QUB | Archaeology and Palaeoecology | The 14Chrono Centre

Manual for psimpoll and pscomb

Menu

Menu M: data analyses

Menu Mh: independent splitting
Walker & Wilson (1978) argued that individual curves for pollen types should be considered independently of each other, especially where the data were pollen accumulation rates, and therefore statistically independent. They presented an approach, with FORTRAN programs by Y. Pittelkow, for making independent splits of the records for individual taxa in a sequence. The approach involves first identifying portions of the record that are effectively non-zero sequences, distinguishing them from other portions of effectively zero sequences. This may be termed presence-absence splitting. The approach then looks at the non-zero sequences, and splits them quantitatively with a maximum likelihood method (Walker & Wilson 1978). Pittelkow's programs implemented the approach in two stages. In psimpoll I have combined these, making presence-absence splits first, and then looking for quantitative splits within non-zero sequences. The detection of either sort of split is a statistical matter, depending on a level of significance that is a function of the number of samples (Walker & Wilson 1978, Table 2) and there may be no splits of either sort for any given taxon. Splits are ignored if they would create `zones' of less than 4 samples. Walker & Pitelkow (1981) present results using the method from three late Quaternary sequences. Walker & Wilson (1978) present curves fitted to portions of the record of particular taxa, identified from presence-absence and quantitative splitting. However, this was not included in Pittelkow's programs, and I have not implemented it in psimpoll. Instead, I give means and standard deviations for each identified segment of the record for each taxon.

Results are printed to the screen (possibly too fast to read when plotting interactively), and to the data analysis file (which must have been specified: menu Ec). Lines marking the location of splits are marked across the plot of each taxon. These are normal lines for quantitative splits, but dashed lines for presence-absence splits. If pollen zones are also being incorporated, these are indicated only by boxes at the righthand end of the plot: marker lines for the zones are not drawn across the plot.

Independent splitting is carried out as the program is `plotting' each taxon to the main output file. It is thus necessary to go through with this stage in order to get any output in the data analyses file. As presently set up, the process will be carried out on all `taxa' except summary data, pollen sums, and rate-of-change results. I leave it to the user to decide whether it is useful or reasonable for items such as palynological richness!

This option toggles `on' and `off'.

Return to main menu.


Back to contents page
Copyright © 1995-2007 K.D. Bennett

Archaeology and Palaeoecology | 42 Fitzwilliam St | Belfast BT9 6AX | Northern Ireland | tel +44 28 90 97 5136


Archaeology and Palaeoecology | The 14Chrono Centre | URL http://www.qub.ac.uk/arcpal/ | WebMaster