FAQ - How reliable and valid is the AQme assessment?

Short answer

Reliability and validity are two different questions, and they need two different answers.

Reliability asks whether the assessment measures consistently. AQme has published reliability evidence: internal consistency across the sub-dimensions, most in the acceptable to excellent range, with three scales currently marginal and under active revision.

Validity asks whether it measures what it claims to measure. AQme has strong content and structural evidence for this. It does not yet have published criterion validity, meaning studies linking AQme scores to later workplace outcomes. Those studies are planned.

If a client asks you the validity question, the honest and credible answer is that we can evidence what the instrument measures and how consistently, and that outcome-linkage research is underway. Saying that plainly builds more trust than overclaiming.

What reliability evidence exists

Our published reliability analysis was conducted by Dr Nicolas Deuschel, Assistant Professor of Management at Universidad Carlos III de Madrid, on 4,849 AQme completions.

Cronbach's alpha measures internal consistency, which is how closely the items within a single scale relate to one another as a group. Conventional interpretation is roughly:

Across the AQme sub-dimensions, most scales fall in the acceptable to excellent range. Work Stress and Mental Flexibility are the strongest. Three scales sit at or below the conventional .70 threshold and are the subject of an active item improvement programme: Unlearn, Emotional Range and Extraversion.

Two of those three, Emotional Range and Extraversion, are two-item facets. Alpha is not the appropriate statistic for a two-item scale, and the correct measure is the Spearman-Brown coefficient. The Unlearn figure in the published report was calculated on a reduced sample during the beta phase.

We name these openly because a reviewer will find them, and because an assessment provider that reports its weakest scales is more credible than one that reports only its strongest.

What validity evidence exists

Validity is not one thing. Test standards recognise several types of evidence, and it is worth knowing which we can demonstrate today.

Content evidence: strong

AQme was built over two years of research across hundreds of studies, in collaboration with universities, professors and specialists in psychology, people analytics and human behaviour, including partnership with IE Business School and IE University in Spain.

Each sub-dimension is anchored in an established research literature. Grit draws on Duckworth and Seligman. Hope draws on Snyder. Motivation Style draws on Higgins' Regulatory Focus Theory. Mindset draws on the optimism and expectancy literature. The environment sub-dimensions draw on perceived organisational support, psychological safety and job demands-resources research.

An important distinction to hold: drawing on a literature is evidence that the construct is well defined and that our items target the right territory. It is not the same as having used another author's validated instrument, and we do not claim their validation results as ours. AQme items are our own.

By design, roughly 30% of the diagnostic content was developed from interviews with senior HR professionals, authors and practitioners rather than from peer-reviewed studies. This was a deliberate choice to keep the instrument connected to current workplace reality, and it is documented publicly in Decoding AQ.

Structural evidence: strong

If the 15 sub-dimensions genuinely measure different things, they should relate to each other moderately rather than either not at all or almost perfectly. Published correlation analysis across 3,569 assessments shows exactly that pattern. Related constructs correlate as theory predicts, for example Team Support with Work Environment, while remaining clearly distinct from one another. No sub-dimension is redundant with another.

Further structural analysis on the current global dataset is in preparation.

Relations to other measures: partial

The sub-dimensions relate to each other in the directions and approximate strengths the underlying theory predicts. This is meaningful internal coherence.

What we do not yet have published is concurrent validation against established external instruments, meaning studies correlating AQme scales with independent published measures of the same constructs. This is on the research roadmap.

Criterion and predictive validity: not yet evidenced

We do not currently have published studies linking AQme scores to subsequent workplace outcomes such as performance, retention or change adoption. Prospective studies with client partners are planned.

Until those studies exist, we do not describe AQme as predictive of outcomes, and partners should not either.

What we do not claim

Being precise here protects you in client conversations.

• We do not claim AQme predicts individual job performance.
• We do not claim the assessment is validated for selection, hiring, promotion or redundancy decisions, and AQai does not license, support or train it for those uses.
• We do not claim AQme is free of self-report bias. It is a self-report instrument measuring self-perception. Coach-led debriefs and free-text responses provide qualitative triangulation, and certification training frames interpretation accordingly.
• We do not claim other researchers' validation results as evidence for AQme.

Intended use

AQme is designed and supported for individual development, coaching, team insight and organisational understanding of adaptability. It is a conversation starter and a development tool, not a gate.

Scores are best read as ranges rather than exact points, and band boundaries as zones rather than lines. Any single sub-dimension score should be interpreted in the context of the whole profile and the individual's role, goals and situation. A profile that looks lower on one dimension is not a deficit. As covered in certification, some contexts reward caution and consistency as much as others reward exploration.

How to answer this in a client conversation

Several partners have asked how to handle this question live. A structure that works:

1. Separate the two questions. "Those are actually two different things, and the answer is different for each. Can I take them separately?" This immediately signals you know the field.

2. Answer reliability with the evidence. Internal consistency analysis on nearly 5,000 completions, conducted by an academic third party, with most scales in the acceptable to excellent range and a named improvement programme for the weakest.

3. Answer validity by type. Content evidence is strong and traceable to established literature. Structural evidence shows the sub-dimensions are distinct and relate as theory predicts. Criterion studies linking scores to outcomes are planned and not yet published.

4. State the intended use before they ask. "It is built and supported as a development and coaching instrument, not a selection tool." Clients who were quietly worried about that will relax, and it demonstrates that we understand where the evidence bar sits.

5. Offer the documentation. Share the Cronbach Alpha Reliability Report, which is cleared for partner and client distribution. If the client has an internal occupational psychologist who wants to go deeper, route the request to hello@aqai.io rather than improvising.

The instinct in these conversations is to defend. Resist it. Naming a limitation before the client finds it is the single most credibility-building move available, and it is usually the moment the conversation turns.

What is coming

An expanded AQme Technical and Scientific Manual is in preparation, covering the full current dataset. It will add refreshed reliability including McDonald's omega, confirmatory factor analysis, measurement stability over repeat assessments, fairness and group-difference analysis, and full documentation of the norming and banding methodology.

We will notify partners through this support centre when it is available and confirm what may be shared with clients.

Related articles

• Academic Research
• FAQ - What research structures were used in developing the AQ Assessment?