Deductive vs inductive reasoning in psychology: the real difference, and what the experiments show

In deductive reasoning, the conclusion must be true if the premises are true. In inductive reasoning, true premises make the conclusion probable, but not certain. That is the real difference between the two: necessity against probability. Many textbooks give a shorter rule: deduction goes from the general to the specific, and induction goes from the specific to the general. The rule fits the usual examples, but it does not fit all arguments.

Most summaries stop at the definitions and two examples. This article also gives you one question that sorts any example, and it shows what psychologists found when they tested how people actually reason.

• Deduction: if the premises are true, the conclusion cannot be false. A good deductive argument is valid.
• Induction: if the premises are true, the conclusion is likely, and it says more than the premises do. A good inductive argument is strong.
• The sorting question is: can the premises be true and the conclusion false at the same time?
• In experiments, people often accept a conclusion because they believe it, not because it follows. In a 1983 study, 92% accepted an invalid argument when its conclusion was believable.
• People also test their ideas badly. In Peter Wason's 1960 number task, only 6 of 29 students found the rule at their first attempt.

The difference in one table

The table shows the main differences between the two types of reasoning.

Why "general to specific" is not the definition

Many introductory sources define the two terms by direction. The APA Dictionary of Psychology describes inductive reasoning as reasoning in which "inferences and general principles are drawn from specific observations and cases". The open textbook Psychology 2e from OpenStax says that deductive reasoning "begins with a generalization" and that inductive reasoning "uses empirical observations to construct broad generalizations".

These descriptions fit the textbook cases. They are not a reliable test. The Internet Encyclopedia of Philosophy calls the direction rule "much too crude" to separate the two types, and it gives two counterexamples:

• "All As are Bs. All Bs are Cs. Therefore, all As are Cs." Every statement is general, so the argument does not go from general to specific. But the conclusion must be true if the premises are true. The argument is deductive.
• "Each spider so far examined has had eight legs. Therefore, likewise, the next spider examined will have eight legs." The argument goes from specific cases to one more specific case. But the conclusion is only probable. The argument is inductive.

The same encyclopedia defines a valid deductive argument as one in which, "if the premises are true, the conclusion must be true". A strong inductive argument is one whose premises "render the conclusion likely". Use this test. If an exam asks for the direction rule, give it, but classify the examples with the test.

Deductive reasoning: valid, sound, and still wrong

The APA Dictionary defines deductive reasoning as reasoning in which "a conclusion is shown to follow necessarily from a sequence of premises". The best-known form is the syllogism. A syllogism has two premises and a conclusion, and each statement uses "all", "no", "some" or "some … not".

People frequently mix up two terms.

• Validity is about the form. If the premises are true, a valid argument cannot have a false conclusion.
• Soundness is about the form and the content. A sound argument is valid, and its premises are true.

A valid argument can thus have a false conclusion if one premise is false. "All psychologists are surgeons. Sam is a psychologist. So Sam is a surgeon" is valid, but it is not sound, because the first premise is false.

The opposite also occurs. The APA Dictionary entry on the syllogism gives this invalid example: "Some women are tall; all mothers are women; therefore some mothers are tall." The conclusion is true, but the premises do not guarantee it. The tall women could all be women without children. The entry adds that, of all possible combinations of terms, "only 24 are formally valid".

Inductive reasoning: strong, cogent, and never certain

Inductive reasoning goes from what you observed to what you did not observe. The Stanford Encyclopedia of Philosophy defines inductive inferences as inferences "from the observed to the unobserved, or to general laws". You use induction when you predict tomorrow's traffic from the traffic of the last month, or when a researcher generalizes from a sample to a population.

The conclusion of an inductive argument says more than its premises. The philosopher Wesley Salmon, quoted in the Internet Encyclopedia of Philosophy, wrote that the inductive argument "expands upon the content of the premises by sacrificing necessity". Even a very strong inductive argument can have true premises and a false conclusion. In the 18th century, David Hume asked what justifies this type of inference. The Stanford Encyclopedia summarizes his answer: "our tendency to project past regularities into the future is not underpinned by reason".

