Setting weights for fifteen CHNRI criteria at the global and regional level using public stakeholders: an Amazon Mechanical Turk study.

Authors:
Kerri Wazny
Kerri Wazny
the Hospital for Sick Children
Canada
John Ravenscroft
John Ravenscroft
UniversityS. Environmental Protection Agency
Kit Yee Chan
Kit Yee Chan
The University of Edinburgh Medical School
United Kingdom
Diego G Bassani
Diego G Bassani
Li Ka Shing Knowledge Institute
Canada
Niall Anderson
Niall Anderson
University of Edinburgh
United Kingdom
Igor Rudan
Igor Rudan
Usher Institute of Population Health Sciences and Informatics
United Kingdom

J Glob Health 2019 Jun;9(1):010702

Centre for Global Health Research, Usher Institute of Population Health Sciences and Informatics, University of Edinburgh, Edinburgh, UK.

Introduction: Stakeholder involvement has been described as an indispensable part of health research priority setting. Yet, more than 75% of the exercises using the Child Health and Nutrition Research Initiative (CHNRI) methodology have omitted the step involving stakeholders in priority setting. Those that have used stakeholders have rarely used the public, possibly due to the difficulty of assembling and/or accessing a public stakeholder group. In order to strengthen future exercises using the CHNRI methodology, we have used a public stakeholder group to weight 15 CHNRI criteria, and have explored regional differences or being a health stakeholder is influential, and whether the criteria are collapsible.

Methods: Using Amazon Mechanical Turk (AMT), an online crowdsourcing platform, we collected demographic information and conducted a Likert-scale format survey about the importance of the CHNRI criteria from 1051 stakeholders. The Kruskal-Wallis test, with Dunn's test for posthoc comparisons, was used to examine regional differences and Wilcoxon rank-sum test was used to analyse differences between stakeholders with health training/background and stakeholders without a health background and by region. A Factor Analysis (FA) was conducted on the criteria to identify the main domains connecting them. Criteria means were converted to weights.

Results: There were regional differences in thirteen of fifteen criteria according to the Kruskal-Wallis test and differences in responses from health stakeholders vs those who were not in eleven of fifteen criteria using the Wilcoxon rank-sum test. Three components were identified: improve and impact results; implementation and affordability; and, study design and dissemination. A formula is provided to convert means to weights for future studies.

Conclusion: In future CHNRI studies, researchers will need to ensure adequate representation from stakeholders to undue bias of CHNRI results. These results should be used in combination with other stakeholder groups, including government, donors, policy makers, and bilateral agencies. Global and regional stakeholder groups scored CHNRI criteria differently; due to this, researchers should consider which group to use in their CHNRI exercises.

Download full-text PDF

Source
http://dx.doi.org/10.7189/jogh.09.010702DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC6445564PMC
June 2019
1 Read

Publication Analysis

Top Keywords

chnri criteria
16
regional differences
12
criteria
9
chnri
9
mechanical turk
8
rank-sum test
8
stakeholders
8
stakeholder groups
8
kruskal-wallis test
8
global regional
8
priority setting
8
chnri methodology
8
fifteen criteria
8
public stakeholder
8
stakeholder group
8
amazon mechanical
8
wilcoxon rank-sum
8
stakeholders health
8
health
6
stakeholder
6

Similar Publications