Strong inductive arguments are typically found in cases where the conclusion is highly probable based on the premises (i.e., strong evidence for the conclusion is provided).
As mentioned earlier, this is because for inductive arguments even if each premise of the sequence of statements used to demonstrate the sweeping statement (or generalization) were true, the conclusion can still be false.
The problem with Arguments by Analogy is that they are inductive arguments. And, this means that the certitude of the conclusion varies and is probabilistic at best.
(22) In inductive arguments, the premises are intended to provide some (strong or weak) evidence for the conclusion and so the conclusion follows with some uncertainty; in deductive arguments, the premises are intended to prove the conclusion and so the conclusion follows with certainty.
Similarly, if one wishes to object to the inter-level circularity found in inductive arguments for induction, the objection rests on the fact that a request for evidence that the conclusion of an inductive argument is justified can never be satisfied so long as one pushes the problem one level back by using an inductive form of argument to "defend" induction.
If we were to accept Alston's model, an inductive argument for induction could be "justificatory" in virtue of the meta-principle at the next level that induction is truth-conducive, which principle could, in turn, be supported by the same inductive argument and "justified" in virtue of a still higher-level statement that induction is truth-conducive, ad infinitum.
When consideration is given to the distinguishing characteristics of
inductive arguments, this interpretation seems to fit their position quite well.
More on deductive and
inductive arguments. Informal Logic Newsletter, 2(3), 7-8.
That would lead us to judge a vast number of obviously bad inductive arguments to be rational, since all that would be required for an inductive argument to be rational would be the positive relevance of its premises to its conclusion.
We may want to say, with Carnap, that inductive arguments must be based on the total available evidence.
3 Note that for Stove inductive arguments are distinguished from deductive arguments not by the nature of their inference (as with Carnap), but by the nature of their premises and conclusions.
The thesis that nature is uniform is contingent, so the claim that the premises of an inductive argument are reasons to believe its conclusion if nature is uniform is an example of misconditionalisation.
