Skip to main content

Developing and validating measures of self-reported everyday and healthcare discrimination for Aboriginal and Torres Strait Islander adults

Abstract

Background

It is well established that racism is a fundamental contributor to poor health and inequities. There is consistent evidence of high exposure to discrimination among Aboriginal and Torres Strait Islander (Indigenous Australian) peoples, but impacts have not been fully quantified, in part due to limited measurement tools. We aim to validate instruments developed to measure interpersonal discrimination.

Methods

Instruments were discussed at five focus groups and with experts, and field tested in developing Mayi Kuwayu: The National Study of Aboriginal and Torres Strait Islander Wellbeing. Data from 7501 baseline survey participants were analysed. Acceptability was assessed according to extent of missingness, construct validity using exploratory and confirmatory factor analysis, and reliability using Cronbach’s alpha. Associations between each instrument and outcomes conceptually understood to be closely (community-level racism) or less closely (family wellbeing) related were quantified to test convergent and discriminant validity.

Results

An 8-item instrument captures experiences of discrimination in everyday life and a 4-item instrument experiences in healthcare, each followed by a global attribution item. Item missingness was 2.2–3.7%. Half (55.4%) of participants reported experiencing any everyday discrimination, with 65.7% attributing the discrimination to Indigeneity; healthcare discrimination figures were 34.1% and 51.1%. Items were consistent with two distinct instruments, differentiating respondents with varying experiences of discrimination. Scales demonstrated very good reliability and convergent and divergent validity.

Conclusion

These brief instruments demonstrate face validity and robust psychometric properties in measuring Aboriginal and Torres Strait Islander adults’ experiences of interpersonal discrimination in everyday life and in healthcare. They can be used to quantify population-level experiences of discrimination, and associated wellbeing consequences, and monitor change.

Introduction

Racism is a fundamental cause of ill health and health inequities globally [1]. Racism is a system of oppression based on the social ranking of groups of people into categories of race, with those lower in the social strata considered inferior and denied access to rights, resources, and opportunities—processes that have clear health consequences [2]. Discrimination is a manifestation of racism at the intrapersonal, interpersonal, and institutional (systemic) level. Given the difficulty in measuring systemic forms, quantitative research has predominantly focused on experiences of interpersonal discrimination [2,3,4]. Discrimination can occur on the basis of ethnicity, race, Indigeneity, or other characteristics; it is difficult to identify the basis for discrimination, but regardless of the perceived reason for the discrimination, evidence indicates that there are negative outcomes associated with exposure [3, 5,6,7,8].

Aboriginal and Torres Strait Islander (Indigenous Australian) stakeholders have identified as a high priority research on the experiences of discrimination, overall and specifically within healthcare [9, 10]. Regardless of the measure used, there is consistent evidence of high exposure to discrimination in this pouplation [11,12,13,14,15,16,17]. This is a consequence of ongoing colonisation, which has contributed to the systematic oppression, disempowerment and exclusion of Aboriginal and Torres Strait Islander peoples.

There is extensive international evidence that experiences of discrimination are associated with poorer wellbeing outcomes, such as psychological distress, depression, anxiety, general health, and physical health—including blood pressure and allostatic load [6, 18]. Discrimination is hypothesised to impact wellbeing through multiple direct and indirect pathways [18]. Racism instigates and perpetuates discrimination and trauma as well as influencing the levels, clustering, and impacts of stressors such as those related to employment, housing, and neighbourhood quality. Discrimination therefore influences access to societal resources and socioeconomic opportunities that promote health. Physiological, behavioural, and psychological responses to discrimination also impact health [19].

Discrimination’s prevalence and strength of association with health varies between and within population groups and by outcome [18]. There is some Aboriginal and Torres Strait Islander-specific evidence on negative outcomes linked to discrimination experiences in everyday life, [11,12,13,14, 20, 21] and within healthcare [15, 16]. Given the high exposure prevalence and known association with numerous negative health outcomes, discrimination is likely a substantial contributor to ill health for the population, and to inequities compared to the non-Indigenous population. However, discrimination’s full impacts have not been robustly quantified.

