Lecture 2 - Norms Flashcards
What is the raw score?
The score of an individual on a test - not meaningful, not going to know whether good/bad score
What are derived scores?
Allows us to ascertain an individual’s position relative to a standardisation (or normative) sample i.e. by reference to norms
What are the three types of derived scores?
- percentiles
- standard
- age scores
What is a percentile score & what is it represented by?
(P) - percentage of people in the standardisation sample who fall below a particular raw scores
What are the advantages of a percentile score?
easy to compute, readily understood, univeral
What are the disadvantages of a percentile score?
inequality of units at the extremes of the distribution - space between the points on a percentile scale is not the same as a space between raw scores
What is a standard score & what is it represented by?
linear z - express an individual’s score in terms of the distance from the mean. That distance is measured in standard deviation units.
In a linear z or z score formula, what does x mean?
z= (x) individuals test score minus (-) the (M) mean from the standardisation sample divided by (/) (s) the standard deviation for the standardisation sample
Why is linear z nice statistically?
retains the exact numerical relationship btw raw scores because we’ve subtracted by a constant and divided by a constant. So the distance between individuals is retained in the z score distribution.
What is the normal distribution range for z?
-3 to +3
What is the relationship between a z score and a percentile score?
if you had a z score of 2 (look at the SD numbers below the bell curve) it would correlate to a percentile score just below it (97.7%)
How do you get the T score (transformation of Linear z)?
Z x 10 + 50
When are T scores commonly used?
personality testing
Linear standard scores will be comparable only if?
They come from similar distributions
What is the deviation IQ score?
to ensure comparability of IQ scores
How do you find the deviation IQ score?
z + mean of standard score + standard deviation
A test score (x) is made up of what two things?
true score (t) + errors (e)
What do confidence errors do?
We put them around an obtained score to aid in its interpretation
What is the formula for a 95% confidence interval?
score (x) is equal to plus or minus 1.96 x standard devation
What is the formula for test scores re 95% confidence interval?
true score (x) is equal to plus or minus 1.96 X the standard error of measurement
What is the assumption of true score theory?
If you gave the same test to someone an infinite number of times, the scores because of error in measurement, would fall out in a normal curve.
In true score theory, the mean of the normal curve is equal to
true score
In true score theory, the SD of the normal curve is equal to
standard error of measurement
What do we use the reliability coefficient for?
estimate the parameters that we need in order to find a 95% confidence interval
What does a z score of 0 mean?
the score is at the mean
What does a z score of -1 mean?
score is one point below the mean
What does a z score of +1 mean?
score is one point above the mean