Searching for Sensory Research Excellence Sensory quality system based on traditional techniques provides flexibility and sound conclusions JACQUELINE PEARCE BECKLEY AND DANIEL R. KROLL

QUIETLY, A CHANGE IS OCCURRING. At some of the large and pro- gressive companies,sensory research is becoming a strategic business unit, playing a part equal to any other valued component of the marketing system.

In the most progressive companies, the known components of sensory research methods, protocols for scientific experiments, experimental design, strategic planning, input from experienced professionals are being restructured into some fairly exciting forms. The research offers the company a competitive edge in the new product phase, in the development of existing products, and in product monitoring, as well as in the overall new business development function. Who wouldn't find the prospect exciting? But the awesome potential of sensory research to give a company's marketing an ad vantage leads to a problem.

The problem lies not in failing to give sensory the recognition it deserves, but rather in mindlessly searching for some exciting technique or methodology. It is then that the real objective of sensory research can be lost.

Several years ago, for example, many manmanagements were beginning to use a process called total quality management (TQM). While its primary focus was related to the function of managing a corporation, the concept had a profound effect on the quality assurance function of many firms. Today, TQM is no longer fashionable as a management tool. However, some quality assurance departments have benefitted from the use of the word ”quality.” Quality has not gone out of fashion.

Nor have the fundamentals of strong sensory work that is conclusion oriented gone out of style. Work that is sound from both a structure and function basis continues to serve us well.

We continue to go back to the work of David Peryam, dedeveloper of the hedonic scale and several other techniques on which sensory science is based. That heritage also drives us to constantly reexamine and innovate, to customize basic systems to the needs of the client and/or project. An excellent example of how successful this can be is demonstrated by a sensory evaluation approach we developed, utilizing fundamental sensory techniques. It can be used in its original form or modified for a given company's needs. It is presently used by a number of large multinational companies and has helped to deal with their quality control issues.

Sensory Quality System

The strength of sensory research is the underlying process. A common feature of individuals who consider themselves passionate about sensory research is a commitment to the structure and details of a test and its strategic plan. There is always a great deal of disagreement about the right or wrong ways to address a question requiring a sensory conclusion. There is generally common ground, however, when it comes to the attention given to the test execution and implementation.

Together with a client, Marianne Gillette of McCormick and Co., Peryam & Kroll developed a test methodology based on the fundamentals we knew were the components of food systems and the people interacting with the systems. The components of this sensory quality system (SQS) are that variability is a constant; more illustrations of references are better than fewer; simple numerical or color systems make rating tasks easier; preestablished decision criteria simplify the task after data are collected; and development of a system that can be used in both the lab and the plant allows for much more straightforward exchange of data.

We based the system on rational decision making. Prior to development of the method, we defined the criteria we believed to be mandatory for a successful sensory quality control program. These were that the program and method must be acceptable to all suppliers; be acceptable to customer management; describe a permissible range of deviation; clearly identify unacceptable product; provide for monitoring and auditing by the customer; and be based primarily on human sensory response.

Features identified as desirable but not totally necessary for the program were that it be modular to enable it to be implemented partially or totally; be eventually transferable to instrumental methods; provide quantitative data; reflect consumer opinion; give direction for on-line correction of defects; be applicable to studies such as shelf life and ingredient replacements; and be accomplished at minimum cost.

SQS as applied to production monitoring is represented by an established ”gold standard” that represents the target for production; a defined range of acceptable deviation from the standard; a selected panel of judges, calibrated on the range of acceptable and unacceptable deviation, to compare production samples to the standard using a rating scale; and a total quality score arrived at by averaging individual ratings.

• The Standard. Defining and maintaining a standard is the sine qua non of the program. The standard must be a physical representation of the end-users' concept of the product, and should be readily available for examination by the panel members.
  
  A key point to be recognized is that, in practice, the target flavor profile is not a single point, but a range. One can define the ideal, but for practical reasons it is necessary to establish the acceptable range of deviation from the standard. This range can be described by the customer, consumer, product development department or committee, or the quality control department. The range is also defined and communicated by means of actual samples.
  
