Our preferences do not determine what’s true.
—Carl Sagan [1]
We have learned how to create good data items and dataset formats, focusing primarily on the characteristics of the individual data items and on the organization of the datasets. But there is a big issue not yet discussed – how to choose what data items to include in your research and how to collect appropriate values for those items. This process is fraught with danger – the danger of bias. Bias is the data collection killer; nothing will compromise the quality of your data more than bias. Biased data leads to incorrect results ...
