This finding provided evidence of universal cognitive processes across these cultures and also indicated that the WISC-III can be val- idly administered with the same interpretations regarding its cognitive structure across these countries. The subtest scores also showed simi- lar reliability in both samples, although slightly higher latent variable means were found in the Canadian normative sample. These analyses were conducted with the raw scores from each country and not the scaled scores, nor the actual verbal, performance and full-scale IQs FSIQs , because comparisons of means of scaled scores across countries would result in zero differences.
Culture and Children's Intelligence : Cross-Cultural Analysis of the WISC-III
Although, as described above, these analyses were conducted with the raw scores, in order for the results to be meaningful to the reader, the results are presented in Figure This means that the influence of the countries on the observed differences is low. For each difference between a country mean and the global mean, the effect size was calculated dividing each difference by the average standard deviation.
According to Cohen , an effect size of 0. This proportion of variance is always low, with the exception of the digit span subtest where larger differ- ences between countries were observed. As the subtest was literally trans- lated in both countries, without any modification, these differences cannot be explained by the non-equivalent difficulty level of modified items as they could t TABLE Clinical Chapters for the high score on information in Sweden or the low score in comprehension in Japan.
According to Baddeley , we can store more short words than long words in our short-term memory because speech-based information is held in memory through an articulatory control process based on inner speech. As various East Asian languages, such as South Korea, have short words for dig- its, people speaking these languages can store more digits in their short-term memory. On the other hand, Lithuanians have longer words for digits and, con- sequently, people speaking Lithuanian can store fewer digits in their short-term memory. This observation is very important for a culturally based interpretation of the digit span subtest.
Thus, Lithuanians do not have a weaker short-term memory than the South Koreans; rather the observed difference between Korea and Lithuania can be explained by the specificity of each language. It is not a consequence of a poor test adaptation, and there is no psychometric procedure to erase this difference. Therefore, translating the digit span subtest to assess a child speaking a different language is not a magical solution to eliminate cul- tural bias. This proportion of variance is very low, with the exception of the symbol search subtest where larger differences between countries were observed.
Georgas et al. However, no suggestion was offered for explaining the lower score of children from Greece. Another interpretation of the high score for South Korean children could be p related to Korean writing.
Korean is written using an alphabetic system called Hangeul and a system of characters close to the Chinese one called Hanja. Learning to read and write using both systems could improve the ability to analyze and recognize symbols such as those presented in the symbol search subtest. If the mastery of the Korean writing system improves the performances on the symbol search subtest, it should also improve the performances on the Coding subtest since this subtest also requires children to analyze, memorize and write symbols.
Unfortunately, this information was missing for Korea in the Georgas et al.
- SearchWorks Catalog;
- Bestselling Series.
- Legend (Legend, Book 1)?
- The Collected Short Stories of Roland Mathias.
- Culture and Children's Intelligence: Cross-Cultural Analysis of the WISC-III;
- On the Road, op. 53, no. 1: Running (En courant) (piano).
However, a study conducted by Chen and Zhu on Taiwanese children p using Chinese letters provided a strong support to this hypothesis. These authors postulated that children who routinely read and write Chinese may perform faster on both the coding and symbol search tasks than peers who primarily read and write English because the symbols used in the subtests are more similar to Chinese characters than to English alphabet. To facilitate this comparison because Taiwan is a more homogeneous society than the United States, minority children of the US normative sample were excluded from the study.
Except for the language difference, the items, instructions, administration procedures and scoring rules were the same for the Taiwan and the US editions of the non-verbal subtests. Mean raw total scores of the matched Taiwan and US samples on block design, object assembly, picture completion, picture arrange- ment, mazes, coding and symbol search subtests were compared directly. The results revealed that on the picture completion, picture arrangement and p object assembly subtests, Taiwanese children had lower scores than their US peers.
However, Taiwanese children did significantly better than their US peers on coding, symbol search, block design and mazes subtests. The average effect sizes across the 11 age groups were 1.
- Coastal Change, Ocean Conservation and Resilient Communities.
- Quaternary Glaciations - Extent and Chronology: Part III: South America, Asia, Africa, Australia, Antarctica (Developments in Quaternary Science)?
- Shop now and earn 2 points per $1.
- A Cross‐Cultural, Comparative Study of the American, Spanish, and Mexican Versions of the WISC‐IV?
- Account Options.
The effect sizes were particularly large for coding and symbol search. However, further research is needed before drawing a strong conclusion on this issue.
