Anything that can be counted or measured is called a variable. Knowledge of the different types of variables, and the way they are measured, play a crucial part in choice of coding and data collection. The measurement of variables can be categorized as categorical (nominal or ordinal scales) or continuous (interval or ratio scales).
Categorical measures can be used to identify change in a variable, however, should you wish to measure the magnitude of the change you should use a continuous measure.
A nominal scale allows for the classification of objects, individual and responses based on a common characteristic or shared property. A variable measured on the nominal scale may have one, two or more sub-categories depending on the degree of variation in the coding. Any number attached to a nominal classification is merely a label, and no ordering is implied: social worker, nurse, electrician, physicist, politician, teacher, plumber, etc.
An ordinal scale not only categorizes objects, individuals and responses into sub-categories on the basis of a common characteristic it also ranks them in descending order of magnitude. Any number attached to an ordinal classification is ordered, but the intervals between may not be constant: GCSE, A-level, diploma, degree, postgraduate diploma, higher degree, and doctorate.
The interval scale has the properties of the ordinal scale and, in addition, has a commencement and termination point, and uses a scale of equally spaced intervals in relation to the range of the variable. The number of intervals between the commencement and termination points is arbitrary and varies from one scale to another. In measuring an attitude using the Likert scale, the intervals may mean the same up and down the scale of 1 to 5 but multiplication is not meaningful: a rating of ‘4’ is not twice as ‘favourable’ as a rating of ‘2’.
In addition to having all the properties of the nominal, ordinal and interval scales, the ratio scale has a zero point. The ratio scale is an absolute measure allowing multiplication to be meaningful. The numerical values are ‘real numbers’ with which you can conduct mathematical procedures: a man aged 30 years is half the age of a woman of 60 years.
Categorical measures can be used to identify change in a variable, however, should you wish to measure the magnitude of the change you should use a continuous measure.
A nominal scale allows for the classification of objects, individual and responses based on a common characteristic or shared property. A variable measured on the nominal scale may have one, two or more sub-categories depending on the degree of variation in the coding. Any number attached to a nominal classification is merely a label, and no ordering is implied: social worker, nurse, electrician, physicist, politician, teacher, plumber, etc.
An ordinal scale not only categorizes objects, individuals and responses into sub-categories on the basis of a common characteristic it also ranks them in descending order of magnitude. Any number attached to an ordinal classification is ordered, but the intervals between may not be constant: GCSE, A-level, diploma, degree, postgraduate diploma, higher degree, and doctorate.
The interval scale has the properties of the ordinal scale and, in addition, has a commencement and termination point, and uses a scale of equally spaced intervals in relation to the range of the variable. The number of intervals between the commencement and termination points is arbitrary and varies from one scale to another. In measuring an attitude using the Likert scale, the intervals may mean the same up and down the scale of 1 to 5 but multiplication is not meaningful: a rating of ‘4’ is not twice as ‘favourable’ as a rating of ‘2’.
In addition to having all the properties of the nominal, ordinal and interval scales, the ratio scale has a zero point. The ratio scale is an absolute measure allowing multiplication to be meaningful. The numerical values are ‘real numbers’ with which you can conduct mathematical procedures: a man aged 30 years is half the age of a woman of 60 years.
Thanks
0 comments:
Post a Comment