Despite research use of over 100 discrimination instruments [22], a 2010 systematic review identified only 24 instruments with published evidence of psychometric validity [7]. Of the identified instruments, all but one were developed within the United States (US) [7]. These instruments may not capture common forms of discrimination experienced by Aboriginal and Torres Strait Islander peoples, given historical and contextual differences [3, 23, 24]. The review [7] identified one instrument developed for Aboriginal and Torres Strait Islander peoples: Measuring Indigenous Racism Experiences (MIRE) [25]. MIRE demonstrated good content and psychometric validity, and reasonable construct and convergent validity and acceptability in a localised sample [25]. Further validation of the instrument in a more heterogeneous sample has not occurred [25]. Other discrimination items have been used in surveys with Aboriginal and Torres Strait Islander peoples; however, to our knowledge, no other discrimination instrument has been validated for use with this population.

The Everyday Discrimination Scale (EDS), a nine-item scale developed to measure exposure to overt discrimination among African-American populations, is among the most commonly used measures internationally [7, 22]. Respondents are asked if they have experienced a set of specific manifestations of interpersonal discrimination, and asked to attribute the discrimination to one or more of their personal characteristics (e.g. race, gender, age). A modified EDS was identified as a reliable and valid measure of discrimination for a diverse sample of US First Nations peoples [24]. This indicates a modified EDS has potential validity for use with other Indigenous populations, noting differences in the discrimination experiences between (and within) Indigenous populations. We are unaware of evidence on the validity of the EDS for any other Indigenous population, or of any healthcare discrimination instruments validated for use with an Indigenous population.

There is a clear need for valid measurement of Aboriginal and Torres Strait Islander peoples’ experiences of discrimination, to inform program and policy development to reduce discrimination and associated harms. The aim of this paper is to conduct a psychometric validation of the instruments developed to capture experiences of overt discrimination in Mayi Kuwayu: The National Study of Aboriginal and Torres Strait Islander Wellbeing.

Methods

Study population

The Mayi Kuwayu Study is a large-scale, national longitudinal study of adults aged ≥16 years. Rolling baseline data collection commenced in 2018, following multiple years of study development. The study employs multi-mode recruitment, with participants recruited via a postal questionnaire, through word-of-mouth or study advertising, and through on-the-ground community researchers in defined sites. Surveys were self-completed on paper or online, completed with an Aboriginal Mayi Kuwayu Study team member over the phone, or completed with the support of an Aboriginal and/or Torres Strait Islander community researcher. The initial data release (Data Release 1.1) includes 7501 Aboriginal and Torres Strait Islander adults whose survey data were processed by 2 July 2019. Details of the study design are provided elsewhere [26].

Face validation process

The Mayi Kuwayu Study questionnaire includes a combination of (1) established instruments, where existing instruments had evidence of validity with this population; (2) modified instruments, where instruments existed but had no evidence of, or established limitations to, validity with this population; and (3) new instruments developed through an iterative community consultation process [26].

The drafted instruments for capturing experiences of discrimination and racism were trialed and discussed at five focus groups, spanning major cities to remote areas. As part of these focus groups, participants completed the current draft of the Study questionnaire and provided feedback on the questionnaire items verbally and/or in writing on the questionnaire (Lovett R et al.: Aboriginal and Torres Strait Islander culture and wellbeing indicator development in Australia, in preparation).

In parallel with the focus groups, the questionnaire was further refined through field testing and input from other stakeholders with lived experience and/or content expertise (e.g. field researchers, study Chief Investigators, independent researchers).

Data

All data analysed in the current study are based on self-reported responses to the final baseline questionnaire, with the exception of remoteness, which was derived based on postcode.

For the final discrimination instruments (Fig. 1), response options included “not at all” (coded as 0), “a little bit” (1), “a fair bit” (2), and “a lot” (3). Responses were summed across items to form a total score, which was categorised as no, low, moderate, or high discrimination. Participants with missing data on any item had a missing total score.

Fig. 1
figure1

Final version of discrimination measures used in the Mayi Kuwayu Study

After each instrument, participants were asked “When these things happen, do you think it is because you are Aboriginal/Torres Strait Islander?”, with the same response options. For each instrument, participants were categorised as experiencing no discrimination (based on total score, coded as 0), or any discrimination attributed to Indigeneity “not at all” (coded as 1), “a little bit” (2), “a fair bit” (3), or “a lot” (4).

To capture participants’ perceptions of the impacts of discrimination, participants were asked “How much do these things affect your life?”; this variable was treated the same as the above.

