Misunderstanding Statistics
The post “Nine out of ten statistics are taken out of context” is a guest post on the website, “Speaking of Research.” The article is by Professor Robin Lovell-Badge who works at the Medical Research Council National Institute for Medical Research in London as the head of the division of Stem Cell Biology and Developmental Genetics. Prof Lovell-Badge’s post tackles animal rights myth that proposes that animal tests for human drugs should cease because they do not work. His post clarifies the misinterpretation of statistics used by animal rights activists to purport their claim. Statistics used by animal rights activists are factual however the information is used in the wrong context.
Individuals who disagree with animal testing for human drugs claim that the testing does not work because the majority of the drugs that go through the essential toxicology tests in animals do not succeed in human clinical trials. Prof Lovell-Badge says that their statements are deceptive because such individuals do not comprehend the purpose of animal safety tests. The professor claims that to understand the fallacy in their statements, one ought to understand the drug development process. Non-animal tests are carried out before animal testing and is done on various research tools such as cell cultures and computer models. During the pre-preclinical trials, the drugs undergo testing to ensure toxic drugs do not move to the animal testing phase. He claims that preclinical animal tests are done to check if the drugs are safe enough to continue to phase 1 of human trials. The first phase of the human trial is meant to ensure that the drug can be used on a greater number of people, not to prove its effectiveness. He further says that the licensing of a drug is done so by looking at the results in the clinical trials on people and not on the preclinical tests done on animals. Therefore, animal rights activists should not claim that animal research on human drugs does not work.
I find the fact that “only 8% of those drugs passing animal testing stages would go on successfully to be FDA approved”(Lovell-Badge,2017) surprising. I thought that the success rates of drug trials would be higher than this. I have also come to realize that the low success rate of drug trials has little to do with the animals because “animal experiments remove 36% of the potential drugs from moving onto the next stage” (Nine out of Ten Statistics Are Taken out of Context | Understanding Animal Research, 2019) and a lot to do with the entire drug development process which includes the preclinical phase, animal trial, three phases of human trials and the registration phase.
Statistics are important because they offer information on current trends in different aspects of life that aide in research. Individuals however misuse statistical data both intentionally and unintentionally. People tend to believe reported statistics because they assume that the author has statistical knowledge on the subject matter. Statistics taken out of context such as the one discussed are a common phenomenon. Aside from the example discussed above, one other situation was in 2011 where the American Beverage Association(ABA) criticized a report that connected energy drinks to a growing number of hospital emergency visits. The report by the Drug Abuse Warning Network claimed that data indicated a “tenfold increase in the number of emergency visits involving energy drinks between 2005(1128 visits and 2008(16053) and 2009(13114)” (Bouckley,2011). However, DAWN failed to mention that the number of emergency visits by individuals who partook “in energy drinks represented less than one 1/100th of one percent of all emergency visits “(Substance Abuse and Mental Health Services Administration, Center for Behavioral Health Statistics and Quality2013). In yet another example, The Council of Conservative Citizens used a report by statistician Lynn Langton, who works for the Bureau of Justice Statistics to push their agenda. The council used the report to claim that “83 percent of violent, interracial crimes were committed by blacks against white victims”(Evans,2018). The number seems high however, what the council failed to mention is that the 83% is because whites experience more violence than blacks. The 83% also represents violence against whites by all races and that “67% of white victims had a white assailant”(Evans,2018). It is important to check the credibility of the use of statistics before taking it as the truth.
Professor Lovell-Badge is negating the idea that animal testing should be stopped because of the majority of clinical animal trials finals in human trials. Animal rights activists should first understand the drug development process before faulting the animal trials. Animal trials are carried out after success in the pre-pre-clinical trials to ensure they are safe enough to proceed to phase 1 of human testing. Statistics used by animal rights activists may be true but are unfortunately used in the wrong context. Organizations and individuals misuse statistics by taking them out of context as shown in the examples discussed above. As a consumer of information, it is important to read the fine print and ensure that the statistics are used correctly to avoid receiving false information.
Part 2
Misuse of Statistics to push Agendas
The article “Nicholas Kristof: Gun deaths more numerous for preschoolers than police” is a critical look at Nicholas Kristof’s claim by Jon Greenberg. Kristof, in an article in the New York Times in 2015 where he wrote that in the US, “more preschoolers are shot dead each year (82 in 2013) than police officers are in the line of duty (27 in 2013)” (NewYork Times,2013). The numbers are starling and led to the investigation to find out if the information was factual by the website The Politifact. Greenberg found Kristof’s claim mostly true. It is important to look at facts and ensure statistics are not manipulated to push an agenda.
