Measurement Scales:
Not All Numbers Mean the Same Thing
Behavioral scientists commonly refer to four types of measurement scales: Nominal, Ordinal, Interval, and Ratio. Although all four use numbers, the numbers do not have the same meaning or allow the same kinds of calculations.
For example, the number on a football player's jersey, a customer's satisfaction rating, an IQ score, and a person's weight are all expressed as numbers. Yet these numbers represent very different kinds of information. Failing to recognize these differences can lead to misunderstandings and incorrect conclusions.
This information is adapted from Applied Statistics: Concepts for Counselors
Nominal Scale
The nominal scale is often called the naming scale. Numbers or labels are used simply to classify people, objects, or events into categories. The numbers have no mathematical value and cannot be added, averaged, multiplied, or arranged from highest to lowest.
If we were studying treatment outcomes for people with depression, we might assign the number 1 to participants receiving counselling and the number 2 to people on a waiting list. The numbers merely identify group membership. They do not indicate that one group is better, larger, or greater than the other.
Other examples of nominal scales include:
Gender categories in a survey
Blood types (A, B, AB, O)
Marital status
Political party affiliation
Sports jersey numbers
Postal codes or zip codes
Although numbers may be used, they function only as labels. No meaningful arithmetic operations can be performed on nominal data.
Ordinal Scale
The ordinal scale goes one step beyond classification by introducing rank order. Scores indicate whether one person or object ranks above, below, or equal to another.
However, ordinal scales do not tell us how much difference exists between ranks.
For example, in a race, the runners finishing first, second, and third can be ranked accurately. What we do not know from the rankings alone is the exact time difference between competitors.
Other examples include:
Class rankings
Customer satisfaction ratings (poor, fair, good, excellent)
Military ranks
Ratings of restaurant quality
Judging in gymnastics, diving, or figure skating
Suppose two students are ranked first and second in a class. The top student may exceed the second-ranked student by only a fraction of a point or by a substantial margin. The ordinal scale tells us who ranks higher but not the size of the difference.
This limitation is often overlooked when people discuss rating scales. For instance, the difference between ratings of 3 and 4 may not represent the same amount of change as the difference between ratings of 8 and 9.
Interval Scale
The interval scale is one of the most important forms of measurement in the behavioural sciences. On an interval scale, the distances between scores are assumed to be equal. This allows researchers to perform a wide range of statistical analyses.
Many educational and psychological test scores are treated as interval measurements.
Examples include:
Intelligence test scores (IQ)
Achievement test scores
Standardized educational assessments
Many personality and psychological inventories
Temperature measured in degrees Celsius or Fahrenheit
Because interval scales have equal intervals, researchers can calculate averages (means), standard deviations, correlations, and many other statistics.
For example, a student with an IQ score of 110 is ten points higher than a student with a score of 100. Likewise, a score of 120 is ten points higher than a score of 110.
However, interval scales lack a true zero point, which means ratio statements are not meaningful.
This point often causes confusion.
A person with an IQ of 50 is not half as intelligent as a person with an IQ of 100. Similarly, a person with a reading score of 130 is not twice as skilled as a person with a score of 65.
Temperature provides another useful illustration. A temperature of 20°C is not twice as hot as 10°C because the Celsius scale has an arbitrary zero point rather than an absolute absence of heat.
The key feature of interval scales is that differences between scores are meaningful, but ratios are not.
Ratio Scale
The ratio scale possesses all the properties of an interval scale but adds one crucial feature: a true zero point. A zero on a ratio scale indicates the complete absence of the characteristic being measured.
Because ratio scales have a true zero, ratio comparisons are meaningful.
Examples include:
Height
Weight
Age
Distance
Income
Number of children in a family
Time elapsed on a stopwatch
Consider weight. A person weighing 200 pounds is genuinely twice as heavy as a person weighing 100 pounds. Likewise, a 10-foot object is twice as long as a 5-foot object.
The same principle applies to age. A 20-year-old person has lived twice as long as a 10-year-old child.
Because ratio scales permit meaningful ratios as well as all common statistical operations, they provide the highest level of measurement precision.
 |
| Four Measurement Scales |
Why Measurement Scales Matter
Understanding measurement scales is more than an academic exercise. The type of scale determines what conclusions we can legitimately draw from data.
A researcher would not calculate an average jersey number for a sports team because jersey numbers are nominal data. Likewise, it would be misleading to claim that a person with an IQ of 120 is twice as intelligent as someone scoring 60 because IQ scores are measured on an interval scale rather than a ratio scale.
The lesson is simple: numbers do not automatically have the same meaning. Before interpreting any statistic, ask what kind of measurement scale produced it. The answer will determine what the numbers can, and cannot, tell us.
Read more about statistics and basic projects in
Creating Surveys on AMAZON
Resource Links:
All Measures A – Z Test Index
Spirituality Measures A - Z
Statistical Concepts A - Z
Post Author
Geoffrey W. Sutton, Professor Emeritus of Psychology at Evangel University, holds a master’s degree in counseling and a PhD in psychology from the University of Missouri-Columbia. His postdoctoral work encompassed education and supervision in forensic and neuropsychology and psychopharmacology. As a licensed psychologist, he conducted clinical and neuropsychological evaluations and provided psychotherapy for patients in various settings, including schools, hospitals, and private offices. During his tenure as a professor, Dr. Sutton taught courses on psychotherapy, assessment, and research. He has authored over one hundred publications, including books, book chapters, and articles in peer-reviewed psychology journals.
Find chapters and essays on Substack. [ @GeoffreyWSutton ]
Comments
Post a Comment