mental abilities Flashcards
construct
theoretical entity that cannot be directly observed, we infer from observable behaviour, called a LATENT VARIABLE
mental abilities
capacity to perform the higher mental processes of reasoning, remembering, understanding and problem solving
manifest variables
observable behaviours that tell us about the latent variable - how quickly you solve a task, how many solutions you come up with - use to infer an underlying construct
implicit theories of intelligence
informal definitions of intelligence + beliefs you have around it –> entity mindsets, incremental
explicit theories of intelligence
use data collected from people doing tasks that require intelligent cognition
binet’s scale purpose
isolate natural ability independent of school context, questionnair which could measure sort of competence kids of different ages could have; assigning mental ages, IDENTIFY KIDS IN NEED OF REMEDIAL EDUCATION
contributions of HH goddard
goddard the racist eugenicist!!
* made classifications from idiot to moron, made a testing station to monitor US immigrants, undesirable
contributions of L Terman
- updated Binet’s thing into Stanford-Binet test, as he thought mental age might seem restricting
IQ
intelligence quotient , introduced in stanford-binet test
ratio iq + calculation
considers both mental and chronological age, proposed by William Stern
* allows comparison of intellectual performance across levels
RATIO IQ = (MENTAL AGE/CHRONOLOGICAL AGE) X 100
problems w ratio iq
difficult to make comparison –> is 13yr w MA 15 doing better than 8yr w MA of 10??
difficulty of application in adults - what’s an activity appropriate for 45 yr old but not 46 yr old
deviation scores
- most mental abs form bell curve
- z-score: Z = (X-M)/SD
iq score calculation
- find the z-score
- take the z-score and multiply by SD
four domains tested by stanford-binet
- verbal reasoning - vocab, comprehension
- abstract.visual reasoning - patterns, copying
- quantitative reasoning - number series, digits
- short-term reasoning - sentences, object remembering
stanford-binet format
- multuple sep tests of progressive difficulty
- avoid boredom = intermixed tests
- 30-90mins
- accessible formats
standford binet test process
est basal + ceiling level for tasks
* basal = lowest level, could have done anything that came before
* ceiling: fail 3 items out of 4 in a row
stanford-binet scoring
50-100
70-79 = borderline impaired/delayed
90-109 = average
130+ = gifted v adv
problems Raven’s Progressive Matrices solves
- can be done w groups en masse (save time, money)
single underlying factor vs multiple abilities
- single underlying factor = because all results on intelligence tests are at least somewhat positively correlated ,suggests singular underlying –> spearman’s g, positive manifold
- multiple abilities - correlation matrix leads to patterns of grouped correlation leading to different factors
middle ground theory
there is a hierarchical structure
* g factor splits into verbal reasoning + reasoning
* verbal reasnoning –> vocab, verbal comp, gen knowledge
* reasoning –> abstr reasoning, matrices, series comprehension
who came up with Gf-Gc theory and when
raymond cattell 1941
what’s the strongest evidence gf and gc are different constructs + what are patterns
- they show different developmental trends
- gf: peaks in early adulthood, then drops gradually through life
- gc: rapid increase until early adulthood, then plateau
validity
extent to which our tests tests what they’re supposed to measure, and whether the answer is used in an appropriate way
reliability
how consistently does that/accuracy
* if a test measures a consistent trait it should consistently produce the same answer, any change in performance due to true differences in that ability
how to test reliability
test-retest –> test on same person and if the test gives a different result it’s not super reliable
estimating reliability
- test-retest reliability: same group of people measured twice w same test
- equivalent/alternative forms: measure same phenomenon using 2 different forms of test
- internal consistency: uses cronbach’s alpha
classical test theory
observed score on every test has two components: true score (real level of ability) and error component (random variance)
X (observed score) = T (true score) + E (error)
content validity
what the test covers: does it cover all the relevant parts, does it test the kind of behaviour relevant to what we want to measure, focus too much on one aspect
construct validity
how well is construct measured by the test
convergent validity
is construct related to other theoretically similar constructs/tests
degree to which two measures that should be related are actually related
discriminant validity
degree to which measures that should not be related are actually unrelated.
who’s head honcho of heritability
sir francis galton
heritability definition
the proportion of the total variation in a given characteristic in a given population that can be attributed to genetic differences between members of that population (proportion 0-1)
how to estimate heritability
- family resemblance studies (problem: genetic relatedness usually close w environmental similarity)
- adoption/twin studies
genotype
underlying genetic factors Gv
phenotype
expression of underlying genetic factors (Pv) that is due to genetic factors (Gv)
problems with heritability estimates
calculating H assumes genetic + environmental contribution are independent, which might not be true
* heritabiltiy seems to vary with SES
* age of comparison: correlation change from childhood to adulthood
* selective placement of adoption - children tend to be adopted out to higher SES
why use monozygotic twins to determine heritability of mental abilities
share 100% of DNA, compared to 50% (dizygotic)
If MZ twins show more similarity in mental abilities than DZ twins, it indicates a significant genetic influence
gender differences in IQ over time
- old: females perform better, but now gap closing
race differences in IQ
avg white population 15 points higher than black population. high degree of overlap, high variability. asian americans score better than white
hernstein + murray’s theory about race and IQ
argue intelligence scores are genetically inherent, should be taken into account into school/social policy, welfare policies encouraging poor women to reproduce should be scrapped
Flynn’s theory
agreed to an extent that environmental differences aren’t enough to explain IQ differences
* but said the truth can’t be racist/held suspect for telling the truth UNLESS their assessment of evidence falls below minimum level expected from a scholar (and it did)
* GRADUAL CHANGE IN EVERYDAY LIFE FROM THINKING/REASONING TO ABSTRACT THROUGHT so explanation in IQ changes are due to MODERNITY
bell curve premises
- there must be a meaningful single number given to intelligence
- we must be able to rank people linearly, predicting social outcomes
- IQ must be highly heritable
- IQ must be effectively unchangeable
IQ and Bias
- doesn’t have statistical bias
- does have systematic statistical unfairness
mindset + stereotype influence on IQ
- believe your group has lower IQ = do worse –> Steelee + Aronson 1995 Stanford experiment –> race difference only manifests when told the test was diagnostic of intellect
Steinberg + adolescent attitudes towards school
- white students - conspicuous effort frowned upon, also underperformance
- AAs - high achievement valuable but not threatened by low, study not as important
- Asian-Am - high achievement important, requires sustained effort