Other variables used for validation are described in Additional File 1.

Statistical analysis

The sample was characterised in relation to age group, gender, and remoteness. For each discrimination instrument, the response distribution across categories was explored for individual items and total scores. Missing data are presented for acceptability assessment. Mean scores (and 95% Confidence Intervals, CIs) are presented overall and by demographic characteristics.

Construct validity was assessed with exploratory factor analysis and Confirmatory Factor Analysis (CFA) for categorical items, each on a random half of the sample. Characteristics of the two sub-samples are summarised in Additional Table 1; chi-squared tests were used to test for differences between groups [27].

Table 1 Distribution of responses to discrimination items (N = 6775)

Principal Axis Factor (PAF) was applied to the “developmental” sub-sample to test the factor structure and identify poorly fitting items for potential removal. Based on the exploratory findings, CFA was conducted in the “validation” sub-sample to confirm unidimensionality and test model fit. Indicators of model fit and appropriateness are described in Additional File 1 [28,29,30,31,32,33].

The relationship between the latent variable and item responses was assessed using a generalised partial credit model for ordinal items. Item slope coefficients 0.5–2.0 are expected [34], with higher values indicating greater discernment between participants on the latent variable; coefficients >4.0 indicate too much covariation with other items [35]. Difficulty thresholds identify the point on the latent variable at which a response category is more likely to be endorsed than the previous category. Non-overlapping 95% CIs between successive difficulty thresholds indicate a linear progression of response categories.

In each sub-sample, reliability of each instrument was assessed using Cronbach’s alpha (α≥0.7 and <0.90) [36].

We tested convergent and discriminant validity by quantifying the association between each discrimination instrument (categorical) and an outcome conceptually understood to be closely related (problems with racism in the community), and an outcome conceptually understood to be less closely related (family wellbeing). We expected to see a stronger association between the discrimination scales and community-level racism (positive), compared to family wellbeing (negative). We calculated prevalence ratios (PRs) and 95%CIs using log-binomial models. We calculated correlation coefficients to assess the continuous relationships, excluding participants missing discrimination scores. Correlations of magnitude 0.10–0.29 were considered “weak”, 0.30–0.49 “moderate”, and 0.50–1.00 “strong” [37].

An alpha level of 0.05 was the threshold for statistical significance.

Results

Face validity

The initial racism instrument tested through the Mayi Kuwayu Study was adapted from the first question of the MIRE (Table S1). MIRE uses a one-stage approach to ask about experiences of unfair treatment due to being Aboriginal and/or Torres Strait Islander [25]. This instrument was tested in two remote settings and discussed in-depth during one focus group. Researchers conducting the focus group identified that many participants described experiences where they were treated unfairly, but because these experiences were so common, it was perceived as "normal" treatment and not identified as discrimination. Alternative instruments were tested in the next survey version, aiming to elicit responses about specific circumstances.

A two-stage approach was employed next (Table S2), first asking about any experiences of discrimination, and then asking if each experience was attributed to Indigenous status. It included items capturing experiences of everyday (13 items) and healthcare (7 items) discrimination. Questions were adapted from previous instruments, including the EDS, or developed based on literature or personal experiences. The response options were changed to match those received positively throughout Mayi Kuwayu Study engagement/feedback processes (“not at all”, “a little bit”, “a fair bit”, “a lot”). Respondents were asked to tick “Yes” for each item if they considered the treatment was due to their Indigeneity. After each set of items, participants were asked two questions about the discrimination’s perceived impact. This version was tested in two major cities. There was heterogeneity in responses, but very few respondents assigned attribution. Field researchers expressed concern that respondents were inadvertently skipping the attribution question due to the multi-column formatting.

The items were restructured into a single-column format in the subsequent survey version. Due to space constraints, a single global attribution question was used, covering all items in the instrument. The question “How stressful is it when these things happen?” was removed, and the wording for the final question was changed to “How much do these things affect your life?”. This version was trialed in one inner regional area. The instruments were subsequently reduced through consultation; the final survey included an 8-item everyday and 4-item healthcare discrimination instrument (Fig. 1).

