Assessment of Pre-school and School-age Children with Language Impairment Flashcards
Introduction to Assessment
Teamwork:
○ Thorough assessment, weigh various options, thoughtful
recommendations for therapy
○ Complex procedure:
○ Finding baseline, determining PLOP
◉ Assessment is ongoing
○ Not sequential! Not assessment→treatment
○ SLP: why/what/how of assessment
○ Continuum: formal→informal; structured→less structured
○ Data collection: unbiased, objective
Psychometric vs. Descriptive Procedures
Two major philosophical approaches to effective communication assessment:
○ Normalist philosophy
○ Neutralist/Criterion referenced
Normalist/Norm referencing
Norm (average)
○ Average performance level (i.e. zero) on a normal distribution
○ Addresses a broad spectrum of content
○ Performance Measured: Comparisons across norm–standard scores, percentile scores
Neutralist/Criterion referenced
Comparing present performance to past performance/specific criteria
○ Descriptive in nature
○ Addresses a clearly defined criteria, aspect of language
○ Performance Measured: summarized meaningfully; measures such as percent correct
Standardized:
Standardized: consistent, standard way to administer the test, to present
and to mark the child’s responses
■ Specific directions to child
■ Repetitions: Number/allowed or not allowed
Normed:
Normed: Test created/administered to group of children (sample
representing all children)
■ Gender, racial, ethnic, geographic, and socioeconomic differences
represented in norming group
■ Test appropriate for all? Linguistic/cultural differences?
Reliability:
Reliability: repeatability of measurement
■ Accuracy/precision –consistently a test measures a characteristic
■ Does the assessment give consistent/dependable results?
■ Internal consistency: measure of reliability
● Degree of relationship between items & overall test
● *High internal consistency-(i.e. children who do well should have
same performance on individual items & perform similarly among
themselves)
Reliability: repeatability of measurement
Test-retest reliability
Test-retest reliability: child administered same test (with time interval
between the administration)
● Two test scores are comparable; consistency of scores
Interjudge reliability
Interjudge reliability: 2 scoring same behavior the same way
● Criterion: specific responses as correct
● SLP needs to use these kinds of tests
Validity
Validity: Tests ability to assess what it purports to assess
■ Effectiveness of the test!
■ Must be proven
○ Three Types:
■ Criterion validity
■ Content validity
■ Construct validity
Three Types of validity
Criterion validity: Effectiveness or accuracy with which a measure predicts
performance
○ Content validity: Faithfulness of the sample/measure representing an
attribute or behavior
○ Construct validity: Extent to which a measure describes or measures some
trait (construct)
Normed tests: Test specific attributes/skills;
Normed tests: Test specific attributes/skills; less complex than functional
language
○ Cautiously….not exclusively! ONE tool!
○ Can ONLY use scores when administered in standard form
◉ SLP must research tests carefully before administering
◉ SLP must be careful not to misuse testing instruments–BE CAREFUL NOT TO:
○ Use scores as a summary of child’s performance
○ Use inappropriate norms
○ Make inappropriate assumptions based on test results
○ Use specific test items for planning intervention goals
○ Use tests to assess therapy progress
Standard Error of Measurement
◉ Misuse of scores–SLP must be familiar with test and testing manual.
◉ Standard Error of Measurement (SEm): Confidence of test scores
○ Standard score of 75 on 2 tests; Test A: SEm of 2, Test B: SEm of 6
○ Test A: confidence 73-77; Test B: confidence 69-81.
○ Test A: SLP can be more confident of score on Test A closer to child’s
performance
◉ Error scores: Two children with same error scores could have responded completely
differently (i.e. no response vs. syntactic error)
◉ Test items may carry the same weight for scoring, but have great differences
developmentally (i.e. functional language)
◉ Normal distribution: score on the bell-shaped curve
○ ⅔ of population should score within 1 standard deviation (SD) on either side of
the mean score
Issues of Standardized Testing (
Inappropriate norms: represent population for whom you are using the test
◉ Cultural/linguistically diverse groups
◉ Test manuals: need to reflect differences; SLP needs to understand/check
◉ Modifying administration procedures—can’t compare scores to norms
○ Clinical judgement!
○ **Must state changes/modifications in written evaluation report!
◉ Incorrect Assumptions: Performance overall assessed, not just test scores
○ No global language assumptions from one test
■ PPVT: receptive vocabulary; only tests this!
◉ Identifying Intervention Goals: Variety of assessments, more than one standardized test
◉ Measuring Therapy Progress:
○ Single, few items on each skill; unable to determine progress
○ Child learning the test–artificially high results
■ Different forms of test, spacing between tests, **highly correlated tests
■ **Criterion-referenced tests-better choice for individual progress monitoring
Variables in Test Selection
Choosing Test:
○ Characteristics: Check Test Manual for accuracy of scoring
○ Does this test:
■ Correctly detecting impairment when it is present (true positive result)
■ Detect impairment when not present (a false positive result)
■ Correctly identifying typical language; no LI (a true negative result)
■ Identify the child as TLD, but LI is present (a false negative result)
◉ SLPs:
○ Familiarity with test/procedure
○ Overall opinion of measure
○ Clinical experience test/task selection
○ Relative importance from data from different measures
