Can experiments detect differences that matter? This question refers to the sensitivity of technology and its ability to detect significant differences. Hanage states that even within species, bacterial strains’ genes differ greatly.
Does the study show causation or correlation? Hanage used the example of a study which explored the diets of seniors, altering and causing dysbiosis of the gut microbiota, thus, causing poor health. Yes, the conclusion fit the results, but Hanage argues that due to their age, seniors have poor health which altered the gut microbiome. The reverse causality was not explored or mentioned. We must keep other factors in mind when discusing results and planning experiments.
What is the mechanism? Hanage argues that we must “pinpoint both whether the microbiome effects human health, and excactly how it does so” by focusing on the actions of specific components of the microbiome. We must be aware of how said components effect the microbiome in order to assess the importance of the results.
How much do experiments reflect reality? We must always keep this question in mind when discussing results. With microbiome related experiments, it is almost impossible to recreate the gut enviornment in a lab. Considering this, lab results may not always be reflected in clinical trials, or real-life uses.
Could anything else explain the results? Similarly to causation vs correlation, this question refers to other factors potentially affecting the results. Hanage explains that there are sometimes other more important facorsthat may explain results.
When discussing controversy, I think the most helpful question is “Does the study show causation or correlation?” This is because the human body is immensly complex and contains an plethora of interconnecting mechanisms. Yes, results may back up conclusions, but there is usually more to the picture. Such as in Hanage’s example, the results clearly backed up the conclusion, but the experimental designer made a mistake by not stepping back and exploring other potential causes of the results.
