Hays, Cha. 5: Measurement Concepts Flashcards
What is reliability?
Looking to see if you get the same results from the same test from the same person during a short amount of time
Instrument assessments look for _____, while personality tests look for _____.
growth; stability
Why do we need to be careful when assessing children for personality disorders?
They are still developing their personality
What is measurement error?
The goal is to reduce the error, but there is going to be some error in your test (known as “error score”)
What are the two parts (scores) of each test?
True score & error score
(T/F) No score will be perfectly reliable or perfectly without error.
True
What are correlation coefficients (aka reliability coefficients)?
When assessing reliability, the statistical measures that determine degree of relationship b/w two factors
What is the range of values for correlation coefficients?
-1 to +1
(T/F) No relationship will be perfect (score of -1 or +1)
True
(T/F) The bigger the number, the stronger the relationship
True
(T/F) If the number is “-“, there is a negative effect; as one factor increases, the other factor decreases
True
What are the base correlation coefficients for achievement tests an personality tests?
Achievement Test: .85-.90
Personality Test: .50-.60
What are the types of reliability?
Test-retest
Split-half
Alternate form
Inter-rater reliability
What is test-retest?
Give test on one occasion, give test again later to same group of people and correlation the scores
What is alternate form in reliability?
Measuring same construct same way, but with different questions. Give test to a group of people then give the alternate version and correlate scores
What is the best way to assess reliability?
Test-retest combined with Alternate form.
What is usually used to assess reliability?
Split-half and inter-item (cheap and simple)
What is inter-rater reliability?
Whether different raters will have same results on same assessment
What is validity?
Does the test measure what it says it’s going to measure? Is the test measuring a construct adequately?
Validity Notes
More difficult to assess than reliability
Assessments are usually only good for one of two assessments at most
What are the different types of validity?
Validity coefficients Face validity Content validity Criterion related validity Construct validity Treatment validity
What are validity coefficients?
The statistical measures that determine degree of relationship b/w two factors.
- Usually lower than reliability coefficients.
What is face validity?
Does it look like the test is measuring what it is supposed to measure?
- Should increase motivation for test-taker to do well
What is content validity?
Is the test assessing the appropriate content for the body of study that a group of students has gone through
- Applies mostly to achievement tests
- A group of experts looks at a group of items to make sure they fit
What is criterion related validity?
Comparing scores with performance
What are two types of criterion related validity?
- Concurrent validity: Give a new test at the same time as an older one measuring the same thing, then correlating their scores. The results should be equal
- Predictive validity: Prediction of how well a person will do with a certain construct (ex: pilot training, ASVAB)
Can we predict low base-rate behaviors (suicide, murder)?
Can be done (Suicide and homicide prediction can be done, but has false-positives because it’s a low base-rate behavior (doesn’t occur very often), so the assessments can’t be used)
What is a low base-rate behavior?
A behavior (suicide and murder) that does not occur often.
What is a false-positive?
Saying someone is a certain way when they aren’t.
What is a false-negative?
Saying someone is not a certain way when they are.
What is construct validity?
Are you really measuring what you’re trying to, or you measuring something else?
Can you have a reliable test that is not valid?
Yes
Can you have a valid test that is not reliable?
No
Reliability is _____ the concept of validity.
subsumed under
What is treatment validity?
Do these tests and their results make any difference in treatment?
- If the person tested concurs and is motivated to take the test, and the results are shared: then it will aid in treatment
What is a response-set (aka response style)?
- When you’re giving an instrument, and the person responds to that instrument in a way that are not what is looked for (through distortion or deception)
- We want people to be honest on answers, but some people unintentionally distort subconsciously (ex: checking all “no” at doctor checklist)
What are validity scales used for?
Assessing for distortion
What tests use validity scales?
Any large, broad scale personality measure
If a person denies any negative behavior, they are probably…
lying. Every one has “chinks in their armor”
What will validity scales show?
Distortion - denial of all negative behaviors
Inconsistencies
A lot of blank answers
A lot of “cannot say” answers
“Yes” responses to extremely odd or infrequent behaviors