310

Thurstone primary mental abilities

Series of tests of 7 "primary mental abilities incl verbal comprehension, verbal fluency, number, spatial visualization, inductive reasoning, memory, and possibly percpetual speed.

Eugenics movement

Social changes, such as demographic transition, immigration, etc. Not all families following demographic transition. Assumed those that moved up social scale practiced birth control, so less successful were producing larger families. Discovery of negative correlation btwn social class and IQ added to picture and negative correlation between chidren's IQ scores and number of children in the family. IQ must be declining in each civilization then, right. Limited data. Endogeineity. Small negative correlation btwn IQ and fertility. this correlation was lower than the children and children in family study. Probably overestimate, though.

Raven's Matrices

Spearman stated most important ingredient of intelligence was ability to see relationships btwn objects, events or ideas, and draw inferences frm those relationships. Penfield and Raven tok this seriously. 3x3 Matrix of diagrams with third diagram in row missing. Select from alternatives which will complete the third row. 3 versions: Standard progressive matrices, Coloured progressive matrices for young children and advanced progressive matrices for students/adults above average IQ. Paper and pencil, can be administered to large group of people at once. Start simple with mimimum amount of instructions. Can be used for deaf or those with diff first languages who may find examiner's instructions difficult during adminsitration of WAIS or WISC. Correlates with IQ tests such as Standford-Binet and WAIS and WISC.

CHC Theory

Spearman's G >>>>>> Fluid intelligence/reasoning (Gf) Crystallized intelligence/knowledge (Gc) Visual-spatial abilities (Gv) Auditory processing (Ga) Short-term memory (Gsm) Cognitive processing speed (Gs) Decision/reaction time or speed (Gt) Quantitative knowledge (Gq) Reading/writing (Grw) Long-term storage and retrieval (Glr) >>>>>> Narrow CHC abilities, over 70. g correlates most with gf.

Wechsler's Deviation IQ

Standford Binet's formula, Mental age x 100/chronological age only made sense when applied to children whose mental age increases as they grow older. It does not increase indefinitely. Could not rely on grade A-D like army tests. New formula was: Actual test score x 100/Expected Score Expected score was average of large representative sample population. If it was equal to average of one's age group, you got a 100. Deviations from the mean were distributed approximately normally. Deviation is 15. Ex. Test score = 100. That is +1 deviation above average of 80. That means an IQ of 115. Or if you get a 40, two standard deviations below the mean, it means a IQ of 70, and so on.

speed of nervous system and iq

Tmany believed nervous system speed would correlate with IQ and g factor. if true, diff ways of measuring speed should all correlate with another. they do not. IT does not correlate well with performance on others. auditry and visual IT for ex do not correlate. much of variance in IQ scores is modality specific and can't be attributed to any gen process of intelligence small correlation between IT and RT. factor analysis: IT loads on visualization speed, RT onto decision time. oMO RT onto both factors. fact of slowest response indicates speed is not the only thing.

measures of gf in wais

WAIS 4: perceptual reasoning index > matrices reasoning, figure weights, arithmetic. abstract reasoning, non verbal. involves visual analyses: coding, storing part solutions in working emory, and goal monitors to plan overall strategy and setting up sequences of sub-goals, satisfying them and moving onto the next appropriate step

Goddard and Feeble Mindedness

concluded feeble-mindedness was a unit character, controlled by a single recessive gene, by studying case of Deborah Kallikak who had family history of such. Stated no feeble minded person should ever be allowed to marry or become a parent.

jensen's apparatus

console containing central home button above which semicurcular array of eight response buttons with small light above it. start with index on home button, when one light is illuminated, press button below that light as rapidly as possible. releasing home button is measure of RT or decision time. time taken to press appropriate target is movement time. negative correlation with IQ, concluded g is more highly correlated with g factor common to all Hick RT and MT variables than with any particular measure. reliable correlation but not large higher iq,fewer slow responses. increase over time

associative learning

correlates weakly with g. pairings of two events. overshadowing (blocking effect). complex associative learning: simple tasks low g loading, complex high g loading. ex. williams and pearlberg: more complex cued recall associative learning task. each stim paired with 3 possible answers, participants cued as to which Y value they should recall. correlated with matrices. not correlated with processing speed of working memory.

if various verbal skills are distinct, why do psychometric analyses uncover a single dimension of verbal ability?

could be that all primitive processes of language comprehension are derived from a single underlying brain process. correlations btwn measures of diff aspects of language comp could also be explained by interactions between them as they are developed. being able to consolidate info into permanent memory rapidly would aid in acquisition of lexical knowledge and increasing one's vocab would increase one's ability to develop text and situation models which could be used to increase lexical knowledge by defining new words in context.

