%% Last Updated: - [[2021-02-22]] - [[2020-12-13]] %% Abductive reasoning is difficult because it draws conclusions from data in a nonlinear, non-systematic way. Instead, this type of reasoning leverages things that are difficult to reproduce, such as specific personal experiences or an informed intuition developed over seeing patterns repeatedly. Abductive reasoning is similar to [[Daniel Kahneman]]'s [[System 1 Thinking]], or what I call [[Fast Thinking]]. It is useful for situations where it would be difficult to systematically reproduce lines of thought. For example, when attempting to determine if someone is lying, there are many possible "tells" that could be ennumerated (such as inability to make eye contact, or a twitch of the mouth)-- but it would be too difficult to ennumerate them and encapsulate them in a unified model. By contrast, our minds do this instinctively when watching a person we know quite well. Abductive reasoning also sounds like applying [[Heuristics]] to a problem, which has always seemed to me to be a fancy word that really just means "educated guess". ## Disadvantages of abductive thinking - It's a type of analysis that is difficult to communicate, teach, and learn. - [[A lack of sample data makes conclusions difficult.]]: It takes a lot of datapoints (experience) to make connections. - [[The quality of the conclusion relies on the quality of the data.]]: You need the __right__ kind of experience to be able to successfully reason abductively. - [[Insufficient but convenient samples can lead to biased output.]] ## See also - [[Thinking]]