外语实证研究方法课第5讲.doc

Main contents of Chapter 5 Part 1 Measurement Scales 5.1 Two kinds of variables in terms of their value 5.1.1 Discrete variables: (1) They are all in the form of integer and no decimal fraction value can be inserted into two integers (2) There is no definite distance between two adjacent numbers 5.1.2 Continuous variables: 1. They can take any values or can have decimal fractions ; 2. There are definite distances between any two values. Exercise 5.1 Judge whether each of the variables listed below be measured with discrete data or continuous data? 1. ( ) Cognitive style 2. ( ) Scores on a ten-point biology quiz. 3. ( ) Grade level 4. ( ) Anxiety level average 5. ( ) Weight in pounds 6. ( ) Education level 7. ( ) Education years 8. ( ) Age 9. ( ) Age shown with senior, middle aged, young 10. ( ) Family income in RMB 11. ( ) Family income in rank 5.2 Categories of Scales 5.2.1 Nominal scales S 1. They are the lowest type of scale. 2. They are used for naming and categorizing data in a variable. 3. They may be labelled as discrete, discontinuous, categorical, or even dichotomous scale. 4. Membership in such groups may occur either naturally or artificially 5. The label’s number is arbitrarily assigned. 6. The purpose of a nominal categorization is to group subjects or instances of data according to some attribute . 7. They do not lend themselves to mathematical manipulation . 5.2.2. Ordinal scales 1. In ordinal scale numbers are used to order events, objects and people along a continuum. 2. Besides showing a difference, they also show an order. 3. Ranks themselves have no numerical value. They have an ordinal value only when they are relative to other ranks. 4. The top rank will always be ‘one’ regardless of what that score is. 5. The difference between the 1st and the 2nd will always be exactly one rank regardless of how different the scores are. 5.2.3. Interval scales Interval scales not only represent the ordering of things but al