Measure what matters most: the human side of performance
Ready to use libraries + easy integration of your prefered models
Technical skills get people hired. Soft skills determine how far they go and how well teams, leaders, and cultures actually function. Synethos is designed from the ground up to make soft skill assessment as rigorous, flexible, and evidence-based as any competency evaluation.
Thanks to the openness of the questionnaire builder, any soft skill assessment can be integrated, whether you bring your own framework, use an established instrument, or build something entirely new with AI-assisted psychometric guardrails.
Validated assessments, ready to deploy
Synethos comes with a library of pre-integrated, scientifically validated soft skill instruments — drawn from the most widely used and replicated measures in occupational and personality psychology. Rather than reinventing the wheel, you can deploy proven instruments directly within the platform and benefit from decades of psychometric validation. Usage of some instruments may be subject to royalty agreements with their authors or publishers — our team can advise on licensing requirements for each.
A 30-item self-report instrument developed by K.V. Petrides measuring global trait emotional intelligence across four broad dimensions: emotionality, self-control, sociability, and well-being. Covers 15 EI facets including empathy, emotion regulation, adaptability, and impulse control. Widely used in occupational and organisational settings for leadership selection and development.
The most scientifically established model of personality, assessing five broad dimensions: Openness to experience, Conscientiousness, Extraversion, Agreeableness, and Neuroticism (OCEAN). The IPIP-NEO is the public-domain implementation developed by Lewis Goldberg, available in 50-item and 120-item versions. Extensively validated across cultures and strongly predictive of workplace performance, leadership, and team fit.
Developed by Angela Duckworth and Patrick Quinn, the 8-item Grit-S measures perseverance of effort and consistency of interest toward long-term goals — a construct shown to predict achievement in demanding professional contexts beyond IQ and conscientiousness alone. Validated across West Point cadets, Spelling Bee finalists, and diverse professional populations. Particularly relevant for roles requiring sustained high performance under pressure.
Developed by Schaufeli and Bakker, the UWES is the most widely used and internationally validated measure of work engagement. The 9-item version assesses three dimensions: vigor (energy and resilience), dedication (enthusiasm and sense of significance), and absorption (full concentration and immersion). Validated in over 30 countries and highly predictive of individual performance, retention, and organisational wellbeing outcomes.
Developed by Shalom Schwartz, the PVQ measures an individual's hierarchy of 10 universal motivational values — including self-direction, achievement, benevolence, security, and universalism — using an indirect portrait-comparison format. Validated across 80+ countries, it provides a stable predictor of attitudes, decision-making styles, and cultural fit. The revised PVQ-RR extends coverage to 19 more finely differentiated values.
A 10-item scale developed by Morris Rosenberg measuring global self-worth and self-acceptance. One of the most widely used and translated personality instruments in the world, with over 20,000 published studies. In professional contexts, self-esteem is linked to resilience, leadership presence, conflict management, and the ability to recover from setbacks — making it a relevant complement to performance-based assessments.
Julian Rotter's foundational instrument assesses the degree to which individuals attribute outcomes to their own behaviour (internal LOC) versus external forces such as luck or powerful others (external LOC). Internal locus of control is consistently associated with higher job performance, proactivity, career success, and leadership effectiveness. A 29-item forced-choice format reduces social desirability bias.
Rooted in Deci and Ryan's Self-Determination Theory, the BPNS assesses satisfaction of the three fundamental psychological needs: autonomy, competence, and relatedness. Highly predictive of intrinsic motivation, wellbeing, and sustainable engagement at work. Used to diagnose whether team or role contexts support or undermine motivation — relevant for development conversations, job design, and retention risk assessment.
The Synethos questionnaire builder is fully open. If you use a proprietary tool, a framework developed internally, or a licensed instrument not listed here, it can be integrated with the same AI-supported scoring, psychometric guardrails, and reporting as the pre-built instruments. Bring your methodology — we provide the infrastructure.
Beyond self-perception: performance-based soft skill testing
Most soft skill assessments ask people how they see themselves. While this provides useful data, it introduces a well-documented and serious source of bias: social desirability — the tendency of respondents to answer in ways that make them appear more capable, agreeable, or emotionally mature than they actually are in practice.
Synethos addresses this bias directly by enabling performance-based soft skill assessments — questionnaires in which respondents must demonstrate a skill through their response, rather than simply report how skilled they believe themselves to be. This is the same logic that distinguishes a coding test from a question about coding confidence.
AI-driven psychometric guardrails — quality before the data is collected
Synethos is built on a simple but powerful principle: a questionnaire that is poorly designed cannot be fixed after the fact. The damage is done the moment a respondent encounters an ambiguous question, a double-barrelled item, or a set of questions that bleed into each other’s constructs. Synethos addresses this at the source with AI-powered psychometric guardrails that act ex ante — before any data is collected — to identify and correct the most common measurement flaws automatically.
AI reviews each item for clarity, neutrality, and structural soundness. It flags double-barrelled questions (asking two things at once), leading phrasing that biases responses, ambiguous wording that respondents interpret differently, and items that are too abstract or too concrete to capture the intended construct reliably.
Items that are supposed to measure the same competency should cohere — they should be theoretically and semantically related. AI checks that each item in a competency scale genuinely reflects the underlying construct being measured, and flags items that are outliers within their own scale, reducing noise and improving internal consistency before the assessment goes live.
Items measuring different competencies should be clearly distinct from each other. AI identifies cross-loading risk — cases where an item theoretically assigned to one competency (e.g. resilience) is semantically so close to another (e.g. stress management) that it effectively measures both. This cross-contamination inflates apparent correlations and distorts the competency scores you actually rely on for decisions.
For open-text items, the quality of the AI scoring outcome depends entirely on the quality of the scoring instruction. Synethos reviews the scoring prompt alongside the question itself, ensuring the criteria are specific, anchored to observable response characteristics, and calibrated to the difficulty level of the item — so AI scores are consistent, explainable, and defensible.
Beyond individual items, AI evaluates the overall measurement model: does the questionnaire as a whole have an appropriate number of items per competency? Are the competencies well-differentiated? Is the overall cognitive load on the respondent reasonable? This holistic review ensures the instrument functions as a coherent whole, not just a collection of individually acceptable questions.
A valid assessment should discriminate across the full range of the trait being measured. AI flags items that are likely to be too easy (answered the same way by almost everyone) or too hard (producing random responses rather than meaningful signal), and suggests reformulations that improve discrimination without compromising accessibility or fairness.
Ex-ante quality assurance — not ex-post damage control
Most organisations discover questionnaire flaws through poor data — low inter-item correlations, floor or ceiling effects, or results that simply do not predict what they should. By then, the assessment has already been administered, and the decisions have already been influenced by flawed data. Synethos guardrails run before anyone completes the questionnaire, giving HR professionals the confidence that what they deploy is fit for purpose from day one.