1. Which of the following is NOT true?

1. Summary measures of location capture important features of the whole distribution of data for a variable.
2. Summary measures of location are easier to remember than the original data.
3. Summary measures of location indicate the ‘centre’ of a set of data.
4. Every data point contributes equally to a summary measure of location.

2. Which of the following measures of location is guaranteed NOT to change by adding a single extreme data point?

1. mode
2. median
3. mean
4. mid-mean
5. none of these

3. Suppose that one data point in a dataset is altered. Which of the following is guaranteed to change?

1. mode
2. median
3. mean
4. mid-mean

4. Which of the following is NOT true?

1. Summary measures of spread capture important features of the whole distribution of data for a variable.
2. Summary measures of spread are easier to remember than the original data.
3. Summary measures of spread indicate the amount of variability around a measure of location.
4. Every data point contributes equally to a summary measure of spread.

5. Which of the following is TRUE?

1. Robust measures are completely unaffected by measurement error.
2. The median is more robust than the mean.
3. Using robust measures means that the researcher does not have to worry about the assumptions of statistical methods.
4. Robust measures are always less efficient.

6. In hypothesis testing, a type I error is ______.

1. failing to confirm the researcher’s preferred hypothesis
2. missing a real effect
3. claiming an effect which is not actually real
4. finding an effect which is not consistent with existing literature

7. Which of the following is TRUE about reference distributions?

1. There is always a single correct reference distribution for a specific significance test.
2. The normal distribution is what we should expect data to look like.
3. Using the right reference distribution means that you do not have to check the quality of your data.
4. Reference distributions are approximations which may be useful in practice.

8. Testing for a difference between three groups requires ______.

1. normally distributed data within each group
2. estimates of a summary measure of location for each group
3. the assumption that all three groups are different from each other
4. a true causal relationship between the grouping variable and the dependent variable

9. Testing an association between two variables requires ______.

1. variables measured at least on an ordinal scale
2. normally distributed data
3. data without measurement error
4. a true causal relationship between the two variables

10. Which of the following is appropriate for using the Kendall’s tau correlation coefficient?

1. two variables measured on nominal scales
2. two variables measured on continuous scales
3. one variable measured on an ordinal scale and one variable measured on a continuous scale
4. one variable measured on an ordinal scale and one variable measured on a continuous scale