‘Intelligence’ is most commonly though of as quickness of understanding or mental power and ability. The testing of these capabilities can be limited according to the race and culture of the individual under assessment, consequently controversy regarding the viability of IQ tests has raged incessantly. Do culture fair tests exist? Or are they merely speculative? Culture fair tests were first developed prior to the First World War in order to assess ability levels of immigrants and other individuals who did not speak English, over the last two decades “culture-fair” tests of mental ability have gained in visibility and also popularity. In 1968 Taylor argued that “there are ‘culture free’ tests which measure intelligence without putting a premium on education or other cultural factors”. Is this statement true? Aspects that should be analysed carefully In consideration of a thorough answer to the question posed, include; the workings of current IQ tests and the regional as well as cultural problems which arise, past research looking at cultural differences, methods posed to overcome these cultural biases and the validity of various current culture fair tests. Whilst deciding whether an intelligence test can be described as culture fair, we must also decide whether we agree or disagree with Taylor.
The aforementioned problems with intelligence tests relate to regional as well as cultural (race differences) endeavours and both these factors should be scrutinised carefully to come to a conclusion regarding the ‘culture fair’ validity of any intelligence test. Initially however, in order to distinguish whether any current commercially available IQ tests are culture fair, it is imperative to understand how they work. IQ tests were first derived by Galton (1869), who used an existing theory of intelligence to develop a test that was deigned to measure “intelligent behaviour” Galton’s tests were found erroneous though as they failed to agree with independent speculations that we might rightly make about intelligence. Consequently a different approach was developed by Binet (1905-1911), an American psychologist. A relevant milestone during this time was the United States Army’s intelligence testing team exposing a very large number of American males to standardized testing for the first time. This served as a training ground for many psychologists, including Binet. His approach assesses relative intelligence by testing the subject using a set of questions, of differing difficulty. Binet argued that the content of the questions was irrelevant as long as their results correlated with another measure of cognitive capacity, i.e. one that increases with age. Hence, IQ is the ratio of mental age to the child’s physical age. The problem apparent from this approach is that these tests should not be regarded as direct measures of cognitive capacity, as the variables under study can all be affected independently by other factors, such as early education and physical environment. Thus arises the problem under consideration.
This problem was realized when researchers attempted to interpret behavioural differences between racial groups. Complication occurred as such groups also differed in culture. The recent interest in culture fair tests stems from various sources, including: research indicating racial differences in test performance on cognitive tests and research showing test differences as a function of socio-economic levels. Therefore before looking in deeper detail at ‘culture fair’ tests, firstly some of the principal ways in which cultural differences between racial groups may be reflected in psychological test results should be examined. These causes include: general cultural milieu, culture conflicts, socio-economic level, and education.
It is clearly palpable that the particular culture in which the individual is reared may influence his/her behavioural development. The experiences of people living in different cultures may lead to different perceptual responses, lead to a different meaning of their actions, and stimulate different standards of behaviour. Porteus recorded a particularly interesting illustration of this in 1979. He was studying Australian aborigines and had great trouble attempting to convince his participants that they were supposed to solve the problems individually and without assistance. He stated that “the subject of a test was evidently extremely puzzled by the fact that I would render him no assistance . . .”
Culture convicts have been analysed through studies of immigrant groups and the problems that they experience adjusting to a new culture. This maladjustment is greatest in the case of second-generation immigrants, as the two frames of referance offered them can serve to be extremely confusing Socio-economic level is also an important cultural factor. Generally, minority groups are associated with low socio-economic status, because the latter is usually related to intellectual development this factor must be taken into account when testing their IQ. Lastly, it is rather self-explanatory that differences in the amount and nature of education are reflected in intelligence test performance.
A lesser side of this problem is apparent though regional factors. For instance, it is assumed that an intelligent person is more likely to recall factual knowledge about their own region, and their own language. An illustration of this is exemplified through a question from WAIS, the Wechsler Adult Intelligence Scale. The question, “What is the distance between New York and London?” was reasoned to be easier for the average American population than for the British, because this information is included in the American equivalent of the National Curriculum. Whilst, “what is the distance between Edinburgh and London?” i.e. the regionalised version of this question is unlikely to be known unless you have actually made that journey. A possible solution to this problem could be to internalise the question in a multiple-choice form. However this modification does not test what it originally set out to, i.e. the ability to recall random pieces of information. Another way of avoiding this regional bias (which isn’t possible in cultural bias) is to put in a mixture of questions, however the assessment of difficulty problem is then still unsolved. Ideally, a test should be used which isn’t regionally biased, i.e. involves abstract shapes unseen by the participants and bases all its information on that given on the question paper.
