A psychometric assessment is a process for measuring, using standardized instruments, what a job requires of a person, and using that data in conjunction with the CV and the interview. It is organized into six steps: defining what to measure, choosing the tests, establishing norms or scales, administering the test, interpreting the results, and providing feedback to the candidate. If one step is missed, the score loses value even if the test is performed well.
Table of Contents
What differentiates a psychometric assessment from a standalone test?
The test is the instrument. The evaluation includes everything surrounding it: why it's administered, who each person is compared to, under what conditions they respond, and what is done with the results. That's why two companies can use the same test and obtain very different quality information.
The Standards for Educational and Psychological Testing AERA, APA, and NCME explain it in terms of validity: it refers to the degree to which evidence and theory support interpretations of scores for their intended uses. What is validated is a specific use. A well-constructed test can be misused.
If you're looking for the catalog, we already reviewed it in the guide. Psychometric tests for employment: types and how to choose them. Here we go with the organization, which can be summarized as follows:
| Passed | Question that must be answered | What should be put in writing |
|---|---|---|
| 1. Define | What decision do you support and what needs to be measured? | Job criteria |
| 2. Choose | What tests cover those criteria? | Battery and justification |
| 3. Standards | Who is each person compared to? | Reference group and date |
| 4. Apply | Do they all respond under comparable conditions? | Notice, instructions and procedure |
| 5. Read | What does the score say and what doesn't it say? | Report and weight of each source |
| 6. Return | What should be explained to the person? | Who spoke to her and when |
Step 1: What do you want to measure and for what decision?
It starts with the decision. Pre-selecting in a mass process, choosing between finalists, or planning employee training require different information, and the same set of data rarely works for all three.
The Guidelines on the use of tests from the International Test Commission (ITC) requires a reasoned justification for using tests, a job analysis, and proof that what is being measured relates to the behaviors that matter in that context. In practice, this fits on one page:
- Write down the tasks or behaviors that make a difference in the position.
- Separate what is required from day one from what can be learned through training.
- Mark what is best observed using other sources, such as the interview, references, or a work sample.
- Write down how the result will be used and what will happen to those who fall short of expectations.
If you already have one competency matrix, This is a good starting point, because the work of describing the position has already been done.
Step 2: How to assemble the battery without making it too long?
Each test must answer a criterion from step 1. A good check is to ask yourself what decision would change with that result: if you can't find any, that test is unnecessary.
Three ways to keep it short and useful:
- Avoid measuring the same thing twice. Two scales that point to the same trait lengthen the session and contribute little.
- Ask for the manual before deciding. The ITC expects that the technical documentation will allow for judging normative groups, reliability and validity for relevant populations, and potential bias, as well as user acceptance and the time required.
- Do it yourself first. Answering the full battery tells you how much she weighs, whether the instructions are understandable, and whether she feels connected to the position.
In Hirint, for example, the Psychometric and technical tests for companies in Mexico They are combined into a battery for each position and provide comparable results across candidates. Deciding who gets in remains your team's job.
Step 3: What standards or benchmarks should be used for comparison?
A raw score alone tells us little. A norm, or scale, is the table that translates that score into a position within a reference group. A 70th percentile, for example, means that the person scored at or above the 70th percentile of that group. If you change the group, the interpretation changes.
The AERA, APA, and NCME Standards require that norms refer to clearly described populations and include the groups with which the user intends to compare. They also distribute responsibility: the test developer describes the norms, and the user applies them appropriately.
What to check in a scale
- Population: who makes it up and whether it resembles your candidates.
- Country and language: if there is data from Mexico or a comparable population.
- Date: when the data was collected.
- Sampling: how the people were selected and how many participated.
- Level or position: Comparing a manager to the general population can inflate your result.
What if there is no Mexican scale?
Ask the provider what evidence they have for using that standard with your population. The ITC cautions against drawing conclusions from comparisons with standards that are irrelevant to the people being assessed or that are outdated. If there is no evidence, state this in the report and give less weight to that comparison.
There are alternatives, too. For verifiable skills, such as Excel or English, it usually makes more sense to compare against a level required for the position. And an internal benchmark, developed with your own employees, shows how someone ranks within your company, though not necessarily in the market.
Step 4: How to apply so that the results are comparable?
Before inviting: notice and consent
The ITC recommends informing people before the test about its purpose, nature, who will receive the results, and how they will be used. In Mexico, the Federal Law on the Protection of Personal Data Held by Private Parties It regulates the privacy notice and requires express written consent to process sensitive data. If a test could reveal such information, or if you use a job reliability test, Review the applicable regulations with your legal department before launching. This guide does not replace that advice.
