Nominal or Categorical Data, Differences between Nominal, Ordinal, and Ratio Scales - Essay Example

EXAM PREPARATION (MARKETING RESEARCH) Question Category/nominal data Nominal or categorical data consist of categories with numbers assigned to each group. Nominal data require neither ranking nor ordering. Besides, they use words as opposed to numbers. Examples of studies using nominal data are surveys (Babbie, 2012, p. 162). Surveys have multiple-choice questions that require the respondent to select. Some of the issues may require simple answers like yes or no. For instance, when assessing whether gender has any influence on the consumer preference, one should include an issue that seeks to identify the sex of the respondent by either selecting male or female in the questionnaire. Another example of categorical data is the use of geography and demographic data to predict behaviour of the consumer. In demography and geography survey, the consumers respond to their state of residence or the city, or the country of origin (Texas, California, and Chicago). However, the study should limit the states or the cities from where it operates from or have subsidiaries. To assess the specific geographic market segmentation, business composition, growth patterns, and the demand difference among the zones or areas of operation, one should also use categorical data. These data can be collected using three criteria. The common approach is the use of open-ended questionnaire that would require coding after data collection (respond allowed to write their thought). Another example involves the use of lists of items in a form to enable the respondent select the options. The categorical data allows the use of multiple response questions (respondent selects). The questions must be coded to make the selection and subsequent analysis easy. The reason for assigning numbers to labels facilitates data analysis. Analysis is possible through the assessment of modal frequencies and percentages. The nominal data are not suitable for studies that seek to establish the comparison and to rank between products in different geographical and demographic zones. Besides, one cannot use these data in studies that require averages and comparison. Example of survey question using nominal data Ordinal data Studies seeking to establish ranking or ordering of products and other business variables use ordinal data. In simple terms, the ordinal data determines natural order hence referred to as the rating scale. One can rate (excellent, good, fair, unsatisfactory, and poor) services or products. Such data is useful when one seeks to leverage on the respondent willingness to rank or make their preferred list of items through their selection strategies. Unlike categorical data, which measures the modal values and their frequencies, ordinal data takes into account the analytical option. These types of data are useful when measuring consumer preference for an individual product. However, this kind of measurement is not suitable when seeking to establish accurate distances between the lists of items selected by the respondent because it only provides the preference. Therefore, choices should never be treated as equal. In fact, this is where the interval data becomes necessary because it helps in the establishment of distances between the intervals in a list of items. The survey question forces the respondent to make selection/choices based on the available level of agreement (strongly disagree, disagree, somewhat agree, agree, disagree, strongly disagree). A good example of ordinal data is when undertaking attitude survey. The measurement is appropriate when ranking the order of preference for similar products from different companies. The established order of preference helps in identifying the most liked product and the company manufacturing the item. When using this measurement, one should ensure complete randomness to avoid biases of the order. Analysis that seeks to establish averages should not use ordinal data because of the difficulties in ascertaining the accurate distances between the variables. Example of ordinal data Example of ordinal data Interval and ratio scale The usefulness of interval and ratio scale is their ability to take into consideration a wide range of statistical transformation and tests. However, the difference between ratio and interval data is that the ratio has a defined point of zero (crime rate, purchasing time, product defect rate, market share, annual sales, and income). It is possible to transform ratio data using square root or logarithms when one aims at creating standard data. Contrary to ratio data, the interval data lacks absolute zero points. One cannot use interval data to compare the magnitude of companies offering a particular product or services. For example, one cannot compare product A and B using this data. It makes it inappropriate in most comparative surveys. The measurement uses Likert scale to assign values to the listed categories. Therefore, this data measures the variables like perceived economic situations in the business, the willingness, and satisfaction of customers to buying a particular product or their preference for an individual company. Besides, one can establish the likelihood, satisfaction, and degree of agreement to a product or service. The data provides equal spacing between the attributes because when spacing is not equal in the interval scale it changes to the ordinal scale. Example of an interval scale Question 2 Differences between nominal, ordinal, and ratio/interval scales Nominal data are categorical data because they consist of categories with numbers assigned to each class. Nominal data require neither ranking nor ordering. Besides, they use words as opposed to numbers. Examples of studies using nominal data are surveys. Surveys have multiple-choice questions that require the respondent to select. Some of the issues may require simple answers like yes or no. For instance, when assessing whether gender has any influence on the consumer preference, one should include an issue that seeks to identify the sex of the respondent by either selecting male or female in the questionnaire. Another example of categorical data is the use of geography and demographic data to predict behaviour of the consumer. In demography and geography survey, the consumers respond to their state of residence or the city, or the country of origin (Texas, California, and Chicago). However, the study should limit the states or the cities from where it operates from or have subsidiaries. To assess the specific geographic market segmentation, business composition, growth patterns, and the demand difference among the zones or areas of operation, one should also use categorical data. These data can be collected using methods like open-ended questionnaire that would require coding after data collection (respond allowed to write their thought). Another example involves the use of lists of items in a form to enable the respondent select the options. The categorical data allows the use of multiple response questions (respondent selects). The questions must be coded to make the selection and subsequent analysis easy. The reason for assigning numbers to labels facilitates data analysis. Analysis is possible through the assessment of modal frequencies and percentages. When ranking or ordering products and other business variables ordinal data are appropriate. In simple terms, the ordinal data establishes natural order hence referred to as the rating scale. One can rate (excellent, good, fair, unsatisfactory, and poor) services or products. Such data is useful when one seeks to leverage on the respondent willingness to rank or make their preferred list of items through their selection strategies. Unlike categorical data, which measures the modal values and their frequencies, ordinal data takes into account the analytical option. These types of data are useful when measuring consumer preference for an individual product. However, this kind of measurement is not suitable when seeking to establish accurate distances between the lists of items selected by the respondent because it only provides the preference. Therefore, choices should never be treated as equal. In fact, this is where the interval data becomes necessary because it helps in the establishment of distances between the intervals in a list of items. The survey question forces the respondent to make selection/choices based on the available level of agreement (strongly disagree, disagree, somewhat agree, agree, disagree, strongly disagree). A good example of ordinal data is when undertaking attitude survey. The measurement is appropriate when ranking the order of preference for similar products from different companies. The established order of preference helps in identifying the most liked product and the company manufacturing the item. When using this measurement, one should ensure complete randomness to avoid biases of the order. Analysis that seeks to establish averages should not use ordinal data because of the difficulties in ascertaining the accurate distances between the variables. The usefulness of interval and ratio scale is their ability to take into consideration a wide range of statistical transformation and tests. However, the difference between ratio and interval data is that the ratio has defined a point of zero (crime rate, purchasing time, product defect rate, market share, annual sales, and income). It is possible to transform ratio data using square root or logarithms when one aims at creating standard data. Contrary to ratio data, the interval data lacks absolute zero points (Babbie, 2012, p. 162). One cannot use interval data to compare the magnitude of companies offering a particular product or services. For example, one cannot compare product A and B using this data, making it inappropriate in most comparative surveys. The measurement uses Likert scale to assign values to the listed categories. Therefore, this data measures the variables like perceived economic situations in the business, the willingness, and satisfaction of customers to buying a particular product or their preference for an individual company. Besides, one can establish the likelihood, satisfaction, and degree of agreement to a product or service. The data does provide equal spacing between the attributes because when spacing is not equal in the interval scale it changes to the ordinal scale. Example of survey question using nominal data Example of survey question using ordinal data Example of survey question using ordinal data Example of survey question using an interval scale Reference list Babbie, E. 2012. The Practice of Social Research. London: Cengage Learning. p. 162. Read More