Examining the Relationship Between Flexible Resources and Health Information Channel Selection

Abstract

This study examines how variations in flexible resources influence where individuals begin their search for health information. Access to flexible resources such as money, power, and knowledge can alter the accessibility of channels for health information, such as doctors, the Internet, and print media. Using the HINTS 3 sample, whether information channel utilization is predicted by the same factors in two groups with distinct levels of access to flexible resources, as approximated by high and low levels of education, is investigated. Differences in access to flexible resources are hypothesized to produce variations in channel utilization in bivariate analyses, as well as changes in coefficient strength and statistical significance in multivariate models. Multinomial logit models were used to assess how a number of variables influence the probability of using a specific information channel first in either flexible resource group. Results suggest that individuals with higher levels of education, a proxy for flexible resources, are more likely to report seeking information from the Internet first, which is consistent with research on the digital divide. It appears that diminished access to flexible resources is also associated with heightened utilization of offline channels, including doctors. A handful of differences in predictors were found between the low and high flexible resource groups when multivariate models were compared. Future research should take into account the distinctions between different offline channels while also seeking to further understand how social inequality relates to the utilization of different channels and corresponding health outcomes.

Notes

¹ Initially distinctions were made based on if a respondent had dial-up, a cable modem, or an “other” connection to the Internet such as a phone at home. However, the rarity of data for people with “other” Internet sources in the data meant that it was not possible to keep this distinction while also dividing the sample based on FR. Furthermore, running the model without disaggregating by FR revealed that all the Internet sources had nearly identical coefficient strength, suggesting that what matters most is having any access to Internet at one’s home, rather than a faster connection.