Cultural Biases in Interpretation and Meaning of Words in Assessment
What is considered wise in one society may not be considered wise in another; the value and meaning of intelligence depends on cultural norms. Demonstrating the culturally-specific nature of knowledge and intelligence, Cole, Gay, Glick, and Sharp (1971) conducted an experiment in which Western participants and Kpelle participants from Liberia were given an object-sorting task. Participants were asked to sort twenty objects that were divided evenly into the linguistic cat- egories of foods, implements, food containers, and clothing. Westerners tended to sort these objects into the groups for food and implements, while Liberian partici- pants would routinely pair a potato with a knife because, they reasoned, the knife is used to cut the potato. When questioned, Liberian participants justified their pairings by stating that a wise person would group the items in this way. When the researchers asked them to show what an unwise person would do, they did the taxonomic sort that is more familiar to the Western culture.
In addition to biases due to culturally specific interpretation of test items, lan- guage impacts the understanding of test items. Cultural and social norms affect
Cultural Bias in Assessment | Kim, Zabelina | 131
how test-takers understand and interpret the wording of test questions. How they make sense of the test items can be influenced by their values, beliefs, experiences, communication patterns, teaching and learning styles, and epistemologies of their cultures and societies (e.g., Solano-Flores & Nelson-Barber, 2001).
Further, test item interpretation can be affected by test questions written in a language other than the native language of the test taker. It is important to con- sider a non-native English speaker’s language proficiency before deciding whether to test her/him in English or the native language (Geisinger, 2003). For example, a Latino might appear acculturated to the test administrators, but may be none- theless more proficient in Spanish than in English (Padilla, 1992). If the test is ad- ministered in English and is timed, it is likely that such a student will have more difficulties than if the test is in English, but untimed. To overcome these testing biases, researchers suggest that acculturation should be measured in addition to psychometric tests (Gopaul-McNicol & Armour-Thomas, 2002); however to date, there lacks a consensus on the most effective ways to measure acculturation (Cabassa, 2003).
Sometimes even the same word may have different meanings for different cultures. For instance, the meaning of educacion in Spanish is different from that of education in English (Reese, Balzano, Gallimore, & Goldenberg, 1995). The social skills of respectful and correct behavior are important to the Spanish when they state educacion, whereas only cognitive processing is important to education in many Western societies. Another well-documented instance is when Native American students are asked, “Who is the son of your aunt?” “Brother” is selected by all of the Native American students in the fifth grade (even though the of- ficially accepted/ expected answer would be “cousin”) because all relatives of the same generation are called “brothers” in Native American culture (Shields, 1997).
The way students use English sentence structure may also depend on cultural backgrounds. Native American students use a different sentence structure from that of English speakers when forming negative questions. When they are asked, “You don’t like eating this, [do you]?” they respond, “Yes” while they actually mean “Yes, you are right, I don’t like eating it” (Shields, 1997). Asian students use the same sentence structure as Native Americans. Thus, tests must take into account students’ ways of knowing and demonstrating their knowledge through their use of language (Solano-Flores & Trumbull, 2003; Swisher & Deyhle, 1992).
Several culturally sensitive tests have been developed to address the issue of this type of bias. Williams’ (1972) Black Intelligence Test of Cultural Homo- geneity (BITCH) is a better predictor of learning ability for African American students than other ethnicity groups. The results of the BITCH show that Afri- can American students perform better on the test than Caucasian students (Wil- liams, 1975). In addition, the Naglieri Nonverbal Ability Test (NNAT) (Naglieri, 1991) and the Comprehensive Test of Nonverbal Intelligence (CTONI) (Ham-
132 | International Journal of Critical Pedagogy | Vol. 6 No. 2, 2015
mill, Person, & Wiederhold, 1997) have been developed as culture-free ability assessments. The NNAT does not use words or language in any of the items on the test, and the figures that make up each item are not specific to any particular culture. In addition, the NNAT has been standardized on an English-speaking sample and a Spanish-speaking sample. The CTONI provides oral or pantomime instructions, and the examinee answers by pointing to the response that he or she thinks is correct. A review of the mean standard scores on the CTONI shows that all African American, Asian American, Hispanic American, Native Indian, and Caucasian students scored well within the normal range (Hammill et al., 1997). This may be because both the NNAT and the CTONI have used standardization samples with ethnic profiles that mirror the ethnic profile of the 1990 U.S. Cen- sus (Zurcher, 1998).