Research methods y13 content
case studies. Content analysis and coding. Thematic analysis.
CS: Detailed, in-depth analysis of an individual. Qual data. Longitudinal.
Ao3 of CS: offer rich, detailed insights on behaviour. Contribute to our
understanding of functioning e.g. patient KF finding MSM to WMM. But its
difficult to generalise findings with such a small sample size. Possible low
validity considering personal accounts from family may be prone to inaccuracy.
Content analysis - Used to analyse qualitative data, and transform into
quantitative. E.g. interview transcripts, films, audio recordings are summarised
and described in a systematic way so overall conclusions can be drawn. Coding
is used to categorise information into meaningful units. E.g. counting up no of
times a word appears, no of times a positive word appears etc, to produce
quantitative data.
Watch film/read transcript, devise coding units e.g., watch again, tally.
Ao3: can circumnavigate many ethical issues normally associated with
psychological research as the materials needed are in the public domain, thus
no issues w gaining permission. However because people are studied indirectly
it loses the content of their material, so interpretation may not be the
intended purpose.
Thematic analysis: analyses qualitative data that involves identifying patterns
within the material e.g. diary, tv ad etc. Once data is transcribed, its reviewed
repeatedly so that the researcher can identify trends in the meaning
conveyed by language.
Reliability. Ways of assessing reliability – test-retest, inter-observer. Improving
reliability.
Ways of assessing reliability:
Test-retest reliability – do one test, same students do the same test at an
appropriate time afterwards (not too close but not too far), compare results
to see if +0.8 correlation = reliability is high.
, Inter-rater reliability - 2 observers agree on criteria e.g. behavioural
categories. Separately they observe and record their findings. They compare
results, an +0.8 correlation = high reliability.
Improving reliability
Questionnaires: If not questions should be altered to be clearer, closed q’s,
pilot study.
Interviews: use same interview each time. If not make sure they are trained
properly
Experiments: use same interviewer, operationalise behavioural categories,
pilot study
Observations: make sure behavioural categories are operationalised to make
sure they don’t have to make their own judgements.
Types of validity: face, concurrent, ecological, temporal. Improving validity.
External validity = what extent can findings be generalised beyond the setting
conducted in/ can transfer findings outside of the study.
Ecological is transferring to another situation
Temporal is transferring to another time period E.g. Asch’s findings was in a
conformist era.
Internal validity = has the researcher measured what they intended to
measure?
Face validity – does it measure what it is supposed to measure.
Concurrent validity – results obtained are close to/match those obtained on
another established test. Close agreement between the 2 sets of data would
indicate high concurrent validity – correlation +0.80.
How to assess research for validity
Face validity = researcher/expert will review the DV measurement to check if it
accurately tests the variable.
case studies. Content analysis and coding. Thematic analysis.
CS: Detailed, in-depth analysis of an individual. Qual data. Longitudinal.
Ao3 of CS: offer rich, detailed insights on behaviour. Contribute to our
understanding of functioning e.g. patient KF finding MSM to WMM. But its
difficult to generalise findings with such a small sample size. Possible low
validity considering personal accounts from family may be prone to inaccuracy.
Content analysis - Used to analyse qualitative data, and transform into
quantitative. E.g. interview transcripts, films, audio recordings are summarised
and described in a systematic way so overall conclusions can be drawn. Coding
is used to categorise information into meaningful units. E.g. counting up no of
times a word appears, no of times a positive word appears etc, to produce
quantitative data.
Watch film/read transcript, devise coding units e.g., watch again, tally.
Ao3: can circumnavigate many ethical issues normally associated with
psychological research as the materials needed are in the public domain, thus
no issues w gaining permission. However because people are studied indirectly
it loses the content of their material, so interpretation may not be the
intended purpose.
Thematic analysis: analyses qualitative data that involves identifying patterns
within the material e.g. diary, tv ad etc. Once data is transcribed, its reviewed
repeatedly so that the researcher can identify trends in the meaning
conveyed by language.
Reliability. Ways of assessing reliability – test-retest, inter-observer. Improving
reliability.
Ways of assessing reliability:
Test-retest reliability – do one test, same students do the same test at an
appropriate time afterwards (not too close but not too far), compare results
to see if +0.8 correlation = reliability is high.
, Inter-rater reliability - 2 observers agree on criteria e.g. behavioural
categories. Separately they observe and record their findings. They compare
results, an +0.8 correlation = high reliability.
Improving reliability
Questionnaires: If not questions should be altered to be clearer, closed q’s,
pilot study.
Interviews: use same interview each time. If not make sure they are trained
properly
Experiments: use same interviewer, operationalise behavioural categories,
pilot study
Observations: make sure behavioural categories are operationalised to make
sure they don’t have to make their own judgements.
Types of validity: face, concurrent, ecological, temporal. Improving validity.
External validity = what extent can findings be generalised beyond the setting
conducted in/ can transfer findings outside of the study.
Ecological is transferring to another situation
Temporal is transferring to another time period E.g. Asch’s findings was in a
conformist era.
Internal validity = has the researcher measured what they intended to
measure?
Face validity – does it measure what it is supposed to measure.
Concurrent validity – results obtained are close to/match those obtained on
another established test. Close agreement between the 2 sets of data would
indicate high concurrent validity – correlation +0.80.
How to assess research for validity
Face validity = researcher/expert will review the DV measurement to check if it
accurately tests the variable.
