I did a meta-analysis for my masters thesis and I was convinced that 2 of the 39 papers in my dataset had fraudulent data.