20 Scaling techniques: Scaling in research, Meaning of measurement scales and scaling, scale classification bases, Important scaling techniques, scale construction techniques

G. Santhiyavalli

Introduction

Measurement of Variables is an integral part of research and an important aspect of research design. The characteristic of individual and business vary from individual to individual and from entity to entity. In the case of human beings, there are certain physical or quantitative characteristics like height, weight and there are certain abstract or qualitative characteristics like intelligence, integrity, attitude creativity, etc., In case of business organization also there are physical characteristics like employees, sales, profit, etc. which are easily measureable. However, there are certain abstract characteristics like reputation, image of the entity, motivation, customer perceptions. The perceptions and feelings of customers and employees are extremely important because they help the company to stay afloat and grow.

Measurement means assigning numbers or symbols to the characteristics of certain objects. We do not measure the object but some characteristics of it. Therefore in research people or consumers are not measured; what is measured only are their perceptions, attitude or any other relevant characteristics. There are two reasons for which numbers are usually assigned. First of all, numbers permit statistical analysis of the resulting data and secondly, they facilitate the communication of measurement results. The assignment of numbers to the characteristics must be isomorphic. Scaling is an extension of measurement. It involves creating a continuum on which measurement of objects are located. Suppose you want to measure the satisfaction level towards coffee day and scale of 1 to 11 is used for the said purpose. This scale indicates the degree of dissatisfaction, with 1=extremely dissatisfied and 11=extremely satisfied. Measurement is the actual assignment of a number from 1 to 11 to each respondent. Scaling describes the procedures of assigning numbers to various degrees of opinion, attitude and other concepts. This can be done in two ways viz, (i) making a judgment about some characteristic of an individual and then placing him directly on a scale that has been defined in terms of thatcharacteristic (ii) Constructing questionnaires in such a way that the score of individual’s responses assigns him a place on a scale.

2. Characteristics or goodness of instruments / measurement scales

A measurement scale has to have certain desirable characteristic or criteria to judge its “goodness” so that one could have faith or trust in the scale that it will measure what it is intended to measure. The following are the main characteristics of measurement scales. They are

• Accuracy and Precision
• Reliability
• Validity
• Practicality

2.1 Accuracy and Precision

The characteristics of accuracy in measurement scale means it should be a true representative of the observation of underlying characteristic. The precision, however, means the power to discriminate/distinguish and indicate the extent of accuracy that can be achieved with the measurement scale.

Eg: The examination is conducted to measure the knowledge and understanding of the student. The marks scored out of say 100, would provide better accuracy and precision than simply grading the students A+, A, B+, B and C

2.2 Reliability

Reliability indicates the confidence one could have in the measurement obtained with a scale. It tests how consistently a measuring instrument measures a given characteristic/attitude is measured again and again, leading to about the same conclusion. However, it may be emphasized that reliability does not necessarily imply that the measuring instrument is also accurate. All it means is consistency in drawing conclusion.

2.3 Validity

The validity of a measuring instrument indicates the extent to which an instrument scale tests or measures what it is intended to measure. For example, if we intent to measure intelligence, the instrument, say question paper, ought to be such that it results in measuring true intelligence; if the paper tests only general knowledge, the instrument is not valid.

2.3.1 Types of validity

• Content validity
• Criterion validity
• Construct validity

￭Content validity

It indicates the extent to which a measuring instrument provides adequate coverage of the issues that are under study.

￭Criterion validity

These are two types. One indicates the success of the measuring instrument used for predicting. The other, also called concurrent validity, is used to estimate the present status.

￭Construct validity

It is one of the most significant aspects in the development of measurement theory and practice. It links psychometric notions and practices to theoretical notions. It attempts to explain the variation observed on several individuals.

Eg: If a test of intelligence is conducted on individuals, and if the test scores obtained by a measurement scale vary from individual to individual, one would like to know the factors to construct behind this phenomenon.

