Use RADAR to Evaluate Information

Did you know penguins can fly? Neither did we until Terry Jones discovered the colony of airborne penguins featured in the video below. 

This BBC video was created as an April Fool’s Joke, but it’s an excellent example of the need to evaluate information found on the open web. 

This BBC video was created as an April Fool’s Joke, but it’s an excellent example of the need to evaluate information found on the open web. Here are a few more:
  • In the late 1990s a chain email was circulated with medical advice about what to do during a heart attack.1 The email encouraged potential victims of a heart attack to cough rhythmically. Unfortunately the lack of information in the email could actually lead to unintentioned medical outcomes.
  • In 2011 satirical news source The Onion posted a story called, “Study Finds Every Style Of Parenting Produces Disturbed, Miserable Adults”. Readers began sharing the article without attributing it to the original publication. The humorous intent of the piece was lost on the concerned parents and educators who called the California Parenting Institute to learn more about the study.2
  • In 2012, a 30-second video of a pig rescuing a trapped goat from a pond went viral on YouTube. The video was featured on Good Morning America and NBC Nightly News. Some animal rights advocates heralded the video as proof of animal intelligence. Pigs may be smart creatures, but the video was faked.3

RADAR

Fortunately there’s a simple acronym that can be used as a tool for evaluating information. In the same way that a ship captain uses radar to navigate the open waters, researchers can use RADAR to navigate the oceans of information on the web4.Each letter in the word represents a criteria by which to measure the reliability of sources.

Relevance – Is the source relevant to your information needs?

Authority – Who is the author(s) of the work and what are the person’s credentials?

Date – Is the information current? Does your research require timely sources?

Appearance – Does the source look professional? Are there advertisements, typos, or biased language?

Reason – What is the purpose of the information: to inform, advertise, entertain, or persuade?

Imagine that you’re writing a paper on the impact of the BP oil spill on birds. Now apply RADAR to the following two webpages.  

BP’s Environmental Assessment Page

Screenshot of a page from the BP website with the following text: Environmental assessment: NRDA More than four years of environmental data suggests that most of the impact was of limited duration and geographical reach Some of the natural resource damage assessment (NRDA) studies being conducted by the state and federal trustees and BP are included here, with access to more detailed information describing the studies via the links below. The studies are based on data collected by scientists to support the environmental assessment and accident response activities. Birds Analyses and field observations conducted to date indicate that the spill had limited effects on bird nesting and populations. For example, no visible oil was seen in 99% of the approximately 500,000 live bird observations made from May 2010 to March 2011. Survival studies of six bird species found that the survival rate was high for oiled birds. The birds were fitted with transmitters that provided data about survival rates. Five species showed no impact in survival rates between oiled birds and unoiled birds in the potential area of impact, as well as birds in areas not affected by the spill. The results of tests from the sixth species studied were inconclusive due to transmitter and methodological issues. In addition, studies conducted in 2011 indicate that bird productivity – the number of fledglings per nest – showed no spill-related impact.

TreeHugger’s Oil Soaked Birds Post

Screenshot of TreeHugger post with advertisements an image of dead bird covered in oil and the following text: Less Than 1% of Oil-Soaked Birds Survive. "Kill, don't clean" oiled birds No, that's not the opinion of a heartless bird-hater, or BP CEO Tony Hayward letting fly another tactless gaffe. It's the actual recommendation of one oil spill expert and animal biologist who says that once birds are thoroughly oiled, the best course of action is to put them out of their misery. Even if all the crude is scrubbed from their feathers, she says, oiled birds are all but certain to die a long, painful death. This may shock many, and the advice certainly appears contrary to that of the myriad conservationists who have set up centers around the Gulf to care for oiled birds.

Initially the content on both pages appears relevant to the topic of the BP oil spill and birds. The authors of the pages cite expert sources, and the content is relatively timely. The TreeHuger site has a number of advertisements. The image of the oil-covered bird – as well as the graphic language about the status of the birds – is intentional in order to evoke emotions in the reader. Conversely the language in the BP site plays down the impact of the oil spill. Both examples exist to persuade readers about the impact or lack thereof of the oil spill. A page from a peer-reviewed journal article on this is listed below. This article, found in the library’s databases, was co-written by three scholars and reviewed by at least one expert before it was published. Of the three sources, it contains the most reliable information.

[Screenshot of journal article with the following text: 6 Summary Using non-proprietary cloud-based and traditional software we were able to investigate and answer two primary questions of where and which birds were most impacted. Google Fusion Tables and Google Maps provided us with telling density maps to locate these centers along the Louisiana coasts and in its bays. Tableau Public’s functionalities (similar to stacking and dodging in ggplot2 (Wickham 2009)) allowed us to visually identify the birds most impacted by frequency. This result is in contrast to what we found by our logistic regression where the results confirmed that those birds with the largest counts (as seen on bar charts) were those birds with a significant p-value. We took our investigation further to follow the progression of the oil spill at the condition and oiling levels in a time series. We found that for the condition levels, dead birds beat out live birds for nearly all days. In addition, for the oiling levels, the frequency of oiled and not visibly oiled birds changes substantially from one week to the next, with overall counts for not visibly oil higher early in the collection period, and after the well was capped. Acknowledgments The authors are grateful to San Francisco Chapter of the American Statistical Association (ASA) for their generous Student Travel Award. We are also grateful to the organizers of the 2011 Joint Statistical Meetings and the ASA sections of the Statistical Graphics and Statistical Computing for the opportunity to create and share our work. We also owe a special thanks to the reviewers at COST, and Professor Bruce Trumbo for all their time and careful guidance in preparing this manuscript. References Cook D (2013) The 2011 Data Expo of the American Statistical Association, Computational Statistics; this issue Diggle PJ, Liang KY, Zeger SL (1994) Analysis of longitudinal data. Clarendon Press, Oxford Draper NR, Smith H (1998) Applied regression analysis. John Wiley & Sons Inc., New York Kutner MH, Nachtsheim CJ, Neter J (2004) Applied linear regression models, 4th edn. McGraw-Hill, New York Lesnoff M, Lancelot R (2010) aod: Analysis of overdispersed data. R package version 1.2, http://cran. r-project.org/package=aod, Accessed 20 April 2011 R Core Team (2012) R: A Language and Environment for Statistical Computing, R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0 (http://www.R-project.org/) Wickham H (2009) ggplot2: elegant graphics for data analysis. Springer, New York]

References

1King, S. B. (2009). If you have a heart attack, just cough. JACC: Cardiovascular Interventions, 2(10), 1032-1033. doi:10.1016/j.jcin.2009.08.011

2Fassnacht, J. (2014, Jul 03). When fantasy is mistaken for reality. McClatchy – Tribune Business News. Retrieved from http://search.proquest.com/docview/1542293145?accountid=28933

3Itzkoff, D. (2013, Feb 26). Really cute but totally faked. The New York Times. Retrieved from http://www.nytimes.com/2013/02/27/arts/television/pig-rescues-goat-and-the-video-is-really-cute-but-totally-faked.html?pagewanted=all&_r=0

4Mandalios, J. (2013). RADAR: An approach for helping students evaluate internet sources. Journal of Information Science, 39(4), 470-478. Tran, T., Yazdanparast, A., & Suess, E. A. (2014). Effect of oil spill on birds: A graphical assay of the deepwater horizon oil Spill’s impact on birds. Computational Statistics, 29(1), 133-140. doi:10.1007/s00180-013-0472-z