CSAT score: top-box formula, scales and margin of error
CSAT is the share of survey responses in the top boxes of the scale. The number only means something once you state the scale, the boxes counted and the sample size.
The formulas
CSAT = responses in top boxes ÷ all responses DSAT = responses in bottom boxes ÷ all responses MoE (95%) = 1.96 × √(p(1 − p) ÷ n) × 100 points n for ±5 = ⌈1.96² × p(1 − p) ÷ 0.05²⌉
The default is top 2 boxes on 5-, 7- and 10-point scales, and top box only on a 1–3 scale, where "top 2" would include the neutral midpoint. Under 30 responses, the library returns no margin of error because the result is directional only.
Worked example
400 responses on a 1–5 scale: 12, 18, 40, 150 and 180 for scores 1 to 5.
- CSAT = share of responses scoring 4–5 on a 1–5 scale (top 2 boxes): 82.5% ± 3.7 points (78.8–86.2%).
- DSAT 7.5%, average score 4.17 out of 5.
- Responses needed for ±5 points: 222.
Try it
Usage
import { csatStats, parseCsatScores } from 'contact-center-math/csat'
const { counts } = parseCsatScores('5 4 5 3 5 4 2 5', 5)
csatStats(counts, 5) // top-2-box by default