Obsession for Detail—

October 2001

I began my life in sensory at a soybean-processing company, where my primary responsibilities included preparing samples for laboratory-panel assessment and recording data. The tasks, much to my dismay, were tedious and routine. I’d have preferred doing what I saw as a much more challenging job of interpreting data and writing reports. It didn’t take long, however, for me to see that my work had profound importance. I owe this enlightenment to my immediate supervisor, Toni Trinchese, and consultant David Peryam, a sensory guru who is best known as the father of hedonic scaling. These sensory pioneers instilled within me a deep respect for sample integrity and documentation that became the foundation of my professional career.

Spotlighting the design
The sensory spotlight is generally on trained panelists and carefully recruited consumer participants. Like the cast of a theatrical production, their work is only as good as the script. In research language, that script is known as the design of experiment (DOE). The DOE is a strategy used to solve a research problem based upon a clearly stated objective. Included in the DOE are experimental variables, sampling frequency and sample presentation sequence.

The DOE’s success depends greatly upon the preparation and integrity of the samples, which must be handled so that no extraneous factors influence the outcome. There is no detail too small to go unexamined, no preparation too minor to be left to chance. An oversight or mistake will turn the work of the sensory analyst to naught and invalidate thousands of dollars worth of research.

Experimental planning for sensory testing is best done in collaboration with a sensory statistician. At Food Marketing Support Services (FMSS), Oak Park, IL, we work with Carr Consulting, Wilmette, IL. Tom Carr has played a major role in the emergence and recognition of sensory evaluation as a key component of product development, enhancement and optimization. Carr takes a broad-enough look at the sampling options to assure research that gives findings that are not a fluke. For example, in target-matching exercises using descriptive analysis, he recommends evaluating multiple production codes of a given product. Sampling multiple codes gives an analyst and developer an idea of the sensory space occupied by the product at a given point in time.

In supply-quality initiatives, Carr Consulting advocates exposing consumers to descriptively validated samples that span a broad range of the sensory space. Designing consumer tests with products that reflect a sensory spectrum results in a broader distribution of liking scores. (“I like this better than that.”) The preparation of consumer samples is every bit as rigorous as the preparation for descriptive sensory testing. Carr Consulting links consumer and descriptive data. The “linking” determines which sensory attributes are the key drivers of liking. FMSS uses key driver data to adjust formulation, processing and packaging variables.

A rose may not be a rose
Pretest procedures are to sensory testing what an off-Broadway run is to a major theatrical production. The first step is to verify or develop a sampling plan outlining a preparation protocol that includes specific times for each step (e.g., thaw, cook, hold), weights, equipment and handling procedures, ingredient codes, seasoning and assembly requirements, and serving-temperature measurement points.

In addition, samples and equipment must be examined for potential factors that could impact the experiment’s outcome. Samples may, for instance, come from production plants in different parts of the country. Therefore, before the study is initiated, a decision must be made about sampling one plant with multiple runs, or multiple plants with multiple runs. Agricultural products, picked at different times of the year, reflect climate and growing cycles. For example, New Mexico chiles can take on a grassy flavor when picked after a frost. Texas-grown cabbage has a different flavor profile than the Californian-grown variety, and they are not interchangeable until proven so. A testing facility makes many calls to field personnel to document factors that could impact a study.

No two agricultural products are the exact same size or shape. I once participated in a study on rotisserie chicken where the project supervisor was a real stickler for uniformity. She wanted all of her chicken samples at the same degree of doneness at the same time. After examining what could be done to decrease variability, a co-worker and I donned heavy thermal gear and stationed ourselves in the walk-in cooler where we measured the cooler temperature from top to bottom and front to back. We racked the chickens on trays so they would be the same distance apart. The careful racking resulted in similar precook temperatures, and more predictable cooking behavior.

Poultry and meat present special challenges for sensory preparation. At Swift & Company, Greeley, CO, where I was director of the taste laboratories, research home economists supported the internal staff of sensory specialists. The research home economist in charge of directing the preparation procedures was extremely diligent. She made sure that panel samples came from the same muscle with the same grain rotation and were of the same thickness and degree of doneness. There would often be as many as 10 treatments within a test.

For work with a major international quick-serve restaurant (QSR) chain, FMSS obtained ingredients — such as frying oil, breading, spices and condiments — that were used to prepare the product as served in the international setting. The goal was to emulate the menu item as served. Substitution of domestic ingredients would have compromised the experiment.