Thurstone; Primary Mental abilities theory

decmpose mind into number of distinct components or faculties. Good IQ tests would reveal these independent faculties and guide construction of better tests designed to be factorially pure, that is, measure of one factor at a time. find a series of separate factors, one of which accounted for the intercorrelation between one set of tests, second for intercorrelation between a second set, and so on. Imagine as vectors. If tests are highly correlated, they will lie close together, if uncorrelated, far apart. If angles are small between, they correlate with each other. Draw vertical axis, labelled g, through cluster which will represent the principal factor of a factor analysis, i.e. the factor best correlated with all six tests. administered battery of 56 different tests to 240 students, revealed 7 different factors (7 primary mental abilities) and eliminated need for g factor.

correlation with g

does IT correlation with IQ mean a correlation with g? not necessarily. factor analysis of earlier versions of WAIS make it clear that verbal scale loads more strongly onto g than performance scale, so IT scores do not correlate particularly well with g factor of WAIS. there is a negative correlation of -.65 between tests' g loading and their correlation with IT. digit symbol has lowest loading onto g. use with words? no, stil little correlation with verbal IQ.

compontental model of analogical reasoning (encoding, inference, mapping, application)

encoding: listing attributes of each of five terms inference: seeing relation between A and B, listing transformations that must be performed on A's attributes to turn A into B mapping: same process as inference, but applied to relationship between A and C, application: applying A:B transformation to C.

positive manifold

finding of positive correlations between wide variety of different tests. Spearman.

Horn & Cattell: Fluid & crystallized intelligence

fluid: raven's matrices, Cattell's culture fair tests crystallized: tests of knowledge, such as nfo and ocab tests of the WAIS. others incl. gv, spatial visualization. Gr, retrieval. Gs, speed factor.

Vernon

general factor g verbal education v:ed perceptual/mechanical/spatial k:m

digit symbol

loads onto speed factor. means IT correlates strongly with Gs, not g.

Terman and feeble-minded

main use of IQ test to identify feeble-minded so that they could be brought under surveillance and protection of society to curtail reproduction of feeble-mindedness and eliminate amount of crime, paperism, and industrial efficiency.

confirmatory factor analysis

A procedure in which researchers, using factor analysis, consider the theory associated with a test and propose a set of underlying factors that they expect the test to contain; they then conduct a factor analysis to see whether the factors they proposed do indeed exist.

Weschsler tests: WAIS and WISC

Filled gap in Standford-Binet by providing norms for adults of diff ages. sampled over 2k people from 16-75 and a test suitable for those aged 5-16, the WISC test. Have been revised to WAIS-R and WISC-R, WAIS III, WAIS-IV AND WISC III, AND WISC IV. Standford binet consisted of number of different tests which were grouped by age, reflecting concept of mental age. Weschler devised a single test suitable for all ages with each test ranging from easy to difficult items.

inspection time

2 vertical lines briefly shown side by side, one longer, followed by mask to obscure original lines. report whether longer line was on left or right. exposure duration gets shorter for subsequent trials. avg correlation with IQ was -.3. perhaps reflects differences in attention/vigilance. improves with practice. IT-IQ correlation may also depend on differences in rate at which high and low IQ participants show improvement.

Goddard

Advocate of Binet scale. Used it to measure children in America.

General intelligence (g) Spearman's two factor theory

All tests should load onto this gen factor, will depend on size of postivie correlations found btwn various tests. each test measures its own specific factor, but also, to a greater or lesser extent, a general factor that is comon to all tests in the battery. It will explain why tests are correlated. however, IQ tests will reveal clusters of high correlations between different tests of abstract reasoning, vocab, or spatial visualization separated by lower correlations btwn tests in one group and those in another, verbal and spatial for ex. Extracting the general factor, then, leaves a residual matrix that will still contain six high numbers resulting from high correlation between three tests within each group.

