analyzing evidence

When reading, it is important to analyse the proof structure and evidence provided by an author to support their claims. The proof structure is whether the argument is deductive (starting from a rule or heuristic and proceeding to apply it in a given case) or inductive (starting from a body of case evidence and trying to synthesise a rule or heuristic that explains it). For example;

Deductive reasoning – Deduction moves from a general law to imagine what necessarily applies to a given case.

All humans are mortal (premise #1)
Jack is human (premise #2)
_____________________________
∴ Jack is mortal (conclusion, from 1-2)

Inductive reasoning – Induction seeks to provide enough case evidence to say that a law is very probably true.

Rum and soda makes me ill  (case #1)
Rum and lemonade makes me ill  (case #2)
Rum and coke makes me ill  (case #3)
_____________________________
∴ It is probably rum that makes me ill  (provisional conclusion)

Here are six types of evidence common in essays and research papers (ordered from probably most persuasive to probably least persuasive in an academic context):

  1. experimental: from reproducible observed results
  2. inferential: from statistical frequencies/trends/correlations
  3. testimonial: from accounts or interpretations of an expert and/or witness
  4. anecdotal: from observation/interpretation of a single case or experience
  5. analogical: from similarities to a comparable or corresponding case
  6. hypothetical: from “thought experiments” in imagined conditions

Most premises of an argument will require some form of proof or evidence, though this may not be necessary with some a priori or well-established premises. For instance, “all humans are mortal” is very unlikely to require experimental proof, and simply mentioning the anecdote of Socrates drinking hemlock or imagining what would happen if philosophers could live forever would be enough.

On the other hand, even “common sense” assumptions may be disproved by experimental or probabilistic evidence. For example, if I say “everyone knows that health is the most important thing,” I would then have to explain the finding from longitudinal studies that close social relationships correlate with better mental and physical health in old age (e.g., Waldinger 2015).

See also: