Why Traditional Assessments Often Fail

You’ve taken one of these tests: you know how it goes. Twenty minutes of multiple choice on a Tuesday afternoon. An email confirms you completed it. Then silence. The HR team had your results before you finished the coffee. You never saw them. You’re not entirely sure what they measured. And the hiring decision, six weeks later, was made on grounds that were never quite explained to you.

The “traditional assessment” from the candidate’s side.

From the HR side it looks similar — a number arrives in a folder, with a percentile that nobody on the team can fully interpret, used to nudge a decision that has already been mostly made by the CV.

Both sides wonder if it is useful.

Hiring asks three questions about a person. Always the same three, whether or not the hiring manager knows they are asking them.

How will this person work?

Some people thrive with structure and follow-through. Others need flexibility, variety, room to experiment. Neither is better. But the mismatch between someone’s natural style and the role they walk into is one of the biggest predictors of early attrition.

How will this person handle pressure?

Work isn’t just tasks. It is relationships, stress, feedback, ambiguity. How someone regulates themselves under load — how they read other people, how they manage disagreement — determines whether the role survives the first hard quarter.

How will this person grow?

Job descriptions are now obsolete faster than they take to write. The ability to absorb new context, update mental models, and keep moving when the role moves is no longer a leadership-track concern. It is a baseline selection criterion for everyone.

Most assessments answer one of these questions, badly. And you neither could really operationalise what you may understand from the answer. The personality inventory tells you something about how the person works. The cognitive test tells you they can think. The leadership-potential questionnaire offers a glance at growth, usually based on self-report. The pieces sit on separate elements and you are left to staple them together by feelings.

When the assessment literature looks at what actually predicts hire quality, it gives the same answer it has been giving for fifty years. Cognitive ability plus personality plus structured behavioural evidence — the composite — produces validity that no single test can touch. Schmidt and Hunter’s foundational meta-analysis showed this in 1998. Sackett and colleagues have refined the estimates since. The single-instrument approach has never been the best the science can offer. It is the cheapest. It is what the apparatus could deliver before AI made comprehensive assessment affordable.

Share

The self-report problem. Most learning-agility and emotional-intelligence inventories ask the candidate to rate themselves. Every candidate says yes to “I learn from experience.” Every candidate says yes to “I manage my emotions well.” The signal you most need — actual behavioural evidence — is precisely what the format cannot deliver. Researchers like Hoff and Burke have made the point bluntly: what most commercial learning-agility instruments measure is not behaviour. It is the candidate’s account of their behaviour.

The fragmentation problem. The three questions hiring actually asks cannot be answered by three separate tests bolted together. They need an integrated reading. A high-conscientiousness candidate with low emotional regulation under pressure is a different bet from a high-conscientiousness candidate with high regulation — but only the integration tells you that. Most platforms sell you three dashboards and call the integration your problem.

The candidate-experience problem. The candidate completes the assessment and never sees the result. Their data went somewhere. A decision was made. They are left guessing why. This is not a minor ethical complaint. It is a structural one — an assessment process that does not return value to the candidate trains every candidate to give the assessment as little authentic information as possible. The system rewards the strategic answer, not the honest one. We then complain about gaming.

What comes next is an assessment that answers the three questions together, returns the result to the candidate, and treats authenticity as the goal rather than the enemy.

For NoesisHiring this is TRIO. Three assessments designed as one system. They are not bolted together — they were built to talk to each other.

• COB — Conscientiousness-Openness Balance — maps where the candidate sits on the spectrum between structure and flexibility. Built on the Big Five tradition, it surfaces the archetype, not a score. There is no “best” placement on the spectrum. There is only the question of fit between the person’s natural balance and what the role actually asks of them.

• EQ — measures the five dimensions Daniel Goleman set out: self-awareness, self-regulation, motivation, empathy, social skills. The candidate’s behaviour under pressure is read, before asked.

• Learning Agility — draws on Lombardo and Eichinger’s research, looking at mental, people, change, and results agility. This is the dimension that distinguishes the candidate who matches today’s role from the candidate who will match the role six months from now, after it has changed.

Three assessments. Three answers. One integrated reading.

And — here is the move that the traditional assessment never makes — every candidate gets their results. The full profile, not a verdict. A description of how they work, how they handle pressure, how they grow, written for them to read. They are required to validate them: how do you feel about yourself in this description? Face Validity is the stronger, and more operational confirmation you want when you have to truly understand someone.

The ultimate value is then visible: If they get the role, they walk in already knowing what the role will demand of them and how they themselves may be prone to. If they do not get the role, they walk away with something most assessments have never given them: a deep insight into themselves that someone else paid for.

Authentic answers become possible because authenticity is now rewarded. There is no “right” placement on COB. There is no reward for a specific level of EQ in every role. There is no learning-agility threshold that gates the candidate. There is only a profile, and the question of how this specific person’s profile can bring value in this role.

The science that supports this is not new. The integration that has been missing was operational, not theoretical. The cost of running three deep assessments on every applicant used to be prohibitive. It no longer is. The cost of returning the results to every candidate used to be impossible at scale. It is now a default. What was never affordable became affordable, and the assessment apparatus has not yet caught up with what it can do.

If you want to know what your own profile looks like — what kind of structure-and-flexibility balance you sit in, how you tend to handle pressure, what your learning pattern looks like — you can take it yourself. It takes about half an hour. You’ll get the results back the same day. We’re in the validation phase, so it’s now free.

[noesishiring.ai/NH-ass/assessments.html

Antonio Specchia is a researcher, Routledge author, and the creator of the Applicant Relationship Management (ARM) framework.

He is the founder of NoesisHiring, an AI-powered recruitment platform. His first book, Customer Relationship Management (CRM) for Medium and Small Enterprises (Routledge, 2022), laid the foundation.

A forthcoming book, Applicant Relationship Management (ARM): Human–AI Collaboration in the Agentic Economy, is in preparation for Routledge. His research on Applicant Relationship Management has been presented widely accepted among academics and HR professionals.