Psychometric validation

The sample included 6775 adults after excluding those missing age, gender, or remoteness (9.7% excluded). 61.3% of the sample was female, 45.8% from major cities, and 63.8% aged ≥46 years; the distribution of the sample differs from that of the total Aboriginal and Torres Strait Islander adult population (Table S3).

Across items, 2.2–3.7% of respondents did not provide a response, indicating the items were broadly acceptable (Table 1). The majority of the sample (58.5–84.8%) reported no experience of each discrimination item.

For the everyday discrimination instrument, the most highly endorsed items were "People act like I am not smart" (mean 0.59;95%CI:0.57,0.61) and "I am treated with less respect than other people" (mean 0.53;0.51,0.55). The mean total score was 2.76 (2.66,2.86), with 41.8% experiencing no, 42.5% low, 6.6% moderate, and 1.9% high discrimination. For the healthcare discrimination instrument, the highest endorsed item was “Health care providers do not listen to what I say” (mean: 0.39;0.38,0.41). The mean total score was 1.06 (1.00,1.11), with 63.1% experiencing no discrimination, 26.7% low, 4.3% moderate, and 1.7% high discrimination. For both instruments, the majority of those experiencing discrimination made some attribution to Indigeneity, and around two-thirds reported that their experiences affected their life “a little” to “a lot”.

The mean total score for everyday and healthcare discrimination varied significantly by age group (lowest among those aged ≥66 years) and remoteness (highest in remote and very remote areas); scores did not vary significantly by gender (Table S4).

PAF results were consistent with two distinct factors, aligned with the intended instruments, with all factor loadings >0.3 and negligible cross-loadings (Table 2). Within this sub-sample, Cronbach’s α was 0.894 for everyday and 0.853 for healthcare discrimination, indicating very good internal consistency.

Table 2 PAF analysis for discrimination items combined (two factors), and individually, within the “developmental” sub-sample

The CFA results confirmed the unidimensional nature of each instrument. For the everyday and healthcare discrimination instruments, the standardised coefficients loaded significantly onto a single dimension and ranged from 0.60–0.79 and 0.71–0.86, and fit was deemed to be acceptable (SRMR = 0.04 and 0.03, CFI = 0.94 and 0.98, TLI = 0.92 and 0.94). The RMSEA indicated model misfit (0.10 and 0.15), but the SRMR was prioritised given item skewness.

For everyday discrimination items, no coefficients were indicative of excessive item covariation; the lowest coefficient was for “Police unfairly bother me” (1.59) (Table 3). The response categories for the item “People act like I am not smart” showed an upward progression with non-overlapping CIs between category thresholds. Category threshold parameters were not in sequential order for the police item, indicating a potential issue with participants’ response grading. The remaining six items had overlapping CIs between category thresholds “a little bit and a fair bit” and “a fair bit and a lot” suggesting these response categories may have been difficult to distinguish.

Table 3 Item discrimination and difficulty parameters for everyday and healthcare discrimination items, within the “validation” sub-sample

For the healthcare items, the slope coefficient for “I receive poorer health care than other people” (5.21) indicated excessive covariance (Table 3). Response categories for this item showed a linear upward progression with non-overlapping CIs; for the other three items, CIs overlapped across categories “a little bit and a fair bit” and “a fair bit and a lot”, as above.

In the “validation” sub-sample, Cronbach’s α was 0.891 for everyday discrimination and 0.857 for healthcare discrimination, indicating very good internal consistency.

Both instruments demonstrated convergent validity with community-level racism (Table 4). The prevalence of any community-level racism was significantly higher among those experiencing low, moderate, or high, compared to no, discrimination in everyday life (PR = 3.30;2.95,3.69 for high) or healthcare (PR = 1.88;1.62,2.16 for high). Correlation with community-level racism was moderate (0.41) for everyday, and weak (0.26) for healthcare discrimination total score. Associations were of reduced magnitude, though still significant, for family wellbeing, supporting divergent validity.

Table 4 Association between discrimination scales and psychological distress (convergent validity) and family wellbeing (divergent validity)

Discussion

These instruments, capturing discrimination experiences in everyday life and in healthcare, were assessed on their face value to capture experiences of Aboriginal and Torres Strait Islander peoples, using robust and iterative consultative processes. The instruments have robust psychometric properties: acceptability, internal consistency, and construct validity. They are relatively brief, and can be used independently or in combination. The instruments capture broad experiences of discrimination, and additional items enable researchers to identify attribution to Indigeneity. They complement the current single other validated discrimination instrument for Aboriginal and Torres Strait Islander peoples, the MIRE [25].

