The behavioral science glossary

Plain-language definitions for the instruments, models, and terms behind Mappa's behavioral assessment for hiring.

See your own behavior map.

A behavioral assessment scored on published, peer-reviewed instruments. Thirty seconds of speech is all it needs — you get the full report, not a teaser.

Start the free assessment

Free · 30 sec of recording · Report is yours to keep and share

Jammil Handal
Client
“Mappa's process is smart, efficient, and consistently delivers the right fit for Priority's specific needs.”
— Jammil Handal, Executive Vice President · MXBuild Priority

Foundational concepts

Psychometrics

Hiring outcomes

Company culture

Assessment instruments

Speech & model terms

  • Speaker embeddingSpeech & model terms

    A speaker embedding is a numerical signature of a voice, used to identify which segments of a recording belong to a specific person and to confirm consistency across a conversation.

    A speaker embedding is an identity check, not a personality measure. It's built to answer "is this the same voice as before," robust to what's being said, which makes it useful for isolating a target speaker's segments but not informative about who that person is.

    How it shows up

    A speaker embedding can confirm the same person is speaking across a recording that includes background noise, a phone hold, or a change in microphone, none of which should change who the system thinks is talking.

    Why it matters

    Keeping identity verification separate from trait measurement is a meaningful design boundary: one answers "who," the other answers "how," and conflating them would be a mistake.

    See what stays constant about your voice, no matter the setting.

    Take the free assessment →