During: modality and instructions
The ITC guidelines for computer and internet testing They distinguish four modalities according to the level of supervision:
| Mode | Supervision | Where it usually fits |
|---|---|---|
| Open | None; anyone can access it. | Orientation, low importance |
| Controlled | Without direct supervision; only known people, with their own access | Remote pre-selection |
| Supervised | Someone grants access and confirms that it was applied correctly. | Final stages |
| Managed | High control of conditions, like an assessment center | Very important decisions |
Whichever method is chosen, everyone must receive the same instructions, a clear timeframe, and a channel for resolving technical questions. If you apply without supervision for a significant decision, the ITC recommends a subsequent verification under supervised conditions, comparing the scores.
It also provides for adjustments for disability. According to the ITC, these should be tailored to each individual case and affect the validity of the score as little as possible, and the interpreter must be aware that adjustments were made when concealing the disability could bias the interpretation.
Step 5: How to read the results without overinterpreting them?
Here, a good piece of evidence can lose its value. Before opening the reports, agree on the weight each source will have: the Standards require a clear explanation of the role of each component when integrating scores with other information to make a decision.
When reading, keep these points in mind, which summarize the ITC guidelines on the use of tests:
- All scores have a margin of error. A small difference between two people may not mean anything, so it's worth checking the reliability of the scale.
- Don't generalize. A result in numerical reasoning does not speak to traits that the test does not measure.
- A cut-off point needs evidence. It should only be used if there is evidence of validity to support it for that use.
- Read the person in their context. Schooling, culture, or prior experience with the test can influence the result.
- Review the unlikely results. The ITC recommends procedures for detecting them before interpreting.
In terms of personality, there's also the social desirability bias: people tend to respond in ways they believe are expected of them. Consistency indicators help identify this, and any such indicator should be treated as a topic to explore in the interview, never as proof that someone is deceiving the interviewer.
According to the ITC, the interpreter must be qualified to do so, and the report must clearly state that the data is only one source of information. The decision is made by a person, based on the CV and the interview.
Step 6: How to give feedback to the candidate?
It's an easy step to skip when the process is rushed. The ITC points out that the person being assessed should know they will receive an explanation of their results as soon as possible and in easy-to-understand terms, and that the oral feedback should be constructive.
How it's done depends on the size of the application. In a large process, a written summary might suffice for those who don't advance, while the conversation can be reserved for the finalists. The important thing is to decide this from the beginning and mention it in the invitation.
A good return includes:
- What was the evaluation used for and what did it measure?.
- Strengths and areas for development, in everyday language.
- What weight did it have in the decision, along with the other sources?.
- Who to contact if the person has doubts or wants to exercise their rights regarding their data.
Avoid providing the items or answer keys, using labels that sound diagnostic, and comparing the person to other candidates. A helpful phrase might be: "Your results show an aptitude for organizing information and less of a preference for negotiating under pressure; for this position, the latter was more important.".
Next: data and review
Define who has access to the results and how long they are stored, as required by the ITC. With internal development or mobility, it is also advisable to clarify whether the results can be reused: the 360-degree evaluation and other sources help to avoid relying on a single piece of data.
Finally, review the process periodically. The ITC recommends monitoring for signs of adverse impact and reassessing the evidence if you change the purpose, form, or delivery method of the test. If you'd like to see how a battery of tests is organized by job position, you can explore Hirint's platform and request a meeting.
When it comes time to choose instruments and tools, these can be useful. What does the Raven's test measure? and how to choose psychometric testing software.
Frequently Asked Questions
What is psychometrics?
Psychometrics is the discipline that studies how to measure psychological characteristics, such as aptitudes, traits, or values, and how to verify that these measurements are reliable and valid. A psychometric assessment applies these principles to a specific decision. In recruitment, that decision might be whom to interview or which candidate best fits the requirements of a position.
What do they ask in a psychometric exam?
It depends on what it measures. Aptitude tests present reasoning problems, personality tests ask you to indicate how accurately certain statements describe you, and technical tests assess job knowledge. Specific questions are not usually published because disclosing them would compromise the test's validity. The person administering the assessment should explain beforehand what kind of tasks the person will encounter.
How long does a psychometric assessment take?
It depends on how many tests the battery includes and what they are like. Each instrument's manual usually indicates its administration time. Rather than looking for an ideal number, it's better to ask whether each test justifies its time with a piece of data that would change the decision, and to inform the candidate of the total duration in the invitation.
Can psychometric tests replace the interview?
No. The assessment provides comparable data, and the interview explores context, motivation, and career path. ITC guidelines recommend presenting the results as one more source of information, alongside others. Use must comply with local law, informed consent, and data protection regulations. The final decision rests with the individual.
What is the difference between psychometric and psychometric tests?
Especially the vocabulary of each country: in Mexico they talk about psychometric tests and in Chile about psychometric tests. In both cases, the aim is to assess skills, traits, or competencies for work. What matters is the same: evidence of validity and reliability for the intended use and norms appropriate to the population being assessed.