As a crude simplification, inductive reason happens when the facts and data are “incomplete” with respect to the actual domain of the decision being made. In many cases, we create abstract rules and procedures based on “lessons learned”, then apply those generalizations to something else. Because there can be gaps1 in the train of logic, the relevance of those facts and data to the conclusion being drawn might be less than “certain”2.
The use of “heuristics” often turns out to fall in this category, as do some uses of similarity parameters. See also reasoning. Contrast with abductive reasoning and deductive reasoning.
Inductive reasoning is not “wrong headed”, but it can be extremely difficult to detect cognitive biases in the process. In most cases, it would be better to convert inductive processes into prospective deduction and, eventually, into true deduction. Unfortunately, that is often a cost-ineffective approach, and “Engineering Judgment” has to suffice.
Footnotes