Data types & Measurement scales: Data types & Measurement scales Chee-Wee Tan Module P1112 Data is all around us: Data is all around us Collect data everyday.
Where do we get the data?: Where do we get the data? Values Define the concept /construct: Define the concept /construct Any concept/construct can be conceptually defined.
Pain is “an unpleasant sensory & emotional experience associated with actual or potential tissue damage, or described in terms of such damage” (International Association for the Study of Pain, 1994)
This is a conceptual definition for ‘pain’ Operational Definition: Operational Definition Abstract construct may require specific definitions to measure it.
Example: Pain – ‘pain threshold’
‘pain unpleasantness’ Process of measurement: Process of measurement Characteristics of the person/object, not the person/object we’re measuring.
Assign of numbers to characteristic. Measurement scale: Measurement scale The way we assign the numbers and the context will affect the measurement scale.
Levels of measurement
Properties of the measurement scale Four measurement scales: Four measurement scales Nominal
Nominal Scale: Nominal Scale ‘In name only’
Labels for identification
Mathematical operators cannot be used here.
Example: License plate numbers, gender. Ordinal scale: Ordinal scale Numbers reflect the ordered relationship
Takes on features of the nominal scale
But since there is order, we can do more in analysis, e.g. finding the median (Week 3)
Example: Oxford scale of muscle strength Interval scale: Interval scale Has all the characteristics of ordinal scale.
But allows inferences to be made on the extent of differences.
Example: Celsius/Fahrenheit scale (temperature)
Arbitrary zero point Ratio scale: Ratio scale Has all the characteristics of the interval scale.
Absolute zero point
Or ‘true zero’, or ‘natural zero’.
Example: Muscle strength measured in Newtons Discrete or Continuous variable: Discrete or Continuous variable Discrete
Varies in discrete steps
Example: number of goals, number of children in a family
Example: Height, weight & time Types of data: Types of data Data types Modified from Fleming & Nellis (1994), p8. Why are measurement scales important?: Why are measurement scales important? Progression Transform higher to lower scale. Reverse not generally true. Why are measurement scales important?: Why are measurement scales important? Metric data (interval & ratio)
Use of parametric statistics
Non-metric data (nominal & ordinal)
Use of non-parametric statistics
For this course, only parametric statistics will be mentioned. Summary: Summary What is a variable & a value?