Glossary

The vocabulary used across this site.

Cronbach’s alpha
How consistently the items on a scale agree with each other, from 0 to 1. Roughly .70 and up is usually considered adequate. Short scales have lower alpha because they have fewer items, not because they measure worse things.
Domain
One of the broad traits — Extraversion, Agreeableness and so on. Five of them in most models here, six in the HEXACO.
Facet
A narrower component of a domain. Conscientiousness contains Orderliness and Achievement-Striving, among others. Two people can share a domain score and differ sharply underneath it. The shorter inventories have no facets at all.
Inventory
A set of items measuring one or more traits, answered about yourself, with no correct answers. This site offers 6. A test, by contrast, has correct answers and you try to do well on it — why the distinction matters.
Item
One statement you rate. IPIP items are phrases rather than sentences by design — each completes an implied “I …”.
Keyed / reverse-keyed
Whether agreeing with an item counts toward the trait or against it. “Keep things tidy” and “Leave a mess in my room” both measure Orderliness, in opposite directions. Reverse-keyed items are flipped when scored, which is how an inventory detects someone agreeing with everything.
Level
The words low, average and high, cut at the 30th and 70th percentiles. Bands of a distribution, not kinds of person.
Norms & reference group
A table of how a named population scored, and the slice of it you are being ranked against. Every percentile on this site is relative to one.
Percentile
The share of that reference group scoring at or below you. A comparison, not a measurement — change the group and it changes.
Prorating
Scaling a partly-answered scale up to full length so it stays comparable. Used only when at least 80% of its items were answered.
Public domain
Free of copyright, usable by anyone for any purpose. Every personality item on this site is public domain, which is unusual — most well-known personality instruments are commercial products with licence fees. The MFQ-2's items are not public domain but openly licensed: free to use with credit to their authors.
Raw score
The sum of your keyed answers on a scale. This is the measurement. It never changes with who you are compared against.
Scale
Any group of items summed into one score. Both domains and facets are scales; so is the response scale you answer on, which is an unrelated use of the same word.
Self-selected sample
A group that chose to take part rather than being sampled at random. Every reference population here is one, which is the single most important caveat on this site.
T-score
The same comparison on a different ruler: mean 50, standard deviation 10. A T-score of 60 is one standard deviation above the group average.

Log in

To see the results you have kept.

Create an account

So your results are still here next time. Taking an inventory never needs one.