  
• The Panel. Initially, 10-25 people are selected from the available population, based on their general interest and availability. These people are given a brief generalized training, including what is expected from them as panelists, and are instructed in the operation of the method. Most important, they are familiarized with the product, its primary ingredients and characteristics, and the acceptable range of quality. Training and calibration are a matter of demonstrating the meaning of these ranges for a specific product.
  
  Calibration of the panel is the cornerstone of this kind of sensory quality system. In the reality of production, a perfect sensory match cannot be expected. Even if it were, a range of training samples allows panel management to identify more and less acceptable product, and teaches the panelists to distinguish among them.
  
  Usually 6-10 calibrated samples are examined by each panel member during the training course. Because the target is a range and not a fixed point, target drift and minor deviations are acceptable, but at some point, differences become borderline, then totally unacceptable.
  
  In practice, the ranges are defined by having product development prepare 10 15 laboratory samples designed to vary in minor, moderate, and major ways from the product target profile. They can be obtained by specially formulating them, modifying standard formulas, using production rejects, or intentionally abusing the product. The objective is to illustrate effects of minor production errors and demonstrate the possible range of deviation. Preparation of the calibration samples should be carefully documented so they can be replicated in the future to train new panelists, or to transfer the panel system to another site.
  
  Ideally, the calibration samples should be graded and approved by the customer. This provides for greater involvement and acceptance of the program.
  
  
• Method of Measurement. Product quality is judged in terms of the degree of difference from the standard. A typical ballot is shown in Fig. l. It uses a 10-point rating scale with successive categories representing judged degrees of difference from the control. A score of 10 would indicate that the sample was identical to the standard, while a 1 or 2 would indicate that the sample was vastly different.
  
  The scale is divided into four ranges that are marked off for convenience, although the panel members should already have become very familiar with them during training. The names of the ranges imply the decision function; i.e., they direct how the results should be used.
  
  The overall quality score on which action is taken is the average of the ratings assigned by the participating panel members, together with a close consideration of the range of scores.
  
  The second section of the bal- lot collects further information about the test sample beyond the quality rating. The intent is basically to have a diagnostic. If a sample has been rated as less than a match (a score of 8 or lower), the panelists are asked to examine the list of attributes and indicate why the sample was not a direct match to the standard. The direction of deviation (stronger/weaker) is also indicated. Such auxiliary information does not determine the immediate fate of the sample, but is very useful in trouble-shooting and tracking down drift.

Sound Sensory Research

Although we felt that SQS was a progressive program, we had no illusions of its being anything but an adaptation of traditional methods. That was, in fact, a point in its favor. We began with basics that had been proven through the years, so we knew that results could be relied on. The system also met a number of principles proven sound over the years:

• Good sensory research does not overpromise. It outlines issues and attempts to address them. SQS acknowledges that variations exist, so it begins with a definition of acceptable and unacceptable boundaries.
• There are no absolutes, because food systems are inherently variable. SQS takes this into account.
• The scaling task should be as simple and straightforward as possible. The 1 10 rating system is simple and easy to understand and use.
• The greatest challenge in sensory research is defining the objective and then determining the process that will best meet the objective. The objective and decision criteria were established before any development work was done on SQS.
• The purpose of sensory research is to provide actionable data. Without an underlying applied component, there is little value in any system. SQS has highly tangible results, and the data collected are directly usable by product development and production. A poor match could shut down a line in the short term, and protect the product's integrity in the long term.

Every sensory professional needs to stay alert to contemporary trends and radical new methods, because they may offer opportunities to his or her clients or employer. However, few radical methods turn out to be more than a fad. When we concentrate on building on classic techniques proven over time, even when designing contemporary systems, we are generally rewarded with a successful system that offers conclusions with a high degree of certainty.

---

Author Beckley, a Professional Member of IFT, is Vice President, Marketing & Sales
Peryam & Kroll Research Corp., 4175 E. LaPalma, Anaheim, CA 92807.
Author Kroll is Executive Vice President, Peryam & Kroll Research Corp.,
6323 N. Avondale Ave., Chicago, IL 60631. Send reprint requests to author Beckley.

---

--Edited by Neil H. Mermelstein, Senior Associate Editor

 

Reprinted from Food Technology 50(2) 61-64
©1996 Institute of Food Technologists