Greenberg claims that indeed there were fewer officers who died in shootings in 2013compared to preschoolers although getting the number of preschoolers is difficult. He found that Kristof’s age range of preschoolers is aged between 0 to four years old. In the US, “majority of preschoolers are four years” (US Department of Education, 2015). Kristof also included both accidental and intentional killings of preschoolers in his article where he included gun deaths in all situations.
The misuse of statistical data is a common phenomenon and result in the receiving of misinformation by the consumer who is likely to take the information as the truth without further investigation. There are numerous ways to lie with statistics to further an agenda. I would lie using statistics to improve sales in a toothpaste company. For instance, the use of sample bias which is the collection of samples in a non-random resulting in either a higher or a lower sampling probability in some instances. For example, If I carry out a survey of an area where my toothpaste company is the only supplier, and ask the respondents what toothpaste they use, then the results will be biased. I could also use faulty polling which is the paraphrasing of questions to get specific answers from the audience. For example, the use of loaded questions such as “Which toothpaste do you normally use? X or Y?”. Single questions are also faulty polling because, at times, a respondent’s answer may be complicated for instance, the respondent might want to clarify that they use X toothpaste because Y toothpaste is unavailable but the surveyor will not accord them the opportunity to explain. Thirdly I would use percentage change with small sample size. If I survey insignificant sample size, then my results will be misleading. For instance, if I ask a question of a sample size of fifteen people and fourteen answers yes, which is 93% and if I ask the same question to a sample size of two thousand and one thousand eight hundred and sixty people answer yes which is 93% as well, the validity of the percentage differs between the two. Lastly, I would use purposeful and selective bias. Purposeful bias is the intentional attempt to influence data findings in the form of adjustments data omissions. Selective bias on the other hand involves bias on the sample size. For instance, asking republicans about Donald Trump’s administration will automatically have an error of selective bias.
Kristof’s article “Gun deaths more numerous for preschoolers than police” wanted to pass the message of the need for safer gun rules. However, the attempt to lie using statistics was not the correct action to take. There are various ways individuals and companies can use statistics to lie such as the use of sample bias and faulty polling. The general public ought to careful not to receive deceitful information brought about by misuse of statistics.
References
Bouckley, B. (2011, November 29). Energy drinks are a casualty of statistics ripped ‘out of context’, ABA. Beveragedaily.Com. https://www.beveragedaily.com/Article/2011/11/25/Energy-drinks-are-casualty-of-statistics-ripped-out-of-context-ABA
Evans, Glenn (2018, March 9). Stats “taken out of context.” Longview News-Journal. https://www.news-journal.com/news/local/stats-taken-out-of-context/article_a639759f-b43e-5f88-a190-4c53cdaa952b.html
Greenberg, J. (2015, October 6). PolitiFact – Nicholas Kristof: Gun deaths more numerous for preschoolers than the police. @politifact. https://www.politifact.com/factchecks/2015/oct/06/nicholas-kristof/nicholas-kristof-gun-deaths-preschoolers-police/
Lovell-Badge, R. (2017, April 5). Nine out of ten Statistics are taken out of context. Speaking of Research; Speaking of Research. https://speakingofresearch.com/2013/01/23/nine-out-of-ten-statistics-are-taken-out-of-context/
Nine out of ten statistics are taken out of context | Understanding Animal Research. (2019, February 14). Understanding Animal Research. https://www.understandinganimalresearch.org.uk/news/communications-media/nine-out-of-ten-statistics-are-taken-out-of-context/
Substance Abuse and Mental Health Services Administration, Center for Behavioral Health Statistics and Quality. (2013, January 10). The DAWN Report: Update on Emergency Department Visits Involving Energy Drinks: A Continuing Public Health Concern. Www.Samhsa.Gov. https://www.samhsa.gov/data/sites/default/files/DAWN126/DAWN126/sr126-energy-drinks-use.html
U.S. Department of Education. (2015). A Matter of Equity: Preschool in America — April 16, 2015 (PDF). https://www2.ed.gov/documents/early-learning/matter-equity-preschool-america.pdf
