Test and Measurement Chapter 9: Theories of intelligence and the ...
Test and Measurement Chapter 9: Theories of intelligence and the Binet Scales The problem of defining Intelligence
Alfred Binet defines intelligence as: the tendency to take and maintain definite direction the capacity to make adaptations for the purpose of attaining a desired end and the power of auto criticism Spearmen: the ability to enduce wither relations or correlates There are many different definitions of intelligence given all by different researchers T.R Taylor identified three independent research traditions that have been employed to study the nature of human intelligence: o The psychometric: examines the element of structure of a test , and then examine the properties of a test through an evaluation of its correlates and underlying dimensions o The information processing approach: we examine the process that underlie how we learn and solve problems o Cognitive tradition approach: focuses on how humans adapt to real world demands People have said that intelligence tests are biased because they do not consider socioeconomic backgrounds There are many critiques of intelligence test The Binet scale has been revised at least 5 times Formal intelligence testing began with a French minister of public instruction around the turn of the 20th century created a test to remove the lower scoring students from the classroom to receive special education experiences o 1904- he gave appointed a commission to which he gave definite assignment: to recommend a procedure for identifying so called subnormal (intellectually limited children) o Binet was a part of this commission Experiment: o Wissler indicated that simple functions such as reaction time and sensory acuity failed to discriminate well among individuals of high and low scholastic ability
Binets principals of Test Construction
Defined intelligence as the tendency to take and maintain definite direction the capacity to make adaptations for the purpose of attaining a desired end and the power of auto criticism Binet wanted to measure judgment, attention and reasoning He used trial and error, and was guided my two major principals Principal one: Age differentiation
o Refers to the simple fact the one can differentiate older children from younger by the formers greater capabilities o Binet searcher for tasks that could be completed by 66.67-75% of the children of a particular age group and by a smaller proportion of younger children o He did come up with these tasks, and could therefor estimate the mental ability of the child o The mental age is independent of the chronological age Principal two: General Mental Ability o Guided by his decision to also measure only the total product of the various spate and distinct elements of intelligence, that it general mental ability o This meant that he freed himself from having to find the total or final product of intelligence o He would judge the value of a score and correlate with the total
Spearmen’s Model of General Mental Ability
Advanced the notion of a general mental ability factor underlying all intelligent behaviour- before Binet He said intelligence consists of one general factor (g) plus a large number of specific factors Referred to as psychometric g (general mental ability) ^ Based on the phenomenon that when a set of diverse ability test are administered to large unbiased samples of the population, almost all of the correlations are positive (called positive manifold) Refer to the hydro example (238) Factor analysis: is a method for reducing a set of variables or scores to a smaller number of hypothetical variables called factors. One can determine how much variance a set a set of tests or scores has in common This common variance represents the g factor Half of the variance in a set of diverse mental ability tests is represented in the g Implications of general Mental intelligence o Concept of general intelligence implies that a persons intelligence can best be represented by a single score g, that presumably reflects the shares variance underlying performance on a diverse set of tests o True performance of any given task can be attributed to g as well as a unique variance The gf-gc theory of intelligence o There are two basic types of intelligence; fluid and crystallized o Fluid: abilities that allow us to reason, think and acquire new knowledge o Crystallized: represents the knowledge and understanding that we have acquired
The Early Binet Scales
Collaboration by Binet and T. Simon Eventually called Stanford-Binet intelligence Test The 1905 Binet-Simon Scale o 30 items shown in an increasing order of difficulty o Individual test o Idiot- described the most ever from of intellectual impairment o Imbecile- moderate levels of impairment o Moron- mildest level o Item 6- upper limit of adult idiots o Item 8- would rule out sever intellectual impairment o Item 16-upper limit for adult imbeciles o First measure of human intelligence o Lacked adequate measuring unit, normative data and validity documentation o Based on only 50 children The 1908 Scale o Was an age scale- the items were grouped according to age level rather that simply increasing the difficulty o Used in only hybrid fashion in modern times o It makes comparing a child’s performance on different kinds of tasks near impossible o Produced only one score exclusively related to verbal, language and reading ability o Was better than the 1905, but improvements were still needed o Introduction to the mental age
