The vocabulary of personality assessment, defined plainly — including the terms that make this category look worse rather than better.
This glossary defines common terms in personality assessment and hiring assessment in plain language. It covers the Big Five model, how results are read, what assessment quality means, and how these ideas are used in recruitment. The terms are definitions, not claims that any test can remove human judgement.
The Big Five, also called the five-factor model, is a way of describing personality across five broad areas. It does not sort people into boxes. It describes patterns of behaviour, preference and tendency on continuous dimensions.
HyperPersonal AI uses the Big Five as the underlying model for its personality work. The site treats those scores as decision support — not automated decisions.
To see the candidate-facing assessment that uses this model, read the app page. Big Five personality test app
A domain is one of the broad areas in a personality model. In the Big Five, the domains are the five main headings under which more specific patterns sit.
A domain is useful for orientation, but it can be too broad for a hiring conversation. HyperPersonal AI therefore reads both domains and more specific facets.
A facet is a narrower part of a broader personality domain. For example, a broad domain may contain facets related to organisation or assertiveness.
Facets help a manager ask better questions. HyperPersonal AI measures 30 facets, but it does not turn them into labels for a person.
IPIP stands for International Personality Item Pool. It is a public-domain pool of personality questionnaire items used by researchers and test builders.
Public-domain does not mean every use is equally good. The value still depends on how items are presented, scored, interpreted and used.
IPIP-300 is a 300-item questionnaire drawn from the IPIP item pool. It is designed to measure five broad domains and 30 narrower facets.
HyperPersonal AI is built on the public-domain IPIP-300 item pool. In the iPhone app, a candidate can answer a few questions a day rather than complete the whole questionnaire in one sitting.
For the fuller explanation of the model and its limits, read the science page. Science behind HyperPersonal AI
OCEAN is a memory aid for the five Big Five domains: Openness, Conscientiousness, Extraversion, Agreeableness and Neuroticism. Some versions use different wording, such as Emotional Stability instead of Neuroticism.
The acronym is useful, but it can make the model look simpler than it is. A hiring conversation often needs the facet level, not just the five letters.
A trait is a relatively stable tendency in how a person thinks, feels or behaves. It is not a promise that someone will act the same way in every setting.
Good use of trait information keeps the job context in view. A trait result is a starting point for a question, not the answer to whether to hire.
A type system puts a person into a category. A trait system places a person along a dimension, where most people sit between the extremes.
HyperPersonal AI uses traits, not types. It does not say a candidate is one kind of person and therefore suited or unsuited to a role.
For a plain comparison of common workplace models, read the guide. Big Five versus MBTI versus DISC
A categorical result puts a person into a named group, such as a type, colour or style. It can be easy to remember and easy to misuse.
Categorical results often hide variation inside the category. HyperPersonal AI does not report personality as types or categories.
Confidence is the degree to which a result should be treated as stable enough to discuss. It is not the same as certainty.
In assessment, confidence can be affected by missing answers, careless responding, unusual response patterns or a weak connection between the measure and the decision. A cautious result can still be useful if it is handled as a question to explore.
A continuous dimension is a scale with degrees rather than fixed boxes. Someone can be higher, lower or near the middle on the same underlying trait.
HyperPersonal AI reports scores as continuous dimensions. This is why the output is better read as evidence for a conversation than as a pass or fail.
A percentile describes where a score sits compared with a reference group. A higher percentile means the score is higher than more people in that comparison group.
A percentile is not a mark out of 100. It is a relative position, so it depends on who the person is being compared with.
A reference population, or norm group, is the group used to interpret a person’s score. It gives the comparison that makes a raw score meaningful.
Changing the reference group can change the interpretation. Any report that uses comparative language should make clear what comparison is being made.
Self-report means the person answers questions about themselves. Many personality questionnaires are self-report instruments.
Self-report has limits. People may misunderstand an item, answer in a way that presents them well, or lack a clear view of their own behaviour. That is why results should be combined with structured interviews, work evidence and human judgement.
