Chapter 11 Testing Flashcards
Standardization
Uniform procedures used in administration and scoring of test
Francis Galton
Pioneer in intelligence testing, believed intelligence was genetic
Test Norms
Where your test score is relative to everyone else’s
Psychometrician’s
People who make tests
Standardization sample
- group of people a cluster of test questions is being tested on for future examinations
- if everyone got it wrong or everyone got it right, they will not include it next time
Reliability
Consistency of the test as a means of measurement
Types
- split half reliability
- test-retest reliability
- equivalent form reliability
Split half reliability
- divide test into 2 sections, correlating ppls performance on the 2 halves
- closer correlation is to +1, more reliable it is
Equivalent for reliability
Correlation bw performance on diff forms of a test
Test retest reliability
Correlation bw persons score on the test the first time they take it with those taken subsequent times after
Test validity
- when a test measures what it is supposed to measure
Types
- content
- criterion-related
- construct
Content validity
How well an exam reflects the entire range of material it is testing
Type
- face validity
Face validity
More obvious form of content validity
Ex: if your testing if you’ll be a good chef, don’t ask dating/relationship questions
Criterion related validity
Correlating scores on the test with other established examinations to see if they have similar scores
Concurrent- test scores and criterion measurement made at same time
Predictive- criterion tests in the future; like in the sat test- does getting a good score predict doing good in college?
Construct validity
- Scores of trait that test is supposed to assess is compared with scores on tests measuring other closely related skills or traits to see, there should be a correlation.
- scores of trait that test is supposed to assess is compared with scores on tests measuring traits or skills that are not generally clustered or associated with the trait of the first test, should be little correlation. If there is correlation, test is not measuring what it is supposed to measure
Types of tests
- aptitude
- achievement
- speed
- power
- group
- individual
Aptitude test
- measure potential or ability
- ex: intelligence tests
Achievement tests
- Measures what one has learned
Intelligence theorists
- Charles spearman
- Howard Gardner
- Daniel goleman
- Robert sternberg
Charles spearman
- With factor analysis, thought diff types of intelligence s could be expressed as single factor, g
Howard Gardner
- thought there was more multiple intelligences
- linguistic
- logical
- spatial
- musical
- kinesthetic
- interpersonal
- interpersonal
- naturalist
Daniel goleman
- Supporter of importance of EQ (emotional quotient) test
- high IQ and EQ needed to succeed
Robert sternberg
- triarchic theory of intelligence
- analytic, creative, practical
Alfred Binet
- made test designed to see which children in schools needed special attention, were in step with peer group, and were ahead of peers
- used concept of mental age
Louis Terman
- used Alfred Binet’s test to create the IQ test (Stanford-Binet IQ)
- assigned all adults chrono age of 20
- mental age over chronological age times 100
David wechsler
- made IQ test for adults, children, and preschool/primary kids
- score based on deviation IQ
Deviation IQ
- scores standardized so mean is 100, deviation is 15, and scores make normal distribution
Normal distribution
- 68% scores within one standard derivation
- 95% scores within 2 standard deviations
- 98 or 99% within 3
Test bias
- ppl think its biased and favors white, middle and upper class rather than other minorities that may have not been exposed to some of the vocab on the sat or IQ tests
Heritability
- measure of how much of a traits variation is genetic
- from 0-1, closer to one, more genetic
Intelligence: Nature or nature
- Since gene pool is stable, Flynn effect suggests nurture
- identical twin’s scores have higher correlations than fraternal twins suggesting nature but identical twins are also treated more similarly than fraternal twins confounding the nature variable
- identical twins seperate day birth have strong correlations of IQ, but some argue they are still placed in similar environments
- racial diff in IQ scores suggest nature, but many argue that the bigger level of poverty for minorities and testing bias is the environ cause for disparity
- gov programs like head start for minority groups correlates with higher scores on intelligence tests
Flynn effect
IQ scores have increased over time