The final instruments performed relatively well in terms of unidimensional fit. The healthcare discrimination item “I receive poorer health care than other people” may have been too broad and somewhat redundant. The additional response burden of answering this item was considered minimal and the item was therefore retained.

A strength of these discrimination instruments is the multiple approaches employed to generate, modify, and validate the instruments—with active involvement of Aboriginal and Torres Strait Islander peoples at all stages. Items were informed by the literature and adapted from previous instruments. New items specific and acceptable to the Aboriginal and Torres Strait Islander population were generated through consulting community members and researchers with content expertise. Face and content validation approaches were iterative and integrated input from the target population and “expert judges” [30], and included focus groups, pilot testing with diverse members of the target population, and review by multiple stakeholder groups. These qualitative processes identified issues with item interpretation and format, and culminated in instruments considered acceptable to the stakeholders involved.

The two-stage approach of measuring all forms of discrimination and identifying experiences attributed to Indigeneity offers data users flexibility in data use. Given difficulties accurately attributing discrimination experiences, the true prevalence of interpersonal discrimination due to Indigeneity likely lies between the prevalence of any discrimination and that attributed to Indigeneity [5]. Some experiences of discrimination due to Indigeneity would likely be missed using a one-stage approach.

The final item about discrimination’s impacts enables examination of participants’ self-identified impacts of their discrimination experiences.

Strengths and limitations

These scales are promising as measures of self-reported experiences of discrimination in everyday life and in healthcare among the Aboriginal and Torres Strait Islander adults (≥16 years). Generalisability of these findings beyond the cohort is unknown; however, the Mayi Kuwayu Study cohort is heterogeneous. The large sample size, and more than satisfactory subject-to-item ratio (>500), [7, 27, 30] would enable examination of the measure by different groups (e.g. by age group, gender, geography, or other characteristic) in future research, including further exploration of the value of including the item on “poorer treatment” in healthcare.

While the scales used in the Mayi Kuwayu Study were adapted from the EDS and other measures, modifications were substantial in order to be relevant and appropriate for the population of interest; this reduces comparability across populations.

In the absence of a gold standard measure of experiences of discrimination, precluding examination of criterion validity [30], scale scores were correlated against measures theorised to be associated with the construct [27].

The instruments used in the Mayi Kuwayu Study do not refer to experiences of discrimination over a specified time period; rather, they capture lifetime exposure. The instruments do not use objective measures of frequency (i.e. weekly or daily), as previous research has demonstrated that subjective assessment of frequency (as occurs here) can be easier than quantification [38].

Future work could explore the implications of using the instruments in continuous versus categorical form, and identify relevant cut-offs specific to this population. The four-tiered response category may have been unnecessary in this sample; future research could explore if fewer response categories would suffice.

While the terms “racism” and “discrimination” were not used within the instruments, these words were made salient through their use in the survey section header [7]. This could have contributed to increased reporting of experiences of discrimination and/or increased attribution to Indigeneity, compared to a neutral header [7].

These instruments are based on self-reported experiences of discrimination, and self-identification of attribution, which may induce biases such as minimisation bias and vigilance bias [6]. However, experiences of discrimination are subjective (this does not mean they are not real), and therefore reliance on self-report is appropriate [3, 6,7,8, 35].

The instruments developed are not exhaustive. As with any measure, they will not capture all experiences of discrimination in everyday life, or in healthcare [7]. Further, these scales do not capture experiences of systemic discrimination, which are difficult to capture by asking about specific individual experiences, [3, 7, 8] but are acknowledged to be common [4, 39]. Methodological advances are required to enable survey research to capture systemic discrimination.

Conclusions

Racism is a fundamental contributor to ill health and to health inequities for Aboriginal and Torres Strait Islander peoples. High quality measurement of experiences of discrimination is therefore essential to underpin action to improve health and reduce inequities.

These instruments capture Aboriginal and Torres Strait Islander peoples’ experiences of interpersonal discrimination. They can be used to enable valid measurement of discrimination’s prevalence, in order to identify priority targets for action, quantify discrimination’s contribution to health and health inequities, monitor trends, and evaluate interventions.

