# What is the difference between reliability and validity?

> Written with AI · [Psychology: research methods and statistics](https://lessoons.com/en/psychology-research-methods) · Grade 10 · 2026-11-13 · https://lessoons.com/en/psychology-research-methods/what-is-the-difference-between-reliability-and-validity

The difference between reliability and validity is consistency versus accuracy. A measure is reliable when it gives the same result again under the same conditions, and valid when it measures what it claims to measure. A measure can be reliable without being valid, like a scale that is always two kilograms off, but it cannot be valid without being reliable.

## The topic explained

### Two different questions about a measure

Every questionnaire, test or observation scheme in psychology is judged by two questions. Reliability asks whether the measure is consistent: would the same person, measured again in the same conditions, get the same result? Validity asks whether it measures what it claims to: does a stress questionnaire measure stress, or only how willing people are to complain? So the difference between reliability and validity is the difference between consistent and accurate.

A target makes the difference visible. If every shot lands in a tight group away from the center, the shooting is reliable but not valid. If the shots land in a tight group on the center, it is both. If they are scattered, it is neither, and then it cannot be valid either, because a measure that jumps about cannot be hitting the right spot on purpose.

### How reliability is checked

There are three common checks. Test-retest reliability: the same people take the test twice, some weeks apart, and the two sets of scores are correlated. Inter-rater reliability: two observers or markers score the same behavior or the same answers on their own, and their scores are compared; it matters for open questions and observations, not for multiple choice. Internal consistency: the items of one test should agree with each other, because they are meant to measure the same thing.

Each check gives a coefficient from 0, no consistency, to 1, perfect consistency. Israel's National Institute for Testing and Evaluation uses the rule of thumb that a high-stakes test should reach at least 0.80. Its 2016 report gives the psychometric entrance exam a median internal consistency of 0.95 for the general score, so the exam is very reliable.

### Kinds of validity

Face validity asks whether the measure looks right to the people taking it. Content validity asks whether it covers the whole topic, usually judged by experts. Construct validity asks whether scores behave as the theory says they should, for example rising with other measures of the same trait. Predictive validity asks whether scores forecast a later outcome. Ecological validity asks whether a finding from a study holds in everyday life.

The psychometric exam again gives real numbers. Its general score correlated 0.43 with first-year university grades, against 0.36 for the bagrut average, and 0.46 for the two combined, in a sample of 100,863 students. These are moderate correlations, the usual size for predicting human behavior, and evidence that the exam adds something beyond the bagrut.

### Reliable is not the same as valid

The most common mistake is to write that a study is valid because it was repeated and gave the same result. Repetition shows reliability only. A bathroom scale that always shows two kilograms too much is perfectly reliable and never valid. A personality quiz can give each person the same type every time and still not predict anything about them. Reliability is a condition for validity, but it is not enough: a good answer names which kind of each is meant and gives the evidence for it.

## Practice: reliability and validity

For 10th grade psychology students. Answer each question from the page, then check your work against the answers.

1. **In one sentence each, define reliability and validity.** — Reliability: a measure gives the same result again under the same conditions. Validity: a measure measures what it claims to measure.
2. **Describe each target: reliable, valid, both or neither?** — First: reliable but not valid (tight group off the center). Second: reliable and valid (tight group on the center). Third: neither (scattered shots).
3. **Write the question each kind of validity asks.** — Face: does it look right to the people taking it? Content: does it cover the whole topic? Construct: do the scores behave as the theory predicts? Predictive: does it predict later results? Ecological: does the finding hold in real life?
4. **Two teachers mark the same 20 essays and give very different grades. Which kind of reliability is low, and how could it be raised?** — Inter-rater reliability is low. It can be raised with a clear marking scheme, examples of essays at each level, training the markers together, and checking a sample marked by both.
5. **Read the chart. Which predicts first-year grades best, and what kind of validity does the chart show?** — The combined score, 0.46, predicts best, then the psychometric general score, 0.43, then the bagrut average, 0.36. The chart shows predictive validity: how well a score forecasts a later outcome.
6. **A bathroom scale always shows 2 kilograms more than your real weight. Is it reliable? Is it valid? Explain.** — It is reliable, because it gives the same reading every time, but it is not valid, because the reading is not your real weight. Consistency alone does not make a measure accurate.
7. **A website offers a 10-question quiz that claims to measure how good a leader you are. Suggest how you would test its reliability and its validity.** — A good answer suggests test-retest reliability (the same people take the quiz twice, weeks apart, and the scores are correlated) and predictive or construct validity (correlate scores with real leadership ratings by others, or with an established leadership measure). It notes that looking sensible, face validity, is not enough.

## Common questions

### Can a test be valid but not reliable?

Not really. If a test gives different results each time, it cannot be measuring the right thing consistently, so its validity is limited by its reliability. That is why test developers check reliability first.

### What is the difference between internal and external validity?

Internal validity asks whether a study really shows that the independent variable caused the change, with no confounding variable. External validity asks whether the result applies to other people, places and times, of which ecological validity is one part.

### What is a good reliability coefficient?

For a test used to make important decisions, a common rule is at least 0.80, the rule Israel's National Institute for Testing and Evaluation applies to its exams. Research questionnaires are sometimes accepted a little lower.

### Is a laboratory experiment low in validity?

It is often high in internal validity, because conditions are controlled, but may be lower in ecological validity, because the task and setting differ from everyday life. A field experiment trades some control for more realism.

### How is reliability calculated?

Usually as a correlation. For test-retest, the two sets of scores are correlated; for inter-rater, the two raters' scores; for internal consistency, a statistic compares all the items. Each gives a value from 0 to 1.
