Clinical Practical - Evaluation Flashcards
Construct Validity
Moderately high
Manifestos are great place to look to find the views of the parties
Operationalised behavioural categories so MH mentioned were sorted into correct groups
BUT
We searched for term ‘mental health’ so could have missed relevant information
e.g. - ‘world leading suicide prevention’ in conservative manifesto, did not measure attitudes thoroughly
Ecological validity
Moderately high
Real world documents are not produced for purposes of the research, so are in no way artificial
However
What they claim/ say in the manifestos may not accurately reflect their real life actions (as revealed by voting records)
E.g. - conservative ministers voting to retain adversion therapy for homosexuals
Population validity
Somewhat low
Target pop = mainstream political parties
Sample = conservative, labour, lib dems, green party
Left out UCIP (3rd highest votes in that election) because they were a one issue party (Europe) so had little interest to say about mh, nevertheless the exclusion of the UCIP left the ring-leaning parties underrepresented with 1 party as appose to 3, not accurately representing RL parties.
Replicability
Very
Use tally table and exact same manifestos to repeat research
Our ‘behavioural categories’ were well defined so any sufficiently well trained observer should obtain roughly the same data
However
inter-rater reliability was found to be low for the lib dems; it was very dense and people may have missed observations since there were as many as 5 tally marks for a single (long) sentence
Objectivity
Subjective
Behavioural categories well defined
BUT
There was some opinion as regards what data should go where in the table
Once we had the data in the quantitative analysis require no opinion (e.g. 21:0 mentions for left wing)
Thematic analysis entirely subjective
Similarities/ differences we noticed and the quotes we used were all a matter of personal opinion and therefore open to bias