Careless responding is answering without proper attention to the item. A person might click quickly, choose randomly, or stop reading the questions closely.
This matters because the report may look precise while resting on weak input. Any serious assessment process should have ways to notice unusual or low-quality response patterns.
A construct is the thing an assessment is trying to measure. In personality assessment, a construct might be a broad trait or a narrower facet.
The construct should be named before the result is interpreted. If no one can say what is being measured, the score should not drive a hiring decision.
A longstring is a run of identical answers across many items. For example, someone may choose the same response option again and again.
A longstring can be a sign of careless responding, fatigue or deliberate non-engagement. It should not automatically be treated as dishonesty.
Protocol validity concerns whether a completed assessment record is usable. It asks whether the response pattern looks attentive enough to interpret.
This is separate from whether the person is a good candidate. A weak protocol says something about the data quality, not the whole person.
Reliability is about consistency of measurement. A reliable measure is less likely to change because of noise in the items, timing or scoring process.
Reliability does not mean the result is important for a job. A measure can be consistent and still irrelevant to the role being hired for.
Response style is a pattern in how someone uses the answer scale. Some people tend to agree with statements, avoid extremes, or choose strong answers often.
Response style can affect how results are read. It is one reason not to treat a single score as a complete account of a candidate.
Retest refers to taking the same or similar assessment again after time has passed. It is used to understand how stable results are.
A change in retest results can come from real change, context, mood, memory, misunderstanding or measurement noise. The meaning depends on the assessment and the situation.
Straight-lining is choosing the same position on a response scale across many items. It is one visible form of a longstring.
It may mean the person was tired, disengaged or rushing. It may also reflect confusion with the task, so it should be handled carefully.
Adverse impact is a pattern where a selection process disadvantages a protected or legally sensitive group. It can arise from tests, interviews, referrals, scorecards or any other hiring step.
This glossary does not state legal requirements for any jurisdiction. If adverse impact is a concern, it is the thing to raise with a qualified adviser before using an assessment in selection.
Alignment means the assessment is connected to the work that will actually be done. The role should be defined before the candidate is interpreted against it.
In HyperPersonal AI, the organisational portal reads candidates against a role the company has defined. That role definition is the anchor for the discussion.
Decision support gives information to a human decision-maker. It does not make the decision itself.
HyperPersonal AI is decision support, never automated decisions. A human makes the hiring decision, using the report alongside the CV, interview, work sample and references where relevant.
Facet weighting means treating some facets as more relevant than others for a specific role. The weight comes from the role, not from a general idea of a good employee.
Weighting should be explicit before results are read. Otherwise, people tend to notice the scores that confirm what they already think.
A scorecard is a structured way to record evidence against role requirements. It helps interviewers compare candidates on the same criteria rather than on memory or general impression.
A scorecard does not remove judgement. It makes the judgement more visible, which makes it easier to challenge before an offer is made.
A structured interview asks candidates the same planned questions for the same role. It uses defined criteria rather than a loose conversation that changes with each person.
Personality results can help shape follow-up questions, but they should not replace the interview. These are starting points for a conversation between people.
A work sample is a task that resembles part of the job. It shows how a candidate approaches actual work, not just how they talk about it.
For a small company, a simple work sample can be more useful than another unstructured interview. Personality data should sit beside that evidence, not above it.
No. It defines general terms used across personality assessment and hiring assessment. Where HyperPersonal AI uses a term in a specific way, the entry says so plainly.
No. A Big Five result is one source of evidence. It can help a manager ask sharper questions, but the hiring decision remains with a human.
Yes. A candidate’s profile is theirs, and an organisation sees it only through an explicit, revocable grant. Access is not assumed because the candidate once applied for a role.
Pricing is USD 10 per employee seat. Candidate profiles are separate from that seat price because the profile belongs to the candidate, not the organisation.
Operated by Alano Tech Pte. Ltd., Singapore.
Terms of Service Privacy Policy
Also from us: Alano.ai Opptymizer Omu Labs