples (sometimes just one participant, as in a case study), allowing data to be rich, in-depth and detailed

• the more a method reflects real experience, the more valid it is, particularly in terms of ecological validity

However, qualitative data is difficult to analyse and often fairly subjective

• e.g. if a participant speaks in a particular tone of voice, does that mean they are bored, angry, upset, or something else entirely?

• A researcher's own expectations can influence how they interpret this kind of data, which can reduce the validity of the conclusions drawn