Research Methods 2

Question 1

Coding

Answer 1

Systematic analysis of data like interviews and documents so code made with categories to count the number of times a certain thing appears in a source, so quantitative data to analyse

Question 2

thematic analysis

Answer 2

notice common or repeated themes from data collected - qualitative ideas which are more descriptive and can be expressed in different ways e.g transcriptions of different interviews. can lead to follow up analysis based on 1 or 2 themes

Question 3

content analysis eval

Answer 3

+ thematic analysis and coding combines for both quant and qual data
+can use both time and event sampling
+ leads to follow up data collection
+ high validity and can be standardised
- time consuming
- rigid but subjective categories, unclear IV/DV
- sampling limitations e.g withdrawal and representativeness

Question 4

Case studies

Answer 4

detailed investigation of specific individual/group with interesting trait to investigate a certain topic with systematic methods
can be qualitative and quantitative data e.g interviews and experiments etc
can be a short term case study or a longitudinal study over a long period of time to test for behaviour and changes

Question 5

Case studies eval

Answer 5

+ investigating a real life situation so ecologically valid
+ holistic as uses both qual and quant data and takes into account multiple aspects of the persons experience
- hard to find a sample if a trait is rare and then to get permission esp if they cannot make decisions for themselves
- hard to generalise small sample as others may experience the same thing differently so unreliable
+- hard to conduct statistical analysis between patients but conducting analysis between the same patient over time is good as minimal extraneous variables

Question 6

Ways to check reliability

Answer 6

Test retest - repeat test a 2nd time with same ppl and compare results gained both times - correlate +0.8 to be reliable
Pilot study - conduct small trial run - issues with procedures, operational is action and objectivity (esp with observation and behavioural categories
Internal reliability - whether methods/instruments are consistent with time (e.g a ruler has more ir than a 1-10 rating scale)

Question 7

Improving reliability

Answer 7

Inter and intra observer reliability - can also include training people what to observe/ categories
changes following a pilot study - e,g standardisation, removing ambiguity e.g instructions or interview questions, reduce investigator effects/ remove investigator, operationalise all variables.
inter researcher reliability - avoid bias in reports and data and ask for peer review/ correlation

Question 8

Assessing and improving external validity

Answer 8

i.e ecological and temporal validity - how well results can be applied to real life
assess via meta analysis + comparison with other studies, consider date published + environment (lab is low field is high), also how DV measured and if mundane realism or e,g demand characteristics
Improve with double/single blind characteristics so no demand characteristics or investigator bias, or reconduct old experiments to be more valid/relevant, and do more naturalistic e.g field experiments or covert observation

Question 9

Face and concurrent validity

Answer 9

Face validity - if other people (probably experts or other psychologists) were to look at your methods at first look (i,e at face value) would it seem like an accurate way to measure what you want - can improve by making modifications to data collection so more relevant
Concurrent validity - usually for new tests or new methods, where data compared to previous establish tests e.g new iq test against old iq test, for similar scores and a correlation coefficient of +0.8 indicates validity

Question 10

Assessing and improving internal validity

Answer 10

i.e whether the methods and collection in the actual experiment are good enough to measure what is required and to establish a cause effect relationship
all types of validity minus eco and temp
experimental validity e.g removing investigator effects or double blind, no demand characteristics or social desirability bias, abide by ethics to avoid behaviour changes, avoid confounding variables, ensure all variables are operationalised
also assess and improving via face and concurrent validity