4. What is the psychometric approach to intelligence PDF

Preview

Summary

Essay on psychometric approaches to intelligence...

Description

What is the psychometric approach to intelligence? Intro In defining intelligence, there is considerable overlap between lay people and experts definitions, both believing that individual differences in intelligence can come about due to differences in biology, environments and mental processes. However, defining intelligence is much harder than this as emphasised by researchers, this is because intelligence is a latent concept, it is not directly observable and can only be inferred by the use of mathematical models based on variables that are measurable, such as differences in biology, environment and mental processes. One particularly supported notion on the definition of intelligence was proposed by Gottfredson, who put intelligence as down to the ability to reason/problem solve/plan/think abstractly etc. However, there seems to be less agreement in the ways that intelligence should be measured and conceptualised, this unresolved issue is one that shall be discussed in the contents of this essay. Parah 1 Intelligence tests are first of all tested to ensure there is a positive correlation between different tests, if there is, then we would assume that those who do well on one test would also do well on other tests. Correlations are often used to assess whether there is a relationship between test scores/academics/occupation and IQ scores if there is a positive correlation, then we would assume that intelligence is a predictor of such measure. As intelligence tests were initially created as a way to test academic performance by Binet, who was commissioned by the French ministry to create a way to identify those with learning/educational difficulties, we would assume that the two would share a positive correlation. The test initially derived from Binet was to get children

to respond to 30 items of increasing difficulty. This test also had a different number of levels, each of which corresponded to a certain year group, for instance, reaching level three would be average for a three year old. At this stage, children may be tested in things such as their ability to follow a match or point to their own nose etc. at level 7, this may begin to become mre complex, such as having to describe a picture or come up with a complex sentence. The number of correct responses a child gave would signify a person’s mental age, for instace, if they got all questions in level 7 correct but only 3 in level 8, they would be classified as a mental age of 7.5. This was groundbreaking because it meant a person’s academic success and learning potential could be assessed using meaningful stimuli. We could argue that this just measures a childs verbal fluency and seems to ignore practical intelligence, which Sternberg highlighted as important to intelligent. It may even also need more language fluency than they have in order to understand the quetions, leading to underestimations of a childs intelligence/abilities due to their lack of comprehension, hence, basing this more on visual/practical intelligence may have been a better approach. Nonetheless, it was this approach that led to the creation of correlational studies such as Deary, who assessed childrens cognitive abilities and general intelligence at age 11 and once agin during their GCSE phase. It was found that there were moderate-strong correlations between general intelligence and their 25 GCSE subjects (.69) and a strong correlation between general intelligence and their 7 best subjects (.72). Therefore, making it clear that academic success can be predicted via correlations with intelligence. However, we must note that in this study, some correlations were stronger than others, for instance, strong correlations were found between intelligence and math scores (.77) but moderate correlations between intelligence and chemistry (.46), therefore it may be the case that there is an underlying factor that predicts success in these subjects more and we may put this down to mediating variables such as parental success or socioeconomic status.

Research has also found moderate correlations between intelligence and occupational success (.43), which is higher than correlations with potential mediating variables such as parental occupation (.28) or socioeconomic status (.33). However, what isn’t considered in the use of correlational studies is that intelligence can be influenced by culture, for instance, evidence suggests that Asian cultures tend to teach that academics are the only way to success, but, this is not the case in western cultures. To account for culture, it may be best to use a test of fluid intelligence, such as raven’s matrices, which is a culture fair test. However, another factor such correlation studies do not account for are differences in personality/motivation, reducing the impact of culture and mediating the relationship between academic success and intelligence. For instance, if we are highly motivated to pursue a reward, then we are more likely to work harder for it. This may explain why the correlation between academic success and intelligence reduces with age, presumably due to the increasing influence of personality. Therefore, although correlations can be positive, it is not a causal link and it is possible that there are mediating factors, hence, this measure may not be as useful, but surprisingly correlations are still widely used even in today’s research. Parah 2 Spearman took a slightly different approach to this and instead used a concept known as factor analysis and data-reduction processes to show that different intelligence tests were intercorrelated and all predicted one innate factor, which he named as the ‘g factor’ or general intelligence. There is evidence to suggest this in the fact that some researchers have found that those who score highly on tests of verbal reasoning also seemed to score highly in other areas such as spatial reasoning. It also has support from academics in the fact that those who score highly in one subject, such as English, also tend to score highly in other written subjects.

