Mid-Semester Test Flashcards
(40 cards)
Describe Psychometrics
The study of theories and techniques in the measurement of psychological attributes, eg. language and speech.
What makes a test standardised?
Designed to be administered and scored in a consistent manner.
What does norm-referenced mean?
score is compared to a sample of matched peers who completed the same test. Norms gathered in process of standardisation.
What is a criterion-referenced test?
Compares performance to a pre-established level of mastery, states whether a task can be performed. Can be standardised.
What kind of data is retrieved from a descriptive test?
Qualitative data - may or may not be standardised.
What is validity?
A measure of whether the test does what it says it does. Is the purpose explicitly stated? Does the construct or model relate to the stated purpose?
What are the 6 types of validity?
Construct Validity
Content Validity
Criterion-related Validity
Face validity
Structural Validity
Cross-cultural Validity
What is construct validity?
Whether the test measures only the concept/idea stated and nothing else
What is Criterion-related Validity?
Do the results reflect results from ‘gold standard’ measurements?
What is content validity?
Are the test items relevant to the skills being assessed? Does it measure all of the skill or only part of it?
What is Face Validity?
The judgement of informed stakeholders that the items that make up an instrument are representative.
What is Structural Validity?
Everything in the test is there for a reason
What is Cross-cultural Validity?
Is the population valid for context and norms?
What is Sensitivity?
The test’s ability to correctly identify people with a language difficulty.
100% sensitivity = test correctly identifies every person with a disease
What is Specificity?
The test’s ability to identify people that DO NOT have language difficulties.
100% specificity = everyone without a disorder is identified
Describe the relationship between sensitivity and specificity. What is a good %?
As one goes up, the other goes down.
good = 90%
fair = 80-90%
unacceptable <80%
What is reliability?
Whether the test produces the same score when done at different times.
What is intra-tester reliability?
Does the same tester produce the same results each time? Recommended 90%
What is inter-tester reliability?
The test produces the same score when done by different testers. >90% good
What is test-retest reliability?
Whether the person would receive the same score if given the test at different times.
What is instrument reliability?
Measures how different the test items are from each other, whether they measure the same characteristics or skills.
What is SEM?
Standard Error of Measurement - confidence interval. Used to provide an indication of how confident you are that the test score represents the person’s ability.
What are the typical confidence intervals used? What happens to the spread of the scores as confidence interval goes up?
68%
90%
95%
The spread of the scores increases as the confidence interval goes up - takes into account more of the scores. eg. 90% score the child would score within this range.
What is a raw score?
The number of points that the client scored correctly. Cannot be used to compare performance to anyone else.