Modernizing a comprehensive measure of genetic literacy
Health Services and Implementation
-
Primary Categories:
- General Education
-
Secondary Categories:
- General Education
Introduction:
Measuring genetic literacy (GL) is a complex task, and current methods to quantify it lack modernity and consistency. This is of particular importance as research shows the United States general population is lacking critical GL skills. Higher GL levels, which correlate with increased confidence and autonomy in medical decision-making, are needed as genetic and genomic medicine moves to the forefront of care. The Genetic Literacy Survey (GLS), developed by the NHGRI’s Social Behavioral Research Branch, is the only commonly used GL measure to include knowledge comprehension as a cognitive domain, despite this domain generally being cited as an important component of GL. The GLS debuted in 2013, and has run longitudinally in several iterations since then, accumulating more than 6000 participant responses. Over time, questions have been updated to accommodate new research understandings, language preferences, and additional subscales. For example, as genetics education shifts from Mendelian concepts to more complex conditions and our understanding of gene x environment models improves, new questions are needed to accurately gauge GL levels today. Our measure has previously illustrated that a sample involved in genetic research had only slightly higher GL scores than a general population, and that several social and psychological factors have direct bearing on GL scores, including education, political ideology, and religious beliefs.
Methods:
After updating the GLS measure to reflect the rapid advancement of genomic research while prioritizing inclusivity and community preferences, we validated the survey using a mixed-methods approach and a nationally representative general population sample of 1005 individuals. We performed Confirmatory Factor Analysis (CFA), a statistical test used to see if a set of observed variables accurately reflect and therefore measure underlying latent variables, on the entire measure to seek best model fit. All five subscales of the measure – Subjective Knowledge, Objective Knowledge, Applied Knowledge, Situational Knowledge, and Knowledge Comprehension – were examined for model fit and covariances through CFA. We then examined correlations between variables of interest and identified questions with low or negative holdings. Finally, we analyzed nearly 1000 responses to an open-ended question about the most important genetics concepts for this population.
Results:
Our CFA results indicate a good measure of model fit, with discrete subscales and individual questions that accurately measure those latent variables of genetic literacy. The knowledge comprehension section was the most distinct and well-fitting subscale, solidifying the utility of that subscale as an integral component of genetic literacy. While overall GLS achieved a successful model of fit per CFA substrates [Comparative Fit Index (CFI) = 0.936 and Root Mean Square Error of Approximation (RMSEA) = 0.019], six questions had low or negative holdings, indicating less accurate measurements of underlying latent variables and thereby prompting room for revision. Qualitative coding for the question “What is the most important idea or fact about genetics that you feel you need to know?” revealed a high prevalence of responses centered on hereditary diseases and family history.
Conclusion:
The modernized, validated GLS is available for use by any interested collaborators. It is already serving as the basis for comparisons of urban and rural GL levels and a study of the relationships between GL and information integrity, as well as pointing to the most requested educational interventions in genetics. Given the more comprehensive range of knowledge domains tested, GLS can serve as a standardizing instrument for measuring GL, in keeping with commonly used definitions of the term. Areas of strength or weakness identified by the GLS are also facilitating more precise development of relevant educational materials for multiple audiences.
Measuring genetic literacy (GL) is a complex task, and current methods to quantify it lack modernity and consistency. This is of particular importance as research shows the United States general population is lacking critical GL skills. Higher GL levels, which correlate with increased confidence and autonomy in medical decision-making, are needed as genetic and genomic medicine moves to the forefront of care. The Genetic Literacy Survey (GLS), developed by the NHGRI’s Social Behavioral Research Branch, is the only commonly used GL measure to include knowledge comprehension as a cognitive domain, despite this domain generally being cited as an important component of GL. The GLS debuted in 2013, and has run longitudinally in several iterations since then, accumulating more than 6000 participant responses. Over time, questions have been updated to accommodate new research understandings, language preferences, and additional subscales. For example, as genetics education shifts from Mendelian concepts to more complex conditions and our understanding of gene x environment models improves, new questions are needed to accurately gauge GL levels today. Our measure has previously illustrated that a sample involved in genetic research had only slightly higher GL scores than a general population, and that several social and psychological factors have direct bearing on GL scores, including education, political ideology, and religious beliefs.
Methods:
After updating the GLS measure to reflect the rapid advancement of genomic research while prioritizing inclusivity and community preferences, we validated the survey using a mixed-methods approach and a nationally representative general population sample of 1005 individuals. We performed Confirmatory Factor Analysis (CFA), a statistical test used to see if a set of observed variables accurately reflect and therefore measure underlying latent variables, on the entire measure to seek best model fit. All five subscales of the measure – Subjective Knowledge, Objective Knowledge, Applied Knowledge, Situational Knowledge, and Knowledge Comprehension – were examined for model fit and covariances through CFA. We then examined correlations between variables of interest and identified questions with low or negative holdings. Finally, we analyzed nearly 1000 responses to an open-ended question about the most important genetics concepts for this population.
Results:
Our CFA results indicate a good measure of model fit, with discrete subscales and individual questions that accurately measure those latent variables of genetic literacy. The knowledge comprehension section was the most distinct and well-fitting subscale, solidifying the utility of that subscale as an integral component of genetic literacy. While overall GLS achieved a successful model of fit per CFA substrates [Comparative Fit Index (CFI) = 0.936 and Root Mean Square Error of Approximation (RMSEA) = 0.019], six questions had low or negative holdings, indicating less accurate measurements of underlying latent variables and thereby prompting room for revision. Qualitative coding for the question “What is the most important idea or fact about genetics that you feel you need to know?” revealed a high prevalence of responses centered on hereditary diseases and family history.
Conclusion:
The modernized, validated GLS is available for use by any interested collaborators. It is already serving as the basis for comparisons of urban and rural GL levels and a study of the relationships between GL and information integrity, as well as pointing to the most requested educational interventions in genetics. Given the more comprehensive range of knowledge domains tested, GLS can serve as a standardizing instrument for measuring GL, in keeping with commonly used definitions of the term. Areas of strength or weakness identified by the GLS are also facilitating more precise development of relevant educational materials for multiple audiences.