Measuring the bias against low-income country research

An Implicit Association Test

Matthew Harris, James Macinko, Geronimo Jimenez, Pricila Mullachery

Research output: Contribution to journalArticle

Abstract

Background: With an increasing array of innovations and research emerging from low-income countries there is a growing recognition that even high-income countries could learn from these contexts. It is well known that the source of a product influences perception of that product, but little research has examined whether this applies also in evidence-based medicine and decision-making. In order to examine likely barriers to learning from low-income countries, this study uses established methods in cognitive psychology to explore whether healthcare professionals and researchers implicitly associate good research with rich countries more so than with poor countries. Methods: Computer-based Implicit Association Test (IAT) distributed to healthcare professionals and researchers. Stimuli representing Rich Countries were chosen from OECD members in the top ten (>$36,000 per capita) World Bank rankings and Poor Countries were chosen from the bottom thirty (<$1000 per capita) countries by GDP per capita, in both cases giving attention to regional representation. Stimuli representing Research were descriptors of the motivation (objective/biased), value (useful/worthless), clarity (precise/vague), process (transparent/dishonest), and trustworthiness (credible/unreliable) of research. IAT results are presented as a Cohen's d statistic. Quantile regression was used to assess the contribution of covariates (e.g. age, sex, country of origin) to different values of IAT responses that correspond to different levels of implicit bias. Poisson regression was used to model dichotomized responses to the explicit bias item. Results: Three hundred twenty one tests were completed in a four-week period between March and April 2015. The mean Implicit Association Test result (a standardized mean relative latency between congruent and non-congruent categories) for the sample was 0.57 (95% CI 0.52 to 0.61) indicating that on average our sample exhibited moderately strong implicit associations between Rich Countries and Good Research. People over 40 years of age were less likely to exhibit pro-poor implicit associations, and being a peer reviewer contributes to a more pro-poor association. Conclusions: The majority of our participants associate Good Research with Rich Countries, compared to Poor Countries. Implicit associations such as these might disfavor research from poor countries in research evaluation, evidence-based medicine and diffusion of innovations.

Original languageEnglish (US)
Article number80
JournalGlobalization and Health
Volume13
Issue number1
DOIs
StatePublished - Nov 6 2017

Fingerprint

Research
Evidence-Based Medicine
Diffusion of Innovation
Research Personnel
Delivery of Health Care
United Nations
Motivation
Decision Making
Learning
Psychology

Keywords

  • Bias
  • Implicit association test
  • Research evaluation
  • Reverse innovation
  • Stereotypes

ASJC Scopus subject areas

  • Health Policy
  • Public Health, Environmental and Occupational Health

Cite this

Measuring the bias against low-income country research : An Implicit Association Test. / Harris, Matthew; Macinko, James; Jimenez, Geronimo; Mullachery, Pricila.

In: Globalization and Health, Vol. 13, No. 1, 80, 06.11.2017.

Research output: Contribution to journalArticle

Harris, Matthew ; Macinko, James ; Jimenez, Geronimo ; Mullachery, Pricila. / Measuring the bias against low-income country research : An Implicit Association Test. In: Globalization and Health. 2017 ; Vol. 13, No. 1.
@article{d5ca05bc0a6d424daed60d82f44bf1c2,
title = "Measuring the bias against low-income country research: An Implicit Association Test",
abstract = "Background: With an increasing array of innovations and research emerging from low-income countries there is a growing recognition that even high-income countries could learn from these contexts. It is well known that the source of a product influences perception of that product, but little research has examined whether this applies also in evidence-based medicine and decision-making. In order to examine likely barriers to learning from low-income countries, this study uses established methods in cognitive psychology to explore whether healthcare professionals and researchers implicitly associate good research with rich countries more so than with poor countries. Methods: Computer-based Implicit Association Test (IAT) distributed to healthcare professionals and researchers. Stimuli representing Rich Countries were chosen from OECD members in the top ten (>$36,000 per capita) World Bank rankings and Poor Countries were chosen from the bottom thirty (<$1000 per capita) countries by GDP per capita, in both cases giving attention to regional representation. Stimuli representing Research were descriptors of the motivation (objective/biased), value (useful/worthless), clarity (precise/vague), process (transparent/dishonest), and trustworthiness (credible/unreliable) of research. IAT results are presented as a Cohen's d statistic. Quantile regression was used to assess the contribution of covariates (e.g. age, sex, country of origin) to different values of IAT responses that correspond to different levels of implicit bias. Poisson regression was used to model dichotomized responses to the explicit bias item. Results: Three hundred twenty one tests were completed in a four-week period between March and April 2015. The mean Implicit Association Test result (a standardized mean relative latency between congruent and non-congruent categories) for the sample was 0.57 (95{\%} CI 0.52 to 0.61) indicating that on average our sample exhibited moderately strong implicit associations between Rich Countries and Good Research. People over 40 years of age were less likely to exhibit pro-poor implicit associations, and being a peer reviewer contributes to a more pro-poor association. Conclusions: The majority of our participants associate Good Research with Rich Countries, compared to Poor Countries. Implicit associations such as these might disfavor research from poor countries in research evaluation, evidence-based medicine and diffusion of innovations.",
keywords = "Bias, Implicit association test, Research evaluation, Reverse innovation, Stereotypes",
author = "Matthew Harris and James Macinko and Geronimo Jimenez and Pricila Mullachery",
year = "2017",
month = "11",
day = "6",
doi = "10.1186/s12992-017-0304-y",
language = "English (US)",
volume = "13",
journal = "Globalization and Health",
issn = "1744-8603",
number = "1",

