Testing inductive reasoning

Though inductive reasoning leads to probable and not absolute truth, you can assess a conclusion’s likely probability by asking three questions. This chart shows how to apply those questions to a sample conclusion based on a survey.

CONCLUSION The majority of students on our campus would subscribe to wireless Internet access if it were available.

EVIDENCE In a recent survey, 923 of 1,515 students questioned say they would subscribe to wireless Internet access.

1. Is the evidence sufficient? That depends. On a small campus (say, 3,000 students), the pool of students surveyed would be sufficient for market research, but on a large campus (say, 30,000), 1,515 students are only 5 percent of the population. If that 5 percent were known to be truly representative of the other 95 percent, however, even such a small sample would be sufficient (see question 2 ).
2. Is the evidence representative? The evidence is representative if those responding to the survey reflect the characteristics of the entire student population: age, sex, level of technical expertise, amount of disposable income, and so on. If most of those surveyed are majoring in technical fields, for example, the researchers would be wise to question the survey’s conclusion.
3. Is the evidence relevant? The answer is yes. The results of the survey are directly linked to the conclusion. A question about the number of hours spent on the Internet, by contrast, would not be relevant, because it would not be about subscribing to wireless Internet access.