The instruments may be meaningful for use with other Indigenous populations with similar discrimination experiences; however, cross-cultural validity would need to be explored, and local adaptation may be required.

Availability of data and materials

The datasets analysed during the current study are available on application to the Mayi Kuwayu Study Data Governance Committee. This governance body oversees and approves applications for data use, in order to maintain the confidentiality of participants, and ensure that all studies using the Mayi Kuwayu data are protective of Aboriginal and Torres Strait Islander data and cultures. The data application process is detailed here: mkstudy.com.au/overview/.

Abbreviations

US:

United States

MIRE:

Measuring Indigenous Racism Experiences

EDS:

The Everyday Discrimination Scale

PAF:

Principal Axis Factor

CFA:

Confirmatory Factor Analysis

PRs:

Prevalence Ratios

HRECs:

Human Research Ethics Committees

SE:

Standard Error

CIs:

Confidence Interval

RMSEA:

Root mean square error of approximation

SRMR:

Root mean squared residual

CFI:

Comparative fit index

TLI:

Tucker-Lewis Index

References

  1. 1.

    Phelan JC, Link BG. Is racism a fundamental cause of inequalities in health? Annu Rev Sociol. 2015;41:311–30.

    Article  Google Scholar 

  2. 2.

    Williams DR, Lawrence JA, Davis BA. Racism and health: evidence and needed research. Annu Rev Public Health. 2019;40:105–25.

    PubMed  PubMed Central  Article  Google Scholar 

  3. 3.

    Williams DR, Mohammed SA. Discrimination and racial disparities in health: evidence and needed research. J Behav Med. 2009;32(1):20–47.

    PubMed  Article  Google Scholar 

  4. 4.

    Paradies Y, Harris R, Anderson I. The impact of racism on indigenous health in Australia and Aotearoa: towards a research agenda. Darwin: Cooperative research Centre for Aboriginal Health; 2008. p. 1–38.

    Google Scholar 

  5. 5.

    Shariff-Marco S, Breen N, Landrine H, et al. Measuring everday racial/ethnic discrimination in health surveys. Du Bois Rev. 2011;8(1):159–77.

  6. 6.

    Lewis TT, Cogburn CD, Williams DR. Self-reported experiences of discrimination and health: scientific advances, ongoing controversies, and emerging issues. Annu Rev Clin Psychol. 2015;11:407–40.

    PubMed  PubMed Central  Article  Google Scholar 

  7. 7.

    Bastos JL, Celeste RK, Faerstein E, et al. Racial discrimination and health: a systematic review of scales with a focus on their psychometric properties. Soc Sci Med. 2010;70(7):1091–9.

    PubMed  Article  Google Scholar 

  8. 8.

    Paradies YC. Defining, conceptualizing and characterizing racism in health research. Crit Public Health. 2006;16(2):143–57.

    Article  Google Scholar 

  9. 9.

    Prepared for the Australian National University by the Lowitja Institute and Abstarr Consulting. ANU Initiative in Aboriginal and Torres Strait Islander health and wellbeing: Stakeholder workshop research question prioritisation. Summary report. Canberra: The Australian National University; 2017.

  10. 10.

    Anderson P. Racism is killing us: statement by pat Anderson AO, chairperson, Lowitja Institute. The Lowitja Institute: Carlton, Australia; 2020.

    Google Scholar 

  11. 11.

    Paradies YC, Cunningham J. The DRUID study: racism and self-assessed health status in an indigenous population. BMC Public Health. 2012;12:131 [published Online First: 2012/02/16].

    PubMed  PubMed Central  Article  Google Scholar 

  12. 12.

    Ziersch A, Gallaher G, Baum F, et al. Racism, social resources and mental health for Aboriginal people living in Adelaide. Aust N Z J Public Health. 2011;35(3):231–7 [published Online First: 2011/06/02].

    PubMed  Article  Google Scholar 

  13. 13.

    Larson A, Gillies M, Howard PJ, et al. It's enough to make you sick: the impact of racism on the health of Aboriginal Australians. Aust N Z J Public Health. 2007;31(4):322–9.

    PubMed  Article  Google Scholar 

  14. 14.

    Priest NC, Paradies YC, Gunthorpe W, et al. Racism as a determinant of social and emotional wellbeing for Aboriginal Australian youth. Med J Aust. 2011;194(10):546–50.

