Concept of Measurement
Assigning numbers or other symbols to any product or event or issue or characteristics as per the certain pre-specified rule is known as measurement. The number is a symbol that can be provided in different ways like 1, 2, 3 … or I, Il, Ill … Numbers should be explained to get the information. Symbols are the means Of finding out the characters of any events. It is easy to measure physical products like length, breadth, height, weight, etc. but measurement of qualitative fact is very difficult. Thus, it is essential to provide numbers or symbols to measure characteristics of qualitative facts which is known as measurement. In business research, numbers are usually assigned for two reasons. First, the number permits statistical analysis of data. Second, the number facilitates the communication of measured results.
Measurement is very important in social science research because such research considers motivation, satisfaction, commitment, etc. These factors are qualitative in nature. So, assigning the number is essential to measure such variables. Thus, a researcher provides numbers to measure the perception of the people. For example, if a researcher intends to find out the satisfaction level of employees of an industry then they can give ‘1’ to absolutely dissatisfied perception and 5 to absolutely satisfied perception. A researcher finds out the satisfaction level of employees analyzing those numbers.
Various experts have defined measurement in different ways. Some o they are given below:
According to Goode and Hatt, “Measurement is the method of turning the series of qualitative facts into a quantitative series.“
According to S. Steven, “Measurement is the equipment of providing
numbers to objects or events according to rule.“
According to Cooper and Schindler, “Measurement in research consists of assigning numbers to empirical events, objects or properties or activities in compliance with a set of rules.“
From the analysis of the above definitions, it is found that measurement is that sort of tool which provides series of numbers to the series of
characteristics of event, product, or things. As per the definition of measurement, measurement is the composition of the following processes.
a. Selection of observable empirical events.
b. Developing a set of rules: A scheme for assigning a number of symbols to represent the characteristics of the events that are to be measured.
c. Applying the rules to each observation of that event.
Nature or Characteristics of Good Measurement
Generally, measurement should be able to measure the things which a researcher intends. The tools, which are used, should be simple and able to increase the efficiency of a researcher. The main criteria for testing the goodness of measures are validity, reliability, and practicability. These features of measurement are described below:
It is related to the rationality of measuring tools. Validity refers to the ability of a measuring tool to measure what it intends to measure. If it does not measure what it is designated to measure, there will be a problem. For example: When we ask a set of questions measuring instrument hoping to tap the concept of the consumer then such question should be able to tap that concept of consumer about the product or service otherwise such questions will be meaningless. The tools that are used by the researcher for the collection and analysis of data should be able to collect and analyze the data. Tools that are able to do so are considered valid and the results drawn from the use of such instrument are also considered as valid. Various experts have defined validity. Some important definitions are given below:
According to Goode and Hatt, “A scale processes its validity when it actually measures what it claims to measures.“
According to Seltiz, “The validity of measuring instrument is defined as the property of a measure that allows the researcher to say that the instrument measures what he says it measures.“
From the analysis of the above definitions, it can be said that the validity of an instrument depends on the ability of the instrument to measure what was expected to measure. The importance of research depends on the validity of the instrument so; the researcher should check the validity of the instrument and data before further processing the research work. Validity can be classified as content validity, criterion-related validity, and construct validity. They are described below:
i. Content validity:
It is also known as face validity. It refers to the adequate coverage of the concept. In other words, content validity ensures that the measuring tools include an adequate and representative set of items that would tap the concept. The more the scale item represents the concept of the research topic; the instrument has the greater content validity. If the instrument contains a representative sample of the universe then the content validity is high. Its determination is judgmental and intuitive but it can also be determined by using a panel of judges. They will judge the capacity of the measuring instruments that they can measure every facet of the concept of the subject or issue under study. Their opinions are evaluated using statistical tools. The instrument which scores rational value, such instruments are considered as valid instruments.
ii. Criterion-related validity:
Criterion-related validity refers to the success of measures used for prediction or estimation. This validity is used when measures differentiate an individual on a criterion (dependent variable) it is expected to predict. It helps to establish a correlation between actual and standard work. If the correlation seems high in the prediction and outcomes generated in the future then the validity of such instruments is high but if correlation seems low then the validity of the instrument is also low. The criterion should possess the following qualities:
a. Relevance: A criterion is relevant if it is defined in a term that is taken as a proper measure.
b. Freedom from biases: When the researcher gives equal opportunity to each measure to score well.
c. Reliability: Criterion is stable or reproducible.
d. Availability: The information specified by the criterion should be available.
Actually, it is a broader term. It refers to predictive validity and concurrent validity. Predictive validity refers to the usefulness of a test in predicting some future performance. Concurrent validity refers to the usefulness of a test is closely related to other measures.
The above two validities are external validities but construct validity is internal validity. If a measure confirms the predicted correlation with other theoretical propositions then such measure possesses construct validity. It wants the agreement between a theoretical concept and a specified measuring instrument.
It is also classified into two groups. One is convergent validity and the next is discriminate validity. Convergent validity refers to the agreement among ratings gathered independently. In other words, the degree to which scores on one scale correlate with scores on other scales designed to assess the same construct. Discriminated validity refers to the degree to which scores on a scale do not correlate with scores from scales designed to measure different constructs.
