casonjc1's version from 2015-06-24 13:33


Question Answer
What makes an instrument or exam reliable?An instrument is reliable when it consistently measures the target attribute and does so accurately. A result that does not change with additional measurements or exams. Results are consistent under consistent conditions. An exam is consistent if the same or comparable results are obtained with repeated administration.
How do you evaluate the reliability of an intervention?Chronbach’s alpha? or Manipulation check Why is this important? To ensure that the treatment was delivered and received as it was intended.
Be able to describe the four types of reliability and validity relationships of an instrument.
1.Test retest reliability-Researchers administer the same measure at different times then compare the results. Under the principal of stability of the instrument ****YOU NEED TO PLAN FOR THIS bc they need to be given at the same time
2.Inter-observer/rater reliability-Equivelence Get everyone to independently rate according to your instrument and see if the results are consistent
3.Parallel forms (alternate forms)Equivelence I want to see if my new instrument is equivelennt to the gold standard *** compare it to a GOLD standard tool – in developing a new instrument I want to see if its reliable I can give it concurrently with another instrument (the gold standard) to see if it results correlate with the instrument findings. ****YOU NEED TO PLAN FOR THIS bc they need to be given at the same time
4.Internal consistency reliability-how good does one item predict the response in another item if you have a bad item in your instrument they will not be internally consistent

Recent badges