    PubMed  Article  Google Scholar 

  15. 15.

    Kelaher MA, Ferdinand AS, Paradies Y. Experiencing racism in health care: the mental health impacts for Victorian Aboriginal communities. Med J Aust. 2014;201(1):44–7.

    PubMed  Article  Google Scholar 

  16. 16.

    Awofeso N. Racism: a major impediment to optimal indigenous health and health care in Australia. Aust Indigenous Health Bull. 2011;11(3):1–8.

    Google Scholar 

  17. 17.

    Ziersch AM, Gallaher G, Baum F, et al. Responding to racism: insights on how racism can damage health from an urban study of Australian Aboriginal people. Soc Sci Med. 2011;73(7):1045–53.

    PubMed  Article  Google Scholar 

  18. 18.

    Paradies Y, Ben J, Denson N, et al. Racism as a determinant of health: a systematic review and meta-analysis. PLoS One. 2015;10(9):e0138511.

    PubMed  PubMed Central  Article  CAS  Google Scholar 

  19. 19.

    Priest N, Williams DR. Racial discrimination and racial disparities in health. The Oxford handbook of stigma, discrimination, and health. New York, NY, US: Oxford University Press; 2018.

    Google Scholar 

  20. 20.

    Priest N, Paradies Y, Stevens M, et al. Exploring relationships between racism, housing and child illness in remote indigenous communities. J Epidemiol Community Health. 2012;66(5):440–7.

    PubMed  Article  Google Scholar 

  21. 21.

    Priest N, Paradies Y, Stewart P, et al. Racism and health among urban Aboriginal young people. BMC Public Health. 2011;11(1):568.

    PubMed  PubMed Central  Article  Google Scholar 

  22. 22.

    Paradies Y. A systematic review of empirical research on self-reported racism and health. Int J Epidemiol. 2006;35(4):888–901.

    PubMed  Article  Google Scholar 

  23. 23.

    Lewis TT, Yang FM, Jacobs EA, et al. Racial/ethnic differences in responses to the everyday discrimination scale: a differential item functioning analysis. Epidemiol Rev. 2012;175(5):391–401.

    Google Scholar 

  24. 24.

    Gonzales KL, Noonan C, Goins RT, et al. Assessing the Everyday Discrimination Scale among American Indians and Alaska Natives. Psychol Assess. 2016;28(1):51–8 [published Online First: 2015/07/07].

    PubMed  Article  Google Scholar 

  25. 25.

    Paradies YC, Cunningham J. Development and validation of the measure of indigenous racism experiences (MIRE). Int J Equity Health. 2008;7(1):9.

    PubMed  PubMed Central  Article  Google Scholar 

  26. 26.

    Jones R, Thurber KA, Chapman J, et al. Study protocol: our cultures count, the Mayi Kuwayu study, a national longitudinal study of Aboriginal and Torres Strait islander wellbeing. BMJ Open. 2018;8(6):e023861.

    PubMed  PubMed Central  Article  Google Scholar 

  27. 27.

    Morgado FF, Meireles JF, Neves CM, et al. Scale development: ten main limitations and recommendations to improve future research practices. Psicol Reflex Crit. 2018;30(1):3.

    Article  Google Scholar 

  28. 28.

    Field A. Discovering statistics using IBM SPSS statistics: and sex, drugs and rock 'n' roll. 4th ed. Singapore: Sage; 2013.

    Google Scholar 

  29. 29.

    Tabachnick BG, Fidell LS. Using Multivariate Statistics. 5th ed. Sydney: Allyn & Bacon/Pearson Education; 2007. p. 131.

    Google Scholar 

  30. 30.

    Boateng GO, Neilands TB, Frongillo EA, et al. Best practices for developing and validating scales for health, social, and behavioral research: a primer. Front Public Health. 2018;6:149.

    PubMed  PubMed Central  Article  Google Scholar 

  31. 31.

    Hu L, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Struct Equ Modeling. 1999;6(1):1–55.

    Article  Google Scholar 

  32. 32.

    Bentler PM. Comparative fit indexes in structural models. Psychol Bull. 1990;107(2):238.

    CAS  PubMed  Article  Google Scholar 

  33. 33.

