Question

1. At Banff National Park, it is necessary to take a random sample (without replace- ment) of 25 of the 809 tagged bears which live near the town of Ban↵. The person who normally does the statistics is hiding, and so someone suggests going to the visitor centre parking lot, looking at the rightmost 3 digits in the licence plates of the first car in the lot, and repeating this process until they have 25 distinct values from 1 to 809.

   What would you need to assume, in order for this sample to be credible as an as-if random sample? Explain, briefly.

2. It is claimed that salmon found downstream from logging operations have become discoloured. For one week, people fishing on the river downstream of the logging are asked to bring any fish caught to a field station for a brief inspection. At the end of the week, 18% of the 243 salmon caught are found to be discoloured.

   a) Is this a survey or an experiment? Explain, briefly. b) Would it be safe to generalize from the 234 salmon examined here to all salmon in this river downstream of the logging operations? Briefly explain why it would be safe to generalize, or provide a reason for why generalizing would be unwise.

3. Suppose that your instructor decides to compare two di↵erent methods of getting students using R. Each student can choose to attend one of two sessions (W 7-8pm or Th 6-7pm) and is subject to Method A in the Wednesday session or Method B in the Thursday session. Each student can only attend one session, and at the end they are tested (same test in both cases) to see how much they have learned to do.

   a) Is this a survey or an experiment? Explain, briefly. b) This would not be a very good way to compare two methods of introducing R. Suggest two problems with sampling which would make it dicult for me to generalize from any results here about which method is most e↵ective.

Answer 1

To consider the sample selected by using licence plates of the first car in the lot, we should assume that the licence plates have digits 0 to 9 and the number of lots should not be very small because if the number of lots will be very small then there is a high chance that the same number will be sampled.

a) This is a survey because we are not applying any treatments.

b) It would not be safe to generalize from the 234 salmon examined at the field station to all salmon in the same river downstream of the logging because the speed of the downstream may vary at other stations of logging operations.

Good luck with your studies!!!