Types of selection assessments: a guide to choosing by role

When someone decides to professionalize their hiring process, they usually run into a huge catalog of assessments and the same question: which one do I use? The answer isn’t about finding “the best assessment”, but about understanding what question each type answers and combining the ones the role needs. A poorly chosen assessment isn’t just a waste of time: it’s noise that clouds the decision.

Selection assessments group into four psychometric families —cognitive, personality, competencies, and integrity— plus the technical assessments that measure a concrete skill. Each family answers something different, and the key to a good process isn’t the quantity, but the right combination for the position’s profile.

The four psychometric families

1. Cognitive (or aptitude). They measure reasoning capacity: numerical, verbal, abstract, learning speed. They answer “how fast does this person learn and solve problems?”. They’re especially useful in roles where you have to learn fast or face new situations.

2. Personality. They describe behavioral styles —models like DISC or the five major factors—. They answer “how does the person tend to behave?”. They don’t measure ability; they provide context about the style. To choose between models, it’s worth reviewing DISC vs Big Five.

3. Competencies. They assess concrete behaviors relevant to a role: customer orientation, leadership, teamwork, communication. They answer “does the candidate show the behaviors this role needs?”. They’re the bridge between general personality and what the position specifically asks for. We explain it in what behavioral competencies are.

4. Integrity. They estimate the propensity toward counterproductive behaviors at work. They answer “what risk is there in roles of trust?”. They matter where there’s handling of money, inventory, sensitive data, or security. More in integrity controls.

Beyond psychometrics: technical assessments

Alongside the four psychometric families are the technical assessments, which measure a concrete skill: Excel proficiency, command of a language, knowledge of a programming language. They don’t describe how the person reasons or behaves; they confirm that they know how to do a specific task of the role. They complement the psychometric ones: the technical assessments verify the “knows how to do”, the psychometric ones describe the “how”.

How to choose according to the role

The golden rule is to define the role before the assessment. Some examples of combinations by profile:

• High-volume operational role (cashier, operator): a basic cognitive measure + key competencies + integrity if there’s handling of money or inventory.
• Commercial role (salesperson, account executive): sales competencies + personality + cognitive.
• Technical role (developer, analyst): a technical assessment of the skill + cognitive + the role’s competencies.
• Role of trust (finance, security): integrity + cognitive + specific competencies.

In every case, the principle is the same: combine the families that position needs, not apply everything “just in case”. More assessments isn’t better; a better combination is.

How many to apply (without overwhelming the candidate)

Applying too many assessments lengthens the process, worsens the candidate experience, and rarely improves the decision. Applying just one usually leaves important questions unanswered. The reasonable point is in the middle: the usual approach is to combine between two and four assessments that, together, cover what the role needs. A good candidate experience is also part of a good process.

In short

Selection assessments group into four psychometric families —cognitive (aptitude), personality (style), competencies (role behavior), and integrity (risk in roles of trust)— plus technical assessments, which measure a concrete skill. No family is “the best”: each answers a different question, and a good process combines the ones the role really needs, usually between two and four, without overwhelming the candidate. The decision starts by defining the position’s profile and only then choosing what to measure, always reading the results in context and with the team’s judgment.