Nonlinear Principal Component Analysis of Food Hygiene and Safety Knowledge and Practices Among Street Food Vendors

Many researchers have reported street food vendor characteristics including their food safety and hygiene awareness, attitudes, and practices. Data on street food vending are usually qualitative and include a large number of variables. This presents difficulties to researchers in trying to find the best way to reduce such data to a limited number of composites for use in further studies. Nonlinear principal component analysis (NLPCA) is a multivariate analysis technique that is most ideal for data reduction to yield uncorrelated summary variables while taking into account non-numeric measurement and the probable nonlinear relationships in the data. The objective of this study was to reduce and summarize the large heterogeneous data obtained from 345 street food vendors (SFVs) from six study locations in Kiambu County, Kenya, containing ordinal and nominal data into one or more components that may be used to elucidate the associations within the SFVs characteristics. NLPCA identified seven components of knowledge and practices related to food safety and hygiene among SFVs, which included: the cleanliness of the vending area, individual SFV hygiene and food handling practices, initiatives put in place to ensure food hygiene and safety, washing practices and food safety knowledge, food safety standards and regulations, storage practices for raw and ready-to-eat food, adherence and attitude toward food safety stipulations and SFVs’ mobility. The seven components also represent the categories of concern in enhancing the quality and safety of street vended foods. These components may be used as the basis for evaluating street food quality and safety.