Table 1
Justifications for including race (R) and ethnicity (E) variables in research
|
Role of the R & E variable |
Purpose of R & E variable |
Sample medical education research question using R & E variables |
|---|---|---|
|
Grouping |
To examine similarities or differences between R or E groups and/or subgroups based on a dependent (outcome) variable |
Is there a significant difference in medical students’ access to professional mentors by R or E group? |
|
Mediating |
To examine whether R or E explains the relationship between an independent (predictor) and dependent (outcome) variable |
Is the association between socioeconomic status and students’ perceptions of the medical school learning environment reduced when R or E are considered? |
|
Moderating |
To examine whether the strength of the relationship between an independent (predictor) and dependent variable (outcome) varies by R or E groups |
Does the relationship between social support and well-being vary by R or E group? |
Table 2
Advantages and disadvantages of various data collection methods
|
Category type |
Advantage(s) |
Disadvantage(s) |
Example |
|---|---|---|---|
|
Multiple-response (exclusive) categories: Multiple options provided; respondent can only select ONE pre-established category [33] |
Maintains original unit(s) of analysis Provides more complete and accurate data [34] Aligns data with most statistical analyses [33] Permits respondents to self-report identity and allows researchers to collect rich data [35] |
Provides less data per category which increases the risk of error in interpreting outcomes [36] Forces respondents into discrete category that does not allow for fluid or broad self-identification [37] |
Respondent must select one option from White, African American, American Indian, Alaska Native, and Native Hawaiian |
|
Multiple-response (inclusive) categories: Multiple options provided; respondent can select MULTIPLE options from pre-established categories [33] |
Introduces issues related to comparability of samples across multiple data sets [38] Forces researcher to decide how individuals fit into certain categories [37] Counts multiracial respondents as members of each individual racial or ethnic group they select which inflates the number of respondents in denominator [33] |
Respondent may select multiple options from White-non-Hispanic, African American, American Indian, Alaska Native, and Native Hawaiian | |
|
Combined categories a: Multiple options combined to define new categories |
Simplifies statistical analysis, interpretation and presentation of results [39] Increases cell size when discrete categories are too small [40] |
Limits conclusions to broad assumptions and generalizations about respondents within groups [41, 42] Perpetuates obsolete majority/minority discourse when using certain binary frameworks (e.g., White/non-White) [34] Uses subjective labels that can perpetuate bias/stereotypes [43] Increases the risk of a false positive result [44] Underestimates the extent of variation between groups by not fully accounting for within group variability [26] |
Respondent must select either URiMb [45] or Non-URiM |
aCombined categories also can represent collapsed and dichotomous categories
bURiM Underrepresented in medicine