Alpha and Beta Tests/Army Tests

Alpha: verbal analogies, series completions, synonyms, antonyms Beta: designed for illiterate. Men req to fill out missing part frm incomplete picture of familiar object like pig without a tail, etc. Visual search task to find same numbers, recording task of matching digits of ten diff patterns. Provided huge body of info about nation's intelligence, tested 1.75 million young men in US in under a year.

Flynn Effect

Are IQ test scores decreasing from one gen to the next? Found higher IQ in children than parents, but needed to test parents at same age. Avg test scores increased at remarkable rate in most industrialized societies in 20th c., the Flynn Effect. Is it just an increase in IQ or real intelligence? Who knows. It's a paradox. Comphrehension tests may seem unnatural to older generations and more easy to those brought up in scientific 20th/21st c.

Feeble-mindedness in US

Brought Binet test to attention of scools. Dr Isaac Kerlin began castrating some inmates of training school for feeble-minded in 1889. Dr. harry sharp pioneered sterilization of criminals by vasectomy 1907 Indiana sterilization law on inmates Cast net to insane, epileptics, drunkards, criminals, prostitutes, tramps and paupers, the undesirables classes. Characteristics said to be hereditary. IQ test, introduced in US in 1910/1911 was not responsible for movement, but did have an effect and justified sterilization in case of mental disability.

Raymond Cattell's tests, 1940: Culture-fair intelligence

Culture-fair intelligence test: critical of standford binet or wechsler models that would not give fair pic of intelligence from children of different backgrounds because they asked meaning of meaningless english words or cultural knowledge that wasn't equally available to all or things that weren't taught in school.

Control of immigration

Goddard admin'd binet tests to newly arrived immigrants in NY, found high proportion tested at feeble minded level or below. In Alpha/Beta tests, English, Scots, Dutch and German scored A/b, but less than 1% of Italians, Poles, or Russ were in these grades. Steps to preserve / increase intellectual capacity, Immigration Act of 1924 to est. national quotas based on national groups beginning 1890 census Marked beginning of new immigration from southern and eastern europe. Objections were more econ, pol, and social rather than intellectual.

Factor analytic approach

Goes back to Spearman's two-factor theory of intellignece, general factor of intelligence and general discriminative ability. Weak empirical basis, but influential. fundamental assumption: corelation between any two tests may be explained as a consequence of each test's correlation with a third, hypothetical measure/factor: r.1.2 = r.1g x r2g the correlation between tests 1 and 2 (r1.2) is equal to the product of each test's correlation with a hypothetical factor labeled "g." the correlation between tests 3 and 4 is explained by correlation with hypothetical underlying factor, which may or may not be the same as that responsible for the correlation between tests 1 and 2. aimed to find a solution to and simplify the correlation matrix. Factor analyses enable one to find out whether there is any evidence for such clusters of different abilities in addition to any general ability. Spearman and Thurstone approaches, both partially right and partially wrong.

Cattell and Wissler

Inspired by idea that intellectual ideas could be measured by sensoral acuity, reaction time, etc. Published mental tests to measure differences in fundamental mental processes. Wissler found no correlation between mental tests and that none of the tests measured intelligence. Now it is clear there is a correlation btwn reaction time and IQ. Wissler's experiment was flawed.

verbal intelligence

Intelligence measured by answering questions involving vocabulary, general information, arithmetic, and other language- or symbol-oriented tasks. reflective in scores of Gc. somewhat correlated set of skills including access to lexical memory, rapid consolidation of info into long-term memory, the posession of knowledge about how to process discourse in general, and knowledge about the topic.

tacit knowledge

Not openly expressed, informal, not directly taught, inaccessable, ill suited for direct instruction. Sternburg and Wagner.

tests of non-verbal reasoning, componential analyses

RPM and RAPM, Cattell's tests, letter-series problems. Requires such operations as coding of items and searching for rules that describe the transformation of one item into another. Simple componential models of analogical reasoning fail to acknowledge the independence of the various components,though. Coding of relevant attributes of each term of an analogy is influenced by the detection of the relationships between the various terms, since those relationships define the relevant attributes. Failure to detect a relationship that will solve the analogy may suggest your initial coding was incorrect, req backtracking and starting again.

hick's law

RT increases as function of number of alternatives btwn which the participant has to choose. linear function. would reaction time and IQ correlation increase with increase in # of alternatives? nah, small effect.

worst performance rule

RT, OMO RT. correlation with IQ -21 for fastest but -.36 for slowest.