The Davis Ells Games was an early attempt to construct a test, which would eliminate the differences between social classes. This was composed of a number of items, which did not require an reading or attention to instructions on the test. Several other tests have also been developed which are ‘culture fair’. Raven’s Matrices is another example of a culture fair test, although experiments demonstrate that there is a correlation between amount of time resident in the UK and the test scores. Japanese children do much better on this test but the reason appears to be due to a significant proportion of schooling time been devoted to the abstract problem-solving tasks that it tests. This was missing from the UK Curriculum, which explains the lower scores. Interestingly, WAIS results show that American blacks, which attend the same school and come from similar backgrounds to the whites from the same area, appear to have significantly lower WAIS scores in some studies. These findings relate the regional to the cultural and make the development of culture fair tests much more imperative. As the regional differences had been standardised, these results must relate to an innate difference, which may be generalised across the entire racial group. Dreger and Miller (1960) concluded similarly that the average IQ of Negroes is approximately 15 points lower than that of whites; Negroes consequently generally do less well on selection tests for employment. Perhaps these results are due to the nature of the test and other tests may be less biased. Raven Progressive Matrices and the Cattell Culture Free Test of Intelligence are examples of “culture-fair Tests” which may prevent bias according to culture. Higgins and Sivers (1958) reported a study using both the Stanford-Binet and the Raven Progressive Matrices. Negro and white children were used in this study and they achieved similar scores on the Stanford-Binet. However, the Negro group scored more than ten IQ points lower on the Progressive Matrices clearly the latter test would have penalised the Negro group had it been a screening device in education or for a job interview.
Conversely, Garrett (1947) proposes that the higher test performance by the white race is an indication of a superior genetic intelligence. Additionally, Jensen (1969) suggested that the test differences were a result of genetic differences between Negroes and whites. Obviously this argument caused great debate. An alternative viewpoint exists which believes that the intellectual ‘superiority’ of the white population stems from environmental factors, rather than hereditary ones. These include; deficient diets, lack of language skills which leads to lower ability. Yet this standpoint denies any real genetic difference between racial groups. Another belief is that the tests themselves are culturally “loaded” and therefore differentially familiar to individuals from different cultural backgrounds. A strong argument is that much of the item content in current intelligence tests is not familiar to the culturally deprived. Jensen (1966) puts forward he proposition that the verbal nature of many IQ tests could account for test differences. He agues that since culturally disadvantaged have had relatively less experience with verbal material this could account for their lower scores.
Anita, Smith, Ray Hays, and Solway (1977) conduced an interesting study in the hope of determining whether the Culture Fair Intelligence Test was more efficient in screening a population than the WISC-R. The format of the Culture Fair test used was evolved in 1949 through the work of Cattell in attempting to measure fluid and crystallised general ability. The findings were that the Culture Fair is a better measure of intelligence than the WISC-R. The reasons they put forward for this are that the effect of cultural bias is reduced and a more accurate picture of intellectual capacity is therefore produced. From this study it is reasonable to say that an intelligence test can be described as culture fair. Yet, problems are not completely dissipated and “the Culture Fair is not et a complete solution for problems of cultural influences in the measurement of intelligence”. Anita e al (1977).
Anastasi (1968) defines a culture fair test as a test that controls the relevant variables that influence test scores. It is possible to construct a test, which presumes only experiences that are common to different cultures; these types of test utilize elements that are common to all cultures and no others. Conversely, no existing test is entirely unrestricted in its cultural reference. Relevant factors which differ from culture to culture must be take into consideration when looking a ‘culture fair’ tests. These include: the use of paper and pencil, the presentation of abstract tasks which have no practical significance, pictures in cultures unaccustomed to representative drawing, intrinsic interest of test content, rapport with the examiner, desire to excel others, and past habits of solving problems in a group or individually (see Figure 1). Evidence even exists that the presence of an examiner of a different racial or cultural group may interfere with rapport during administration and have consequences on test performance. Canady (1936) studied Negro and white school children that were given the Stanford-Binet by both Negro and white examiners. The findings were that in both white and Negro groups, the mean IQ was about six points higher when an examiner of their own race tested the participants.
