July 2001
Volume 8 • Number 3

Contents

Introducing Statistical Thinking to the Food Industry—Facilitating and Inhibiting Factors

by Frøydis Bjerke, Matforsk—Norwegian Food Research Institute, Margrethe Hersleth, Matforsk—Norwegian Food Research Institute

During the 1990s, MATFORSK (the Norwegian Food Research Institute) taught applied statistics to personnel from the Norwegian food industry, focusing on the simple tools of statistical process control (SPC) and experimental design. Through this work, observations revealed that many companies have difficulty applying these methods to their processes and quality improvement projects. Therefore, a study on factors that might facilitate or inhibit the introduction of applied statistical methods in the industry was initiated.

In 1997, eight Norwegian food companies participated in a training program focusing on simple statistical tools. The participating persons and companies are the subjects of this study. The result of this study shows that several factors affect the success of applying statistical methods. For example,

• Management’s involvement and insight into statistical methods

• Systems for internal distribution of competence

• Some aspects of corporate culture

Although this material is from the food industry, these findings likely apply to many branches of industry and manufacturing. This article discusses the results from the study in detail and considers the challenges of introducing statistical thinking to nonstatisticians.

Key words: applied statistics, management, network training program, six sigma, total quality management

INTRODUCTION

The lack of statistical thinking and the need for statistical methods have been observed by research statisticians and statistical consultants working for industry and manufacturing in several countries. These needs have been increasingly recognized by industrial managers around the world. In Norway, however, many leaders and managers from the food industry, as well as other industries, are still ignorant of the ideas behind statistical thinking and continuous improvement (Kroslid 1999, paper B).

The core elements of statistical thinking are generation of data, extraction of relevant information from data, and utilization of information for optimal decision making (Rao and Arthanari 1995). Thus, statistical thinking is well suited as a basis for continuous improvement of processes and products, or continuous quality improvement. Continuous improvement is crucial for the Norwegian food industry in order to meet increasing international competition, as national trade barriers diminish.

MATFORSK—the Norwegian Food Research Institute—serves the Norwegian food industry through various research projects and contract work. In addition, MATFORSK runs courses and training programs within its fields of competence for personnel from the food industry. The field of applied statistics has always been closely connected to the food-related research activities at MATFORSK. Through this work, it has been observed that many companies have difficulty applying methods of statistical process control (SPC) and design of experiments (DOE) to their processes and quality improvement projects. In this context, it is worth mentioning that many Norwegian food companies and processing plants are relatively small (less than 250 employees), and that there still is a lot of manual operation in the food industry.

Consequently, the concept and methods of statistical thinking are not as widely applied as desired, throughout the food industry. Motivated by these observations and considerations, MATFORSK initiated a case study on how knowledge about applied statistics was received by the food industry. The study sought to identify factors that facilitate and inhibit the introduction of applied statistical methods.

Because of limited resources, the size of this study is rather small. The intent is to point out some important aspects of introducing statistical methods to the food industry, as they are observed in this material. Some methodological aspects on case study research on small samples are discussed in Eisenhardt (1989).

The next section of this paper describes a contextual model that is the basis for the design of the case study. This model investigates how some facilitating and inhibiting factors in the companies relate to statistical thinking. The following section describes the case study and the learning program called the network model. Then the results from the study are discussed in view of the contextual model. In addition, some important aspects of the results in view of the network learning model are mentioned. Finally, in the conclusion and further considerations section, the results are put into a broader perspective.

The subject of this paper concerns the interface between two interesting fields: organization and management theory on one side and applied statistics on the other. There are particular challenges in this constellation. The authors would like to emphasize that their background is within applied statistics, where they have ample experience in introducing simple statistical methods to nonstatisticians and practitioners. However, they are eagerly approaching the organizational subjects, motivated by this experience.