|
Stigma Measures |
Measuring Stigma Excerpted from Evidence-based Generic Tools for Operational Research on HIV, Carla Makhlouf Obermeyer (ed.), 2008 (forthcoming), Geneva: World Health Organization. Most of the items presented in this toolkit are designed to be used with Likert-style response categories. Commonly, studies using these items have utilized three-point or five-point response categories. Five-point response categories include: strongly agree, agree, neither agree nor disagree/not sure, disagree, and strongly disagree. Using five options enhances the variability of responses and permits a greater understanding of the nuances in stigma. However, five options can be difficult for survey respondents to respond to, particularly for respondents with limited literacy skills. In these circumstances, using a three-point response scale can be a good alternative. The three-point scale includes: agree, partially agree/not sure, and disagree. This option permits more variability and nuance than a simple yes/no response, but is more intuitive for respondents and easier to administer than the five-point option. It is recommended that multiple survey items be used to measure each dimension of stigma in a population, and then combined into an index. Using a greater number of items is generally preferable, since each item may potentially capture important nuances in attitudes and behaviors. In an index, responses for each item are summed together to form a stigma score. Items can be summed separately for each dimension of stigma, or all the items can be combined into an overall stigma score, depending upon the goals of the research. Alternatively, the items can be analyzed individually to explore specific elements of stigma. It should be noted that, while most of the survey items presented in this section are framed negatively (e.g., “HIV is a punishment for bad behavior”), some are framed positively (e.g., “clients who are sex workers deserve to receive the same level and quality of care as other clients”). During the creation of an index, researchers must be sure to code items in the same direction, such that both negatively framed items and positively framed items have higher scores for more highly stigmatizing responses. Additionally, given that stigma measures are still being refined and tested, it is always a good idea to pre-test a survey instrument in local contexts. Pre-testing allows researchers to improve the appropriateness and relevance of survey items to local contexts. Comparative research across different contexts is also useful to contribute to identifying core quantitative measures, and in turn help program managers design better programs to help reduce HIV-related stigma and its related negative health outcomes.
|
||
|
Measures for inappropriate fear of contagion and resulting avoidance of people living with HIV |
|||
|
How to use the survey items |
|||
|
Working Report Measuring HIV Stigma: Results of a Field Test in Tanzania |
|||