The conversion of original data into interpretations and subsequently insights is the point of being a scientist. It is the struggle that the scientist has with the original data that provides the vision. Without that effort the scientist is either a technician simply concerned with producing data but not understanding its meaning or a philosopher focused on making sense of data and assuming its validity. For the conversion of data into interpretations, the scientist must identify the sources of variance in the data collected. The contribution of technique to the variance observed is particularly important to assess. The effect size is another important attribute that influences the interpretation of the data. In biomedical science modest effect size is likely to have little relevance for pathophysiology. Another important parameter to consider in order to understand the implications of the data collected is sample size. Under-powered studies can lead to weak interpretations, but sample size is often limited by cost and patient availability. Finally, it is essential that investigators understand the statistical analysis used in order to make interpretations with meaning. The unknown trade-offs made in how statistical analysis proceeds in the algorithms of artificial intelligence represent a major issue because they rely on faith instead of inviting doubt. Nevertheless, doubt is the impetus for advancement in biomedical science.