Guttman's Radex Model

Tests represented in circular space with complexity of test inversely related to distance from centre of the circle, referred to as simplex. tests loading onto g would be close to the center, those loading less onto g would be near the circumference. circumplex: describes which direction from the center in which a particular test lies: tests more closely related to one another will lie in neighbouring directions from the center, while those less closely related might lie in opposite directions. ex. raven's matrices/letter series at the center defining Gf while verbal/spatial/memory/perceptual speed test form clusters around this core. Gf loads more strongly onto g than Gs, Gr, Gc Gv, etc.

Galton's contribution to IQ

There is variation among members of any species and that variation is inherited. Studied this with Cambridge students, noting there was a considerable range in the number of marks awarded in the honour examinatinos in mathematics from 250 to over 7500 in a 2 year period. Needed to prove this eminence ran in families. So, he studied judges. Of the 286 judges appointed to English bench btwn 1660 and 1865, 109 had one or more relatives who were also eminent. Conducted similar analysis of statesmen, military commanders, and scientiests, poets, painters, musicians, etc. Endogenity, compare with adopted sons like nephews of popes and dignitaries of Catholic church. Those sons less likely to achieve eminence than natural sons of eminent men. Social helps the same but hereditary gifts not. Ancedtodal, wanted to measure. Thought it could be measured by sensory acuity, reaction time, and etc.

Content of Wechsler tests

WAIS III-13 tests. WAIS-IV 15 tests in 4 groups. Distinctions btwn verbal and nonverbal tests, originatin to alpha/beta tests. Diversity of content ensures test is fair and goes back to Binet's notion that intelligence could manifest in different ways. They should all correlate with each other. Cattell's did not, and it was concluded that they could not all adequately measure intelligence. It would reveal Spearman's g factor. Shouldn't correlate exactly, lest they are redundant.

measures of Gc

WAIS III: verbal scale vs performance scale WAIS IV: verbal comprehension (VCI). Vocab subtest, information subtest, comprehension subtest, similarities subtest. tests of vocab size, high loading in Gc. correlates with global IQ. raw scores improve with age until 60. Artificial vocabularies. learn arbitrary names for stuffed animals. matched for Gf (raven's matrices). children who had lower vocab scores did poorly on memorizing new words with equal opportunity to learn. implies differences are not just opportunity differences. memroy differences? language as a tacit learning. infer meaning. infer info about subcomponents of the word. Ritchie, Bates, Plomin test assessing reading abilities in twins. differences not explained by reading exposure differences, correlated with intelligence. information subtests correlate with vocab. comprehension tests, meaning of sayings/proverbs, what is appropriate thing to do in situations, etc. conceptualization of verbal reasoning/social judgement/knowledge of social rules and crystallized intelligence. raw scores increase with age.

Binet

Why should we need a test to prove intelligence? Needed to find children who were good at school, new classes of children were causing problems for over stretched teachers. Find who needed to be assigned to special classes. Thought common sense, practical good sense, iniative, ability to adapt were essential ingredients to intelligence, not scholastic knowledge. So, his tests measured everyday practical knowledge and skills like pointing out parts of body, name objects seen in picture, give definitions, repeat series of digits or complete sentence, copy a diamond, say what the difference is between objects, find rhymes, tell time on a clock and say what time it would be if clock changed, etc etc. As children grow older, they produce better answers. So he developed a series of test items ranged according to age at which children in sample were capable of solving. Ex. few 4 year olds would be able to complete 6 year old tests, most 8 year olds should be able to. Used to assign a child a "mental age." Produced a relative measure of intelligence. Binet scales formed basis of modern IQ tests just as mental age formed basis for IQ scores.

implicit learning

acquisition of knowledge that takes place largely independently of conscious attempts to learn and largely in the absence of explicit knowledge of what was acquired Reber states it is product of basic associative system while explicit is product of cognitive system which generates tests and hypotheses.

principle of the indifference of the indicator

more diverse the set of tests included in a test battery, the better that battery would measure general intelligence because the idiosyncrasies in particular tests would cancel each other out. But certain tests correlated more highly with some than others, discovered by specialized sets of variables like visuo-spatial abilities.

