Development and validation of a theory based gender norms scale (The GNS)
Summary:
Gender norms are increasingly recognized as important modifiers of behavior change interventions. Despite awareness of how gender norms shape behavioral expectations within a specific social context, current gender norms scales measure individual attitudes instead of beliefs about the social context. Existing scales do not incorporate key theoretical constructs of social norms theory (e.g., reference groups, descriptive, and injunctive norms) shown to influence health behaviors. We used a mixed-methods approach to develop and validate a novel gender norms scale in Odisha, India. To ensure that the items that we created were relevant to the population, we conducted qualitative research in the community (16 focus groups and 25 key informant interviews). We also used two theories to identify items, the theory of gender and power, and the theory of normative social behavior, conducted cognitive testing, and piloted the items in the community. These steps generated a pool of 28 items, which we administered in a survey among 3,110 women. We evaluated psychometric properties using exploratory factor analysis and structural equation modeling. The analysis resulted in two sub scales with one factor each, descriptive gender norms and injunctive gender norms. Both subscales represent high internal validity with polychoric alpha values of .84 and .85 respectively. The GNS, gender norms scale, improves on previous gender norms scales in three important ways. We include both descriptive and injunctive norms, we move beyond aggregating individual attitudes to measuring perceptions of community-level norms, and we expand on past sub domains included in gender norms measurement.
Background/Objectives:
Gender norms are increasingly recognized as important modifiers of health behavior change interventions. Despite growing awareness of how gender norms shape behavioral expectations within a specific social context, current gender norms scales measure individual attitudes instead of beliefs about the social context. Existing scales do not adequately incorporate the key theoretical constructs of social norms theory (e.g., reference groups, descriptive, and injunctive norms) shown to influence health behaviors. This study shows the development of a gender norms scale that identifies a community-level referent group, distinguishes between descriptive and injunctive norms, and includes a broader set of items that represent gender.
Description of Intervention and/or Methods/Design:
We developed the GNS, gender norms scale, in four stages: (1) formulation of survey items for the scale, (2) survey administration, (3) psychometric analysis, and (4) scale validation. To ensure that the items we developed were relevant, we conducted 16 focus groups and 25 key informant interviews. Based on qualitative findings, cognitive testing, and drawing on two theories, the theory of gender and power and the Theory of Normative Social Behavior (TNSB), we generated a pool of 28 items. We then administered these items in a survey among 3,110 women in Odisha, India. To analyze the psychometric properties of the items, we ran exploratory factor analysis. Based on the Theory of Normative Social Behavior, we separated the descriptive norms and the injunctive norms items during factor analysis to allow for separate sub scales. Finally, to test that the model fits the data well, we used structural equation modeling.
Results/Lessons Learned:
The analysis resulted in two sub scales with one factor each, descriptive gender norms and injunctive gender norms. Both subscales represent high internal validity with polychoric alpha values of .84 and .85 respectively. Each subscale contains 10 items that are mirrored descriptive and injunctive norms questions. For example, in the descriptive norms item, In most families you know, taking care of children is only the woman's job is mirrored in the injunctive scale, Most families you know believe that it should only be the woman's job to take care of the children. The two subscales were significantly correlated (r =.72 p < .001) but also distinct. To demonstrate criterion validity, we showed that inequitable gender norms were positively associated with attitudes about the acceptability of violence against women in both subscales (r =.09, p < .001) (r =.14, p < .001).
Discussion/Implications for the Field:
This scale improves on previous scales in three important ways. We include both descriptive and injunctive norms, we move beyond aggregating individual attitudes to measuring perceptions of community-level norms, and we expand on past sub domains included in gender norms measurement. Additionally, this is a parsimonious scale as researchers can use one sub scale, 10 items, or both sub scales, 20 items, to measure the complex construct of gender norms. It is critical to improve gender norms measurement to bring the world closer to sustainable development goal (SDG) #5, achieve gender equality and empower all women and girls.