2.4  Practicability

From theoretical viewpoint, a measure is ought to be reliable and valid. However, from practical viewpoint, the measure should be

• Economical
• Convenient
• Interpretable

The economic consideration leads to a comparison between ideal research project and availability of budget for a study. Thus, the measuring instrument has to take cognizance this aspect and designed accordingly.

The convenience implies the ease with which an instrument like questionnaire could be easily administered to the subjects/participants/respondents. In this one should give due attention to the proper layout of the measuring instrument. This poses more challenge in the situation whereas the concepts and constructs are rather difficult to understand.

The interpretability of an instrument, like questionnaire, is the ease with which the researcher is able to interpret the responses from the subjects/respondents/participants.

3.  Properties of scales

• Distinctive classification
• Order
• Equal Distance
• Fixed Origin

3.1 Distinctive classification

A measure that can be used to classify objects or their characteristics into distinctive classes or categories is said to have this property. This is the minimum requirement for any measure. Eg: Gender classifies the individuals into two distinctive groups, male and female.

3.2 Order

A measure is set to have an order if the objects or their characteristics can be arranged in a meaningful order. Eg: Marks of the student can be arranged in an ascending or descending order.

3.3 Equal Distance

Thedifference between any two consecutive categories of a measured tribute, are equal, then the measure is said to have equal distance. Eg: The time difference between 2 pm to 3 pm is same as the difference between 3 pm and 4 pm.

3.4 Fixed Origin

A measurement for measuring a characteristic is said to have a fixed origin if there is a meaningful zero or “absence” of characteristic. Ex: Income of an individual, Sales of a Company.

4. CLASSIFICATION OR TYPES OF MEASUREMENT SCALES The most widely used classification of measurement scales are

• Nominal Scale
• Ordinal Scale
• Interval Scale
• Ratio Scale

  4.1 Nominal Scale

The qualitative scale without order is called nominal scale. The nominal scale involves classification of measure objects into various categories such as ‘Yes’ or ‘No’, ‘pass’ or ‘fail’. Numeric value is assigned to these classified categories like house number, telephone number, and roll number of the student. The data collected through a nominal measure scale is called nominal data.

4.2 Ordinal Scale

A qualitative scale with order is called an ordinal scale, it tells whether an object has more or less of characteristics than some other objects. It is a scale that does not measure values of the characteristics but indicates only the order or rank like 1st, 2nd, 3rd etc. Some examples are Ratings of hotels, restaurants, and movies. The data obtained using ordinal scale is termed as ordinal data.

4.3 Interval Scale

A measurement scale whose successive values represent equal value or amount of characteristic that is being measured and whose base value is not fixed is called an interval scale. It provides more powerful measurement than ordinal scales. It also incorporates the concept of equality of interval. This is a quantitative scale of measure without a fixed or true zero. It is a quantitative data that can be measured on a numerical scale. However, the zero point does not mean the absence of the characteristic being measured. Some examples are temperature, time, longitude, latitude, etc. The data obtained from an interval scale is termed as interval data.

4.4 Ratio Scale

This is the highest level of measurement and has all the four properties of a scale. Ratio scale represents the actual amount of variables. Ratio scales are quantitative measures with fixed or true zero. The data obtained from ratio scales are referred to as ratio data. Ratio is also a quantitative data that can be measured on a numerical scale but, here the zero point is fixed and implies the absence of what is being measured. In fact, if a scale has all the features of an interval scale, and there is a true zero point, then it is called a ratio scale.

Table 1 Examples of measurement scales

Statistical Analysis based on Scales

Depending on the property of the scales, there is a limitation on the descriptive statistics one can perform on the scales.

Table 2 summarizes the descriptive statistics that can be used on the type of scales

5. SCALING TECHNIQUES

Several scales formats have been developed to enable a researcher in collecting appropriate data for conducting a study. The scales are broadly divided into two categories viz.

• Conventional scaling
• Unconventional scaling

The conventional scales are used in the questionnaire format and are most common. The unconventional scales are used for unconventional collection of data through games, puzzles, etc.

The conventional scales are of two types viz, Comparative Scaling techniques and Non comparative techniques.

