chapter 2 revised
TRANSCRIPT
Basic Assessment Principles
Chapter 2
Nominal
Ordinal
Interval
Ratio
Measurement Scales
Individual’s score is compared to performance of others who have taken the same instrument (norming group)
Example: personality inventory Evaluating the norming group
size sampling representation
Norm-Referenced Instruments
Individual’s performance is compared to specific criterion or standard
Example: third-grade spelling test How are standards determined?
common practice professional organizations or experts empirically-determined
Criterion-Referenced Instruments
Robert 72 Miles 96 Jason 68 Whitney 79Alice 82 Paul 59 Pedro 86 Jane
85Beth 94 John 82 Kelly 92 Michael
81Amy 77 Kevin 85 Justin 72 Rebecca
88Porter 62 Ling 98 Sherry 67 Maria
86
Norm-Referenced: Sample Scores
X 50-59 60-69 70-79 80-89 90-100
f 1 3 4 8 4
Frequency Distribution
Frequency Polygon
0123456789
50-59 60-69 70-79 80-89 90-100
Histogram
012345678
60-69 70-79 80-89 90-100
Measures of Central Tendency
Mode – most frequent score Median – evenly divides scores into two halves
(50% of scores fall above, 50% fall below) Mean – arithmetic average of the scores
Formula: NX
M
Measures of Central Tendency
Example:
Sample scores – 98, 98, 97, 50, 49 Mode = 98 Median = 97 Mean = 78.4
Measures of Variability
Range – highest score minus lowest score Variance – sum of squared deviations from
the mean Standard Deviation – square root of
variance
Formula:
2
NMX
s
Normal Distribution
Skewed Distribution
Raw scores Percentile scores/Percentile ranks Standard scores
z scores T scores Stanines Age/grade-equivalent scores
Types of Scores
X 50 60 70 80 90
f 1 3 4 9 4
% 5% 14% 19% 43% 19%
%ile 5 19 38 81 99*
Percentiles
98th percentile 98% of the group had a score at or below
this individual’s score
32nd percentile 32% of the group had a score at or below
this individual’s score If there were 100 people taking the
assessment, 32 of them would have a score at or below this individual’s score
Interpreting Percentiles
Units are not equal
Useful for providing information about relative position in normative sample
Not useful for indicating amount of difference between scores
Interpreting Percentiles
Types of Standard Scores
z Scores
z score = X-M s Mean = 0
Standard deviation = 1
Mean = 50
Standard deviation = 10
T Scores
Stanines
Standard Scores: Summary
Possible problematic scores Age-equivalent scores Grade-equivalent scores
Problematic because: These scores do not reflect precise performance on an
instrument Learning does not always occur in equal developmental
levels Instruments vary in scoring
Additional Converted Scores
Adequacy of norming group depends on: Clients being assessed Purpose of the assessment How information will be used
Examine methods used for selecting group Examine characteristics of norming group
Evaluating the Norming Group
Methods for selecting norming group:
Simple random sample
Stratified sample
Cluster sample
Sampling Methods
Size Gender Race/ethnicity Educational background Socioeconomic status
Is the norming group appropriate for use with this client?
Norming Group Characteristics