Answers

1.  
   1. No, this statement is not reliable. This statistic is a result of an inappropriate estimation because of its unrepresentative sample. Choosing Oshawa as a representative population neglects regional variations such as type of industry (e.g. resource-based, tourism, manufacturing, high-tech) or employment (e.g. seasonal, permanent). In order to have a representative sample, the statistician must gather information from many different areas across the country, including rural, semi-urban and urban areas. Return to question 1a
   2. No, this statement is not reliable. Participants in a TV station phone-in poll select themselves voluntarily, which can lead to an unrepresentative sample. Return to question 1b
   3. This statement may or may not be reliable. Its reliability is questioned because we do not know the source of the statistics about school leavers. Return to question 1c
   4. Again, this statement may or may not be reliable. We do not know the origin of this information. However, if the survey was conducted properly nationwide, and if 'youth' is defined as those people between the ages of 15 and 24 years, then there is no reason to state that this statistic is incorrect. Return to question 1d
   5. No, this statement is not reliable. If the statistics are conflicting, this could mean that different definitions of "forest cover" are used or that one of the sources may be flawed, or that the results have been misinterpreted by one of the parties. Return to question 1e
2. The two types of non-sampling errors are random errors and systematic (or bias) errors. Return to question 2
3. Sampling error is influenced by the size of the sample and of the population, the variability of the characteristic of interest in the population, the sample design and the estimation method. Return to question 3
4. Systematic errors can result from
   • coverage errors
   • response errors
   • non-response errors
   • processing errors
   • estimation errors
   • analysis errors.
   Return to question 4