5.1 COMPARATIVE SCALING TECHNIQUES

The comparative scale involves direct comparison of the different objects. Comparative scale data are measured on ordinal scale and interpreted in relative terms and such as they generate non metric or non-numerical data.

Types of comparative scaling techniques

• Paired comparison
• Rank order
• Constant Sum

5.1.1 Paired Comparison

In paired comparison scales, the respondent is asked to select one object from the list of two objects, on the basis of some criteria. This forces the respondents to compulsorily select one of the two. Such scales are used when the study requires to distinguish between the two specified objects.

Example

In the study of consumer preferences about two brands of glucose biscuits viz, Parl-G and Tiger.

Select any one of the two brands.

Which Glucose biscuits do you prefer on the basis of ‘Taste’?

1 Parle-G                       1 Tiger

Which Glucose biscuit do you prefer on the basis of ‘Price’?

1 Parle-G                       1 Tiger

Which Glucose biscuits do you prefer on the basis of ‘Taste’?

1 Parle-G                       1 Tiger

This scaling technique is useful when the researcher wants to compare two or more objects. In the above example we have compared two brands over three factors. Hence the number of comparison is three.

5.1.2 Rank order scaling

In the rank order scaling, respondents are presented with several objects simultaneously and asked to order or rank them according to some criterion. It is also termed as forced ranking scale. Unlike paired comparison, rank order scaling technique prompts respondents to rank a given list of objects.

Example

Rank the following services in the order of importance attached by you, while selecting a new mobile services provider. The most preferred can be ranked 1, the next as 2 and so on. The least preferred will have the last rank. Do not repeat the ranks.

5.1.3 Constant Sum Rating Scale

When it is to assess the relative importance attached by a respondent to the objects in a list, the constant sum scaling technique is used. In this technique, a respondent is asked to allocate certain points, out of a fixed sum of points, for each object according to the importance attached by him/her to that object. If the object is not so important, the respondent can allocate zero point, and if an object is most important he/she may allocate maximum points out of the fixed points. Generally, the total fixed points are 100 for simplicity but it may be taken as some other value depending on the study.

Example

Allocate the amount you would like to spend on your birthday on the following items, out of total amount of Rs.10000/- (Please note that total amount allocated should be exactly Rs.10000).

Check your progress:

State whether the following statements are true or false

• Nominal scale can only involve the assignment of numbers. Alphabets or symbols cannot be assigned
• A comparative rating scale attempts to provide a common frame of reference to all respondents

5.2  NON-COMPARATIVE SCALES

In the non-comparative scales, the respondents do not make use of any frame of reference before answering the questions. The resulting data is generally assumed to be interval or ratio scale.

For eg: The respondent may be asked to evaluate the quality of food in a restaurant on a five point scale (1=very poor, 2=poor and 5=very good).

TYPES OF NON-COMPARITIVE SCALES

• Graphic rating scales
• Itemized rating scales:
• Likert scale
• Semantic differential scale
• Stapel scale

5.2.1 Graphic rating scale

This is a continuous scale, also called graphic rating scale. In the graphic rating scale the respondent is asked to tick his preference on a graph.

Eg: Please put a tick mark on the following line to indicate your preference for fast food.

To measure the preference of an individual towards the fast food, one has to measure the distance from the extreme left to the position where the tick mark has been put. Higher the distance, higher would be the individual preference for fast food. The basic assumption in this scale is that the respondents can distinguish the fine shade in differences between the preference or attitude which need not be the case. Further, the coding, editing and tabulation of data generated through such a procedure is a very tedious task and the researchers would try to avoid using it.

Another version of graphic scale could be:

Eg: Please put a tick mark on the following line to indicate your preference for fast food.

This is a slightly better version than the one discussed earlier. For eg: if a respondent had earlier ticked between 5 and 6, it is likely that he would remember the same and the second time, he would tick very close to where he did earlier. This means that the difference in the two response could be negligible.

