Transcript for:
Understanding the Misleading Aspects of Statistics

every day we make life-altering decisions based on statistics statistics constantly manipulate our thoughts and help us decide everything from which car toy to what companies we should invest in and even what types of food we should or shouldn't be eating statistics are a fundamental part of our economy and culture and many people take them as absolute truth yet statistics can be inherently deceiving have you ever been left in disbelief after hearing a statistic that seems too good to be true well chances are you're probably right in his article how statistics can lie James be had stated that statistics do not create themselves people have to create them there is no such thing as a perfect statistic but some are less imperfect than others in marketing especially one strategy that is commonly used to manipulate statistics is to make false correlations between unrelated data sets for example we asked a group of high school students to fill out a long survey and found that 100% of the people who preferred orange jelly beans over others wanted to see the world in a complete Anarchy clearly this shows the eating orange jelly beans is the cause of anarchistic thoughts in teenagers in the same survey we found that 100% of people who prefer pink jelly beans had a fear of overflowing toilets despite the fact that these statistics are technically true anyone will tell you that eating orange jelly beans will not cause you to become an anarchist and eating pink ones will not cause you to develop a fear of overflowing toilets clearly something is a miss here there are many factors that must be considered if you wish to determine if data is accurate in this case cause and effect is important one distinction that also must be made here is the presentation of the statistic the correlation is obviously coincidence but we presented the data in a way that implied that one caused the other the sample size must also be considered statistics with a smaller sample size are generally less accurate and those with a larger sample size are more accurate in this case only 18 students were interviewed the only student who liked orange jelly beans also wanted to see the world end in complete Anarchy the small sample size makes coincidences like these much more likely this strategy manipulating statistics is often used in advertisements for medicine and most miracle cures we must learn to consider the way that the data was collected and what the sample looked like when analyzing statistics one thing that can strongly affect data is the way in which the survey was done for example if you ask a group of people a biased yes or no question like do you support children bullying each other as a result of violent video games most people will say no because this is a more socially acceptable answer even if they don't believe that violent video games cause bullying when viewing statistics the whole story must always be considered for example there are technically 5.88 popes per square mile in the Vatican City but obviously there's only one Pope and you cannot have a fraction of a person the reason that this statistic is technically true is that the Vatican City is less than one square mile large statistics are often represented graphically graphs are common because they allow statistics and Trends and data to be presented in an easy to read way how however graphs can often be misleading for example this graph shows the annual profit for two hypothetical companies in millions of dollars at first glance company a has a much greater profit than Company B however if we look closely at the numerical values we notice that the profits of both companies are very similar company A's profits were less than $1 million greater than company B's profits the scale of the graph was altered to emphasize the difference manipulation of statistics is common in both inferential statistics and probability statistics probability statistics are easily manipulated because many people simply do not understand probability that well for example you could say that there is a 50% chance of rain on both Saturday and Sunday there's a then there is 100% chance of rain over the weekend doesn't seem so bad right however by the same logic if you flip a coin twice there's 100% chance of one heads and 100% chance of one Tails over the course of the two flips but it's obviously possible to have two heads or two tails in a row close analysis of probability statistics is crucial due to their easily manipulated nature overall it is necessary for people especially consumers to learn to better analyze the statistics that affect the decisions they make every day we must learn not to take all statistics at face value especially in advertising and business a better understanding of statistics and how to analyze them is an extremely important skill that everyone should develop