Psychologists study how people judge the strength of inductive arguments. Osherson and his colleagues proposed a model in 1990 with arguments such as "Grizzly bears love onions. Therefore, all bears love onions." In their model, an argument is stronger when the categories in the premises are similar to the category in the conclusion, and when they cover that category well. This model, and later experiments such as the 2010 study described below, suggest that people use similarity to judge inductive arguments. Similarity also drives the representativeness heuristic, which our article on availability vs representativeness explains.

How to classify any example

Exam questions describe an argument and ask which type of reasoning it is. Use these steps.

1. 1.Find the premises and the conclusion.
2. 2.Imagine that every premise is true.
3. 3.Ask if the conclusion can still be false.
4. 4.If the conclusion cannot be false, the reasoning is deductive.
5. 5.If the conclusion can be false, ask if the argument claims that the conclusion must follow. A syllogism makes this claim: it presents its conclusion as certain, not as likely.
6. 6.If it makes this claim, the argument is an invalid deduction.
7. 7.If it claims only that the conclusion is likely, the reasoning is inductive.

The table applies the steps to six examples. The third and fourth examples break the direction rule.

The last row is a special case. The Internet Encyclopedia of Philosophy notes that some definitions make "deduction" a success term, so an invalid deductive argument cannot exist. Psychology experiments still use the label, because their tasks ask people to judge validity.

What experiments show about deductive reasoning

Psychologists test deductive reasoning with tasks that have one correct answer under logic. Two classic tasks show how people depart from that answer.

Belief bias: a believable conclusion seems valid

Evans, Barston and Pollard published three experiments in 1983. Participants read syllogisms and decided if the conclusion "necessarily follows" from the premises. The researchers changed two things: the validity of the argument and the believability of the conclusion.

Here is a valid argument with a believable conclusion: "No cigarettes are inexpensive. Some addictive things are inexpensive. Therefore, some addictive things are not cigarettes." In the next argument, the terms "cigarettes" and "addictive things" change places. The argument is still valid, but the conclusion is now unbelievable: "No addictive things are inexpensive. Some cigarettes are inexpensive. Therefore, some cigarettes are not addictive."

Validity had an effect, but belief had a larger one. An invalid argument with a believable conclusion was accepted as frequently as a valid argument with a believable conclusion. The authors found that belief bias was stronger for invalid arguments than for valid ones. The APA Dictionary defines belief bias as the tendency to accept conclusions "because they are believable rather than because they are logically valid".

The Wason selection task: few people look for the counterexample

Peter Wason introduced the selection task in the 1960s. The APA Dictionary describes a standard version. Four cards each have a letter on one side and a number on the other side. You see E, D, 3 and 8. The rule is: if a card has a vowel on one side, it has an even number on the other side. Which cards must you turn over to test the rule?

The correct answer is E and 3. The E card breaks the rule if it has an odd number on the back. The 3 card breaks the rule if it has a vowel on the back. The 8 card cannot break the rule, because the rule says nothing about cards with even numbers.

Most people choose E and 8, or only E. In a 1995 review in Cognition, Sperber, Cara and Girotto wrote that "only around 10%" of participants choose the correct cards in the standard abstract version. The result changes when the rule is about a duty or a right. In the drinking-age version by Griggs and Cox (1982), you are a police officer, and the rule is "if a person is drinking beer, then the person must be over 19 years of age". With rules of this type, the 1995 review says, "a majority of subjects" choose the correct cards.

Wason also connected the result to Piaget's theory. According to the APA Dictionary, Piaget's hypothetico-deductive reasoning appears in early adolescence and marks the formal operational stage. In his 1968 report on the task, Wason wrote that his participants "did not give evidence of having acquired the characteristics of Piaget's 'formal operational thought.'"