Performance testsPencil and paper tests
Oral instructionsWritten instructions
Familiar item contentUnfamiliar to all item content
Figure 1. Some dimensions along which culture fair tests might vary. Taken from ‘Some comments on culture fair tests’ Richard Arvey. The left side represents test variables which are less culturally biased.
A startlingly interesting idea is that the concept of intelligence may be culturally conditioned. This thought comes into being as the criterion employed in validating intelligence tests is usually success in our society. Scores on the test are correlated with school achievement and if the correlations are high, it is decided that the test is a good measure of “intelligence”. Therefore intelligence tests measure the ability to succeed in our particular culture and nothing else.
The cultural factors that may affect performance on psychological tests include: general traditions and customs, motivation to excel on the sort of tasks sampled by intelligence tests, and social expectancy. As we have seen, culture-fair tests attempt to utilize content common to all cultures. Yet, such tests still tend to favour certain cultures in various ways.
Within a certain culture, particular abilities and certain ways of behaving are selected and therefore any test developed within a particular culture reflects such a selection and consequently favours individuals reared in the culture. It is inherently clear that “tests cannot be developed in a vacuum, but instead are dependant on a subject pool that has been raised in some environment.” Richard Arvey (1972). Dreger and Miller (1960) believe that an intelligence test can never be culture fair. They came to this conclusion through analysing a series of studies, which make comparisons between backs and white on “culture-fair” tests. All the studies they investigated showed whites still performing better on the tests than the black participants. Hence, their viewpoint that “the search for a culture fair test is illusory”. On the other hand, perhaps there is no need to develop a more ‘culture-fair’ test. The reasons why this might be true revolve around the idea that the cultural actors that cause poor performance on an IQ test, may affect the participants capability of actually doing the job. A culturally disadvantaged individual who has a deficit of verbal skills may not do well in a particular job because of the same deficit. Anastasi (1968) argues, “A test constructed entirely from elements that are equally familiar in many cultures might measure trivial functions and possess little theoretical or practical validity in any culture”.
Due to the lack of theory and the subjective basis of the questions used in current IQ tests, it is near impossible to make the current IQ tests culture fair. At present, we have no way of reasonably comparing groups from different regions and/or cultures. No test can be truly culture-fair, as tests cannot be designed in a cultural vacuum.
Matarazzo (1992) argued that biological tests of intelligence might one day become normative. He suggests this as evidence has suggested that members of minority groups score even lower on culture fair tests, thus speculating that biological measures may in fact be culture fair. Richard Davis (1993) stated that biological tests “represent a potential for objectivity typically not achieved in the fields of intelligence and cognitive testing”.
Anastasi, A. (1964) Culture-fair testing. Educational Horizons, Fall.
Anastasi, (1968) A. Psychological testing. New York: MacMillan Co.
Anita, L. Smith, J., Ray Hays, and Solway, K.S. (1977). Comparison of the WISC-R and Culture Fair Intelligence Test in a juvenile delinquent population. The journal of Psychology, 97, 179-182.
Arvey, R.D. (1972). Some comments on culture fair tests, Personnel Psychology, 433-446.
Canady, H.G. (1936) The intelligence of Negro college students and parental occupation, Amer. J. Social, 42, 388-389.
Darlington, R.B. (1971) Another look at “cultural fairness”. Journal of Educational Measurement, 8, 71-82.
Davis, R. (1993). Biological tests of intelligence as culture fair. American Psychologist, pp. 695-696.
Dreger, R.M. and Miller, K.S. (1960) Comparative psychological studies of Negroes and whites in the United States. Psychological Bulletin, 57, 361-402.
Eells, K., Davis, A., Havinghurst, R.J., Herrick, V.E., and Tyler, R.W. (1951). Intelligence and cultural differences. Chicago: University of Chicago Press.
Higgins, C. and Sivers, C.H. (1958) A comparison of Stanford-Binet and Colored Raven Progressive Matrices IQ’s for children with low socio-economic status. Journal of Consulting Psychology, 22, 565-568.
Jensen, O.A. (1966). Cultural bias in selection. Public Personnel Review, April, 125-130.
Matarazzo, J.D. (1992). Psychological testing and assessment in the 21st century. American Psychologist, 47, 1007-1018.
Shuey, A.M. (1966). The testing of Negro intelligence, (2nd Ed.) New York: Social Science Press.
Taylor, V.R. (1968). Control of cultural bias in testing: An action program. Public Personnel Review, July 3-14.