Stanford-Binet intelligence scale was designed to ensure that all items of the test contributed equally to the determination of total IQ score. This implies that an individual might perform well on a test involving the production of geometric forms, but badly on a test of vocabulary (McKenna, 2000). Although the tester might note strengths and weaknesses, they would cancel each other out in arriving at the overall IQ score. McKenna (2000) says that “under the 1986 revised scale, standard age scores are substituted for IQ scores, and it is now possible to obtain scores for different areas of the test” (p. 82). In accordance with the current view that intelligence is an accumulation of different abilities, it is possible to obtain a separate score for each of the broad areas of intellectual ability (McKenna, 2000).
Wechsler adult intelligence scale was developed by psychologist David Wechsler in 1939 as one of the first intelligence tests aimed to determine separate abilities. The test has been widely used in various situations. Wechsler intelligence scale is an extremely reliable and well-standardized instrument. Goldman, Stein & Guerry (1993) say that unlike Stanford-Binet intelligence scale, Wechsler adult intelligence scale has the advantage of providing both verbal and performance measures to the examinee.
Stanford-Binet intelligence scale was developed in France. It is based of the earlier work of Alfred Binet, who was a French psychologist. This test is well known and is widely used. It has been revised on a number of occasions. McKenna (2000) indicates that the index of intelligence used is the intelligence quotient (IQ), which is expressed as a ratio of mental age (MA) to chronological age (CA) as follows: IQ=MA/CA*100. In this context, using 100 as a multiplier means that when MA=CA, the IQ will have a value of 100. If MA is less than CA, the IQ will be less than 100. If MA is greater than CA, the IQ will be higher than 100. McKenna (2000) claims that in Stanford-Binet intelligence scale “an IQ between 90 and 110 is considered to be normal, but above 130 it is considered to be very superior” (p. 82). A person with an IQ below 70 is perceived to be retarded. As it is the case with different features of individuals, distribution of IQ scores in the population approximates the bell-shaped normal distribution curve (McKenna, 2000). In both Stanford-Binet scale and Wechsler adult intelligence scale most cases fall into the mid-value of the curve, with just a few cases at the left and right extreme positions on the curve.
Wechsler adult intelligence scale was developed because Stanford-Binet test was perceived to rely heavily on individual’s language ability and did not cater for the needs of adults. Wechsler adult intelligence scale has two sections, which include a verbal scale and a performance scale. McKenna (2000) claims that each section generates separate scores as well as an overall IQ score. Verbal score is based on tests such as vocabulary and verbal similarities, in which the test taker has to say how two things are similar (McKenna, 2000). For example, Wechsler intelligence scale was compared with Stanford-Binet intelligence scale using a sample of 50 children aged six years. Stanford-Binet intelligence scale was about 2 points lower than Wechsler intelligence scale. Unlike Wechsler intelligence scale, IQ scores in Stanford-Binet intelligence scale cannot be computed below 45 points for verbal or performance scales and below 40 points for the full scale IQ. This applies even when minimal performances are obtained. Goldman, Stein & Guerry (1993) note that since some programs for retarded people look at the lower ranges of IQ for placement, Stanford-Binet intelligence scale may not be the instrument of choice in these cases.
Performance score in Wechsler scale is based on tests such as picture completion, which requires identification of a missing part of an object in a picture. It also includes tasks of picture arrangement, which requires rearrangement of a scrambled set of cartoonlike pictures in an order that tells a coherent story. Overall score in Wechsler intelligence scale, unlike in Stanford-Binet scale, combines verbal and performance scores (McKenna, 2000).
In the performance scale of Wechsler test adult intelligence scale items require manipulation or arrangement of blocks, pictures, or other materials. McKenna (2000) indicates that Wechsler adult scales allow for scores to be recorded for each of the sub-tests. Therefore, the tester gets a better understanding of intellectual strengths and weaknesses of the individual. McKenna (2000) says that both Stanford-Binet and Wechsler scales satisfy the requirements of an accurate test. They have high reliability and validity. Both tests are also fairly valid predictors of achievement, particularly at school (McKenna, 2000).