Abstract submitted by:
Erica Sedlander - George Washington University
Jeffrey Bingenheimer - George Washington University
Michael Long - George Washington University
Minati Swain - Dcor Development Corner
Nadia Diamond-Smith - University of California San Francisco
Rajiv Rimal - Johns Hopkins University
Gender norms are increasingly recognized as important modifiers of behavior change interventions. Despite awareness of how gender norms shape behavioral expectations within a specific social context, current gender norms scales measure individual attitudes instead of beliefs about the social context. Existing scales do not incorporate key theoretical constructs of social norms theory (e.g., reference groups, descriptive, and injunctive norms) shown to influence health behaviors. We used a mixed-methods approach to develop and validate a novel gender norms scale in Odisha, India. To ensure that the items that we created were relevant to the population, we conducted qualitative research in the community (16 focus groups and 25 key informant interviews). We also used two theories to identify items, the theory of gender and power, and the theory of normative social behavior, conducted cognitive testing, and piloted the items in the community. These steps generated a pool of 28 items, which we administered in a survey among 3,110 women. We evaluated psychometric properties using exploratory factor analysis and structural equation modeling. The analysis resulted in two sub scales with one factor each, descriptive gender norms and injunctive gender norms. Both subscales represent high internal validity with polychoric alpha values of .84 and .85 respectively. The GNS, gender norms scale, improves on previous gender norms scales in three important ways. We include both descriptive and injunctive norms, we move beyond aggregating individual attitudes to measuring perceptions of community-level norms, and we expand on past sub domains included in gender norms measurement.
Background/Objectives:
Gender norms are increasingly recognized as important modifiers of health behavior change interventions. Despite growing awareness of how gender norms shape behavioral expectations within a specific social context, current gender norms scales measure individual attitudes instead of beliefs about the social context. Existing scales do not adequately incorporate the key theoretical constructs of social norms theory (e.g., reference groups, descriptive, and injunctive norms) shown to influence health behaviors. This study shows the development of a gender norms scale that identifies a community-level referent group, distinguishes between descriptive and injunctive norms, and includes a broader set of items that represent gender.
Description of Intervention and/or Methods/Design:
We developed the GNS, gender norms scale, in four stages: (1) formulation of survey items for the scale, (2) survey administration, (3) psychometric analysis, and (4) scale validation. To ensure that the items we developed were relevant, we conducted 16 focus groups and 25 key informant interviews. Based on qualitative findings, cognitive testing, and drawing on two theories, the theory of gender and power and the Theory of Normative Social Behavior (TNSB), we generated a pool of 28 items. We then administered these items in a survey among 3,110 women in Odisha, India. To analyze the psychometric properties of the items, we ran exploratory factor analysis. Based on the Theory of Normative Social Behavior, we separated the descriptive norms and the injunctive norms items during factor analysis to allow for separate sub scales. Finally, to test that the model fits the data well, we used structural equation modeling.
Results/Lessons Learned:
The analysis resulted in two sub scales with one factor each, descriptive gender norms and injunctive gender norms. Both subscales represent high internal validity with polychoric alpha values of .84 and .85 respectively. Each subscale contains 10 items that are mirrored descriptive and injunctive norms questions. For example, in the descriptive norms item, In most families you know, taking care of children is only the woman's job is mirrored in the injunctive scale, Most families you know believe that it should only be the woman's job to take care of the children. The two subscales were significantly correlated (r =.72 p < .001) but also distinct. To demonstrate criterion validity, we showed that inequitable gender norms were positively associated with attitudes about the acceptability of violence against women in both subscales (r =.09, p < .001) (r =.14, p < .001).
Discussion/Implications for the Field:
This scale improves on previous scales in three important ways. We include both descriptive and injunctive norms, we move beyond aggregating individual attitudes to measuring perceptions of community-level norms, and we expand on past sub domains included in gender norms measurement. Additionally, this is a parsimonious scale as researchers can use one sub scale, 10 items, or both sub scales, 20 items, to measure the complex construct of gender norms. It is critical to improve gender norms measurement to bring the world closer to sustainable development goal (SDG) #5, achieve gender equality and empower all women and girls.
Abstract submitted by:
Erica Sedlander - George Washington University
Jeffrey Bingenheimer - George Washington University
Michael Long - George Washington University
Minati Swain - Dcor Development Corner
Nadia Diamond-Smith - University of California San Francisco
Rajiv Rimal - Johns Hopkins University
Source
Approved abstract for the postponed 2020 SBCC Summit in Marrakech, Morocco. Provided by the International Steering Committee for the Summit. Image credit: Harsha K R from Bangalore, India via Wikimedia (Creative Commons Attribution-Share Alike 2.0 Generic license)