Terman’s Stanford-Binet Intelligence Scale
They revised it again in 1911 with small improvements Terman’s 1916 is the one that really flourished The 1916 Stanford-Binet Intelligence scale o Increased the size of the standardized sample o Was exclusively white native-Californian people o ^ Was not representative o Went from 50 to 203 individuals o The intelligence Quotient IQ: a unit for expressing the results of intelligence tests, based on the ratio of the individuals mental age (MA) as determined by the test and to actual chronological age CA IQ= MA/CA x 100 o There was a maximum mental age of 16 which was limiting The 1937 Scale o Extended the age range down to 2
Increased max mental age to 22 years, 10 months Scoring standards and instructions were improved Added items decreased focus on verbal skills The test was still not balanced as only 25% of the test was non verbal The sample size went up to 3184 from 11 different us states (still focused on whit urban people) o There was an inclusion of an alternate equivalent form Problems with the 1937 Scale o Its reliability coefficients were higher for older subjects than for younger ones o Reliability figures also varied as a function of IQ (lower the IQ the higher the reliability) o Unique standard deviations were also a problem, this meant that an IQ score at one age did not reflect the same IQ score at a different age 1960 Stanford-Binet revision and Deviation IQ (SB-LM) o Tried to create a single instrument from the best of the old tests o Task that showed an increase in the percentage passing with an increase in age received the highest priority as did tasks that correlated highly with scores as a whole o Instructions for scoring were improved and IQ tables were extended from age 16-18 o Deviation IQ- solved the differential variation in IQ’s, it was simply a standard score of 100 with a standard deviation of 16 o Mean set at 100 o It was ascertained by evaluating the standard deviation of metal age for a representative sample at each level o IQ tables were corrected so that the IQ sores were Comparable o In 1972 there were more people in the standardized sample (2100) and it included non whites o o o o o
The modern Binet scale
Model for the fourth and fifth editions of the Binet Scale o These version incorporated the Gf-Gc theory o They are based off a hierarchal model o At the top is G which reflects common variability of all tasks o At the next level there are three group factors o Crystallized abilities reflect learning-the realization of original potential through experience Has two subgroups: verbal reasoning and nonverbal reasoning o Fluid-analytic abilities- represent original potential or the basic capabilities that a person uses to acquire crystallized abilities o And the last group is short term memory o The role of Thurstones multidimensional model
He argues that intelligence could be best conceptualized as a comprising of independent factors or primary mental abilitiesthese group factors were correlated and from them a g factor could be extracted Characteristics of the 1986 revision o Age scale was entirely eliminated o Items with the same content were now placed together o In the 2003 fifth edition here are now five main factors rather than four o This means that you can now get smaller scores for the individual groups (because grouped by content now) o The drawback is less variety in the items Characteristics of the 2003 fifth edition o Non-verbal and verbal scales are equally weighted o The exam begins with two routing measures- one verbal and one non verbal, they are organized in a point scale (similar content of increasing difficulty) o The point of the routing test is to examine there ability level o The remaining 8 subtests are arranged in an age scale o Equal verbal and non verbal which allows an individual content score o There is mixing within the groups o Items that are too easy are skipped, so you start at the estimated ability from the routing test o Estimated level of ability is called the start point o Basal- level at which a minimum criterion number of correct responses is obtained o They continue until they hit the ceiling in which a certain number of incorrect responses have indicated the items are too difficult o The scale for the subgroups have a mean of ten and an standard deviation of three o In addition there is a standard score with a mean of 100 and a SD of 15 for nonverbal and verbal IQ, full scale IQ and each of the five factors Psychometric properties of the 2003 fifth edition o Age range goes from 2- 85+ o Sample increased to 4800 of all ethnicity’s and gender o 300 addition people were included that were gifted, mental retarded, had ADHD o Possible scores range from 40-160 o Coefficients listed on 252 o Canivez claims that the test does not measure five factors Median Validity o There is construct, content, empirical item analysis and considerable criterion-related evidence of validity
