Customer Satisfaction Pilot Studies and Analysis

Dispersion of Responses

There are also several ways to describe the results as a pattern. Questions that have distinct categories (e.g., strongly agree, moderately agree, strongly disagree) can be described by reporting the frequency with which each response category is chosen. Even the responses to the ACSI questions can be reported as frequencies (see Figure 1.).


The simplest approach to displaying data is to look at frequencies. Each ACSISAT question has 10 possible responses ranging from 1 to 10, with 1 representing the least favorable response and 10 representing the most favorable response. The responses can be reported as a table or in graphic form as illustrated in the examples in Figure 1. 4

Strengths. Frequencies, percentages, and related graphs are easily developed from any spreadsheet program such as Microsoft Excel.

The visual display of data approach has the advantage of allowing you to see the exact pattern of responses. Figure 1 Comments, which appear below, further discuss this point. Looking at the frequencies is often a good way to begin explaining the results. It puts an appropriate context around the reporting of means or other statistics of central tendency. In the examples used above to explain the meaning and use of average, a different level of understanding the data would be immediately apparent if a visual depiction of the scores accompanied the reported means of 6. Another strength is the ease of visually displaying this information in a manner that is familiar to most people.

Weaknesses. Although this approach to data display is more complete than an average, interpreting the visual information from the three questions without a numeric average or similar statistic can be challenging for a general audience, who must attempt to combine the frequencies in their heads. The three examples below (Figure 1) illustrate both the strengths and weaknesses of reporting frequencies and a companion number of the percentages of customers who chose each response. Try covering the means and standard deviation statistics and developing an overall picture of the relationship between the three questions from the frequencies alone.

Figure 1: State A Participant Responses to Three ACSISAT Questions

Figure 1

Figure 1 Comments
An immediate value of looking at the frequency pattern is that all of these general satisfaction questions tend to have a large number of highly positive responses. The patterns also show, however, that there are a substantial number of responses that are at 1, the lowest point on the scale. Such a pattern with a large number of positive responses combined with a substantial number of very low responses forces one to ask the question, what is different about these two groups of respondents? Do different customer groups (e.g., Youth versus Dislocated Workers, East County versus West County One-Stop customers) perceive service quality differently? If additional study indicated that one group of customers was less satisfied with service, then management and staff would be able to examine both service content and service delivery to determine any service quality gaps, and they could then be addressed.

Home | About | Resources | Services | Forums