Week 4 lecture Flashcards
(32 cards)
internal consistency
measured by cronbachs alpha and is measuring how related items within a test are to each other based on their construct.
test-retest reliability
when an assessment produces consistent results when taken at two different time frames.
content validity
the idea that the content of the test is accurately measuring the desired construct
convergent validity
a method used to test the construct validity where the scores of one exam are compared to another with the same construct.
divergent validity
when an exam is compared to that of a different construct to see the correlation. to ensure that construct is not similar to someone else.
norm group
the group that is being evaluated in a study, a sample of the targeted population.
things to do before administration of test
understand manual, test, and take it yourself. have proper environment and all necessary supplies for est. follow instructions verbatim if instructed. note any deviations
after administration of assessment guidelines
Review the results either immediately after or schedule a time to discuss. Consider asking some questions to the client about their experience with the test.
- explore high or low scores and what they can mean
-interpret score and how this can fit into the context of the overall client and their story
3 forms of interpreting scores
norm, criterion, and ipsative
Norm-referenced interpretation
When the scores are compared to the norm group in the studies
ipsative interpretation of assessments
comparing scores against the client or their own previous scores
criterion referenced interpretations
when the scores are compared to criteria scores for various disorders or other things. Based on established cut off scores.
how are scores organized (3 ways)
- distributions: a set of scores
- frequency distributions: distribution categories based on how many points each item has. (5 points for item A, 6 points for item B)
- Group frequency distributions: where distributions are ranged by common scores (10 ppl got 90, 5 ppl got 80)
Explain what positive and negative skews indicate?
positive skew indicates tat there is a greater distribution of scores on the left side of the mean.
negative scews indicate a greater distribution on the right side of the mean.
Kurtosis
a statistic that describes the peakness or flatness of a distribution. as compared to a normal bell curve
what is the measure of central tendency
central tendency is the distribution of average performance where most scores end of close to the mean, making a bell curve.
what is mean, median, and mode
mean is he sum total of items divided by the # of items
median is the middle score of all numbers if ordered from smallest to biggest
mode is the most common score among several numbers
what is variability
variability is the degre to which the scores differ from one another, often measured via standard deviation
what is standard deviation
this is a method used to calculate the distance of scores from the mean. Square root of the variance.
what are the two main ways norm-referenced scores are displayed?
Percentile ranks and standards scores (z-score, t-score, deviations IQs, CEEb scores, stanines, sten scores)
how do we find the standard score from the raw score, especially when comparing two different assessments.
you should use a T score or a z score which can automatically inform you the SD, some test booklets have this.
What is a z score?
this is (x-Mean)/SD. To create a curve from z score, create a mean of 0 with sets of SD of 1. then calculate your scores.
what are t-scores
T=10(Z score)+50. ere you should use a mean of 50 with a SD of 10. Calculate this using a Z score as a base.
How to know if you need to use a t or z score?
If there is no SD known and the sample if less than 30, then use a t score. if the sample is greater than 30 with an SD use a z score.
