Page 51 - Kennemerland VOC ship, 1664 - Published Reports

K. MUCKELROY: WRECK SITES
I 11 111 IVa IVb IVc V VI VIII IX XI
- ~-~ -__ ~___~ -~ -- - -~-
69 0 700x0 X 0 12x X 0 14 X X X
78 X
80 X X X x o 0 0
82XX 84 X
85 X 86 X
0 00 0
88 X
0 , Present in quantity greater than average.
~ ~~~~
~
X, Present in quantity less than average.
Figure 3. Abundance/presence/absence matrix for 11 classes of artefact.
E
F
coefficient of 0.833 between classes IVb and XI), simply because of mutual absence. Since, on any archaeological site, there are likely to be a fair number of sparse artefact classes, there seems to be a stronger case to be made against the Simple Matching Coefficient. However, final judgment on this point is suspended until comparison can be made with a matrix produced by a totally different method.
The idea behind this second approach is that it should be possible to use the actual quantities recorded for each area, rather then the approximations used in the first method. However, one is immediately confronted with the point that, with a recording accuracy of the nearest metre, a grid of metre squares is probably too fine; it is likely that a sub- stantial number of finds are misplaced within this division. Furthermore, such a division with its 90 separate units, involves a very large amount of computation under any quantitative approach, representing so much wasted effort if the increase in definition is in fact spurious. Consequently, the first step on this occasion is to divide the site into 22 squares, each covering 4m”; the details of this division are indicated by the heavy lines in Fig. 4. Two empty squares, numbers 41 and 90, are excluded from the new framework. The new squares are designated by the letters
which identify the areas in Table 1, with numerical suffixes where necessary. On the basis of this material, a matrix of 1I columns (classes) and 22 rows (the new areas) is
~~ ~ ~
D
A
Figure 4. The numbers of the squares.
181