}

TY - JOUR

T1 - Measuring the bias against low-income country research

T2 - An Implicit Association Test

AU - Harris, Matthew

AU - Macinko, James

AU - Jimenez, Geronimo

AU - Mullachery, Pricila

PY - 2017/11/6

Y1 - 2017/11/6

N2 - Background: With an increasing array of innovations and research emerging from low-income countries there is a growing recognition that even high-income countries could learn from these contexts. It is well known that the source of a product influences perception of that product, but little research has examined whether this applies also in evidence-based medicine and decision-making. In order to examine likely barriers to learning from low-income countries, this study uses established methods in cognitive psychology to explore whether healthcare professionals and researchers implicitly associate good research with rich countries more so than with poor countries. Methods: Computer-based Implicit Association Test (IAT) distributed to healthcare professionals and researchers. Stimuli representing Rich Countries were chosen from OECD members in the top ten (>$36,000 per capita) World Bank rankings and Poor Countries were chosen from the bottom thirty (<$1000 per capita) countries by GDP per capita, in both cases giving attention to regional representation. Stimuli representing Research were descriptors of the motivation (objective/biased), value (useful/worthless), clarity (precise/vague), process (transparent/dishonest), and trustworthiness (credible/unreliable) of research. IAT results are presented as a Cohen's d statistic. Quantile regression was used to assess the contribution of covariates (e.g. age, sex, country of origin) to different values of IAT responses that correspond to different levels of implicit bias. Poisson regression was used to model dichotomized responses to the explicit bias item. Results: Three hundred twenty one tests were completed in a four-week period between March and April 2015. The mean Implicit Association Test result (a standardized mean relative latency between congruent and non-congruent categories) for the sample was 0.57 (95% CI 0.52 to 0.61) indicating that on average our sample exhibited moderately strong implicit associations between Rich Countries and Good Research. People over 40 years of age were less likely to exhibit pro-poor implicit associations, and being a peer reviewer contributes to a more pro-poor association. Conclusions: The majority of our participants associate Good Research with Rich Countries, compared to Poor Countries. Implicit associations such as these might disfavor research from poor countries in research evaluation, evidence-based medicine and diffusion of innovations.

AB - Background: With an increasing array of innovations and research emerging from low-income countries there is a growing recognition that even high-income countries could learn from these contexts. It is well known that the source of a product influences perception of that product, but little research has examined whether this applies also in evidence-based medicine and decision-making. In order to examine likely barriers to learning from low-income countries, this study uses established methods in cognitive psychology to explore whether healthcare professionals and researchers implicitly associate good research with rich countries more so than with poor countries. Methods: Computer-based Implicit Association Test (IAT) distributed to healthcare professionals and researchers. Stimuli representing Rich Countries were chosen from OECD members in the top ten (>$36,000 per capita) World Bank rankings and Poor Countries were chosen from the bottom thirty (<$1000 per capita) countries by GDP per capita, in both cases giving attention to regional representation. Stimuli representing Research were descriptors of the motivation (objective/biased), value (useful/worthless), clarity (precise/vague), process (transparent/dishonest), and trustworthiness (credible/unreliable) of research. IAT results are presented as a Cohen's d statistic. Quantile regression was used to assess the contribution of covariates (e.g. age, sex, country of origin) to different values of IAT responses that correspond to different levels of implicit bias. Poisson regression was used to model dichotomized responses to the explicit bias item. Results: Three hundred twenty one tests were completed in a four-week period between March and April 2015. The mean Implicit Association Test result (a standardized mean relative latency between congruent and non-congruent categories) for the sample was 0.57 (95% CI 0.52 to 0.61) indicating that on average our sample exhibited moderately strong implicit associations between Rich Countries and Good Research. People over 40 years of age were less likely to exhibit pro-poor implicit associations, and being a peer reviewer contributes to a more pro-poor association. Conclusions: The majority of our participants associate Good Research with Rich Countries, compared to Poor Countries. Implicit associations such as these might disfavor research from poor countries in research evaluation, evidence-based medicine and diffusion of innovations.

KW - Bias

KW - Implicit association test

KW - Research evaluation

KW - Reverse innovation

KW - Stereotypes

UR - http://www.scopus.com/inward/record.url?scp=85032939852&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85032939852&partnerID=8YFLogxK

U2 - 10.1186/s12992-017-0304-y

DO - 10.1186/s12992-017-0304-y

M3 - Article

VL - 13

JO - Globalization and Health

JF - Globalization and Health

SN - 1744-8603

IS - 1

M1 - 80

ER -