Reducing Bias in Assessment Materials

Being cognizant of common bias pitfalls helps writers communicate with transparency and integrity.

Bias is inherent in all writing

All writers bring inherent bias to their work. “Bias is any trend or deviation from the truth in data collection, data analysis, interpretation, and publication which can cause false conclusions” (Šimundić, 2013). For writers of research and technical materials in the psychological assessment field, bias in publication can manifest both as prejudiced word choice and as lack of proper citation and attribution.

Historically, certain written constructions and phrasings have been allowed to persist even though they reflect bias, which may be based on race, age, gender identity, disability, sexual orientation, and more. For example, the word “wheelchair-bound” presumes that people who use wheelchairs are confined to them or that the wheelchair is a restrictive device. Part of an editor’s job is to catch and point out words that convey bias, even unintentionally. In the example above, a better phrase is simply “uses a wheelchair.”

Over the last few years, the field of technical and academic editing has become quite appropriately focused on ensuring readers and subjects are being written about in a fair, representative, and respectful manner. At PAR, we have embraced newer comprehensive guidance from the American Psychological Association (APA; 2020) about how to reduce assumptions and bias in writing. Our house style guide—a company document that provides direction regarding acceptable writing conventions, appropriate citation, and PAR-specific styling—has been updated to reflect the same tenets, and our data collection practices are being modified to gather more specific information from participants.

Specificity and sensitivity

Writers should keep two concepts in mind to prevent bias: specificity and sensitivity.

When we write with specificity, we eliminate unconscious attitudes that might sneak into overly general language. Being as precise as possible also allows your research or other scholarly work to reflect the most accurate and helpful information for readers. For example, using specific terms for gender identity and sexual orientation (e.g., “cisgender women,” “bisexual people”) allows for deeper and more accurate representation at the individual level. Writing with sensitivity means writing while acknowledging people’s preferences and humanity. For example, overgeneralizing by using adjectives as nouns to label groups—for example, “the poor” or “schizophrenics”— eliminates the individuality of the people in those groups. Instead, use adjectival forms or nouns with descriptive phrases, like “poor people” or “individuals with schizophrenia.”

At PAR, one of the more significant changes we’ve implemented to write better with sensitivity is ceasing the use of “he or she” (and its variants) as generic third-person pronouns in our publications after 2020. Instead, “they” conveys the same concept and encompasses the gender identity of cisgender, transgender, and nonbinary individuals.

Attribution is everything

Another type of misrepresentation can emerge if authors do not appropriately cite their sources. It’s unethical and misleading to publish statements or ideas without providing information about where they originated. Transparency builds trust with readers and helps authors avoid plagiarism—the cardinal sin of writing. Taking this one step further, many believe attribution reform is needed in medical literature.

“Attribution is everything that is conveyed to the reader about the origins and development of the work, including all the parties involved, what they did and what their motives were” (Matheson, 2022; see also Matheson, 2016a; 2016b). Although articles typically list primary and secondary authors, issues of ghostwriting and corporate funding can confound the origins and development of articles. The 2000 Merck VIGOR study is a landmark example. Designed to increase indications for a rheumatoid arthritis drug, the study obfuscated several issues, including a financial conflict of interest for the head of the safety board and irregular data collection and analysis procedures that masked the cardiovascular risk associated with taking the drug (Krumholz et al., 2007).

At PAR, we hold our test-related publications to a strict standard that ensures comprehensive and appropriate citation of sources and clarity about the key players and steps in test development. Refer- ences and citations within each test manual are thoroughly verified, and we dedicate separate chapters to detailing information about standardization/data analysis and test psychometrics.

Many other forms of bias in writing (and research) exist. But through transparency and respectfulness, authors can ensure their writing avoids unfairness and elicits trust in readers.

References

American Psychological Association. (2020). Publication manual of the American Psychological Association (7th ed.). https://doi.org/10.1037/0000165-000

Krumholz, H. M., Ross, J. S., Presler, A. H., & Egilman, D. S. (2007). What have we learnt from Vioxx? BMJ, 334(7585), 120–123. https://doi.org/10.1136/bmj.39024.487720.68

Matheson, A. (2016a). Ghostwriting: The importance of definition and its place in contemporary drug marketing. BMJ, 354, i4578. https://doi.org/10.1136/bmj.i4578

Matheson A. (2016b). The ICMJE Recommendations and pharmaceutical marketing—strengths, weaknesses and the unsolved problem of attribution in publication ethics. BMC Medical Ethics, 17, 20. https://doi.org/10.1186/s12910-016-0103-7

Matheson, A. (2022). Attribution and credit bias in publication ethics. Indian Journal of Medical Ethics, 7(3), 204. https://doi.org/10.20529/IJME.2022.023

Šimundić, A. M. (2013). Bias in research. Biochemia Medica, 23(1), 12–15. https://doi.org/10.11613%2FBM.2013.003