Sensostat Try the demo

Textbook · Module 07

Rapid Methods: Napping, Sorting, Flash Profile

Napping and projective mapping, free sorting, PSP, pivot profile, flash profile: when speed matters more than the full accuracy of a classical profile.

Module 07 of 11~20 min readinteractive + quiz

When the classical profile is too expensive

A full descriptive analysis means months of panel training and weeks of sessions. Rapid methods give a map of similarities and differences between products in a day or two, using untrained or semi-trained assessors. Speed is paid for with accuracy: rapid methods do not replace the quantitative profile where absolute intensities are needed.

1–2 daysinstead of 3–6 months of panel training
10–15experts or 50+ consumers
r ≈ 0.8convergence of rapid-method maps with classical DA

Napping® and projective mapping

The respondent receives all samples at once and arranges them on a sheet (a "tablecloth", nappe) by the principle: the more similar, the closer. Each respondent's coordinates are normalised and combined by multiple factor analysis (MFA) into a consensus map. Adding descriptions to sample groups turns napping into projective mapping (ultra-flash profile).

The napping tablecloth interactive

Imagine yourself as a respondent: arrange the six chocolates by taste similarity, dragging the circles. Press "Read the map" — the system will normalise the coordinates and show which samples you grouped (in a real study these coordinates go into MFA).

Drag the samples around the tablecloth.
How MFA builds one map from 60 "tablecloths". Each respondent gives their own coordinate matrix (products × 2 axes). Multiple factor analysis (MFA) first normalises each matrix — dividing it by the square root of its first eigenvalue — so a respondent with a "stretched" map does not outshout the others, while dimensions shared by all are amplified. The normalised matrices are then combined and subjected to PCA. Similarity between maps is measured by the RV coefficient (0–1): RV > 0.7 between napping and the classical profile is considered a good match of configurations. A typical quality check: compare the RV of the rapid-method map with the QDA map of the same product frame.

Sorting, PSP, pivot profile

Free sorting

The respondent arranges the samples into groups by similarity (any number of groups). The co-occurrence matrix → DISTATIS / multidimensional scaling. The simplest method for large sets (15–20 samples).

PSP (polarized sensory positioning)

Samples are rated relative to fixed reference "poles" ("how similar is it to pole A / B / C?"). This allows comparing data across sessions and weeks — the references stabilise the scale.

Pivot profile

One sample is the "pivot"; the others are described as "more/less than the pivot" on free attributes. A quick negative profile: what in the sample falls below the standard.

Flash profile

Each expert (10–15 people) chooses their own attributes and ranks all samples on each of them — with no lexicon agreement and no scale training. The analysis is generalized Procrustes analysis (GPA). A session takes from 30 minutes to 2 hours, and the resulting map correlates with classical DA at r ≈ 0.8: absolute intensity calibration is lost, the structure of differences is preserved.

Applicability limits

TaskIs a rapid method sufficient?
Competitor mapping, early recipe screeningYes — napping / sorting / flash profile
Consumer description "in their own words"Yes — CATA (Module 06)
Comparing batches on specific attributes with proof of significanceNo — DA with ANOVA is needed (Module 05)
Panel monitoring, scale calibrationNo — only a trained panel
A product sensory passport for a company standardNo — only a quantitative profile

Further reading

Test yourself

module 07 quiz

Three questions on rapid methods

1. In napping, the respondent…

Napping = "tablecloth": spatial arrangement → coordinates → MFA. Scales and temporal attributes are DA and TDS respectively.

2. Flash profile saves time mainly by…

Lexicon and training are 80 % of DA time. Flash profile sacrifices absolute calibration for speed; the analysis is GPA.

3. Can a rapid method substantiate a product sensory passport in a company standard?

Rapid methods give the structure of differences, but not calibrated intensities with confidence intervals. A normative document requires the latter.

And how does taste change over time?

Module 08 — temporal methods: TI, TDS and TCATA, dominance curves and significance lines.

Module 08 →