Within Respondent Comparisons

---

The first test was to determine if respondents discussing multiple food products for one organization give different answers for each product category (“within respondent comparison of one organization”). There were 124 responses about multiple food categories. Overall, the mean of the “average Euclidean distance for one organization” was 1.4 with a standard deviation of 1.8. A one- sample t-test was run to test the hypothesis that the average Euclidean distances for one organization were zero. Based on a t-score of 8.7 significant at 99% confidence, this hypothesis was rejected. It was concluded that the nature of the food product category discussed affected responses provided about the relation- ship with a customer or suppliers; the business environment; the satisfaction with the information system, and if it was perceived that information was exchanged as often as necessary. As a result, each respondent’s answers about each food product category were treated as a separate case for the remainder of the analysis.

The second test was to determine whether each respondent could make meaningful comparisons between each pair of customer/supplier organizations discussed (“within respondent comparison of organizational pairs”). There were 90 responses about pairs of customers/suppliers. Overall, the average between organization Euclidean distance was 5.8 with a standard deviation of 3.7. After removing one extreme outlier, a one-sample t-test was run to test the hypothesis that the average between organization Euclidean distances was zero. Based on a t-score of 14.9 significant at 99% confidence, this hypothesis was rejected. It was concluded that respondents made different responses about pairs of organizations.

There was a possibility that the differences between organizations were due to perceived differences within organizations for each product category. Therefore, a further one-sample t-test was run to test the hypothesis that the average “between organization Euclidean distances” was significantly different than the mean of the average “within respondent Euclidean distance” (1.4). Based on a t-score of 11.37 significant at 99% confidence, this hypothesis was rejected. It was concluded that respondents can make meaningful comparisons between pairs of customer/supplier organizations. In addition, responses vary depending on what food product category is being discussed.

The next research question addressed was whether staff in different departments within an organization had different perceptions about customers and suppliers. A comparison was made within the organization between sales and purchasing staff.

---