On the other hand, the purpose of Stanford-Binet intelligence scale is to provide a measure of person’s general intelligence. Goldman, Stein & Guerry (1993) say that this measure of general intelligence is obtained by presenting an individual with a variety to tasks of known increasing difficulty. It is important to note that Stanford-Binet intelligence scale presents test items by age level rather than by subtests based on separate aspects of intelligence or on specific abilities as it is the case with Wechsler intelligence scale.
Besides creating intelligence test for adults, David Wechsler constructed a set of different intelligence tests for various age groups. These tests are collectively referred to as the Wechsler Scales (Sigelman & Rider, 2011). Wechsler Preschool and Primary Scale of Intelligence (WPPSI) was constructed for children between ages 3 and 8. Wechsler Intelligence Scale for Children (WISC-IV) is appropriate for schoolchildren aged 6 to 16, while Wechsler Adult Intelligence Scale is used for adults (Sigelman & Rider, 2011).
Unlike Stanford-Binet scale, Wechsler adult intelligence test yields a verbal IQ score based on items measuring vocabulary, general knowledge, arithmetic reasoning and the like. Performance IQ measures are based on such determinants of nonverbal skills as the ability to assemble puzzles, solve mazes, reproduce geometric designs with colored blocks, and rearrange pictures to tell a meaningful story (Sigelman & Rider, 2011). As with Stanford-Binet scale, a score of 100 is defined as average performance for the person’s age. Sigelman & Rider (2011) claim that “a person’s full scale IQ is a combination of the verbal and performance scores” (p. 278).
Another major difference is that compared to Wechsler adult intelligence scale, Stanford-Binet scale is basically an age scale, in which content changes with age. Criterion validity of items and their distribution is determined by evaluation of percentage passing items and tests at each age level with regular increase from younger to older ages (Goldman, Stein & Guerry, 1993). It should also be noted that Stanford-Binet performance is explainable by a single common factor.
Studies indicate that for older children whose functioning level is considerably below their chronological age, Stanford-Binet provides substantial information about things that children can do rather than about things they cannot do. Goldman, Stein & Guerry (1993) say that clinically such information is helpful in giving recommendations for the child. For younger children, the game format is an advantage. Furthermore, testing time can be shorter than with Wechsler adult intelligence scale. Unlike Wechsler intelligence scale, Stanford-Binet scale norms are fairly recent providing a wide range of IQ scores, which are useful when special placement of retarded children may be a part of the referral question.
The main drawback of Stanford-Binet scale is that younger children are tested by items that require the use of many small toys and objects, making it difficult for an inexperienced examiner to administer the test easily (Goldman, Stein & Guerry, 1993). The IQ score which is obtained does not permit its easy breakdown into verbal and performance areas or more specific areas of intellectual functioning. Mahmud (2010) says that Wechsler scales differ from intelligence scale like Stanford-Binet scale in that it is a point scale rather than an age scale. The author further says that in Wechsler scale all items of a given type are grouped into subsets and arranged in an increasing order of difficulty within each subset. Mahmud (2010) also says that another major difference is that unlike Stanford-Binet scale Wechsler intelligence scale includes verbal as well as performance items from which separate verbal and performance IQs can be computed (Mahmud, 2010).
Apart from being used to determine intelligence, Mahmud (2010) says that Wechsler intelligence scale is also useful for psychiatric diagnosis in cases of brain damage, psychosis, emotional disorders, and other clinical cases. It is fundamental to note that both Stanford-Binet and Wechsler intelligence scales are administered individually and involve an experimenter, who gives a variety of tasks to the test taker.
Wechsler intelligence scales became known for at least two major innovations when they were first developed. Firstly, they were less dependent on verbal ability than Stanford-Binet scale. They also included many items that required nonverbal reasoning (Goldman, Stein & Guerry, 1993). Wechsler intelligence scale differs from Stanford-Binet scale in that it allows to compute three scores including verbal IQ score, performance IQ score, and overall full scale IQ score. For example, verbal subsets of Wechsler scale include information, comprehension, arithmetic, similarities, digital span, and vocabulary. Performance subsets which can be used in Wechsler intelligence scale include digit symbol, picture completion, block design, picture arrangement, and object assembly (Goldman, Stein & Guerry, 1993). Secondly, with the development of Wechsler intelligence scale, deviation of IQ score based on the normal distribution of intelligence came up. This has caused scientists to abandon the notion of intelligence quotient widely used by Stanford-Binet intelligence scale.
