There are several types of bias in statistics, and avoiding and understanding these statistical biases can help you better interpret data. Here are some of the different types of statistical biases you may encounter.
Confirmation bias is an error that involves allowing a preconceived notion to impact how you prioritize or interpret information. An example of confirmation bias would be if you had a strong opinion that most people preferred vanilla ice cream over chocolate ice cream and, as a result, gave more weight to data that supported that conclusion.
Selection bias is an error that stems from using population samples that don’t accurately represent the entire target group. For example, data taken from one neighborhood would not accurately represent a large city. There are many reasons selection bias arises—some intentional, some not—including voluntary participation, limiting factors for participation, or insufficient sample size.
Outliers can significantly skew data. For example, when analyzing income in the United States, there are a few extremely wealthy individuals whose income can warp any calculation of averages. For this reason, a median value is often a more accurate representation of the larger population.
Observer bias is a type of statistical bias that’s biased as a result of the subjectivity of the observer. No human can be completely unbiased, so observer bias is always going to be an issue. The best you can do is learn to recognize it.
An example of this was a rat test performed in the 1960s where two groups of students tested rats, which were categorized as “bright” and “dull”. The students who had the “dull” rats handled them poorly and reduced their chances of completing the maze, which ultimately affected the results of the study.
Funding bias refers to the likelihood that a study has to favor the person who funded it. These studies tend to provide inaccurate data that can make it difficult to apply that data to your business.
Funding bias is especially popular with product comparisons. If Bounty pays for a paper towel comparison, that comparison is much more likely to favor Bounty than another brand.
Omitted variable bias
With omitted variable bias, the lack of a variable affects the legitimacy of the statistic. For example, a study about cars that doesn’t include the year or mileage may provide inaccurate results.
Omitted variable bias is one of the most common types of bias in statistics. When you’re looking at data, make sure that data takes into account all the relevant variables.
Survivorship bias is when you only take into account surviving data points. By not taking into account every potential source of data, you could be getting a flawed representation of the data.
A classic example of survivorship bias is WWII, when planes that survived were studied so they could be reinforced where they were shot most. In reality, it would have been best to look at downed planes and reinforce future models in the spots where those planes were shot and taken down.