Methodology
How PRISM measures
Every one of the 29 dimensions, the instrument behind it, and how a response becomes a score. What has not been validated yet is named at the bottom.
Measured today: 27 of 29. This page is generated from the scoring model itself, so it cannot drift from what the product actually does.
Cognitive6
- Fluid Reasoning
Your ability to spot patterns and rules in novel material, without relying on prior knowledge.
Matrix completion (Raven-style), procedurally generated per Wang & Su 2015
- Working Memory
How much information you can hold and manipulate in your head at once.
N-back task (Kirchner 1958)
- Processing Speed
How quickly you can take in information, filter what matters, and act on it.
Flanker task (Eriksen & Eriksen 1974)
- Attention & Inhibition
Your ability to focus on the relevant signal and resist responding to distractions.
Go/No-Go task
- Numerical Reasoning
How you handle quantitative information: ratios, percentages, simple inferences from data.
PISA-calibrated quantitative items
- Verbal Reasoning
How precisely you read complex text and distinguish what's stated from what's implied.
Watson-Glaser RED framework: Recognize / Evaluate / Draw
Behavioural5
- Risk Preference
How you trade off potential gains against potential losses under uncertainty.
BART: Balloon Analogue Risk Task (Lejuez et al. 2002)
- Planning & Forward-SearchNot yet measured
How well you think several moves ahead before acting.
Intended: Tower of London (Shallice 1982) — not implemented, and no score is produced for it.
- Impulse Control
Your ability to hold back a quick response when the situation calls for it.
Go/No-Go task
- Cognitive Flexibility
How easily you can shift mental strategies when the situation changes.
Card-sort task (Wisconsin-style)
- Reward LearningNot yet measured
How well you learn from feedback over many trials to find the better option.
Intended: Iowa Gambling Task variant (Bechara et al. 1994) — not implemented, and no score is produced for it.
Personality5
- Openness
How drawn you are to new ideas, aesthetic experiences, and unconventional approaches.
IPIP-NEO-120 Openness facets (Johnson 2014)
- Conscientiousness
How organized, reliable, and goal-directed you are: the strongest single Big Five predictor of job performance per Sackett et al. 2022.
IPIP-NEO-120 Conscientiousness facets (Johnson 2014)
- Extraversion
How energized you are by social interaction and external stimulation.
IPIP-NEO-120 Extraversion facets (Johnson 2014)
- Agreeableness
How cooperative, trusting, and oriented toward others' interests you are.
IPIP-NEO-120 Agreeableness facets (Johnson 2014)
- Emotional Stability
How calm you stay under pressure and how resistant you are to anxious or negative reactivity.
IPIP-NEO-120 Neuroticism facets reversed (Johnson 2014)
Interests6
- Realistic Interests
Your interest in hands-on, practical work with tools, machines, or physical materials.
O*NET Interest Profiler short form (Rounds, Su, Lewis, & Rivkin, 2010)
- Investigative Interests
Your interest in analyzing, researching, and understanding how things work.
O*NET Interest Profiler short form
- Artistic Interests
Your interest in creative expression, design, and aesthetic work.
O*NET Interest Profiler short form
- Social Interests
Your interest in helping, teaching, and working with people in supportive ways.
O*NET Interest Profiler short form
- Enterprising Interests
Your interest in leading, persuading, and pursuing business or organizational goals.
O*NET Interest Profiler short form
- Conventional Interests
Your interest in structured, detail-oriented work with clear procedures and data.
O*NET Interest Profiler short form
Work values6
- Achievement
How much you value visible accomplishment and using your abilities to their fullest.
O*NET Work Importance Locator
- Independence
How much you value autonomy, self-direction, and freedom from close supervision.
O*NET Work Importance Locator
- Recognition
How much you value visible status, advancement, and being seen for your work.
O*NET Work Importance Locator
- Relationships
How much you value co-worker connection and helping others through your work.
O*NET Work Importance Locator
- Support
How much you value supportive management and clear company policies.
O*NET Work Importance Locator
- Working Conditions
How much you value variety, security, and the material conditions of work (compensation, environment).
O*NET Work Importance Locator
Situational judgement1
- Workplace Judgment
How you handle the complex interpersonal and ethical situations workplaces produce: teamwork, learning, customer focus, integrity, and initiative.
MENA-contextualized Situational Judgment Test (Christian, Edwards & Bradley 2010 framework)
How a score is computed
- Correctness items (matrices, numerical, verbal, situational judgement) score as the proportion answered correctly, mapped directly onto 0–1.
- Self-report items (personality, interests, work values) average the answers on their scale, then rescale to 0–1. Reverse-keyed items are flipped before averaging.
- Timed tasks score on the measure the paradigm defines — mean reaction time, the flanker effect, hits against false alarms, adjusted pumps on unexploded balloons, completed categories — then normalise onto 0–1 against the published range for that task.
- Every score carries a confidence derived from how much of its instrument was actually answered. A dimension whose confidence is zero is reported as not measured rather than as a low score — not answering and scoring at the bottom are different things, and PRISM does not confuse them.
The exact band boundaries and the career-matching weights are not published here. The method is what makes the measurement reviewable; the constants would hand over the model without making it any more so.
What we have not done yet
- Bands are fixed thirds of the score range, not percentiles. "High" means the score fell in the top third of what the instrument can produce — it does not mean you scored above some proportion of other people. PRISM has not yet collected a normative sample, so no percentile claim would be true.
- No differential item functioning analysis has been run yet across language, gender or age. It is planned as the dataset grows, and we will publish what it finds — including if it is unflattering.
- Two of the 29 dimensions are not measured yet. They appear in every report marked as such, so nobody is shown a complete map that is missing rows.
- Career matching is a similarity score, not a prediction. It compares your 29-dimension profile against each career family's importance profile. It describes fit as the taxonomy defines it; it does not forecast what you will do or how well you will do it.