Database Contents
Database Contents
Participants
The first version of NZSED includes speech data from six speakers in groups defined by the following variables:
- Region (Wellington and Hamilton)
- Ethnicity (Maori and Pakeha)
- Sex (Male and Female)
- Age ("younger", "mid-age" and "older")
Read more about the speaker profile data being collected.
Materials
The materials are based on those recorded for ANDOSL (the Australian National Database of Spoken Language), and include both read material and dialogue. The components are:
- 200 phonetically rich sentences (as in ANDOSL) (2434 words/speaker; 8685 phonemes/speaker)
- 10 reference sentences
- 19 hVd words (as in ANDOSL)
- Isolated digits (as in ANDOSL)
- Two map-task sessions (as in ANDOSL) (to date c 290 minutes for 34 speakers)
- Word lists related to maps
- Interview component ("Discuss you opinions of the Road Safety ads on radio and television") (to date c 140 minutes for 34 speakers)
Note: the following phonetic variables are targeted in the reference sentences.
| -ing | t-voicing | final glottals |
| NEAR/SQUARE diphthongs | uptalk/High Rising Terminal | l-vocalisation |
| h-dropping |