Classification of Daily Food Sets in an Agro-Pastoral Society in North-Central Namibia: A Comparison of Cluster Analysis and Two-Way Indicator Species Analysis

Abstract

Daily food surveys to clarify the patterns of local food cultures are typically used to record daily food consumption and understand local food culture. Previous studies used survey data obtained mainly by calculating the frequency of occurrence of each food item, dish, or combination of specific items. To understand African food culture by focusing on the combinations of food items and to show the different perspectives that are free from the subjectivity of the empirical study, this study examined the application of two classification methods to a food survey dataset obtained through a food survey in a rural village in northcentral Namibia. To arrive at a better classification method, the author attempted to use two representative classification methods, cluster analysis and two-way indicator species analysis (TWINSPAN), to classify the same food dataset, focusing on the combination of dishes. Differences between the two sets of results were examined. Both methods extracted four major groups with some subgroups that abstracted some features of the food culture of Owambo. Although there were some differences in classification between the two methods, the authors found no obvious trends when using those two methods in this study. The classification of combinations of food items presented in this research could be applied to good effect by other studies on food culture.