IQ tests and implicit/explicit knowledge

more likely to measure explicit system. Reber study found no correlation between IQ and performance on implicit learning but sig. correlation between IQ and explicit letter series completion task. Other studies point to same conclusion.

why a positive manifold?

mutualism: cognitive processes have mutual beneficial or facilitating relations each process supports development of other processes > some studies in development show that scores on different IQ subtests are more correlated in young children then correlation becomes weaker then stronger again..

localization of functions

particular function may be localized in a particular region of the brain, but it doesn't help why people differ in verbal intelligence. whether or not the brain can be divided up into distinct cognitive modules is a question entirely independent of whether or not individual differences in performance on tests involving one module correlate with individual differences on tests of other modules. g may refer to efficiency of central processor that controls performance of all nodules.

Lewis Terman

produced Standford revision of 1911 Binet-Simon scale, called Stanford-Binet test. 40 completely new items added. Tested over 1000 children aged 4-14. Obtained much more accurate info. Defined IQ as mental age/chronological age x 100.

reading span test (daneman and carpenter)

read sentences, report words at end of eah sentence. words crowd WM. with a larger working memory, can hold onto more words.

tacit knowledge correlations

reliable correlations btwn performance on diff tets of tacit knowledge, such as psych and business questionnaires. many load onto factor of general tacit knowledge correlations for implicit learning tasks are weaker

visuo-spatial ability tests - Thurstone 1938

roughly same correlation with raven's matrices that raven's matrices has with verbal tests. provided distinction for nonverbal tests into general abstract reasoning and tests of visuo-spatial skills. Thus, verbal, non verbal reasoning, and spatial IQ. argued intelligence consisted of some half-dozen specific abilities, attempted to devise tests that measured each of these abilities independently of others.

odd man out reaction time test (OMO RT)

same eight light box as Jensen. 3 lights illuminated each trial, two close together, one further apart. respond to odd man out light. correlation between this and IQ, higher than standard RT tasks and IQ.

implicit learning tests

sequence of events, ex. series of lights illuminated on computer screen or letters of alphabet. artifiial grammar learning: memorize apparently arbitrary letter strings that are gen by Markov process. then told there is a set of rules and asked to judge whether some new leter strings are or are not grammatival. people are usually unable to articulate the rules, but perform sig. above chance. can be asked to manage a specific factory and take decisions about purchasing policy or employee relations, whose outcomes are dependent on a large number of variables. their performance indicates they have learned something, but their knowledge is fragmentary.

latent variables

sum of battery tests = IQ score, but if some tests have higher g loadings than others, the sum may not be best way of estimating intelligence. true estimate of g would be obtained only by discarding specifics and calculating based on person's score on g factor or measuring verbal/spatial, but extracting verbal/spatial factors from each of the two groups of tests and calculate scores on these factors.

Standford-Binet

the widely used American revision (by Terman at Stanford University) of Binet's original intelligence test.. Did not provide norms for adults of diff ages.

associationist theory of multiple processes

there are millions of processes involved in the cognitive tasks included in the battery. tests that load into g are simply tasks that sample a high proportion of these millions processes.

central executive and executive control

○ Function of goal monitor was to decide order in which sub problems should be tackled, keep track of which sub goals had been satisfied and modify path being taken when a difficulty arose ○ Decline in gf possibly related to decline in exec function

zahlen verbindungs test (ZVT)

trail finding, 10x9 grid containing semi-random arrangement of numbers 1-90. trace line connecting numbers in ascending order. score can be time taken to complete task or number of items correctly completed in fixed time. Vernon adopted it, and did it in diff ways like descending order, connectin every alternate order, interspersing numbers and letters from 1 to a 2 to b, etc. multiple choice versions. correlates with raven's matrices about .42. performance correlates with IQ in range of .3-.6, reliable and sustainable.

measures of Gv

vary among people. can be independent from IQ. dyspraxia, etc. spatial visualization, orientation and speed of mental rotation. wais iv: picture completion, visual puzzle subtest, block design subtest.

