Are Inferential Reading Items More Susceptible to Cultural Bias Than Literal Reading Items?

Banks, Kathleen

doi:10.1080/08957347.2012.687610

Cited by 9 publications

(19 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Topic familiarity, strongly tied to cultural background, can shape test performance (McCullough, 2013;Moore, 1996;Reynolds, Taylor, Steffensen, Shirey, & Anderson, 1982). In addition, underlying value orientations embedded in passages/questions (e.g., individualism) may favor readers from cultures that share those value orientations (Banks, 2012;Lewis, 1993). Also, whereas cultural bias in tests may be particularly acute in inferential reading tasks (Banks, 2012;C.S.…”

Section: What Kinds Of Readers Should Society Raise?mentioning

confidence: 99%

What Matters Most? Toward a Robust and Socially Just Science of Reading

Aukerman¹,

Schuldt²

2021

Reading Research Quarterly

View full text Add to dashboard Cite

Science of reading is a term that has been used variously, but its use within research, policy, and the press has tended to share one important commonality: an intensive focus on assessed reading proficiency as the primary goal of reading instruction. Although well intentioned, this focus directs attention toward a problematically narrow slice of reading. In this article, we propose a different framework for the science of reading, one that draws on existing literacy research in ways that could broaden and deepen instruction. The framework proposes, first, that reading education should develop textual dexterity across grade levels in the four literate roles first proposed by Freebody and Luke: code breaker (decodes text), text participant (comprehends text), text user (applies readings of text to accomplish things), and text analyst (critiques text). Second, the framework suggests that reading education should nurture important literate dispositions alongside those textual capacities, dispositions that include reading engagement, motivation, and self‐efficacy. Justification is offered for the focus on textual dexterity and literate dispositions, and we include research‐based suggestions about how reading educators can foster student growth in these areas. Finally, we propose that reading education should attend closely to linguistic, cultural, and individual variation, honoring and leveraging different strengths and perspectives that students bring to and take away from their learning. Reimagining a science of reading based on these principles has the potential to make it both more robust and more socially just, particularly for students from nondominant cultures.

show abstract

Section: What Kinds Of Readers Should Society Raise?mentioning

confidence: 99%

What Matters Most? Toward a Robust and Socially Just Science of Reading

Aukerman¹,

Schuldt²

2021

Reading Research Quarterly

View full text Add to dashboard Cite

show abstract

“…Banks (2012) also furthered the item bias conversation with a study that analyzed the results of fifth- and seventh-grade Hispanic, black, and white students’ performances on the reading section of a reading/language arts test. Banks compared students’ responses on literal multiple-choice items (i.e., those asking for basic recall of what students had read) with students’ responses on inference multiple-choice items (i.e., those requiring reasoning beyond what appears in the text).…”

Section: Cultural Bias Researchmentioning

confidence: 99%

“…If distinct correct options must exist for a music assessment, the possibility that distractors may function differently for various cultural groups should at the very least be cause for concern. According to Banks (2012), inference items have increased risk of DIF and cultural bias compared with literal items. This suggests that music questions requiring students to make use of their music experiences outside of school (e.g., knowledge of particular folk or popular songs) are more likely to be culturally biased.…”

Section: Validity and Bias In Music Assessmentmentioning

confidence: 99%

Cultural Bias in Testing

Kruse

2016

Update: Applications of Research in Music Education

View full text Add to dashboard Cite

The findings and discussions related to cultural bias in testing have in no way been unanimous. However, the considerations of this area of inquiry may possess meaningful implications for educators of any subject. In this review of literature, I describe the issues, research, and arguments surrounding cultural bias in testing and discuss implications for the field of music education. A working description of cultural bias in testing for the purpose of this article involves the notions of (a) significantly different results for definable subgroups from apparently similar ability levels and (b) issues with the fair and equitable interpretation and use of test results. Applications of general education scholarship to music education settings include investigations and perceptions of cultural bias as well as suggestions for improved fairness consisting of addressing group differences, offering diverse ways to perform, discouraging misuse, and accommodating for differences.

show abstract

“…Bolt () found some evidence that DIF was associated with item format and the type of problem addressed in the math portion of the SAT using variations of a single item across versions of the test. Banks (), using a seven‐step process to assess DIF in a reading/language arts test for elementary school children, reported some evidence that inferential items were more susceptible to cultural bias than literal items. Penfield () and Suh and Bolt () have explored the use of DDF as a means of explaining DIF.…”

Section: Introductionmentioning

confidence: 99%

“…DIF is defined as differential endorsement by members of different, but equally able, examinee subgroups of the right option in a multiple-choice test item. Researchers have investigated cultural hypotheses regarding the presence of DIF, but research has often not been supportive of cultural explanations or demographic hypotheses (Banks, 2012;Scheuneman, 1987). There are two possible reasons why this *Correspondence should be addressed to Neal Schmitt, 46 Sunset Rock Lane, Reading, MA 01867, USA (email: schmitt@msu.edu). research may have failed to provide support for hypotheses regarding cultural and demographic influences on test item responses.…”

mentioning

confidence: 99%

Culture as a determinant of option choice in a situational judgement test: A new look

Schmitt

Prasad

Ryan

et al. 2018

J Occupat & Organ Psyc

View full text Add to dashboard Cite

Past research on cultural correlates of biased items in multiple‐choice tests have almost always considered post‐hoc explanations that have not been very convincing and they have considered only differential subgroup responses to the correct or preferred answer. In an effort to determine the degree to which culture explains the difference between Chinese and Americans’ responses to a situational judgement test (SJT), 135 options to 25 SJT items were rated on the degree to which they reflected seven dimensions of culture. Ratings of the cultural content of the options were significantly and moderately correlated with Chinese–American differences in Most Likely responses to the options. In addition, a loglinear analysis of differential distractor functioning (DDF) yielded results consistent with the hypothesis that culture was at least partly the reason for subgroup differences in option choices and the resultant findings of DDF. The results suggest that SJTs and perhaps other non‐academic measures must be used with caution across cultural groups. Practitioner points Consultation with experts and test takers in the cultures with which you hope to use an instrument should always precede test use. Review with them both the item stems and options (if any) for items that may have a different meaning across cultural groups. Measurement invariance of instruments used with different cultural groups should be assessed. When sample sizes do not permit sophisticated analyses such as confirmatory factor analyses or item response theory analyses, examine and seek to understand any simple descriptive statistics that reflect a difference across cultural groups. There should be an examination of examine response differences to distractors in multiple‐choice items as well as differences in responses to correct options.

show abstract

Are Inferential Reading Items More Susceptible to Cultural Bias Than Literal Reading Items?

Cited by 9 publications

References 40 publications

What Matters Most? Toward a Robust and Socially Just Science of Reading

What Matters Most? Toward a Robust and Socially Just Science of Reading

Cultural Bias in Testing

Culture as a determinant of option choice in a situational judgement test: A new look

Contact Info

Product

Resources

About