Data Types: Using Nominal Data in Survey Research

Earlier this week, I outlined four types of survey questions, one of which is nominal data. There is nothing marginal about nominal data. In fact the majority of survey questions are nominal in nature. That is they are categories with numbers assigned to them to facilitate analysis. In most research courses they are introduced as variables such as eye or hair color, a person’s name or the state they live in. In consumer or B2B marketing research they would include, for example, the name of the store you purchased your DVD player at or the industry your customers are involved in. Let’s look at a few question types useful for collecting nominal data. Most nominal data is collected via questions that provide the respondent a list of items to choose from, for example: Demographic Survey ExampleDemographic Survey ExampleConsumer Questionnaire Example

There are three ways that nominal data can be collected

In the first example the respondent is given space to write in their home state. This is a form of open-ended question that will eventually be coded with each state being assigned a number. This information could also be provided to the respondent in the form of a list, where they would select one option. The second example is in the form of a multiple response question where each category is coded 1 (if selected) and 0 if not selected. It also incorporates an open-end component allowing the respondent the option of writing in a category not included in the list. These ‘other’ responses’ will need to be coded as well if they are to be included in the analysis. Nominal data is analyzed using percentages and the ‘mode’, which represents the most common response(s). For a given question there can be more than one modal response, for example if olives and sausage both were selected the same number of times. The advantage of using a multiple response question (the pizza topping example) is that it allows you the ability to create a metric variable which can be used for additional analysis. In this scenario, the respondent can select any or all options providing you with a variable that ranges from zero to the maximum number of categories. This becomes a useful tool for consumer segmentation. Nominal data is best used for profiling your respondents. Although limited in it statistical abilities this type of data is critical for gaining a deeper understanding of your survey respondents. Next, we will examine ordinal data. Written by Greg Timpany