The Web as a Research Tool

Posted on: July 25th, 2014 by admin 2 Comments
Print Page

Last month the prestigious journal, Proceedings of the National Academy of Sciencespublished a paper in which the opening sentence of the abstract stated, “We show via a massive (N=689,003) experiment on Facebook, that emotional states can be transferred to others via emotional contagion, leading people to experience the same emotions without their awareness.” Not too surprisingly, this paper elicited a storm of protest because the experimenters had manipulated the emotions of users without apparently seeking their explicit approval.

Although Facebook users, when signing up, are told that their information will be used for “internal operations, including troubleshooting, data analysis, testing, research and service improvement,” commentators have argued this does not constitute informed consent. Despite the negative publicity from this and other dubious practices, Facebook’s financial fortunes have not suffered. Within the past 24 hours its shares hit an all-time high on the news that second-quarter sales had surged by 61% to $2.9 billion. Such is the power of combining 1.3 billion active users and advertising revenue.

Another scientific application of the World Wide Web has just appeared in Academic Radiology, where the authors have reported on novel data sources for women’s health research. Their subtitle, “Mapping breast cancer screening online information seeking through Google Trends,” provides a hint of what they accomplished. Simply by clicking on the link to Google Trends and entering the search term “breast cancer” will bring up a chart that tracks, on a month-by-month basis for the past ten years, the relative search frequency for this term. As we might expect, the chart spikes each October given that this has become breast awareness month around the world.

Dr Soudabeh Dehkordy, a radiologist from the University of Michigan, and her co-authors were interested in finding out whether the Web search query “dense beast” entered in Google Trends would provide insight regarding women’s information-seeking behaviour. Given that breast density has become a hot topic over the past few years – because it can lead to a false negative result with X-ray imaging – the authors demonstrated that newsworthy events and legislative action were highly correlated with the peaks in search volume, particularly in those states where legislation was under consideration.

Google Trends also enables the researcher to explore the geographic distribution of search queries. Aside from the US, Canada is the only other country where women seem to be concerned about breast density. When will other countries take notice?

2 Responses

  1. You might consider the possibility that the majority of the world will never use the word density, since they use a different language.
    The English speaking rest of the world might be less susceptible to this somewhat hysterical and typical US topic. Perhaps even because of :
    1. the lack of a clear definition of who is dense, and inability of doctors to discriminate between 3 and 4.
    2. no evidence from randomized trials,
    3. no evidence that additional testing is cost effective with respect to mortality and interval tumors.

    • Kit Vaughan says:

      Thanks for this feedback, Ard. You are absolutely right that any Google Trends search would be influenced by the person’s home language. I also appreciate your point about the US being very focused at the moment on the topic of breast density. Time will tell if this parameter deserves such attention.