We need to question our information more
In the wake of modern technology, access to information has never been so easy, from websites, news apps, social media sites etc. Everyday we are flooded with a mesh of statistics, graphs and opinions online. 2.5 quintillion bytes of data are created on a daily basis(Jackson, 2013) , to put that in perspective, if we lay quintillion pennies side by side we would be able to cover the surface of the entire earth 2 times over. (MetaFilter, 2010).
But how much of this information is of a qualitative and objective standard to portray an accurate account of the situation? How many times have you read something on social media and just assumed it was true because a friend posted it? How many times have you took something at face value without giving it a second thought? Taking stories/graphs/stats at face value leaves us susceptible to misinformation which in turn leads to bad/inaccurate decision making. I do feel in this era of misinformation exemplified by the recent US election critical thinking skills and a healthy skepticism towards information are now more important than they have ever been. We must also be careful not to fall into the trap of confirmation bias. We should look at opposing views and try to understand them, that we we can make a more informed/objective/accurate decision.
Stats can always be manipulated and cherry picked to paint a certain picture, for example "90% of people rated our product better than our competitors". This is not so impressive if the survey size was 10 people which is why you should not believe a percentage right off the bat. You should ask yourself how many people were surveyed to come up with that stat and is the sample proportionate to the population size? Was the survey conducted by an independent party ? or are there any other studies to prove/disprove the accuracy of this stat?
News outlets often present information like "a new study shows x is linked to y". Again we tend to take this at face value because a reputable news source quoted the the finding. However just because one study found a link does not mean there is a link. Have other studies been conducted using the same method on different data with the same result been replicated? Who undertook the study, are they reputable, is it likely they have an ulterior motive? Has the study been peer reviewed? One factor that is often overlooked especially with regression models is the confidence interval, the academic standard is 95%, anything below that is generally considered a poor model. Biased studies will tend to lower the confidence interval to force the results they are looking for. They may also cherry pick or select certain data that will give them the result or they might transform the data using complex equations etc to force the results they seek. Sometimes transformations are necessary but studies that do not clearly define why a transformation was made is a definite red flag.
Charts can also be used as a tool to distort the real view of the situation. One common method is to alter the scale of they y axis on a chart. If you look below we can see that product C has had significant sales, however if you look at the y axis it does not start at 0 and their are only 3 ticks to increase the size of C. It looks like C has strongly out performed the others because we "zoomed in" on the chart
However when we make the tick size proportional (zoom out) to the data we can see the performance of C is not so significant.
Questioning new information is a good thing, analyzing how it was created and where it came from will leave us less susceptible to misinformation and we will be able to make better, more informed decisions as a result, however this takes a little bit of effort and an open mind!