PSYB32 – Lecture 4 – October 2 2012
Midterm next Tuesday. – MCQ – read chapters, go over important sections mentioned in class
Reliability – consistency of measurement
Test-Retest reliability – some constructs don’t change (e.g.) so they have HIGH Test-Retest
Alternate form reliability – Giving different versions of the test –eliminating practice effects
Internal consistency reliability – ensuring all items in test are asking the same thing – test this by
splitting the number of questions in half and comparing them.
Validity – if a test is not reliable, it is not valid. Reliability is a necessary component
Content Validity – need to measure as much of the construct as possible. The questions in a
survey need to ask enough questions to get a full scope of what is trying to be understood. The
test should measure ALL kinds of symptoms (if testing for a disorder).
Criterion Validity – Test needs to measure what we say it measures.
Convergent Validity – Results on 2 different tests for the same disorder, for example, should
come to the same conclusion (should converge)
Predictive Validity – the test needs to be predictive of what it is measuring.
Construct Validity – In order to test this, we compared the results of patients with and without
the disorder. The difference in results measures the construct validity.
Psych Assessment :-
Questions in a psych assessment are defined by many variables.
Depends on paradigm followed – this will dictate the questions
Structured Interview – Questions given in a planned order (e.g SKID)
Semi-Structured – Questions not as systematically followed ( a bit more free form)
Lying, nervousness, withdrawal
Can be almost diagnostic.
Limitations of Clinical Interviews:-
Honesty by patient/subject is not a guarantee
Negative impression management -> feigning or exaggerating symptoms or feelings
Sometimes the person may have no insight on how they are feeling
Positive Impression Management – Not willing to admit minor faults/symptoms.
Screening Measures – Checklists of symptoms a patient may or may not have.
Beck Scales –
-Situation/Context of interpretation is important
-How do you prove someone is exaggerating?