What experiments show about inductive reasoning

Wason's 2-4-6 task from 1960 tests how people use evidence to find a general rule. Wason compared two methods: a search for confirming evidence only, and a search for confirming and disconfirming evidence.

In the 2-4-6 task, 29 psychology undergraduates heard that the numbers 2, 4, 6 followed a simple rule. Their task was to find the rule. They wrote new sets of three numbers, and the experimenter said if each set followed the rule. When they were "highly confident", they announced the rule. The rule was "three numbers in increasing order of magnitude".

Most participants announced a narrower rule first. Wason gives "increasing intervals of two" as a typical example. A person with that guess tests sets that fit it, for example 8, 10, 12. Each test gets a yes, so the wrong rule seems confirmed. Wason called this enumerative induction: the use of confirming evidence alone. The better method was eliminative induction: tests that can show the guess is wrong. For the guess above, 1, 2, 3 and 6, 4, 2 are such tests.

Wason concluded that the participants with two or more incorrect rules "were unable, or unwilling to test their hypotheses". The 2-4-6 task is an early demonstration of what psychologists now call confirmation bias: a preference for evidence that confirms what you already think. Our list of ten cognitive biases defines it and shows how it relates to other biases.

Are they two different systems in the mind?

Logic separates deduction and induction. Psychologists ask a different question: does the mind use two separate processes, or one process with two thresholds?

• Lance Rips tested this in 2001. Participants evaluated the same arguments under two instructions: is the argument deductively correct, and is it inductively strong? In some conditions, people judged one argument correct more often than a second argument, but they judged the second argument strong more often than the first. Rips concluded that people have "distinct ways to evaluate arguments".
• Heit and Rotello reported two experiments in 2010. Deduction judgments depended more on validity. Induction judgments depended more on the similarity between the premises and the conclusion. Fast deduction judgments were like induction judgments.

Heit and Rotello call the relation between the two types "one of the most important open questions in reasoning research". The current evidence suggests that the distinction is not only a rule of logic. People also seem to evaluate arguments in two ways.

Where abductive reasoning fits

Abduction is an inference to the explanation that best explains what you see. The Stanford Encyclopedia of Philosophy uses this example: two friends had a serious argument, and now you see them jogging together. The best explanation that you can think of is that they made up.

Abduction and induction both give conclusions that can be false. The difference, in the same entry, is that "in abduction there is an implicit or explicit appeal to explanatory considerations, whereas in induction there is not". Another name for abduction is "inference to the best explanation". A researcher who proposes a cause for an unexpected result uses abduction.

Deduction and induction in psychological research

Psychology research uses both types of reasoning in a cycle. OpenStax Psychology 2e says that the two processes are "inseparable, like inhaling and exhaling", and that the scientific process is "circular".

1. 1.Researchers observe patterns in data and use induction to build a theory.
2. 2.They use deduction to derive a hypothesis from the theory. The APA Dictionary says that deductive reasoning "underlies the process of deriving predictions from general laws or theories".
3. 3.They test the hypothesis with new data.
4. 4.If the results do not agree with the theory, they change the theory, and the cycle starts again.

A good test of a hypothesis is the eliminative method from the 2-4-6 task: look for the result that can show the theory is wrong.

How to remember the difference

Two short rules help.

• Deduction: the conclusion is already in the premises. Ask: can I be wrong if the premises are true? For deduction, the answer is no.
• Induction: the conclusion goes beyond the premises. You get new information, and you pay for it with uncertainty.

Then test yourself. Do not read the definitions again. Write six new arguments, cover the answers, and apply the sorting question to each one. Our guide to memorising psychology terminology explains why tests help you remember more than a second reading does.

Psychology World is a dictionary app with more than 18,000 psychology terms, and it supports this type of practice. Search the dictionary for deductive reasoning, inductive reasoning, syllogism, belief bias and confirmation bias, and save the terms to one bookmark list. Then study the list as flashcards, as multiple-choice questions, or type each answer. The multiple-choice mode needs a list with a minimum of four terms. The home page shows the four steps of the learning loop and what each study mode does.