Alfred Binet defines intelligence as: the tendency to take and maintain definite direction the capacity to make adaptations for the purpose of attaining a desired end and the power of auto criticism Spearmen: the ability to enduce wither relations or correlates There are many different definitions of intelligence given all by different researchers T.R Taylor identified three independent research traditions that have been employed to study the nature of human intelligence: o The psychometric: examines the element of structure of a test , and then examine the properties of a test through an evaluation of its correlates and underlying dimensions o The information processing approach: we examine the process that underlie how we learn and solve problems o Cognitive tradition approach: focuses on how humans adapt to real world demands People have said that intelligence tests are biased because they do not consider socioeconomic backgrounds There are many critiques of intelligence test The Binet scale has been revised at least 5 times Formal intelligence testing began with a French minister of public instruction around the turn of the 20th century created a test to remove the lower scoring students from the classroom to receive special education experiences o 1904- he gave appointed a commission to which he gave definite assignment: to recommend a procedure for identifying so called subnormal (intellectually limited children) o Binet was a part of this commission Experiment: o Wissler indicated that simple functions such as reaction time and sensory acuity failed to discriminate well among individuals of high and low scholastic ability
Binets principals of Test Construction
Defined intelligence as the tendency to take and maintain definite direction the capacity to make adaptations for the purpose of attaining a desired end and the power of auto criticism Binet wanted to measure judgment, attention and reasoning He used trial and error, and was guided my two major principals Principal one: Age differentiation
o Refers to the simple fact the one can differentiate older children from younger by the formers greater capabilities o Binet searcher for tasks that could be completed by 66.67-75% of the children of a particular age group and by a smaller proportion of younger children o He did come up with these tasks, and could therefor estimate the mental ability of the child o The mental age is independent of the chronological age Principal two: General Mental Ability o Guided by his decision to also measure only the total product of the various spate and distinct elements of intelligence, that it general mental ability o This meant that he freed himself from having to find the total or final product of intelligence o He would judge the value of a score and correlate with the total
Spearmen’s Model of General Mental Ability
Advanced the notion of a general mental ability factor underlying all intelligent behaviour- before Binet He said intelligence consists of one general factor (g) plus a large number of specific factors Referred to as psychometric g (general mental ability) ^ Based on the phenomenon that when a set of diverse ability test are administered to large unbiased samples of the population, almost all of the correlations are positive (called positive manifold) Refer to the hydro example (238) Factor analysis: is a method for reducing a set of variables or scores to a smaller number of hypothetical variables called factors. One can determine how much variance a set a set of tests or scores has in common This common variance represents the g factor Half of the variance in a set of diverse mental ability tests is represented in the g Implications of general Mental intelligence o Concept of general intelligence implies that a persons intelligence can best be represented by a single score g, that presumably reflects the shares variance underlying performance on a diverse set of tests o True performance of any given task can be attributed to g as well as a unique variance The gf-gc theory of intelligence o There are two basic types of intelligence; fluid and crystallized o Fluid: abilities that allow us to reason, think and acquire new knowledge o Crystallized: represents the knowledge and understanding that we have acquired
The Early Binet Scales
Collaboration by Binet and T. Simon Eventually called Stanford-Binet intelligence Test The 1905 Binet-Simon Scale o 30 items shown in an increasing order of difficulty o Individual test o Idiot- described the most ever from of intellectual impairment o Imbecile- moderate levels of impairment o Moron- mildest level o Item 6- upper limit of adult idiots o Item 8- would rule out sever intellectual impairment o Item 16-upper limit for adult imbeciles o First measure of human intelligence o Lacked adequate measuring unit, normative data and validity documentation o Based on only 50 children The 1908 Scale o Was an age scale- the items were grouped according to age level rather that simply increasing the difficulty o Used in only hybrid fashion in modern times o It makes comparing a child’s performance on different kinds of tasks near impossible o Produced only one score exclusively related to verbal, language and reading ability o Was better than the 1905, but improvements were still needed o Introduction to the mental age