    Maydeu-Olivares A, Shi D, Rosseel Y. Assessing fit in structural equation models: a Monte-Carlo evaluation of RMSEA versus SRMR confidence intervals and tests of close fit. Struct Equ Modeling. 2018;25(3):389–402.

    Article  Google Scholar 

  34. 34.

    Hays RD, Morales LS, Reise SP. Item response theory and health outcomes measurement in the 21st century. Med Care. 2000;38(9 Suppl):II28.

    CAS  PubMed  PubMed Central  Google Scholar 

  35. 35.

    Nguyen TH, Han H-R, Kim MT, et al. An introduction to item response theory for patient-reported outcome measurement. Patient. 2014;7(1):23–35.

    PubMed  PubMed Central  Article  Google Scholar 

  36. 36.

    Streiner DL. Starting at the beginning: an introduction to coefficient alpha and internal consistency. J Pers Assess. 2003;80(1):99–103.

    PubMed  Article  Google Scholar 

  37. 37.

    Cohen J. Statistical power analysis for the behavioural sciences. Hillsdale, N.J: L. Erlbaum Associates; 1988.

    Google Scholar 

  38. 38.

    Reeve BB, Willis G, Shariff-Marco SN, et al. Comparing Cognitive Interviewing and Psychometric Methods to Evaluate a Racial/Ethnic Discrimination Scale. Field Methods. 2011;23(4):397–419 [published Online First: 2011/11/01].

    PubMed  PubMed Central  Article  Google Scholar 

  39. 39.

    Henry BR, Houston S, Mooney GH. Institutional racism in Australian healthcare: a plea for decency. Med J Aust. 2004;180(10):517–20.

    PubMed  Article  Google Scholar 

Download references

Acknowledgements

The authors acknowledge all Aboriginal and Torres Strait Islander peoples and their continuing connection to culture, land, and seas. We would like to acknowledge and thank all contributors to the development of the Mayi Kuwayu Study, and all survey participants. We acknowledge the assistance and guidance of the Mayi Kuwayu Study Data Governance Committee, the Study Chief Investigators and Partners, and all members of the Mayi Kuwayu Study team.

Funding

KT (NHMRC Fellowship, ref.: 1156276), RL (NHMRC Fellowship, ref.: 1122273), NP (NHMRC Fellowship, ref.: 1123677), PB (NHMRC Fellowship, ref.: 1158707), AC (NHMRC Fellowship, ref.: 1122544 & 1173146), RJ (NHMRC Postgraduate Scholarship, ref.: 1189913).

Author information

Affiliations

Authors

Contributions

KT, RL conceived the idea for the manuscript. RL, JC conducted the focus groups. RL, JC, KT, RJ, NP, DW, GG contributed to instrument development. KT, JW conducted the data analysis. KT, RC, RJ conducted the literature review. PB, AR provided statistical advice. KT produced the initial draft of the manuscript. All authors were involved in interpreting results and drafting the manuscript. All authors approved the final version of the paper and take responsibility for its content.

Corresponding author

Correspondence to Katherine A. Thurber.

Ethics declarations

Ethics approval and consent to participate

Participation in the Mayi Kuwayu Study was voluntary and with informed consent. The Mayi Kuwayu Study is Aboriginal-led and governed, and conducted with ethics approvals from national, State and Territory Human Research Ethics Committees (HRECs) and from relevant Aboriginal and Torres Strait Islander organisations [26]. This validation study was conducted under The Australian National University HREC protocol 2016/767, and in accordance with the operational research policies of the Mayi Kuwayu Study Data Governance Committee. Early results and future directions were discussed with the Thiitu Tharrmay Research Reference Group. Aboriginal and/or Torres Strait Islander peoples were involved in every stage of the current study, from the development of the measures, the analysis, interpretation, and write up of results.

Consent for publication

Not applicable.

Competing interests

The authors declare we have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Thurber, K.A., Walker, J., Batterham, P.J. et al. Developing and validating measures of self-reported everyday and healthcare discrimination for Aboriginal and Torres Strait Islander adults. Int J Equity Health 20, 14 (2021). https://doi.org/10.1186/s12939-020-01351-9

Download citation

Keywords

  • Health inequalities
  • Measurement
  • Measurement tool development
  • Psychometrics
  • Social Epidemitology