1what are norms ?
summary of distribution of characteristics in a represntative sample
**established using a normal distribution curve which illustrates the fact that most people, (approx 95%) score between one and two standard deviation above or below the mean average**
2what are relative norms ?
clasification of an individual
3what is the goal of relative norms ?
classify on a continuum
4how are the items in a relative norm ?
maximally discriminating
5what is the most used norm ?
relative norm
6what are absolute norms ?
can an individual do it or not
7what is the goal of absolute norms ?
determine if the criterion is reached
8what is the interpretation of absolute norms ?
previously determined criterion
9where is absolute norms used the most ?
education
10what are norm referenced tests ?
distribution of scores in a large sample
11how can we calcuate norms ?
1. central tendency
2. measure variability
3. skewness
12what is the central tendency of norms ?
mean
median
mode
13what is the variability of norms ?
standard deviation
variance
14what is the skewness of norms ?
positive skewe : right skewed : many difficult items FLOOR EFFECT
negative skewe : left skewed : too high CELILING EFFECT
15what is the floor effect ?
to many difficult items no one can answer correctly
16what is the celiling effect ?
the items are too easy everyone can answer
17when is a score indicative of a disorder ?
>2 SD
18what types of norms scores are there?
1. precentile
2. standard scores
3. other standardized scores
4. normalizing standard scores
19which is the most used norm score ?
percentile
20how do we use precentiles as a norm score?
cumulative frequency distribution
21what kind of transormation are percentile norms ?
non liniar trensformation
22how are the groups in a precetile norm in terms of size?
the same size
23how can we part percentiles ?
1. 100 groups : percentiles
2. 4 groups : quartiles
3. 5 groups : quitiles
4. 10 groups : deciles
5. 20 groups : ventiles
24what are percentiles ( 100 groups ) ?
P1 = first percent
P50 = median
P100 = highesc score / maximum
25what are qurtiles ?
P25 = 1st quartile (q1)
26what are quitailes ?
P20 = 1st quintile
27what are the 5 most imposrtant percentile points ?
1. P50 : median : scores < averege score
2. P16 : score < _2SD
3. P2 : score < _1SD
4. P84 : score < +1
5. P98: score < +2
28what is the order of the most important percentile points ?
2 / 16 / 50 / 8 / 98
29how does the non liniear aspect transalte in the raw score ?
characteristics of raw score changes : not the same distances between the percentiles
30how will the differences be in non linear transformation?
the further from middle of normal distribution show larger discrepancy in underlying raws scores
31how do we use median and standard deviation for percentiles ?
WE DO NOT USE THEM !
32what are standard scores ?
z score
33how are standard scores more useful to tell us norms ?
they reflect raw score
we can use M and SD
34how is the transformation of z scores ?
linear
35what are the M and SD of z scores ?
M= 0
SD = 1
*score can be negative or positive
36what is the z score formula ?
37what do each part of the z score formula represent ?
Zi = standard score of individual i
Xi = raw score of individual i
Mx =. mean of test score
SDx = standard deviation of test scores
38how do liniar transformations advantage us ?
distribution of raw scores remain the same
*the M and SD of raw scores differes
39what is the difference between z score and percentile?
z score : difference (SD) relative to averege raw score (M)
percentile : ranking
40why do we use t scores ?
to avoid negative scores
still arbitrary M and SD + same other charcteristics
41what is the formula of t scores ?
42what do each component of the formula represent ?
Ti = t score of individual
Xi = raw score of individual
Mx = mean of test score
SDx = SD of test score
43is skewness diferentiating the meaning of a score ?
YES ! skewness influences menaing of score relative to norm
44how does skewness influnce the meaning?
more skewed → larger discrepancy between precentiles and z score
45in which situation caan we not compare norm scores ?
when the shapes of their distributions are not the same shape