working memory and exec function

working memory: capacity for controlled, sustained attention in face of interference or distraction. low attention span groups decline greater than high spans, implying greater susceptibility t proactive interference. Low span also susceptible to stroop effect, like red in green ink, they will hesitate more. working memory tests executive function

is g real?

yes, g is a statistical entity. we don't know if it is a single underlying ability. some theories around speed and mental efficiency.

working memory

§ Daneman and carpenter, sentences, remember last work □ Difficult □ Hgh correlation btwn that and ability to answer factual questions about passage □ Remember some information at the same tme processing new information

numerical ability

§ Daneman and tardif, three numbers, join two together to create new number divisble by 3 § Then recall correct answer to each item at end of series, correlated .57 with word span test and .51 with compsosite vocab and comprehension scores § Leather and henry gave 7 yr old 2 wm tasks, one listening to sentecnes and recalling last words and req to count number od dots on each series of cards and recall total number for each card § Also standard vocab and reading compreension test and arithmetic test § Correlation, listen span predicted vocab and reading comp better than count span, count span predicted arithmetic better than verbal tests, but two span tasks correlated well and predicted arithmetic scores equally well

verbal fluency

§ F words, boy's names, animal names in a minute § Frontal lobe damage § Females better § Plan search more efficiently by grouping responses into defined clusters

increasing intelligence test

§ Jaeggi et al, gave participants continuous training on an n-back tasks for 8, 12, 17, or 19 days and compared performance on tests of gf before and after training with control group with no n - back training. Eight days of training, no effect, but 12 days, suggestion of difference btwn experimental and control groups , came stati sig after 17 nd 19 days Groups were tested with diff tests though. Gave particpants only 10 mins, when usually there is 45 mins. Can't measure much in 10 mins, just easy one get tested

spacial visualization

§ Measure ability to form or construct representations of relatively complex visual patterns, shapes, or figures, representations that are accurate, detailed, and durable that various operations can be performed on them § Establishment of representation of target and retention and processing of spatial information, becomes a working memory task § Cube with letters, change to match other cube

spatial ability and visual imagery

§ Operate on or otherwise transform mental representations of visually presented objects § Differences are not caused by visual imagery, blnd ppl can still solve ethese ○ Navigation and other special abilities § Paper and pencil tests of Gv correlate with navigational skill, but not very strongly

speed of lexical access

§ Posner letter matching, stimuli are alphabet letters § Physical identity: two physically idnetical letters A-A or a-a while diff means upper and lower case letter A-a § Letter name identity two A's still count same even if shown in diff type face A-a § Homonym identity condition § Ppl with higher Gc scores repond more rapidly than those with lowers § RT tasks § Sentence verification test: answer as rapidly as possible to true/false description of simple diagram § Why? Differences in lexical access may reflect differences in amount of practice pppl have had with related tasks such as reading § Jackson 1980 taught artifical allphabet and still found sig correlation § Differences in Gc are differences in speed of access to and effieicny of processing and retrieval of lexical material

spatial ability

§ Three dimensional mental rotation, paper folding, form board, DAT space relations test § Second group of tests to oriente self in wider spatial frame of ref or visualize a scene from a diff perspective has sometimes identified Gv, called spatial orientation but little evidence § Diff ppl may adopt diff strategies for same problem and sometimes same strategy for solving what the test constrcutor had assumed were different tests

Gardner's multiple intelligences

○ Expertise prob more important than intelligence ○ Thesis of Gardner states that it is intelligent that underlies all of musician, athlete, etc's achievements, there is more than 1 type of intelligence Linguistic, Logico-mathmatical, visuo spatial, musical, bodily kinaesthetic, naturalistic, inter and intrapersonal question of whether to call them intelligences or skills, talents, abilities mozart for ex. was great at music, but evidence for intelligence is desisory. also, good vs intelligent athletes. § Gardner insists inellgiences indpendent of one another § Lingusitic, loico and spatial same as or closely relate dto gc, gf, and gv § Signs of intelligence he said were islands of ability, exceptional ability in one domain but often below average in others § Most savants have below avg test scores, but particular skills like claendar counting are not independent of IQ § Correlation was .78 with IQ