-
face validity
means that a test looks like it assesses the skill it claims to assess. A lay person can make this judgement. Face validity alone is not a valuable measure of validity because it is based merely on appearance, not on content or outcomes.
-
content validity
means that a test's contents are representative of the content domain of the skill being assessed. For example, a valid articulation test should elicit all phonemes, thereby assessing the spectrum of articulation. Content validity is related to face validity; content validity, thoughm, judges the actual content of the test (rather than superficial appearance) and is judged by individuals with expert knowledge
-
construct validity
means a test measures a predetermined theoretical construct, which is an explanation of a behavior or attribute based on empirical observation. For example, the theoretical construct that preschool children's language skills improve with age is based on language developement studies. Therefore, a valid test of early language developement will show improved langauge skills when administered to preschool children of progressively increasing ages.
-
Criterion validity
refers to validity that is established by use of an external criterion (2 types)
-
Concurrent validity
refers to a test's validity in comparison to a widely accepted standard. For example, the Stanford-Binet Intelligence Scale is already accepted as a valid assessment of intelligence. Newer intelligence tests are compared to the Stanford-Binet, which serves as the criterion measure.
-
predictive validity
refers to a test's ability to predict performance ( the criterion measure) in another situation or at a later time. It implies that there is a known relationship between the behaviors the test measures and the behaviors or skills exhibited at some future time. College entrance exams, such as the GRE, are used because of their predictive validity. The GRE scores are expected to predict future academic performance.
-
reliability (definition of)
means results are replicable. When administered properly, a test gives consistent results on repeated administrations or with different interpreters judging the same administration. There are several types of reliability
-
test-retest reliability
refers to a test's stability over time. It is determined by administering the same test multiple times to the same group and then comparing the scores. If the scores from the different administrations are the same or very similar, the test is considered stable and reliable.
-
split-half reliability
refers to a test's internal consistency. Scores from one half of the test correlate with results from the other half of the test. The halves must be comparable in style and scope and all items should assess the same skill. This is often achieved by dividing the test into odd-numbered questions and even-numbered questions.
-
rater reliability
refers to the level of agreement among individuals rating a test. It is determined by administering a single test and audio- or videotaping it so it can be scored multiple times. There are two types of rater reliability
-
intra-rater reliability
is established if results are consistent when the same person rates the test on more than one occasion
-
inter-rater reliability
is established if results are consistent when more than one person rates the test
-
alternate-form reliability / parallel form reliability
refers to a test's correlation coefficient with a similar test. It is determined by administering a test (Test A) to a group of people and then administering a parallel form of the the test (Test B) to the same group of people. The two sets of test results are compared to determine the test's alternate form reliability.
-
Norm-Referenced tests (defined)
norm-referenced tests are always standardized. They allow a comparison of an individual's performance to the performance of a larger group, called a normative group. "How does a client compare to the average?" The normal distribution then provides a range of scores by which others are judged when they take the same test. The normal distribution is often depicted using a bell-shaped curve.
-
Norm-referenced test (advantages)
- *tests are objective
- *skills of an individual can be compared to those of a large group of similar individuals
- *test administration is usually efficient
- *many norm-referenced tests are widely recognized, allowing for a common ground of discussion when other professionals are involved with the same client
- *administration and interpretation guidelines are clearly specified in the accompanying manual
- *insurance companies and school districts prefer known test entities for third party payment and qualifications for school services
-
Norm-referenced test (disadvantages)
- *do not allow for individualization
- *tests are generally static; they tell what a person knows, not how a person learns
- *the testing situation may be unnatural and not representative of real life
- *the approach evaluates isolated skills without considereing other contributing factors
- *norm-referenced tests must be administered exactly as instructed for the results to be considered valid and reliable
- *test materials may not be appropriate for certain populations, such as culturally and linguistically diverse clients
-
-
CVA
cerebral vascular accident
-
-
dc
discontinue or discharge
-
-
-
-
-
-
OT
occupational therapy or therapist
-
-
prn
whenever necessary (pro re nata)
-
PT
physical therapy or therapist
-
-
-
-
-
SOAP
subjective, objective, assessment, plan
-
-
WFL
within functional limits
-
-
-
-
|
|