Introduction
Hiring is one of the most critical decisions in any organisation.
Every new hire reshapes team dynamics, impacts productivity, and influences culture and climate.
Yet too often, selection is approached with inadequate tools, vague criteria, or — worse — with reliance on gut feeling or “first impressions.”
When the wrong choice is made, the consequences are costly. According to the Society for Human Resource Management, a bad hire can cost up to 30% of the employee’s annual salary. But the damage isn’t just financial: it can lead to demotivation, loss of trust in the process, and team friction.
Artificial intelligence has changed the rules of the game
Impeccable résumés, persuasive cover letters, optimized LinkedIn profiles — today, everything can be generated or refined in seconds with generative AI tools.
Smart candidates don’t just prepare — they rehearse interviews with chatbots, generate AI-enhanced responses, and tailor profiles to bypass automated screening systems. Some tools even identify common recruiter biases, like similarity bias, confirmation bias, or the halo effect.
The risk? The more a recruiter relies on personal experience, the more vulnerable they may become. As shown by Kausel, Culbertson & Madrid (2016), overconfidence is one of the key factors that reduces decision-making accuracy in hiring: when we’re too confident in our impressions, we stop evaluating objectively.
The traditional recruiter is becoming a player at a rigged table.
This isn’t about blame, but about context: AI has raised the bar, and what once seemed like “common sense” can now turn into a systematic error.
What can HR specialists do in the age of AI?
Faced with increasingly optimized candidates and outdated traditional methods, how can recruiters safeguard fairness, ensure future performance, and protect organizational value?
Is it possible to select objectively, consistently, fairly — and cost-effectively?
Yes, but only if the selection process is built on solid, evidence-based foundations. In other words: only if the process is VALID.
VALID: The 5 Key Traits of a Scientific Selection Process
V – Valid
Definition A selection process is valid when the criteria and tools used statistically predict the candidate’s future job performance.
This predictiveness is not subjectively determined by the recruiter, but is the result of empirical research, statistical analysis, and replicated studies published in scientific literature.
Plain explanation Saying a tool is "valid" doesn’t mean it seems to work — it means there’s objective data proving it works.
The recruiter doesn’t decide what works: the studies do.
A structured interview, for example, has been shown to predict performance well. A firm handshake, on the other hand, has not.
📚 Note for HR Specialists: Predictive validity is established through statistical analyses that correlate selection scores with later job performance. It’s not a matter of perception, but of measurable, replicable data.
A – Accurate / Reliable
Definition A selection process is accurate and reliable when it produces the same results under equivalent conditions.
In other words, if the same candidate undergoes the process twice, or is assessed by different recruiters, they should obtain the same (or very similar) score.
Plain explanation Imagine taking the interview today, and then again in a week, or being interviewed by someone else: if your results differ, the issue isn’t you — it’s the tool.
A reliable process is stable and consistent. Otherwise, selection becomes a lottery.
📚 Note for HR Specialists: Reliability is measured with indicators like test-retest reliability or inter-rater agreement. A process that isn’t stable across time or assessors can’t be considered scientifically sound.
L – Logical / Quantifiable
Definition A selection process is logical when it produces a numerical score that allows candidates to be compared — and, if well designed, to be compared against a normative sample.
Plain explanation When deciding among candidates, we need comparable numbers — not just impressions.
A strong selection process assigns each candidate a score derived from valid, reliable tools. Ideally, this score should not only rank candidates but also position them — as a percentile — against a larger reference population, such as similar job roles.
📚 Note for HR Specialists: Quantification ensures transparency, comparability, and documentation. Where possible, using percentiles against normative benchmarks gives meaningful context.
I – Investment-worthy
Definition A selection process is investment-worthy when the benefits it generates (performance, retention, team climate) justify the costs.
Plain explanation Every selection process has a cost: people’s time, tools, tests, follow-ups. But it also has value: hiring the right person can transform performance, collaboration, and morale.
An effective process is sustainable not because it’s the cheapest — but because it’s the most efficient: worth the cost, and delivering results.
📚 Note for HR Specialists: Investment is measured over time. A higher upfront cost may generate far greater long-term savings by avoiding turnover, poor fit, or cultural misalignment.
D – Differentiating
Definition A selection process is differentiating when it reliably distinguishes between candidates.
This means that score differences reflect real differences in future job performance.
Plain explanation If two candidates score very differently, we expect — on average — different performance outcomes.
In this sense, to “discriminate” is positive: it identifies the best match, based on fair and justifiable criteria.
A process that scores everyone similarly, or can’t tell who has real potential versus who just makes a good impression, is of little use.
📚 Note for HR Specialists: In hiring, "to discriminate" means to distinguish objectively and justifiably. A fair process rewards true potential and avoids arbitrary or homogenized decisions.
How VALID Are Today’s Common Hiring Practices?
| Method / Tool | V A L I D | Why |
|---|---|---|
| CV + cover letter | ❌ ⚠️ ❌ ✅ ❌ | Easily manipulated with AI, low predictive value, qualitative only |
| Unstructured interview | ❌ ❌ ❌ ⚠️ ❌ | High bias, no consistency or measurability |
| STAR-based interview | ⚠️ ⚠️ ⚠️ ✅ ⚠️ | Helps candidate structure answers, but not recruiter judgment |
| Structured interview with rubric | ✅ ✅ ✅ ⚠️ ✅ | High validity and reliability if well-designed, costly but sustainable |
| Standardized GMA tests | ✅ ✅ ✅ ✅ ✅ | Strong, replicable predictors, norm-referenced |
| Assessment center with experts | ✅ ✅ ✅ ⚠️ ✅ | Powerful but expensive — best for critical hires |
| AI-based CV screening | ⚠️ ❌ ✅ ✅ ❌ | Cheap but unreliable, easily gamed with smart prompts |
| Generic personality tests | ❌ ✅ ⚠️ ⚠️ ⚠️ | Valid in measuring traits, but often not performance-related or resistant |
| Job-targeted personality tests | ✅ ✅ ✅ ✅ ✅ | Focused on predictive traits, with bias controls |
📌 Note on Personality Tests:
Many personality tests are psychometrically valid — but this doesn’t mean they’re useful for hiring.
A scientific selection process requires role-specific tools, measuring only those traits that predict job performance. Otherwise, the recruiter receives a rich but irrelevant profile — once again subject to personal bias.
Tests must also include lie/desirability scales to detect social desirability and self-deception, ensuring results reflect the real person — not an idealized version.
At HURACT, we’ve developed a system to generate custom personality tests from the job description itself — identifying only the traits that matter for that role.
Conclusion – Returning to Science to Choose Wisely Again
In a world where AI helps candidates appear perfect and traditional methods show their limits, intuition-based hiring is becoming dangerously ineffective.
The VALID model is more than an acronym: it’s a critical filter, a practical guide, and an ethical safeguard.
Any process that isn’t valid, reliable, logical, investment-worthy, and differentiating exposes the organization to costly mistakes and dissatisfaction.
Returning to science, evidence, and rigorous design isn’t a luxury — it’s a necessity for responsible, transparent, visionary hiring.
In today’s complex job market, smart selection is your true competitive advantage.
Want to Design a VALID Selection Process?
In the course “Human Resource Selection & Assessment – Scientific Method & AI” by HURACT, you will learn to:
- Design a selection process based on scientific evidence
- Use validated tools and understand their limits
- Integrate AI wisely, without falling prey to it
- Build structured interviews, tests, and rubrics that really work
- Critically assess and improve your current practices