Locus of control: what Rotter actually measured

Locus of control is the expectation that what happens to you depends either on your own behaviour (an internal locus) or on luck, fate and powerful others (an external one). The term comes from Julian Rotter, whose 1966 monograph set out the idea and the questionnaire that measured it. What that questionnaire records is an expectancy about how the world works, not a verdict on your character, and more internal is not automatically better.

• Locus of control is a generalized expectancy: a belief, learned from experience, about whether outcomes follow from what you do.
• Internal and external are the two ends of one scale, not two kinds of people.
• Rotter’s I-E Scale asks 23 forced-choice questions plus six fillers. The total runs from 0 to 23, and a higher score means a more external locus of control.
• It is not self-efficacy. Bandura drew the line in 1977: locus of control is about whether actions produce outcomes, self-efficacy is about whether you can perform the actions.
• The links to better health are correlational, and “always be more internal” goes further than the research does.

What Rotter meant by locus of control

Rotter described the core of the idea in a short commentary he wrote in 1981, by which time the monograph had been cited more than 2,735 times. The effects of a reward or punishment on the behaviour that preceded it, he wrote, “depend in part on whether or not the reinforcement is perceived as contingent on the person’s behavior”. People learn differently when they think a result came from skill than when they think it came from chance. And, the claim that made the construct useful, “stable individual differences in expectancies for internal versus external control of reinforcement in the same situations can be validly measured”.

Three words in that summary do most of the work. Reinforcement is the behaviourist term for the outcomes that follow an action: praise, grades, money, failure. Contingent means the outcome depends on the action. And expectancy means the thing being measured is a forecast, not a feeling about yourself. Rotter called it a generalized expectancy because it is the forecast a person carries into new situations before they have evidence about how those particular situations work.

The idea did not arrive in one paper. Rotter credits the doctoral dissertations of his students E. Jerry Phares (1955) and William James (1957) with the first versions of the test, and says the questionnaire in the 1966 monograph was its fifth revision. He also says his own interest had been theoretical, and that the wider attention came from the “sociopolitical concerns” of the 1960s, when many people were acutely aware of how little control they had over events that mattered to them.

Internal and external locus of control are two ends of one scale

A person with an internal locus of control tends to expect that outcomes follow from their own effort and decisions. A person with an external locus of control tends to expect that outcomes are decided by chance, fate or other people with more power. Take an exam result:

• Internal reading: “I got a low mark because I started revising too late.”
• External reading: “I got a low mark because the paper was unfair and the marking is a lottery.”

The labels “an internal” and “an external” are convenient shorthand for people who score near either end, and the research literature uses them that way. But the scale is a sum of answers, so most people land somewhere between the ends rather than in one camp or the other. Rotter came back to how the construct was being read in a 1975 paper on “problems and misconceptions”. According to its abstract, it deals with how general or situation-specific the expectancy is, whether it is one dimension or several, and how test scores were being misused.

How locus of control is measured: the I-E Scale

Rotter’s Internal-External Scale, the I-E Scale, is the questionnaire most people mean when they search for a locus of control test or scale. It is a research instrument, not a diagnostic one.

Forced choice, 23 pairs and six fillers

Each item is a pair of statements, one reflecting an internal view and one reflecting an external view, and the person picks the one they believe more strongly. There is no option to agree with both. The published version has 23 scored pairs and six filler pairs, which are there to make the purpose of the test less obvious and are not counted.

What a high score means

The score is the number of external choices, so it runs from 0 to 23, and a higher score indicates a more external locus of control. That surprises people who expect a high score to be the good result. On Rotter’s scale, a high number means you chose the external statement more often, nothing more. Other locus of control questionnaires are scored in other directions, so a number means little until you know which instrument produced it.

One external, or two?

From the 1970s researchers questioned whether a single dimension was enough. Levenson’s 1973 scales split the external end in two: a belief that powerful others control your outcomes, and a belief that chance does. The resulting three scales, Internal, Powerful Others and Chance, are known as the IPC scales. The split matters because trusting a doctor, a parent or an institution to act is a different belief from thinking that results are random.

Locus of control is not self-efficacy

The two terms are routinely swapped, and Albert Bandura, who introduced self-efficacy, noticed. In his 1977 paper he wrote that locus of control “is often treated in the literature as analogous to self-efficacy”, and then separated them: “Rotter’s (1966) conceptual scheme is primarily concerned with causal beliefs about action-outcome contingencies rather than with personal efficacy.”

Bandura’s distinction rests on two different expectations:

• An outcome expectancy is “a person’s estimate that a given behavior will lead to certain outcomes”. This is the territory of locus of control.
• An efficacy expectation is “the conviction that one can successfully execute the behavior required to produce the outcomes”. This is self-efficacy.

You can hold one without the other. You can be sure that practice leads to better piano playing (a strongly internal expectation) and still doubt that you are able to practise enough (low self-efficacy). Or you can be confident in your own ability and still expect that nothing you do will change a decision someone else has already made.

Four terms that get confused with locus of control

Locus of control sits in a small cluster of near-synonyms. Each one has its own founding paper and answers a different question.

The difference between locus of control and attributional style is mostly one of direction. Locus of control is an expectation about outcomes still to come; attributional style is a habit of explaining ones that have already happened. Learned helplessness is closer still, and our article on what Seligman actually found about learned helplessness covers the original experiment and the 2016 paper in which its authors reversed their own explanation. A fifth term lives in a different vocabulary: among the cognitive biases, the fundamental attribution error describes how people explain other people’s behaviour, not how they forecast their own outcomes.

Terms this close together interfere with each other when you learn them, which is why a deliberate approach to memorising psychology terminology treats near-synonyms as a set rather than one by one. In Psychology World you can save terms to a named list and drill that list with the Question game, which needs at least four words. A set like the four above is exactly that size.

What the research does and does not show

Two findings come up most often, and both need their limits stated.

Health: associations, not causes

A 2023 study of 15,288 Australians in the national HILDA household survey found that a more internal locus of control went with higher self-control and better physical and mental health. The authors are explicit about what kind of evidence this is: “Our analyses are descriptive rather than causal.” The study also measured locus of control with a seven-item mastery scale rather than Rotter’s I-E Scale, a reminder that “locus of control” in a headline can refer to several different questionnaires.

Scores became more external between 1960 and 2002

A 2004 meta-analysis by Twenge, Zhang and Im pooled 97 samples of American college students (18,310 people) and 41 samples of children aged 9 to 14 (6,554 people). Between 1960 and 2002, average locus of control scores became more external by about 0.80 standard deviations. The average college student in 2002 was more external than 80% of college students in the early 1960s. That is a change in group averages over four decades, and it says nothing direct about whether one person’s locus of control can change.

Is an internal locus of control always better?

Self-help pages often answer yes. The sources above do not support that as a rule.

Bandura’s paper contains the clearest counter-example. A child who “fails to grasp arithmetic concepts and expects course grades to be dependent entirely on skill in the subject matter has every reason to be demoralized”. Believing that outcomes depend on you is no comfort when you doubt you can do what they depend on.

Bandura also separated two reasons people give up. Some lack confidence in their ability. Others are “assured of their capabilities but give up trying because they expect their behavior to have no effect on an unresponsive environment”. He noted that the two call for different remedies: the first for building competence, the second for changing the environment itself. When a situation really is outside someone’s control, an external reading of it can be accurate rather than a flaw to correct.

If a sense that nothing you do makes a difference has lasted for weeks or is getting in the way of daily life, that is worth raising with a doctor or a psychotherapist rather than working on alone with a questionnaire.