Responses and Scores

When submitting a response or fetching a score, each one has an associated code, associated type, and attached data. The score or response must be a json object whose fields include code, type, and data.

The code (of type string) serves as an identifier for scores and responses and should be used to index a response or score. Each type (of type string) determines what significance the response or score may have and, in some cases, also determines the structure of the data (of type object) field. Details on each type of response or score, what it means, and how its corresponding data is structured are included below.

For specific details on what scores and responses are part of a project and survey, please consult your personalized information packet.

Response Type

The following is a list of recognized response types. The enum value of the response type must be included in every uploaded response.

{
  "duration": 4.2
}

{
  "answer": "example text"
}

{
  "answer": 123.4
}

{
  "answer": "2022-01-15"
}

{
  "answer": "13:30:00"
}

{
  "answer": "2022-01-15T13:30:00Z"
}

{
  "answer": 3
}

{
  "answer": 0.5
}

Score Type

The following is a list of all recognized non-verbal score types. These types will be present when fetching scores via the Canary Speech API.

{
  "result": 50
}

{
  "result": 89
}

{
  "result": 4
}

{
  "result": [1, 2, 3]
}

{
  "result": 11.5
}

{
  "result": {
    "mean": 21.2,
    "median": 24,
    "mode": 17
  }
}

{
  "result": {
    "average": 30,
    "stddev": 1.2
  }
}

{
  "result": {
    "min": 0,
    "max": 100,
    "count": 23,
    "sum": 426,
    "mean": 18.5,
    "median": 23,
    "mode": 17,
    "stddev": 6.5
  }
}

Vocal Scores

In addition to calculated non-verbal scores, there are also scores based on vocal processing. The value and structure of these scores vary depending on the target model used to process the audio files and the score code. Your personalized information packet will contain detailed information, the most common scores are:

Behavioral Scoring Models

{
  "result": "Additional Screening Not Recommended"
}

{
  "result": "low"
}

{
  "result": "low"
}

{
  "result": "low"
}

{
  "result": "excellent"
}

Cognitive Scoring Models

{
  "result": "Additional Screening Not Recommended"
}

{
  "result": "negative"
}

{
  "result": "negative"
}

{
  "result": "healthy"
}

Misc. Scoring Models

{
  "result": 50
}

{
  "result": 50
}

{
  "result": 50
}

{
  "result": 50
}

{
  "result": "Low"
}

If your vocal score code isn't in this list, you may need to consult your personalized information packet for the desired labels and values.

Further Reading

• • Uploading responses
• • Fetching scores