Hints for Quiz 5 on Measurement; this is a 20 question quiz which still counts as 10 points.

1. All scales/measurement devices should be pretested for reliability and validity using people who could be members of the group you want to study (minority groups, children, the elderly). True or False?

2. What's the difference between random error and systematic error?

3. Which reliability test is subject to multiple testing effect? And what experiment can be done to tease out whether multiple testing effect was occurring or whether the test results really show high reliability?

4. A reliable measure must first be valid. True or False.

5. A valid measure must first be reliable. True or False.

Either statement 4 or statement 5 is true. Which one?

6. Accuracy and precision are associated with validity. True or False?

7. What's the weakest form of validity and what's the most complex form of validity?

8. Higher levels of measurement, e.g. interval or ration, are more precise and permit more statistical operations. True or False?

Wherever possible select the highest level of measurement because it's more precise and allows more statistical operations.

9. Which levels of measurement involve a fixed order of response choices?

Some researchers argue that attitude scales can be considered at the interval level though each item in the scale is at the ordinal level. What is their argument?

Remember when talking about level of measurement, the higher levels of measurement have the same properties, and more, than the lower levels. For ex.--Ordinal level variables have the same properties as nominal (mutually exclusive and exhaustive) but also appear in order of magnitude.

10. Socioeconomic class is a continuous variable. True or False?

11. If you move from observing indicators (counting them) and then infer a theoretical concept, e.g., minutes spent crying as a sign of sadness/grief, is this inductive or deductive reasoning?

Remember inductive is from the ground up.

12. What's the major reason for using a scale instead of just one item/indicator?

13. Pick out the definition of spouse abuse that could not possible be operational.