Terman’s Stanford-Binet Intelligence Scale
They revised it again in 1911 with small improvements Terman’s 1916 is the one that really flourished The 1916 Stanford-Binet Intelligence scale o Increased the size of the standardized sample o Was exclusively white native-Californian people o ^ Was not representative o Went from 50 to 203 individuals o The intelligence Quotient IQ: a unit for expressing the results of intelligence tests, based on the ratio of the individuals mental age (MA) as determined by the test and to actual chronological age CA IQ= MA/CA x 100 o There was a maximum mental age of 16 which was limiting The 1937 Scale o Extended the age range down to 2
Increased max mental age to 22 years, 10 months Scoring standards and instructions were improved Added items decreased focus on verbal skills The test was still not balanced as only 25% of the test was non verbal The sample size went up to 3184 from 11 different us states (still focused on whit urban people) o There was an inclusion of an alternate equivalent form Problems with the 1937 Scale o Its reliability coefficients were higher for older subjects than for younger ones o Reliability figures also varied as a function of IQ (lower the IQ the higher the reliability) o Unique standard deviations were also a problem, this meant that an IQ score at one age did not reflect the same IQ score at a different age 1960 Stanford-Binet revision and Deviation IQ (SB-LM) o Tried to create a single instrument from the best of the old tests o Task that showed an increase in the percentage passing with an increase in age received the highest priority as did tasks that correlated highly with scores as a whole o Instructions for scoring were improved and IQ tables were extended from age 16-18 o Deviation IQ- solved the differential variation in IQ’s, it was simply a standard score of 100 with a standard deviation of 16 o Mean set at 100 o It was ascertained by evaluating the standard deviation of metal age for a representative sample at each level o IQ tables were corrected so that the IQ sores were Comparable o In 1972 there were more people in the standardized sample (2100) and it included non whites o o o o o
The modern Binet scale
Model for the fourth and fifth editions of the Binet Scale o These version incorporated the Gf-Gc theory o They are based off a hierarchal model o At the top is G which reflects common variability of all tasks o At the next level there are three group factors o Crystallized abilities reflect learning-the realization of original potential through experience Has two subgroups: verbal reasoning and nonverbal reasoning o Fluid-analytic abilities- represent original potential or the basic capabilities that a person uses to acquire crystallized abilities o And the last group is short term memory o The role of Thurstones multidimensional model
He argues that intelligence could be best conceptualized as a comprising of independent factors or primary mental abilitiesthese group factors were correlated and from them a g factor could be extracted Characteristics of the 1986 revision o Age scale was entirely eliminated o Items with the same content were now placed together o In the 2003 fifth edition here are now five main factors rather than four o This means that you can now get smaller scores for the individual groups (because grouped by content now) o The drawback is less variety in the items Characteristics of the 2003 fifth edition o Non-verbal and verbal scales are equally weighted o The exam begins with two routing measures- one verbal and one non verbal, they are organized in a point scale (similar content of increasing difficulty) o The point of the routing test is to examine there ability level o The remaining 8 subtests are arranged in an age scale o Equal verbal and non verbal which allows an individual content score o There is mixing within the groups o Items that are too easy are skipped, so you start at the estimated ability from the routing test o Estimated level of ability is called the start point o Basal- level at which a minimum criterion number of correct responses is obtained o They continue until they hit the ceiling in which a certain number of incorrect responses have indicated the items are too difficult o The scale for the subgroups have a mean of ten and an standard deviation of three o In addition there is a standard score with a mean of 100 and a SD of 15 for nonverbal and verbal IQ, full scale IQ and each of the five factors Psychometric properties of the 2003 fifth edition o Age range goes from 2- 85+ o Sample increased to 4800 of all ethnicity’s and gender o 300 addition people were included that were gifted, mental retarded, had ADHD o Possible scores range from 40-160 o Coefficients listed on 252 o Canivez claims that the test does not measure five factors Median Validity o There is construct, content, empirical item analysis and considerable criterion-related evidence of validity
