How to Lie With Statistics

Research Fraud

Research fraud undermines the credibility of academic work, with over 10,000 papers retracted in 2023 alone due to fraudulent practices. Gary Arndt highlights the issue, noting that even prestigious institutions like the Dana Farber Cancer Center have faced retractions 1. He explains that while most research is legitimate, the presence of fraudulent data can lead to misleading results, regardless of the statistical analysis applied.

If the data is bad, then it doesn't matter how good your statistical analysis of it is because you will still get a fraudulent result.

--- Gary Arndt

Arndt also touches on p-hacking, where researchers manipulate data to find correlations, further complicating the integrity of research findings 1.

   

Cherry-picking

Cherry-picking data involves selectively using data points to support a desired outcome, often omitting inconvenient data. Gary Arndt describes this as a form of data fraud by omission, where outlier data is ignored to present a more favorable dataset 2. This practice is common in politics, where statistics can be manipulated to show crime rates or stock performance in a misleading light.

Cherry picking data happens all the time. If a politician wants to show that crime is increasing or decreasing, you just have to pick the right crime to look at.

--- Gary Arndt

Arndt also warns about the dangers of small sample sizes, which can skew results and lead to inaccurate conclusions, especially in fields like election polling and rare disease research 2.