Exploring Objective and Subjective Numeracy at a Population Level: Findings From the 2007 Health Information National Trends Survey (HINTS)

Abstract

Numeracy is a critical component of decision making in health, and low numeracy may adversely affect risk comprehension, medical treatment, and health outcomes. Health researchers have developed objective and subjective measures of numeracy that potentially could be used for clinical or health research purposes. To examine the association between objective and subjective numeracy at a population level, data were obtained from the 2007 Health Information National Trends Survey, a health communication survey that collects data from a nationally representative sample of the U.S. population. Associations between items from 2 published subjective numeracy measures (STAT-Confidence Scale; Subjective Numeracy Scale) and a single published objective numeracy measure were examined with respect to one another and with respect to sociodemographic characteristics using a multivariate logistic regression model. Controlling for demographic covariates, both subjective numeracy measures were significantly associated with each other (p < .0001) and with objective numeracy (p < .0001). Compared with respondents who regarded themselves as low in subjective numeracy, those who regarded themselves as high in subjective numeracy had significantly higher odds of answering the objective numeracy question correctly (STAT-Confidence Scale OR = 1.42, CI [1.07, 1.90]; Subjective Numeracy Scale OR = 1.71, CI [1.28, 2.28]). However, the subjective measures performed poorly as diagnostic indicators of objective numeracy. Sensitivity and specificity for the STAT-Confidence Scale and the Subjective Numeracy Scale were 67%/49% and 50%/71%, respectively. More work is needed to elucidate the relation between objective and subjective numeracy, at the conceptual and empirical levels, and to develop more robust measures of subjective numeracy that can better discriminate between individuals with low and high objective numeracy.

Acknowledgments

The authors gratefully acknowledge and thank William Waldron for his statistical expertise.

Notes

^a Percentage of respondents who answered the objective numeracy question correctly.

^b Chi-square p value.

^c Percentage of respondents high in subjective numeracy (i.e., found it “easy” or “very easy” to understand medical statistics).

^d Percentage of respondents high in subjective numeracy (i.e., responded “generally prefer numbers”).

This article not subject to US copyright law.