Academic journal article Visible Language

Making Cancer Surveillance Data More Accessible for the Public through Dataspark

Academic journal article Visible Language

Making Cancer Surveillance Data More Accessible for the Public through Dataspark

Article excerpt


This paper describes findings from an experiment to determine whether visual design could enhance the effectiveness of the presentation of cancer surveillance data online. The research team included designers who created an interface called Dataspark (DS) for California citizens to see incidence rates for colorectal cancer in the state. The design of the display used principles of relative scale, color, shape, and arrangement. In a randomized experiment, this interface was compared to two displays that are hosted by established cancer organizations but do not use principles of scale, color, shape, and arrangement in the same way. Approximately 550 California citizens participated in the experiment, during which they were assigned at random to use one of the three displays and then asked questions about understanding, ease of use, engagement and personal relevance. Results showed that the Dataspark display was significantly more effective in helping participants understand the data and explore the interface. User engagement and personal relevance were modest for all three displays. This paper analyzes the results and introduces some strategies to address engagement and personal relevance in future work.

pproximately 1 out of 2 Americans bom today will be diagnosed with cancer in his or her lifetime (National Cancer Institute, 2011 ). Worse yet, those who are poor, African American, or uninsured are more likely to get cancer, be diagnosed at more advanced stages when cancer is more difficult to treat, and have a shorter period of survival after diagnosis compared to other Americans (American Cancer Society, 2013, 42).

Because cancer is one of the leading causes of death in the U.S., all 50 states have made the reporting of cancer diagnoses mandatory. Cancer registries compile and compute cancer incidence and mortality data, looking for trends over time and variability by geography, race, gender, age, and other variables. This work is part of the larger field of cancer epidemiology, the study of who gets cancer In populations and why. Many cancer registries are organized through the federally funded Surveillance Epidemiology and End Results program (SEER) which began in 1973.

Data collection has grown in virtually every aspect of life. But as information design and programming guru Ben Fry describes,

" We're getting better and better at collecting data, but we lag in what we can do with it, " (Fry, 2008, 2)

For cancer registry data, the gaps between what is collected, what the public understands, and how individuals behave in response, are large,

In their 2009 book Making Data Talk, David Nelson, Brad Hesse and Robert Croyle describe the challenges associated with lay audiences sifting through health data on and off the Internet. (Nelson, Hesse and Croyle, 2009, 4) As they explain, "Unfortunately, examples of poor communication of data abound-on Web sites, in written materials (e.g., reports, brochures), during oral presentations, and during media Interviews, leaving many people awash in a morass of confusing 'data smog.'" (Ibid.) (Shenk, 1997)

The project described in this article explores the development and testing of an interactive display to inform the public about variation In colorectal cancer rates across population sub-groups, using data from the state of California. The display Is a transdlsclpllnary synthesis of principles of data selection, visual design, user interface and communi- cations created to understand how the Impact of cancer statistics on the public can be improved through better presentation.

The project addresses an emerging challenge in public health communications:

as data and their delivery systems become ever more individualized, robust, and search-driven, can they also be used to help people see the connection between individ- ual behavior and population health?

This question is particularly important for addressing the health needs of populations at disproportionately high risk of cancer and other diseases, shown In our project by the comparatively high rates of colorectal cancer-a cancer thought to be preventable through screening-for certain population sub-groups, like African Americans. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.