Anecdotal references are common place throughout many forms of communication and media. This includes news of all formats (print, broadcast, and online), political speeches, marketing, advertising, internal business operations and many others. Anecdotal references serve as supportive evidence to persuade an audience of an intended message, argument, or conclusion. The drawback however is the scope of anecdotal evidence.
It is quite often that politicians reference individuals by name and highlight personal experiences with the purpose of creating supportive evidence to messaging and political strategies. At the last presidential state of the union address in January 2020 President Trump invited 15 guests. The personal references included messages in support of minority assistance programs, advocacy for change in foreign countries, military strategy, and immigration reform. (state of the union reference)
Another common example of this tactic is within broadcast news. Commentators and specialist reporters commonly identify individuals by name and offer extremely personal pieces that again represent supportive evidence to drive home a message or outcome. Think about farms and farmers visited in the heartland of America when watching pieces covering the impacts of tariffs.
In the context of an analyst who has the responsibility of consuming company data and providing information for decision support, anecdotal evidence does not serve that purpose. Per the definition of anecdotal it is an unreliable method of supporting a conclusion.
In many cases anecdotal references present themselves as convenient alternatives to the rigorous application of the scientific method. Common forms range from personal accounts, whether from you or your trusted network, through to drawing conclusions from inadequately sized populations. When considering the pressure to deliver impactful information quickly it is understandable how people fall into this behavior.
An outcome published in a recent study from several universities on the impact of evidence type (anecdotal vs. statistical) on persuasion yielded interesting results. The analysis of 61 studies identified that
“… in situations where emotional engagement is high (e.g., an issue associated with a severe threat, involving a health issue, or affecting oneself), statistical evidence is less influential than anecdotal evidence. However in situations where emotional engagement is relatively low (e.g., an issue associated with low threat severity, involving non-health issue, or affecting others), statistical evidence is more persuasive than anecdotal evidence.”
These findings stress the importance of conducting a scientifically rigid analysis and acknowledging the moral responsibility to not exploit stressed and/or compromised situations.
Analysis in the absence of proper science can lead to the mis-representation of information and therefore incorrect conclusions and actions. The ramifications of such an outcome can have long-reaching impacts such as poorly informed strategy decisions, incorrect public messaging, or detrimental affects of customer experience.
In contrast to the negative position established so far anecdotal references are not entirely without value. When used in conjunction with scientifically proven outcomes anecdotal information serves as a great method for creating relatable messaging.
A key component of every analysis is the communication required to impart knowledge to decision makers. The best method for doing so is through story-telling. The more relatable the story the easier it will be for an audience to hear and process the knowledge.
Thanks for reading.
About the Author: My name is Ion King and I am the Chief Officer at Corsair's Ventures. My focus is on helping others passionate about growing careers in Data Science & Analytics achieve their goals. Connect with me on LinkedIn or find more of my articles on medium
This article is a reflection of my opinion with additional information gathered from the sources linked throughout.
Header photo by Joel Fulgencio on Unsplash