Nagra Blog Post 3

      While reading Professor Johnson’s paper, I noticed many interesting connections and counterpoints to what has been drilled into my biological science education. I think it might be useful to point some of these out and apply them to the sphere of medicine with some questions that the class might discuss. Particularly in the section that connects scientific inquiry to machine learning, I was a bit reluctant to agree with the idea that “ like scientific inductions, machine learning programs use evidence (or known data) to form predictions (or generalizations to new ideas)” (2). While I somewhat agree with this statement, it is essential to discuss the difference between machine learning through observation and scientific study through experimentation. Observational studies, unlike experiments, can only be used to draw correlations. Experiments, on the other hand, can provide causal explanations. Writing this, I already see an interesting application of Johnson’s ideas. What is one to consider controlling in an experiment? Even these factors are inherently value-based. It would be interesting to discuss how a distinction between observational and experimental study could apply to machine learning and further to the discussion on value-based algorithms. 

    When Johnson introduces the flaw in the “value-free ideal,” that “scientists should always be guided by these epistemic virtues only, and not social or ethical values,” I instantly thought of another flaw in this ideal, that even the formation of what is to be tested or not, is inherently biased. In essence, even hypothesis formation is biased from what we think we already know and is infested with social and ethical valuations of what is to be studied. Later Johnson says that “there can be no algorithm for building algorithms,” to which I would reply that aren’t we as humans just complex biochemical algorithms? Perhaps I am missing the point here, but I think it is an interesting discussion to have. 

    Another place I agreed was the discussion of the Kuhnian value of novelty. Kuhn seems to imply that we should accept current theories while trying to discover more, and while this may make sense in most cases, science is made stronger by testing old hypotheses for validation. Furthermore, as ideas build upon one another with the idea of induction, it is imperative to test the old ones for validity. Suppose I base my theory on an inducted hypothesis with a probability of being right of 99%, based on another inducted idea with 99%. In that case, my theory now becomes weaker as (99% times 99% is less than 99%). 

    Lastly, an essential connection between Johnson’s paper and current public health is the idea of weighing how valid we need a study to be to apply it to the general population. For example, at what P-value is it appropriate to implement a vaccine to the people? Do we weigh the possible deaths of people who take the vaccine or the people who will stay unvaccinated and perhaps infected? It’s a delicate balance between the two; it’s an exciting balance to discuss.