5.2.2 Itemized rating scale

In the Itemized rating scale, the respondents are provided with a scale that has a number of brief descriptions associated with each of the response categories. It is widely used in survey research. There are certain issues that should be kept in mind while designing the itemized rating scale. These issues are:

Number of categories to be used

There is no hard and fast rule as to how many categories should be used in an itemized rating scale. However, it is a practice to use five or six categories. It is a fact that the additional categories need not increase the precision with the attitude of being measured.

Odd or even number of categories

By using even number of categories the scale would not have a neutral category and the respondent will be forced to choose either the positive or the negative side of the attitude. If the odd numbers of categories are used, the respondent has the freedom to be neutral if he wants to be so.

Balanced versus unbalanced scales

A balanced scale is the one which has equal number of favorable and unfavorable categories.

Example for balanced scale:

How important is price to you in buying a new car? Very important

Relatively important

Neither important nor unimportant Relatively unimportant

Very unimportant

Example for unbalanced sale:

How important is price to you in buying a new car? More important than any other factor Extremely important Important Somewhat important Unimportant

Nature and degree of verbal description

Verbal descriptions must be clearly and precisely worded so that the respondents are able to differentiate between them.

Forced versus Non-forced scales

In a forced scale, the respondent is allowed to take a stand, whereas in the non- forced scale, the respondent can be neutral if he/she so desires. Paired comparison scale, rank order scale and constant sum rating scales are examples of forced scales.

Physical Form

There are many options that are available for the presentation of the scales. It could be presented vertically or horizontally. The categories could be expressed in boxes, discrete lines or as units on a continuum. They may or may not have numbers assigned to them. The numerical values, if used, may be positive, negative or both.

Eg: Suppose we want to measure the perception about Jet airways using a multi-item scale.

i)  Likert scale

The Likert scale is the most frequently used variations of the summated rating scale commonly used in the studies relating to attitudes and perceptions. Summated rating scales comprise statement that expressed either a favorable or an unfavorable attitude toward the objective of interest on a 5 point, 7 point ot on any other numerical value. The respondents are given a certain number of items (statements) on which they are asked to express their degree of agreement or disagreement. Likert scale is also called a summated scale because the scores on individual items can be added together to produce a total score for the respondent.

Likert scale statements to measure the image of the company

Likert scale has several advantages that make it more popular. It is relatively easy and quick to compute. Further, it is more reliable and provides more data for a given amount of respondent’s time, as compared to other scales. The data gathered is interval data.

ii)Semantic differential scale

This scale is widely used to compare the images of competing brands, companies or services. In semantic differential scale, a respondent is required to rate each attitude object on a number of five-or-seven point rating scales. The difference between likert and semantic differential scale is that, in a likert scale, a number of statements are presented to the respondents to express their degree of agreement or disagreement. However, in semantic differential scale, bipolar adjectives or phrases are used. The advantage of semantic differential scale is that it is versatile and gives multi dimension advantage. It is widely used to compare image of brands, products, services and companies. The data generated from this scale can be considered as numeric in some cases, and can be summed to arrive total scores.

iii) Stapel scale

Stapel scale is used to measure the direction and intensity of an attitude. The scale generally has 10 categories involving numbering -5 to +5 without a neutral point and is usually presented in a vertical form.

Eg: Suppose a restaurant is to be evaluated on quality of food and quality of service, and then the staple scale would be presented as:

• The data generated in staple scale is interval data.
• It can be used to collect data through telephonic interview.
• This method is most applicable were evaluative responses are to be rated on a single dimension.
• The scale is most economical were several items are all to be rated on the same dimension.

Conclusion:

Different types of measurement scales and scaling techniques are presented in a lucid manner and in a simple language. I hope the module on scaling techniques will help you to frame your questionnaire suitably to get the required primary data.

Web links

• https://study.com/academy/lesson/comparative-vs-non-comparative-scales-in-marketing-research.html
• www.statisticshowto.com/scales-of-measurement/
• www.mymarketresearchmethods.com/types-of-data-nominal-ordinal-interval-ratio/
• https://www.scribd.com/doc/28266207/Measurement-Scaling-Techniques