Validity

Validity

What is validity?

Accuracy of an observed outcome of research - does it accurately represent natural behaviour

Validity

What is internal validity and what are some examples?

What is it?

• Questions the cause/effect relationship b/w the manipulation of the IV + the observed change in the DV

• If change effected by ANY OTHER VARIABLE OTHER THAN IV = lacks internal validity

Examples:

• Social desirability bias → Ps hide genuine opinions/behaviours + respond/act in more socially acceptable way to be seen in a positive light

• Demand characteristics → Ps think have discovered aim + behave in way to prove/disprove it

• Investigator effect → Researcher’s behaviour/characteristics influences Ps behaviour (eg: researcher bias → consciously/unconsciously, researcher behaves in way that influences Ps behaviour according to their theory OR biased interpretation of results)

• Uncontrolled extraneous variables → eg: not using standardisation + controlling participant variables

Validity

What is external validity?

• Can the findings be generalised beyond the study itself?

  • From the sample → target population

  • From the experimental set up → real world settings/activities

Validity

What is ecological validity?

• Type of external validity

• The ability to generalise a research effect beyond the particular setting/context in which it is demonstrated

Validity

What is mundane realism?

• Type of external validity

• How similar tasks/materials/activities in the experiment are similar to stimuli experienced IRL

Validity

What is population validity?

• Type of external validity

• Describes how easily the sample is representative of target pop (eg: gender, age, ethnicity etc)

Validity

What is temporal validity?

• Type of external validity

• How well the findings can be generalised to other time periods → generally asked of older studies where social norms/tech/etc were v different, questioning if the findings would be the same if conducted in a modern society (eg: into relationships etc)

Assessing validity

What are the ways validity can be tested?

• Face val

• Concurrent val

Assessing validity

What is face validity?

• Does it look like it's measuring what it is supposed to be?

• Eg: Ps ability to remember their times table would lack face val if testing mathematical ability bcs is more a test of memory

Assessing validity

What is concurrent validity?

• The extent to which data from a newly created test is similar to an established test of the same variable conducted at the same time

• Assessed w/ statistical test of correlation → high concurrent val if strength correlation/relationship is 0.8 or higher

Improving validity

How can internal validity be improved?

• Random allocation - controls participant variables

• Standard procedures - controls extraneous variables

• Counter-balancing - controls order effects

• Single + double blind trails - controls researcher bias/demand characteristics

• Peer review - controls researcher bias

Improving validity

How can external validity be improved?

Replication ensures cause/effect relationship in OG study isn’t limited to just that experimental/observational set up:

• Across multiple settings improves eco val

• W/ diverse groups of people improves pop val

• Across time (replicating historical studies) improves temp val

(or can do first two in the first place)