In this writing exercise, I will be discussing 5 key questions addressed in Hanage’s paper.

Can experiments detect differences that matter?

This question is a practical probe of the study’s findings. If the study can delineate a clear difference in the data between two groups from which research had previously been unable to separate, I think the study will most likely have an application in the clinical setting.

Does the study show causation or correlation?

This question is vital for us as readers to take what we are reading with the proper restraint. Causative associations (collected during an RCT) are significantly more promising than associational data collected during an observational study. Knowing that data being presented is causational can add to its method as something of interest.

What is the mechanism?

Like the question above, this question is a precaution. Do the methods of the research make sense? Is the data being presented in a candid way, or is the author making a statistical significance out of nothing? I think similar to the last question as well, knowing the mechanism is correct adds to the ethos of a study and can make data presented seem more credible.

How much do experiments reflect reality?

While research methodologies may prefer more idealized systems, resultant data from studies constructed with the real-world in mind is better applied to the real world, naturally. While RCTs on other species can give important clues as to where to look for similar causative relationships in the human organism, those studies would never be as applicable to real-world scenarios as a clinical trial with humans would be.

Could anything else explain these results?

The fewer covariates there are, the better. When something being analyzed is affected by less things (independent variables), there are less unpredictable changes in the dependent outcome. When the topic of interest being analyzed has less covariates, the data is likely more impactful and can show a stronger causative association.