Measurement in Research

 Measurement in Research

Measurement refers to the process of assigning numbers (or other symbols) to the objects and events under study. It is the central task of any inquiry. It is a careful and purposeful examination of the real world for the purpose of describing objects and events. In some cases, measurement is easier, like measurement of wage, profit, etc. But, in other cases, like social welfare, externalities, etc. it is somewhat difficult to measure. Measurement plays an important role in the process of research as follows:

1. It helps to understand the problem under study.

2. It helps the researcher to test the hypothesis on the basis of available evidence.

3. It leads to the discovery of new knowledge. It is possible that the available evidence leads to the conclusion that the axioms theory does not represent the real world. Thus, measurement employed for the test of statistical significance. Similarly, one can find one important tool for achieving effective theory.

Levels of Measurement

S.S. Steven has divided the measurement of objects and events into four levels as follows:

a)  Nominal measurement

It is the most elementary method of measurement, which classifies the objects and events into a number of mutually exclusive subgroups. The objects and events are divided into different subgroups in such a way that each member of the subgroup has a common set of characteristics. For example, the population of a district can be divided into male and female - on the basis of gender; Hindu, Buddhist, Christian, Jain, and Islam - on the basis of religion and so on. While classifying into these subgroups numbers or symbols are assigned to each subgroup. But, it should be noted that such numbers or symbols have no quantitative significance, i.e., they cannot be added, subtracted, multiplied, or divided. These numbers and symbols do not show the superiority of one subgroup over others. Only they tell that the subgroups are qualitatively different from one another. For example, different numbers can be assigned to the football players to identify them. Such numbers have no quantitative meaning, i.e., they cannot be used to compare the performance of the players; they cannot be used to take the average of numbers, and so on. Thus, nominal measurement only provides a convenient way of identifying objects and events.

The possible arithmetic operation that can be applied to the nominal measurement is counting. In the case of measures of central tendency, the model can be calculated. There is no generally used measure of dispersion for the nominal measurement Chi-square test can be particular the percentage of the objects and events falling in a subgroup.

b)  Ordinal measurement

The ordinal measurement possesses all the characteristics of nominal measurement, i.e., the objects and events of each subgroup have common features. In addition to this, it ranks the objects and events in an ascending or descending order. However, this measure does not give an idea about how much one object or event is higher or lower than others. Thus, ordinal measurement provides only the relative position of the two or more objects and events on the basis of some characteristics. For example, a consumer ranks different car companies on the basis of his preferences as follows:

1. Toyota

2. General Motors

 3. Ford

4. Kia

5. Nissan

6. Mahindra.

In this example, the numbers assigned to different car companies show the consumer's preference over different companies. But, they do not tell anything about how much amount one company is preferable over others.

c)  Interval measurement

Interval measurement possesses all the characteristics of ordinal measurement, i.e., all the objects and events in each subgroup have common features, and they are arranged in an ascending or descending order. In addition to this, this measurement ranks the objects and events in such a way that numerically equal distance on the scale of measurement represents the equal distance in the characteristics being measured. Thus, in the case of interval measurement, the distance between the objects or events have meaning. By comparing such distances, we can say that by how much amount one object or event is greater or less than the other.

Interval scales are nice as a result of the realm of statistical analysis on these information sets disclose. As an example, the central tendency is measured by mode, median, or mean; standard deviation can even be calculated.

Like the others, you'll keep in mind the key points of an “interval scale” pretty simple. “Interval” itself suggests that “space in between,” that is that the vital issue to remember–interval scales not only tell us regarding the order but however, also regarding the value between each item.

Here’s the problem with interval scales: they don’t have a “true zero.”  For example, there is no such thing as “no temperature,” at least not with Celsius.  In the case of interval scales, zero doesn’t mean the absence of a value; however is truly another number used on the size, like zero degrees Celsius.  Negative numbers even have which means. Without a real zero, it's not possible to compute ratios.  With interval information, we are able to add and subtract, however, cannot multiply or divide.

d. Ratio measurement

Ratio measurement possesses all the characteristics of interval measurement. Besides this, it is based on the true zero points. It means zero represents the absence of particular characteristics in question So, in the ratio measurement, the ratio of objects and events has meant so that such ratio can be used for the purpose of comparison between the objects and events. Measurement of income, sales, costs, number of purchasers, length, etc. are examples of ratio measurement. In this case, for example, we can say that a person Learning Rs. 16,000 per month exams four times the salary of a person Learning Rs. 4,000 per month. Generally, all the statistical techniques can be applied to the ratio measurement.

At last, nominal variables are used to “name,” or label a series of values.  Ordinal scales give good information regarding the order of decisions, like in a very customer satisfaction survey.  Interval scales offer us the order of values + the ability to quantify the distinction between each one.  Finally, ratio scales offer us the ultimate–order, interval values, and the ability to calculate ratios since a “true zero” will be defined.