Scores Flashcards
SEM/Confidence intervals are important because
Remind us that a clients score really represents a range of probable performance rather than a specific point
Also for comparing performance over time
Standard error of measurement
Standard deviation that would be obtained if person took tests multiple times
Their scores overall plotted - because human behavior is never constant
Takes human variability into account
Mean
Most frequency used score on standardized tests
The norming sample should represent..
The population for whole you are using the test for, otherwise the norms are inappropriate
Norms aren’t stable, they change over time
True positive result
Correctly detecting impairment that is present
False positive result
Detect impairment when it’s not present
True negative result
Correctly identifying typical language - no language disorder
False negative result
Identifying child as having typical language when language impairment is present
Norm references tests are appropriate for determining….
If a language impairment exists,
BUT don’t determine specific area of deficit for intervention
Standard comparisons
Raw scores compared with other children
Represent equal units across range of scores
Z score
Mean = 0
Std deviation = 1
T score
Mean = 50
Std deviation = 10
Scaled score
Mean = 10
Std deviation = 3
Stanines
Mean = 5
Std deviation = 2
Normal cure equivalent (NCE)
Mean = 50
Std deviation = 21