Understanding users of a freely-available online health risk assessment
: An exploration using segmentation

  • Corinne Hodgson

Student thesis: Doctoral ThesisDoctor of Health (DHealth)


Health organizations and governments are investing considerable resources into Internet-based health promotion. There is a large and growing body of research on health “etools” but to date most has been conducted using experimental paradigms; much less is known about those that are freely-available.Analysis was conducted of the data base generated through the operation of the freely-available health risk assessment (HRA) of the Heart and Stroke Foundation of Ontario. During the study period of February 1 to December 20, 2011, 147,274 HRAs were completed, of which 120,510 (79.8%) included consent for the use of information for research and were completed by adults aged 18 to 90 years.Comparison of Canadian users to national statistics confirmed that the HRA sample is not representative of the general population. The HRA sample is significantly and systematically biased by gender, education, employment, heath behaviours, and the prevalence of specific chronic diseases. Etool users may be a large but select segment of the population, those previously described as “Internet health information seekers.”Are all Internet health information seekers the same? To explore this issue, segmentation procedures available in common commercial packages (k-means clustering, two-step clustering, and latent class analysis) were conducted using five combinations of variables. Ten statistically significant solutions were created. The most robust solution divided the sample into four groups differentiated by age (two younger and two older groups) and healthiness, as reflected by disease and modifiable risk factor burden and readiness to make lifestyle changes. These groups suggest that while all users of online health etools may be health information seekers, they vary in the extent to which they are health oriented or health conscientious (i.e., engaging in preventive health behaviours or ready for behaviour change). It is hoped that this research will provide other organizations with similar data bases with a model for analyzing their client populations, therefore increasing our knowledge about health etool users.
Date of Award12 Mar 2015
Original languageEnglish
Awarding Institution
  • University of Bath
SupervisorGordon Taylor (Supervisor)


  • Internet use
  • health assessment
  • risk assessment
  • segmentation

Cite this