What Is a Psychometric Test?

A psychometric test is an objective, standardized measurement procedure used to quantify unobservable psychological phenomena based on a sample of behaviors.

Put more simply: it is a tool designed to put a number or a category on things that cannot be seen or touched directly, such as intelligence, anxiety or a personality trait, based on how a person behaves or responds.

Psychology professionals use these tools in clinical, educational and workplace settings to assess people’s characteristics with scientific rigor.

It works by recording a person’s responses to a series of stimuli (questions, tasks or situations designed for that purpose) to arrive at a score. When the test is well built and well used, that score makes it possible to describe, compare or make predictions with a known margin of error.

The foundations of psychological measurement

Psychometrics is the branch of psychology responsible for developing the theories and methods that support building and administering these assessments.

Its main challenge is making measurable something that cannot be observed directly: nobody can see someone’s extraversion the way they can see their height, so it has to be inferred from specific behaviors and answers.

To do this, specialists define the phenomenon they want to study (the construct) very precisely and identify the specific behaviors that serve as observable evidence.

The design also needs theoretical support that justifies the choice of questions or tasks. The goal is for the sample of behaviors analyzed to represent the phenomenon well: the questions should genuinely cover what is meant to be measured, not just a small or skewed slice of it.

Main types of assessment instruments

There are different ways of classifying these tools by format or purpose, but one fundamental methodological divide groups tests by how their answers are treated.

• Maximum-performance tests: assess a person’s competence, knowledge or aptitude, like a math exam or an intelligence test. They distinguish between correct and incorrect answers, and that is the basis for grading performance.
• Typical-performance tests: explore motivations, attitudes or personal preferences, how a person usually is rather than how much they know. There are no right or wrong answers, a structure common in personality questionnaires and inventories. When a test asks whether someone prefers to plan their vacation well ahead or improvise as they go, no answer is better than the other, only one that fits each person’s way of being more closely.

Scientific criteria for guaranteeing quality

The international reference in this field is the Standards for Educational and Psychological Testing, published jointly by the American Educational Research Association (AERA), the American Psychological Association (APA) and the National Council on Measurement in Education (NCME). The 2014 edition organizes test quality around two ideas: reliability and validity.

Reliability: consistency of measurement

Reliability describes how consistently and precisely a test carries out its measurement. If a person took the same test again under similar conditions a few weeks later, a reliable test should give a similar result, not a completely different one each time.

It is estimated in several ways, and each one answers a different question:

• Internal consistency: do the items of the same scale point in the same direction? The most widely used index is Cronbach’s alpha (Cronbach, 1951). An alpha of 0.80 on a ten-item scale means those questions vary in a fairly coordinated way, as if they measured one thing; for personality scales, values below 0.70 are usually considered weak.
• Test-retest stability: does the person get a similar score when retaking the test days or weeks later? This is expressed as a correlation between the two administrations. A correlation of 0.85 after four weeks means the ranking of people barely changes from one time to the next; for a test that assigns types, what matters is the percentage of people who get the same type the second time.
• Inter-rater agreement: when the score depends on human judgment (for example, scoring a drawing or an interview), do two scorers reach the same result?

A reliable procedure minimizes random error, but reliability says nothing about what is actually being measured.

Validity: measuring what you mean to measure

The second pillar is validity. The 2014 Standards define it as the degree to which evidence and theory support the interpretation of scores for their proposed uses.

A simple example: a bathroom scale can be very reliable, always showing the same weight, but if what we actually wanted was to measure someone’s height, that scale would be useless. It would be reliable, but not valid for that purpose.

That is why validity is not a fixed label attached to a test: what gets validated is a specific use. The Standards distinguish five sources of evidence that build up over time:

• the content of the items (whether they cover the construct well),
• the response processes (whether people understand and answer what is intended),
• the internal structure (whether statistical analyses find the dimensions the theory proposes),
• the relations to other variables (whether the test correlates with what it should and not with what it shouldn’t),
• and the consequences of using it to make decisions.

The process of interpreting results

Raw scores from a completed test have no meaning on their own. A “7 out of 10” says very little if you don’t know what to compare it with.

For the measurement to be useful, the professional needs to place that data within a structured reference system.

In norm-referenced assessments, the person’s score is compared with that of a representative reference group (for example, the adult population of a country) and is usually expressed as a percentile: the 70th percentile means the person scores higher than 70% of that group.

Criterion-referenced assessments, on the other hand, compare the result with a fixed standard (such as “can swim 25 meters without stopping,” yes or no) with no need to compare it with anyone else.

What all this means for the enneagram

The enneagram is a personality model, not a specific instrument: there are several different questionnaires for identifying a type, and each one would have to demonstrate its own reliability and validity.

The most comprehensive systematic review to date (Hook and colleagues, 2021, Journal of Clinical Psychology) analyzed 104 independent samples and found mixed evidence. On the plus side, some factor analyses partly match the theory, and the scales relate to Big Five traits in the expected ways. On the minus side, those analyses usually find fewer than nine factors, no study has derived the nine types using clustering techniques, and there is hardly any research supporting wings or movements between types.

As we read it, that places enneagram tests, including the one on this site, as tools for personal reflection rather than validated psychometric tests. If you want to see how to check your own result or explore the model in practice, you can take our enneagram test.

Sources

• American Educational Research Association, American Psychological Association, & National Council on Measurement in Education (2014). Standards for Educational and Psychological Testing. American Educational Research Association.
• Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16(3), 297-334. https://doi.org/10.1007/BF02310555
• Hook, J. N., Hall, T. W., Davis, D. E., Van Tongeren, D. R., & Conner, M. (2021). The Enneagram: A systematic review of the literature and directions for future research. Journal of Clinical Psychology, 77(4), 865-883. https://doi.org/10.1002/jclp.23097