Article Text

Download PDFPDF

Cutting edge methodology
P1-10 Development and psychometric properties of a scale to measure health effects of discrimination
  1. J Bastos1,
  2. E Faerstein2,
  3. R Celeste3,
  4. A Barros1
  1. 1Federal University of Santa Catarina, Florianopolis, Santa Catarina, Southern Brazil, Brazil
  2. 2State University of Rio de Janeiro, Rio de Janeiro, Southeastern Brasil, Brazil
  3. 3Federal University of Rio Grande do Sul, Porto Alegre, Rio Grande do Sul, Southern Brazil, Brazil
  4. 4Federal University of Pelotas, Pelotas, Rio Grande do Sul, Southern Brazil, Brazil


Introduction The development of discrimination scales is an emerging field of enquiry in the area of social determinants of health. However, published scales cannot distinguish health consequences of discrimination as a result of the exposure to differential treatments of any kind from the strict attribution of these events to discrimination. We report the development of a scale that may clarify the relative importance of the effects of these cognitive mechanisms for health outcomes.

Methods Successive versions of the instrument were developed based on a systematic review of racial discrimination scales, focus groups and an evaluation by a panel of experts. Instrument refinement was achieved via cognitive interviews and pilot-testing, so that the final scale version was administered to 424 university students in Rio de Janeiro, Brazil. Structural dimensionality, two types of reliability and construct validity were assessed.

Results Exploratory factor analysis corroborated the hypothesis of unidimensionality, and the experts indicated that scale items were face and content valid. Internal consistency, estimated by Cronbach's α, was 0.8, and test-retest reliability was higher than 0.5 for 14 out of the 18 items, according to the weighted κ statistics. The scale's score was statistically higher among socially disadvantaged individuals and correlated with adverse health behaviours and outcomes. Nevertheless, the low test-retest reliability and the observed invariance of specific items indicate that the scale should be assessed in other population domains.

Conclusion The proposed scale will enable the investigation of aspects of the relationship between discrimination and health not previously documented in the literature.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.