How to Use a Chi Square Test in Likert Scales

In survey research, a Likert scale is an approach to response categories that measures the extent of a person's satisfaction or agreement with a specific set of statements or questions. Using a Likert Scale set of questions to measure customer service reactions on services or products is one way the measurement tool operates. This type of response category makes it easy to quantify survey responses thereby simplifying data analysis.

A variety of options for analyzing Likert scale data exist including the chi square statistic. The chi square statistic compares survey respondents' actual responses to questions with expected answers to assess the statistical significance of a given hypothesis. An example of chi square statistic might be examining whether two groups of people have varying opinions. The greater the level of deviation between actual and expected responses, the higher the Chi Square statistic will be. This deviation level indicates how much less the results fit the original hypothesis.

There are two types of chi square statistic test: the chi-square goodness of fit test and the chi-square test for independence. Combine the response categories in your Likert scale. For example, if your Likert scale uses the response categories of strongly agree, agree, disagree, strongly disagree, neither agree nor disagree, combine the agree and strongly agree responses into one category and the disagree and strongly disagree into another. This gives you three categories of responses: agree, disagree and neither.

Run the chi square statistical test, using your spreadsheet program or statistical software. For example, to find the test in Excel, simply click the Formula tab at the top of your spreadsheet. Then choose More Functions and select Statistical; which displays the variety of available procedures followed by a section of the chitest or chi square procedure. Clicking on a cell and dragging the mouse over the range of data you want analyzed tells Excel the data on which to conduct the chi square test.

Next, examine the results of the chi square test generated by a spreadsheet or statistical program. When reviewing results, pay close attention to the size of the chi square statistic and the level of statistical significance. A higher chi square statistic indicates greater variation between observed and expected responses. Most spreadsheet and statistical programs use a significance level of .05, meaning there is only a 5 percent chance that the statistical significance; if any, resulted from random chance.

Interpret the results of your analysis. Analysis at this point involves looking at whether the chi square indicates a statistically significant relationship that exists without revealing information about the strength of that statistical relationship. Because there may also be a number of survey takers that leave answers blank, the chi square test is a good option because you can add a row for "no answer". While comparative analysis examines the data relationship, the chi square makes it possible to also measure non-reponse sampling errors and their corresponding relationship as well.