An important mathematical subject broadly used but the data and conclusions are often abused and misunderstood. Things to look out for when interpreting the data!
Data gathering methods! For example, when I was studying statistics took part in experiments, one particular was a psychometric study to determine the sexual orientation mean for a population, in a study done by the justice department. Originally the questionnaire was written in French, but poorly translated into English, so several times one had to ask for interpretation what the sentence meant! In a social experiment, where one is supposed to have an emotional reaction to the questions, this can lead to serious errors in the data, that may never show up in the data gathering report! Here is another example of the problems that arise with data gathering. Images/Ego depletion, an influential theory in psychology, may have just been debunked_.pdf
Data point spread! In establishing trend line, I included an article from Discover,Images/Statistics_ When Confounding Variables Are Out of Control - Neuroskeptic.pdf
Lets first look at the data point spread! As an analyst I like the data points a lot closer to the trend line. This article shows great variability in the data points!
For example, when taking time measurement find a consistent point where the element to be measured starts! In other words, avoid the problem of confounded data! Next how to deal with variability in the data this is what the article is about confounding and using SEM.
One other common occurrence is displacing the data, using the data that has no physical relation to another instance and drawing conclusions from it! This is often used in politics and conspiracy theories. Here are some articles referring to this problem. Images/Was This Donald Trump Aide Right About Female Genital Mutilation_ He Made A Shocking Claim _ Bustle.pdf
In project management and estimating where this can slip in is using cost estimate data inconsistently. For example, many years ago reviewed the time estimate data that architects use for initial cost estimates of project. The sample was a time estimate to install counter top. Well if one installs a prefab rolled edge counter top no back splash the time estimate was high, install a stone counter top the time estimate was low. Extrapolate the data over a large project lets say a high rise the difference is huge. Use the data wrong one can find oneself with one frustrated client! In this instance the data was for initial estimates for project pricing. The accuracy is not very good usually around 25 percent plus or minus.