Exploring Validity with Intelligence Testing Flashcards
To understand what it means to say a test is valid. To understand the distinction between reliability and validity To consider why it’s important for tests to be valid To examine whether intelligence tests are valid measures of intelligence.
what is validity?
degree to which the test actually measures what it purports to measure
what is the problem with the definition of validity?
do not really purport to measure anything, instead people do
how does cattell suggest intelligence?
assessed by psychological tasks, such as judging which of two sounds was larger or which of the two weights were heavier
Wissler’s flawed validation of a psychophysical tests of intelligence in the Cattell tradition was..
…largely responsible for the twilight of the psychophysical test era of intelligence testing and the dawn of the era of judgemental tests in the Binet-Simon tradition
test validity
extent to which a test accurately measures what it is supposed to measure
internal reliability
different items within a test correlate positively with one another, suggesting they are measuring the same construct
* high reliability -> all of the items are indicated to be measuring the exact same thing
test-retest reliability
The reliability of a test over time. So, if the same person took an intelligence test at different points in time, you would expect the result to be similar.
how do we test-retest reliability in circumstances?
the split-half method can be used to test this
what are some ways in which intelligence tests have been / are used
- selection within education
- job selection
- managing immigration
- diagnosis
why is test validity important?
need to make sure a test is measuring what we think it’s measuring
what do we risk without test validity?
drawing false conclusions from its results
what is a problem with test validity?
it often creates a closed system -> makes the test look good regardless of what it is
what’s the opinion of new tests departing from old tests?
even if it’s better, the worse the test looks
internal-consistency reliability
commonly important parts of the psychometric analyses, value items within a test to the extent they are too like each other
what does reliability analyse and validity analyse do?
reliability maximises similarties among items and validity maximise similarities among the tests