Threats to validity: training and teaching in studies

I had an interesting discussion with someone the other day about threats to attributing causality in studies with “training” as part of the intervention. This discussion reminded me of another discussion I had with a few other colleagues about a paper that was published in the Journal of Medical Internet Research (JMIR). The article is by Birru et al and is titled “Internet usage in low-literacy adults seeking health information: An observational analysis“.

After reading this study, I had a few problems with the conclusions made by the authors. Basically, the study provided Internet-skills training to low-literacy adults and then had the individuals complete a series of tasks related to using the Internet to find health information. The authors concluded that because the individuals had difficulty completing the tasks, health information on the Internet needed to be made more readily accessible (i.e., written at a lower reading level). Here are some of the issues that I have:

  • Okay, so the individuals had difficulty completing the tasks. But, what exactly were the authors testing: the effectiveness of the Internet-skills training/classes or the ability of the individuals to find health information? Were the difficulties the result of a poor/ineffective training program? There should have been some sort of evaluation of the training program because these individuals were not experienced users of the Internet.
  • Individuals had difficulty understanding the information. Not really unexpected, but the authors based their conclusion on difficulty understanding *text*. But, we’re dealing with low-literacy adults. Just because they have difficulty reading doesn’t mean they can’t understand the concepts if received in some other format. The literature suggests that low-literacy individuals develop alternative coping skils like asking others to read the information to them, asking others for help, or finding alternative information sources.
  • Final point is related to theories of learning. Just because an individual has difficulty understanding text by themselves does not mean that learning does not take place. What if learning occurs in a group setting – that the interaction between individuals is the agent that triggers learning?

Overall, I support the authors conclusion that health information on the Internet needs to be more readily accessible, but is this finding really new? I mean, we’ve known for the past 10-20 years or so that health information needs to be written at a lower level, based on research conducted on pamphlets, brochures, and other printed material. So, why would using the Internet be any different?



, ,