If the above-stated criteria are met, our measuring instruments are valid and will result correctly. Otherwise, we shall look for more information and modify the measuring instruments. The following table shows the key concepts of various types of validity.
||What is measured?
|| Degree to which the content of the issue under study is adequately represented by the instruments.
|| – Judgmental, -Panel evaluation with content validity ratio.
|| Degree to which the predictor is adequate in capturing the relevant aspects of the criteria.
|| – Correlation.
|| Description of the present; criterion data are available at the same time as predictor scores.
|| – Correlation.
|| Prediction of the future; criterion data are measured after the passage of time.
|| – Correlation.
|| Finding out the difference in measures, segregating the ideas that are measured and observed how appropriately the measure represents.
|| – Individual Judgment, – Correlation of proposed test with established one, – Convergent-discriminant techniques, – Factor analysis.
Reliability is related to the results of the research. Reliability refers to the act of generating stable and consistent results when the instruments are used in different samples and situations. In social science and behavioral science research, a researcher should collect and analyze the data using various instruments for drawing the results. Results of the research depend on the nature of data, validity, and reliability of instruments and results. Highly reliable data provides more accurate results. A stable and consistent result is possible with the help of reliable data. The major duty of the researcher is to find out the correct results. Measurement should have the following qualities to be reliable:
If the stable and consistent result is obtained with the use of the same instrument in the same sample is known as stability. A consistent result is considered a reliable result.
It is concerned with variations at one point in time among observers and samples. A way to test the equivalence of measurement is the study of differences in the results developed by different observers or in the different samples by the same researcher. If the results are more equivalent then the measuring instrument is more reliable.
iii. Internal consistency:
If similar instruments are used and the responses are highly correlated then such measuring instruments are considered as internally consistent instruments. The data collected by using such instruments are considered more reliable and the results produced are also considered reliable.
Reliability can be measured using the following methods:
a. Test-retest method:
If some instruments are used to measure the attitude of the people or collect data in the same sample of people, the results is the same or tentatively similar then it can be said that the measuring instrument used by the researcher is reliable enough. For applying this method, a researcher can use the quantitative or qualitative method. Researchers can use correlation to see the relationship between two results. If the correlation is high, it can be assumed that the measuring instrument used in the research is highly reliable but if there is low correlation then the instrument is less reliable so the modification in an instrument is essential.
b. Alternative or parallel form method:
A researcher develops two measuring tools covering the same concept and administers both the forms in the same sample. If the result is highly correlated then the measuring instruments are reliable and the results of the research also considered reliable. In this method, both forms have similar items and the same response format with only the wordings and the ordering of the questions are changed. These two forms are administered at the same time to the same sample; so, there is less chance of having errors due to the pace of time.
c. Split- half method:
It shows the correlation between two halves of an instrument. It shows the degree to which instrument items are similar and reflect the same construct under study. Under this method, the instrument is divided into two equal halves. The division of instruments can be made based on odd or even numbers or on a random sampling basis. These two instruments are distributed to two different groups of samples and the results of such instruments are correlated to each other. If the result is highly correlated, the instrument is highly reliable otherwise modification in an instrument is essential. It is generally checked calculating corn Bach alpha value.
d. Inter-rater method:
The consistency of the judgment of several respondents or rater on the same question or issue is also a measure of reliability. If several respondents rate in a similar way then the instrument is able to convey the message and the reliability of the instrument is high otherwise not. It is especially relevant when the data are obtained through observation, projective tests, or unstructured subjectively interpreted.
Reliability estimation can be summarized as follows:
||What it shows
||What is measured?
|| Reliability of an instrument inferred from examinee score, same test is administered twice to same subjects over an interval of less than six months.
| Parallel form
|| Degree to which alternative form of the same measure produce similar results; administered simultaneously or with a delay.
| Split half
|| Internal Consistency
|| Degree to which instrument items are similar and reflect the same underlying constructs.
|| Specialized correlational formula (corn bach alpha)
|| Internal Consistency
|| Estimate of the similarity of judges/ respondents’ observations or scores.
|| Specialized correlational formula.
Measuring instruments that are developed to measure the attitude of the people must be applied in practice. It should be defined clearly so that it can be operationalized into action. Practicability should have the following qualities:
The instrument should be less expensive. It means there should be a trade-off between the ideal research project and the budget available for the research. More items give more reliability but for shortening the interview or observation time, we need to reduce the measurement questions. As well, we need to change the data collection method due to economic factors. For example, rather than having a personal interview, a researcher uses a telephonic survey to reduce the cost.
A questionnaire with detailed and clear instructions is easier to complete correctly than a questionnaire that does not have clear instructions. We can make it convenient by giving attention to the design and layout of the questionnaire.
When another person than the questionnaire designer requires interpreting the results then a designer should provide detailed information for the interpretation of the results. instruction for administration, scoring keys and instructions, reference group, evidence about reliability, evidence of the relationship of the test to other measures and guides, and describing the results are to be provided by the designer.