Exploration and description

Numbers of corms

Applying similar methods of analysis to measurements of numbers of corms per plot we find that the number recorded for subplot 55 is four times its standard error. Three other observations are also listed but their deviations are smaller. Again no reason could be found for this anomaly and so this measurement was replaced by an * in the 'Edited data' spreadsheet in CS16Data1a.

Running the analysis of variance again and producing the residual plots we see that the histogram shows signs of skewness to the right and that the normal plots are slightly curved and deviate from the 45⁰ line. Measurements of counts, such as in this case, number of corms tend to follow a Poisson distribution. A square root transformation is likely to produce a distribution closer to normality.

Calculating the variable sqrt_corms (Spread → Calculate → Column... and rerunning the analysis of variance we can see that the transformed data fit better the assumption of normality.

For number of corms we shall therefore analyse the transformed values.