What are some best practices you follow to avoid bias in your data analysis?

Before I start a project, I always draw up a detailed analysis protocol. This structured approach helps me to recognise potential problems at an early stage and ensures reliable results. I carefully document all variables and set clear criteria for which data I include or exclude. Quality control is an essential part of my process - I systematically check for missing values, identify outliers and scrutinise data distributions to detect anomalies. To validate my results, I use cross-validation methods and carry out sensitivity analyses for key parameters. Throughout the process, I keep detailed documentation of all analytical decisions. This methodical workflow has proven to be effective in minimising bias while maintaining high standards of analytical quality.