Top, parameters and encoding of the small multiples visualisation components .
Values for the six parameters (a1 to a6) are reported along axes starting from the centre of the square (growing values outwards).
On each axis, the maximum value for a parameter across the collection is reached when reaching the circle. In this example, value is zero for axes a1 to a4, axes which correspond to the expression of a doubt. By contrast, values for a5 and a6 are always positive, and a yellow triangle connecting them to the centre is drawn filled.
The seventh parameter – duration of life (b) – is represented by an arc, with growing dating clockwise (midday is year 1000, 6 o’clock year 1600). A dotted line points out doubts concerning the dating. Finally, in the upper right comer of the square little black circles identify various possibilities concerning the first stage of the artefact’s evolutionary chain (c).

Bottom, the combination of small multiples (ordered by type of construction and estimated duration of life).
The small multiples visualisation gives a synthetic, comparison enabling overview of 8 parameters, with a ninth parameter (construction type) represented through the background colour. Each square icon represents a given edifice’s “knowledge pattern”. Edifices are aligned by construction type, and then ordered by decreasing duration of life.

At first glance, when analysing the visualisation, some edifices do clearly compare to others, in particular inside groups corresponding to a given construction type (colour of background) : observe for instance D1 and F1, E2 and F2, A3 and B3, D3 and D4, D6 and E6.
Similar patterns can also be spotted across groups: C3 compares to F2, D6 and E6, I3 compares to B6, B2 compares to D1 and F1, etc.. Beyond one to one comparisons, clusters can be observed in or across groups: (E3, G3, A4, E4); (A1, D1, F1, H1, B2), etc..

Groups as such do have particularities: typically the filled yellow triangle - corresponding to the overall amount of transformations vs. certain transformations - is significantly smaller in the case of wood constructions (brownish background). Observing the whole collection also underlines a pattern : the duration of life is apparently not the determining factor in the amount of information and of transformations : compare for instance B2 and D2 (B2 is older, but with less transformations than D2).

But besides helping to spot patterns inside a collection, such a visualisation can also be fruitful in underlining significant exceptions, “outliers” inside a group. C2 for instance shows significant differences with the other edifices on line 2 – an indication that the “knowledge pattern” for this object is slightly different.