Research Methods Flashcards
What is internal consistency
If the individual tests high on one item they should test high on another, if they both have the same underlying construct
Inter rater reliability?
No matter who’s doing the marking you should get the same mark
Re-test reliability
Get the same results for individuals over time
Face validity
Does it look like it’ll test the construct
Useful in certain contexts, but actively want to avoid it in others
Content validity
Having construct clearly defined and un controversial
Criterion related validity 2 types
Concurrent validity- time existing test and look at your test on the same sample of individuals
Predictive validity- does my test predict the behaviour based on the construct
Good specificity and moderate sensitivity in detecting psychiatric disorders
SDQ predicted prevalence of child mental health
Construct validity
Most stringent form
Test should have a good agreement between the construct and the data
More recent papers disagree with this, and other cultures say it doesn’t
Discriminant validity
Does it predict behaviour you would expect it to and also not predict behaviour you would not expect it to
What is the teacher version of the SDQ
25 items on questionnaire
Measure core constructs on mental health, emotional problems and conduct problems
Why is the SDQ weak
Evidence that scores seem to change with age, evidence using other questionnaires there is cultural validity
Construct different depending on culture
What is a psychometric approach
Measuring psychological constructs
Define what the construct is
Define how the construct affects behaviour
Not directly observable- no empirical evidence
Take indirect measurements of construct via people’s behaviour
