ESSENTIAL DIFFERENCES AMONG NOMINAL, ORDINAL, INTERVAL AND RATIO SCALES

ESSENTIAL DIFFERENCES AMONG NOMINAL, ORDINAL, INTERVAL AND RATIO SCALES

While observing different real world events, the problem of measuring is overcome by devising different types of scales and mapping or transferring the real world observations in to that scales. We have different type of scales these are as under:-

Nominal Scale

When we use nominal scale, we partition a set in to categories that are mutually exclusive and collectively exhaustive. For example we try to differentiate Pakistani employees of multinational company in to different categories on the basis of their regions. We may define categories like Punjab, Baluchistan, NWFP, FATA and AJK. Any employee of the company can fall into only and only one category, this is called mutually exclusive and there is no other category than these categories this is called exhaustive.

If we use numbers to identify categories they are recognized as labels only and have no quantitative vale. Nominal classification may consist of any number of separate groups if the groups are mutually exclusive and collectively exhaustive.

Ordinal Scale

This approach of scaling includes the characteristics of nominal scales plus an indicator of order. Ordinal scales are possible if the transitivity postulate is fulfilled. It states that if a is greater than b, and b is greater then c, than a is greater than c. When using ordinal scale it implies a statement of “greater than” or “less than” without stating how much greater or less. Using ordinal approach any number or cases can be ranked. While ordinal measurement speaks of “greater than” and “less than” measurements, other relationships may be used “superior to” “happier than” or “above”.

An other extension of this concept occur when we are interested in more than one properties. For example we may ask a car deriver to rank different cars by price, performance, speed, comfort and design. To develop an overall index, the researcher typically adds and averages ranks for each dimension or using a multidimensional scale. Researcher may face difficulty when combining the rankings of several respondents.

Interval Scale

This type of scaling includes properties of nominal and ordinal scaling plus an additional strength. It incorporates the concept of equality of interval means that the distance between 1 and 2 is equal to the distance between 3 and 4. Researchers use intelligence scores, semantic differential scale and many other important scales as interval. When we use interval as scale, we use arithmetic mean as the measure of central tendency and standard deviation as a mean or measuring disbursement of values around central tendency. For example we are interested in the mean age of company labour workers. We can categories interval scale as 15 to 20, 20 to 25, 25 to 30, 30 to 35, 35 to 40, 40 to 45 and 45 to 50. We can easily calculate mean value and standard deviation after collecting and arranging data as per above scale.

Ratio Scale

Ratio scale carries all the properties and powers of the above-mentioned techniques of scaling with an addition of provision for absolute zero or origin. The ratio scale represents the actual amounts of variations between different variables. In business research, we find ratio scales in many areas. There are money values, population counts, distances, return rates and amounts of time in a time-period sense.

All statistical techniques mentioned upto this point are usable with ratio scales. Other manipulations carried out with real numbers may be done with ratio-scale values. Thus, multiplication and division can be used with this scale but not with other scales. Geometric mean, harmonic mean and coefficients of variations can be calculated using this technique of scaling.

 All of the above mentioned scaling techniques are widely using for different types of case studies in the field of research. Researcher can face problems evaluating different variable if these are measured using different techniques of scaling. Certain statistical techniques require the measurement levels to be the same.

Since the nominal variable does not have the characteristics of order, distance or point of origin, we cannot create them artificially after the fact. The ratio-based salary variable, on the other hand, can be reduced. Rescaling salary downward into high-low, high-medium-low, or another set of categories simplifies the comparison of nominal data. The loss of measurement power accompanying such decisions is sometimes costly in that only nonparametric statistics can then be used in data analysis. Thus, the design of the measurement plan should anticipate such problems and avoid them when possible.