Clinical Chapters relatively poor performances of Japan on the coding subtest and of Greece on the symbol search subtest. These differences could be related to cultural spe- cificity in early education, but this interpretation is not currently supported by empirical data. In the cross-cultural research described above, IQ test scores in the 12 different countries were also examined as a function of national indicators of affluence and education Georgas et al. The factors which influence mean IQ test scores between nations i.
While genetics is a significant determiner of intellectual abil- ity, children are not borne with a genetically predetermined, fixed IQ score, but with a range of intellectual potential.
Culture and Children's Intelligence (Enhanced Edition)
Thus we are not contesting the large con- tribution of genetics to measured intelligence but rather reinforcing that intellec- tual potential may be fully or partially actualized depending on qualities of the environment during the critical developmental years.
The United States pro- vides a natural laboratory for this research. It is comprised of African American, Hispanic and Caucasian majorities although these demographics are show- ing change with the influx of people from other countries. Of course the WISC-IV was developed and carefully standard- ized in the United States such that data from all three groups could be analyzed in a comparative manner. After controlling for parental education and income, race explained only 1.
It is not known why the African American — White difference was not completely eliminated by controlling for parent education and income as it was in the Hispanic — White analyses. However, Weiss et al. Parent expectations explained Finally, parental expectations as a function of parent education and income was investigated, and after controlling for parent educa- tion and income, parental expectations continued to account for This means that parent expec- tations are not fully overlapping with socio-economic status.
Weiss et al. The spe- cific parental behaviors and the general characteristics of cognitively enriching home environments need to be better understood by researchers and clinicians alike, but may include an increased amount and variety of linguistic and motoric stimulation in the early developmental years followed by parental monitoring of homework and leisure activities during later childhood and adolescence. Clinicians frequently are requested to assess the intelligence of children with p different cultural and linguistic backgrounds.
In such situations, their main con- cern is to avoid bias and conduct a fair assessment. With this aim, they can act in different ways. Some of their actions can be efficient, but others could be inappropriate. The most common actions to overcome possible cultural bias in cognitive ability tests are reviewed in the next sections, and their advantages and limitations are pointed out.
To what extent can the child be considered as being a member of the population in which the test was developed and standardized? Clinical Chapters To answer to this question, several acculturation scales were developed. Acculturation is not a dichotomous phenomenon: being or not being member of a specific culture. It is continuous and often long process of incorporation of a different culture.
Berry suggested appraising the degree of acculturation referring to two dimensions: the immersion in the culture of the native society and the immersion in the culture of the immigration society. Comparing the levels of immersion in each culture corresponds to four acculturation processes: assimila- tion moving away from the native culture and immersing fully in the immigra- tion culture , integration equal immersion in both cultures , separation complete immersion in the native culture and withdrawal from the immigration culture and marginalization lacking of meaningful immersion in both cultures.
It includes 32 items reflect- ing acculturation in the domains of language, interaction, media and food. However the results cannot be regarded alone or in isolation. They should take into account other relevant information regarding linguistic and school education. Differences between the standard language usually spoken by the main group of the population or in the larger country where this language is spoken and its local variants can be small, but sometimes important.
Culture and Children's Intelligence : James Georgas :
Typically, when develop- ing or adapting a test, only the standard language is taken into account. The test is then considered as the reference for all the variants of this language. But not taking variants and dialects into account when developing or adapting a test can lead to assessment bias and unfair decisions based on the scores in the adapted instrument. Norms were devel- oped for the same test in Canada, United Kingdom and Australia.
Why develop local norms if the test is similar to the original one? Why not use the US norms?
In fact this is an empirical question that can best be addressed when sufficient data are available. While the United States and Canada share a common border and English is the majority language spoken, the question of how well the Wechsler scales travel from the United States, where the tests were developed, standard- ized and normed, to Canada has been addressed.
It was first shown that with only a very few word changes, the items from these tests work very well in both countries DIF analysis, subtest reliabilities, etc. For example, the PRI difference between the United States and Canada was about five-scaled score points thereby necessitating the development of Canadian norms. Developing local norms is not always possible because it is too expen- p sive in small populations.
4 editions of this work
While the same language is spo- ken in both countries, there are some dialectal differences vocabulary, syntax, expressions… that could create bias in the verbal items and, consequently, lead to unfair assessment for the Belgian children. Biased items, showing a differential functioning between the French and Belgium groups, were flagged and deleted.
Although the norms for the French WISC-IV did not includ any Belgian children, the DIF analysis conducted during the item try-out helped to make possible a more fair measure for assessing these children. Unfortunately, DIF analysis conducted between linguistic subgroups of the population for which a test is being developed or adapted is uncommon.