As Higher Statistics students at James Gillespie's High School in Edinburgh, we have recently looked at statistical sampling techniques and would like to take issue with two points raised by Laurence Olins regarding pesticide use in Scotland (Pesticide use not as widely used as suggested, Letters, October 20).

Mr Olins said the relevant survey "only represents 32% of crops grown in Scotland". By our understanding, the random stratified sampling used is a sound statistical technique which can give valid results and it would be unusual, costly and more time-consuming to perform a census of all holdings.

Secondly, we would question the assertion that nearly half of the holdings surveyed were less than five hectares in size. The published data, on page 41 of the report (Soft Fruit Crops 2012, published by the Scottish Government), would seem to show that of all the holdings surveyed, 23 of 495, or about 4%, were in this smallest size category, compared with 13% in the whole of Scotland. It seems that if anything, the smallest holdings are under-represented. We wonder if perhaps the figure is obtained from looking only at results from the north, where of 12 holdings surveyed, five were in the smallest size category.

Umar Amjad, Adam Cheong-Macleod, Kumin Kang, Stuart Kumar, Yang Liu, Andrew MacNish Porter, Rebecca Wu

James Gillespie's High School

Edinburgh