Basic Assessment Principles Flashcards
Nominal
- Most basic
- Classifying by name
- No indication of amount or magnitude
- Good for count or percentage
- Ex: Gender - Male=0 ; Female=1 —> Frequencies
Ordinal
- Measure of magnitude
- Rankings
- No equal intervals
- Ex: Likert scale
- Often used in psychological testing –> Ranges
Interval
- Equal intervals
- Does not include absolute zero
- Ordered categories
- Ex: IQ Tests, golf scores, temperature in Fahrenheit
Ratio
- Equal intervals plus absolute zero
- Must able to ‘see’ a meaningful zero
- Rare in helping fields but can be used for frequency
- Ex: Speed, frequency of behavior, temperature in Celsius
Norm-Referenced Instruments
- Score compared to performance of others who took the same instrument (norming group)
- Commonly used to assess. Intelligence, achievements, perceptual skills, personality, and behaviors
- 2 major methods: percentiles and standard scores
- Need to consider norming group
Criterion-Referenced Instruments
- Score compared to established standard/criterion
- Focus on how individual performs related to set criteria
- Mastery component
- Challenges: Must adequately measure domain; need to be able to ID criterion for mastery
Frequency Distribution
Chart of number range and frequency
Frequency Polygon
Line graph connected and colored in
Histogram
Bar graph
Mode
Most frequent score
Median
Evenly divides scores into two halves - 50% above, 50% below
Mean
Arithmetic average of scores
Range
Highest score minus lowest score
Variance
Sum of squared deviations from the mean
Standard Deviation
Square root of variance
Normal Distribution
Bell-shaped curve
Skewed Distribution
Positively skewed: to the left
Negatively skewed: to the right
Raw Scores
Meaning less without additional info
Percentile Scores/Ranks
- x% of norming group had a score at or below yours
- NOT percentage of items answered correctly
- add percentages as you go
Z Scores
- or -
- M=0, S=1
- 99.5% of z scores fall between +/- 3 standard deviations
- standardize
T Scores
-Same function as z scores, but no negatives
-M=50, S=10
-Can convert to z score by multiplying z by 10 and adding the number 50
T=10(z)+50
Stanines
- Standard nines
- Range 1-9
- M=5, s=2 (approximate)
- Range of scores instead of actual
- Not equal intervals
Age Equivalent Scores
- Based on what an individual can do at a given age
- Ex: If the average score for an 11.0 y.o. on a test is 15/30, age equivalent score would be 11.0 for any examiner obtaining score of 15
Grade Equivalent Scores
- Based on what average grade level students can do
- Ex: Average score of students halfway through 6th grade (6.5) on a math test is 25. Any examiner obtaining score of 25 has math knowledge of GE 6.5
Problems with Developmental Comparisons
- Systematic misrepresentation
- Extrapolation
- Typological thinking
- False standards
Systematic Misrepresentation
Developmentally, not the same
Extrapolation
Guessing outside the bounds of what is known
Interpolation: guessing with what is known
Typological Thinking
Average score represents a statistical abstraction
False Standards
Tests are constructed so at least half of any age/grade group performs at or below level
Evaluating the Norming Group
- Adequacy of norming group depends on client being assessed, purpose of assessment, and how info will be used
- Examine methods for selecting groups
- Examine characteristics of norming group
Sampling Methods
- Simple random sample
- Stratified
- Cluster
- Convenience
Norming Group Characteristics
- Size
- Gender
- Race/Ethnicity
- Educational background
- SES