Trull & Prinstein (2003) say that the fifth edition of Stanford-Binet test is based on a hierarchical model of intelligence. Specifically, Stanford-Binet fifth edition assesses five general cognitive factors. Each factor is tapped by both verbal and nonverbal subset activities. The first subset is known as fluid reasoning. It tests the ability to solve new problems. Trull & Prinstein (2003) say that the second subset assessed by Stanford-Binet scale is quantitative reasoning. This subset determines the ability to solve numerical and word problems as well as understand fundamental number concepts. Stanford-Binet scale is also used to assess visual-spatial processing. This entails testing the ability to see relationships between objects, recognize spatial orientation, and carry out pattern analysis (Trull & Prinstein, 2003).
Another factor of difference is that while Wechsler adult intelligence scale assesses verbal and performance capabilities, Stanford-Binet intelligence scale assesses working memory. Stanford-Binet intelligence scale tests the ability to process and hold both verbal and non-verbal information and then interpret it. Trull & Prinstein (2003) say that this type of scale plays a fundamental role in the assessment of knowledge. Stanford-Binet intelligence scale assessment determines the ability of a person to absorb general information accumulated over time from experiences at home, school, work, or the environment in general. Each subset used in Stanford-Binet intelligence scale is composed of items of varying levels of difficulty that are designed for individuals from age 2 to adulthood.
Subsequently, earlier versions of Stanford-Binet intelligence scale had a number of disadvantages compared to Wechsler adult intelligence scale. Firstly, the test was designed for adults. This implies that it offered items, the content of which was more appropriate and more motivating for adults in comparison to the school-oriented Stanford-Binet scale. Trull & Prinstein (2003) claim that in contrast to Stanford-Binet intelligence scale, items of which were arranged according to age levels, Wechsler intelligence scale grouped its items into subsets. For example, Trull & Prinstein (2003) say that all arithmetic items were put into one subset and arranged in order of increasing difficulty. In addition, there was a performance scale and a verbal scale. A separate IQ score for each scale could be calculated along with a full scale IQ (Trull & Prinstein, 2003).
Systematic inclusion of performance items helped mitigate the overemphasis on verbal skills, which limited the effectiveness of earlier Stanford-Binet scale in special populations (Trull & Prinstein, 2003). Unlike Stanford-Binet scale, Wechsler used a deviation IQ concept. Wechsler approach assumes that intelligence is normally distributed. It also compares individuals with their age peers (Trull & Prinstein, 2003). Wechsler adult intelligence scale involves several subsets, one of which includes vocabulary. In this subset, the examinee must define words that increase in difficulty. Trull & Prinstein (2003) indicate that this subset also correlates highly with full scale IQ. Some scientists feel that it comes close to measuring what is usually termed as g.
Similarities subset used in Wechsler scale consists of a series of items. For each one the examinee must explain how two objects are alike. Trull & Prinstein (2003) say that “the subset requires the basic ability to form abstractions and develop concepts” (p. 208). The digit span subset is a measure of short-term memory and attention. At this stage, three sets of digits are read aloud by the examiner. Trull & Prinstein (2003) further claim that for the first time, the examiner must repeat digits in the order that they were read. In the information subset, short questions are asked to measure knowledge that one would be expected to have acquired as a result of everyday living and cultural interactions.
Comprehension in Wechsler adult intelligence scale requires examinee to explain why certain procedures are followed, interpret proverbs, and determine what should be done in a given situation (Trull & Prinstein, 2003). Items measure common sense and practical judgment in solving a problem. In Stanford-Binet intelligence scale psychometric analysis supports reliability and validity scores. Trull & Prinstein (2003) note that reliability og Stanford-Binet scale across all age groups are generally high: .80s for factor scores and .90s for IQ scores.
In terms of validity of Stanford-Binet intelligence scale composite scores, strong supportive evidence has been obtained. Trull & Prinstein (2003) say that the correlation between Stanford-Binet fifth edition and several Wechsler scales was .84 for children and .82 for adults. Correlation between SB-5 scores and scores on achievement tests was substantial as well. Ayers, Baum & McManus (2007) claim that like Binet, Wechsler recognized that intelligence goes beyond what his own test measures.
