Significance testing
Maps to:
POST /v1/analyses/significance-test,POST /v1/analyses/significance-tests
A significance test tells you whether the difference between two datapoints — for example the same metric at two points in time, or for two brands — is statistically meaningful rather than noise.
Single test
POST /v1/analyses/significance-test
The request compares a baseline datapoint against a test datapoint at a
given confidence level. Each datapoint carries the figures you get back from an
execute result (volume, score, positives,
negatives, neutrals, and the combined perspectives).
{
"meta": { "version": "v1" },
"data": {
"metric": "buzz",
"percent": 95,
"baseline_datapoint": {
"metric": "buzz", "date": "2024-01-01",
"volume": 506, "score": 3.52,
"positives": 11.21, "negatives": 7.70, "neutrals": 72.43,
"positives_neutrals": 83.64, "negatives_neutrals": 80.12
},
"test_datapoint": {
"metric": "buzz", "date": "2024-02-01",
"volume": 521, "score": 4.93,
"positives": 11.81, "negatives": 6.88, "neutrals": 67.75,
"positives_neutrals": 79.56, "negatives_neutrals": 74.63
},
"alias": "jan-vs-feb"
}
}
percent— the confidence level (e.g.95).alias— optional label, echoed back so you can match results.
Response
The result reports a verdict per score type:
{
"meta": { "version": "v1" },
"data": {
"score": 1, "positives": 0, "negatives": -1,
"neutrals": 0, "positives_neutrals": 0, "negatives_neutrals": -1,
"alias": "jan-vs-feb"
}
}
| Value | Meaning |
|---|---|
1 | Significant increase from baseline to test |
-1 | Significant decrease |
0 | No significant change |
Batch tests
POST /v1/analyses/significance-tests
Same as above but data is a list of tests, and the response is a list of
verdicts in the same order. Use alias on each to keep them straight.