Question 1 Answers to part (i) were rather disappointing. Typical problems such as locating the sampling units from the air, the quality of the aerial search, and the accuracy of counting animals in the photos, etc, were seldom noted. Some candidates wrote about the errors caused by major movements of herd taking place during the survey, which was pleasing. Most candidates were able to estimate the total number of caribou in the herd and calculate the estimated variance V using the formula given in part (ii). However, there was some confusion with the standard error to be used in the 95% confidence interval, with some candidates using √ V n , rather than V ½. A few candidates lost marks through calculating an unweighted mean of the stratum means, as an estimate of the mean number of caribou. The merits of using stratified sampling in part (iv), and using optimal allocation in part (v) were well understood, but sometimes candidates wrote too generally, and not in terms of what the data showed for this survey