However, there are still controversies about this claim, for instance, human abilities are all uniquely diverse, hence the term individual differences, so could we explain all intelligence just due to a single innate g factor? This is not what Cattle and Horn believed, instead, taking the view that there are two major divisions of intelligence, known as fluid and crystallised intelligence. Fluid intelligence is the ability to perform well in non-verbal tests, which do not rely on previous knowledge, experiences or culture. This may include measures such as reaction time or raven’s progressive matricies, which a participant must fill in the missing element of a pattern. On the other hand, crystallised intelligence assesses verbal intelligence and is based on our knowledge, education and past experiences. Evidence of such as distinction comes from the fact that crystallised intelligence tends to increase with age, presumably due to having more experience, meanwhile, fluid intelligence tends to decrease with age due to the effects of ageing. From this, we can also see that intelligent individuals tend to do better in non-verbal tests, as seen in fluid intelligence, over verbal tests. Thurstone, did not believe Cattle and Horn’s theory was comprehensive enough and instead, put forward the theory of primary mental abilities. This put intelligence as down 7 primary abilities including verbal comprehension, word fluency, spatial visualisation, reasoning, perceptual speeds associative memory and numerical abilities. This seems to be a more comprehensive theory as it breakdown intelligence, beneficial as we may see some people with good verbal comprehension but poor perceptual speed. Therefore, this gives us a better view of a person’s abilities individually, seemingly more useful than a single g factor and only two divisions of intelligence. However, there is a weakness to this study, which was a strength to Spearman, which is the fact that those who do well on one factor, tend to also do well on other related factors. This may not be the case for everyone, but it is possible that there is some kind of underlying factor.

Parah 4 Carroll was the one research who has seemed to take the most extensive approach to intelligence to date. To do this, he incorperated theories from Sternberg as well as Cattle and Horn into his own theory of the three striatum model of intelligence. To create this model he re-analysed more than 400 sets of data from researchers across the world to form a hierachical model of intelligence. At the top of this hierachy was the general factor of intelligence proposed by Sternberg. This was followed by broader factors, similar to Thurstones primary mental abilities model, such as processing speed or memory, as well as Cattle and Horn’s fluid and crystallised intelligence. The lower level were the narrow levels of intelligence, including more specific factors of the broader factors. For instance, the narrow factors of crystallised intelligence may be comprension/reading/writing abilities. Therefore, in this study, it is clear that he agrees with Sternberg in that there is a general innate factor to intelligence, but that this should also incorperate a more extensive range of intellectual abilities as proposed by Thurstone. This study includes virtually all known cognitive abilities and provides the most comprehsnsive analysis of human abilities in this combination of studies, also making it the most reliable of research. By using a model such as this, it allows us to create a curve of normal distribution, in which we can categorise those with learning difficulties (below 70 IQ), exceptional (115-130) or average abilities (85-115). These studies allow for the creation of self-report measures to assess the general populations intellience, in such tests, participants are given measurements of their verbal, spatial, mathematical, interpersonal and intrapersonal skills as well as just a measure of overall intelligence to assess from 50-150. However, the problem with this is that people can often succumb to biases such as social desirability bias and answer in a way that makes them look better and smarter than they actually are. Furthermore, people can even underestimate their abilties, a common example being underestimations of math abilities due to its complexity, claiming

they hate maths and therefore, must be bad at it. Therefore, it is possible that people do not know themselves as well as we think we do, as shown by the fact that correlations between these self reports and actual intelligence is weak (.3). Therefore, a better approach may be to use self-report measures alongside experimental measures to get the most accurate measure of the intelligence of the general population. Nonetheless, what all of these theories succumb to is the idea of book smarts, which is not what achievements are all about. Individals may have high emotional intelligence or creativity, which can allow them to achieve just as much as someone who is traditionally intelligent. Judging an individual based on their intellectual abilities also gives rise to discrimination against those who are economically disavantages or of a minority race. Therefore, this may be a cost to intelligence tests as it can lead people to believe they are not good enough or smart enough to achieve in life and can also lead to prejudicial beliefs.

To conclude, there are a number of ways that intelligence has been conceptualised and psychometrically measured that can be used in combination with self-report measures. The most comprehensive theories coming from Cattle and Thurstone, but, arguably, the use of factor analysis to identify the g factor, was a good approach as it may underlie these theories....