On the other hand, in Wechsler adult intelligence scale the average full scale IQ spilt half reliability coefficient across age groups was .98. The average split half reliability coefficients across age groups for verbal comprehension, perceptual reasoning, working memory, and processing speed index scores ranged from .90 to .96 (Trull & Prinstein, 2003). Test reset reliabilities over an average of three weeks ranged from .74 to .90 across age groups for various subsets.
In their studies Comer & Gould (2012) mention that Wechsler borrowed a lot from Stanford-Binet scale and other tests. However, his tests included much fewer tasks requiring verbal ability. Each individual who took the test received subset scores, which Wechsler then grouped into two main categories: verbal score and performance score. It is important to note that the individual also received an overall score that Wechsler associated with the g factor. Wechsler discarded the old formula for calculating an IQ score. Comer & Gould (2012) say that although he still referred to a person’s overall score as ‘intelligence quotient’ or IQ, he derived the score from a normal distribution.
Another fundamental difference according to Aiken (1998) is that unlike Wechsler adult intelligence scale Stanford-Binet intelligence scale was an age scale where child’s performance was converted to a mental age and then to intelligence quotient by dividing mental age by chronological age and multiplying by 100. Researchers established that some tests on the Stanford-Binet scale were appropriate for adults. However, it was still not considered to be an adequate measure of intelligence in adults. Aiken (1998) says that Stanford-Binet intelligence test was administered to one examinee at the time, while group tests could be administered simultaneously to a large group of individuals.
Aiken (1998) explains that unlike Stanford-Binet intelligence scale, Wechsler adult intelligence test would prove to be not only a valid measure of adult intellectual functioning, but would also contribute to making clinical diagnoses. Aiken (1998) says that “unlike the age scale format of Stanford-Binet, Wechsler adult intelligence scale was a point scale on which points was earned for passing subsets and was then converted to scaled scores and IQs” (p. 82). Using Wechsler intelligence scale, three types of IQs including verbal, performance, and full scale were determined on a standard score scale having a mean of 100 and a standard deviation of 15 (Aiken, 1998). It was noted that the difference between verbal and performance IQs as well as the pattern of scatter of standard scores in various subsets and qualitative nature of responses made by the examinee were all analyzed to arrive at a diagnosis.
Sue et al (2012) says that current fifth edition of Stanford-Binet intelligence scale is used for individuals aged 2 to 85. Sue et al (2012) further notes that Stanford-Binet requires considerable skill in its use since it is much more complicated to administer and score. He also says that test procedure is designed to determine a basal age and a ceiling age. Generally Sue et al (2012) says that Wechsler adult scale is preferred over Stanford-Binet for school age children. The author also noted that Wechsler test is easier to administer and yields scores on different cognitive skills. Consequently, he says that “Stanford-Binet intelligence scale is the standard to which other tests are compared because of its long history, careful revision and periodic updating” (p. 82). The latest Stanford-Binet intelligence test can now be used for all ages. Nevertheless, Wechsler adult intelligence scale was the first adult IQ test and like Stanford-Binet test, it yields a single overall IQ score.
Like Stanford-Binet test, Wechsler scale comprises of a variety of tasks measuring different kinds of verbal, perceptual, motor, and cognitive abilities. Apart from using performance on these subsets to generate a single overall IQ score, Wechsler claimed that person’s pattern of relative strengths and weaknesses across subsets, which measure different kinds of mental abilities, could be used to identify normal and abnormal variation in numerous cognitive characteristics and coping capacities. This meaningful interpretation of profiles of subset scores resulted in Wechsler scales being widely used in clinical health setting. They are used not only as a measure of intellectual ability, but also of neuropsychological impairment and disordered thinking.
In conclusion, Ollendick & Schroeder (2003) claim that both Wechsler adult intelligence scale and Stanford-Binet intelligence scale yield valuable information about individual’s mental abilities and produce accurate IQ scores. Two intelligence scales reflect basic cognitive ability. Scores also reflect individuals’ personality, motivation, and test taking skills. Wechsler adult intelligence scale presents an opportunity for the examinee to display his or her cognitive skills during a standardized interview with an experienced observer. As a result, despite their fundamental differences, the two intelligence scales play a very important role in the development of IQ testing.