Essential equipment
Equipment is a key factor in both consumer and descriptive testing. Experimental accuracy requires strict adherence to laboratory procedures for cleanliness, storage and functionality — checked and double-checked. If equipment could potentially impact the sensory behavior of a product, then testers must obtain dedicated equipment. In a study of coffee blends, for instance, 10 new identical coffee makers were purchased. To verify an ad claim about product storage under refrigeration, six refrigerators of the same make and model were procured to eliminate an equipment variability factor.

This kind of testing is costly, but not to address the equipment factor can skew research data, which is even more costly. Beware of product evaluation companies that cut corners on equipment integrity.
When dealing with foodservice, use a client’s own branded equipment after rigorous performance verification. With coffeemakers and urns, for instance, calibrate temperatures, and measure and record warming-plate cycles. Perform test runs with deep fryers, surface-temperature measurements of grills, and careful calibration of holding bins, toasters and other equipment. If there is any new equipment, the client must provide training and precise preparation protocol to the technicians who work on the consumer or expert panel tests.

Backstage manager
When all of the equipment and sample variables have been addressed, preproduction rehearsals begin. Put the samples “through their paces” to make sure there are no surprises. Time sample preparation and collect temperature readings. Throughout the trials, consult the client to ensure that all of the procedures are the same as officially prescribed practices.

All of this requires great organization and management skill. “I have to double-check — no, triple-check — every aspect of the sampling plan,” says Peggy Eades, manager of FMSS descriptive sensory services. Her casebook is full of examples where she had to make adjustments for unanticipated circumstances. In one instance, she was surprised to find that a frozen product was shipped in different size cartons, which affected the thaw time. Uneven browning of a pan of frozen biscuits had to be addressed via a sampling redesign. To account for disparities in degree of perceived “browned” flavor, the descriptive sensory panel sampled both interior and exterior biscuits. As the result of pretest observations, an aroma analysis of cooking bacon was included as part of a sensory analysis, when cooking aromatics were not a concern in the original project scope. The data resulted in an alteration from a creosote-like smoke ingredient to a more appealing sweet, natural-wood smoke product.

Eades defends the practice of ordering a 10% product contingency on every product and ingredient as a backup. Is it costly to over-order? Certainly. Is it even more costly to under-order? Absolutely! Something as unexpected as a power outage in the middle of a thaw process could ruin an enormous amount of research if not for the backup product.

Eades also has a system of checks and balances. “Organization, precision and reproducibility are the watchwords of our sensory team,” she explains. Two technicians handle the inventories. Technicians select color-coded product and a supervisor rechecks by code number before preparing for sampling. Temperatures are checked, rechecked and documented.

As a member of the FMSS expert descriptive sensory panel, Eades is responsible for gathering, organizing and maintaining records of all sensory activities. In a time of great industry turnover, these records are, in some cases, the only research history that exists. One of our long-time clients has called me back many times over the past 12 years to recreate a test scenario, replay the findings and explain the rationale — information that would have been impossible to retrieve and too costly to replicate had we not archived the data.

Attention to and the recording of details prepares sensory analysts and technical support personnel when called on as expert witnesses. If the methodology is ingrained, and the witness has supportive documents that substantiate what is true, there is little cause for intimidation.

Proper presentation
Each descriptive sensory session begins with a reference sample. The reference “anchors” attributes and intensities that will be measured in the blind sample series. The reference also is included as a blind sample. “We don’t include the discussion reference in our statistical analysis,” explains Ruta Lesniauskas, a Carr Consulting analyst, “but we do use it as a panel check.”

Experiments take into account a fatigue factor that limits the total number of samples evaluated, determines the amount of time in between samples, and determines the order of sampling. Experienced panelists can evaluate more samples than a consumer panel. Generally, there is a 15-minute interval between samples, though as long as 30 minutes may be necessary if lingering after-flavors or residual mouth sensations persist.

Sample size varies by product. In many cases, a product is presented in a quantity similar to what would be considered a serving. Some descriptive situations, such as a beverage, condiment or white-meat turkey slice, require only a small portion. Other tastings, such as pizza, a cheeseburger or fries, require a full portion. It is imperative that more than enough is available to provide for each panelist’s sensory exploration, as well as a thorough evaluation of the product attributes.

Present samples on plain plates or in Pyrex beakers and mark them with a random three-digit code. In some cases, the sample may be presented “as eaten,” such as in a sandwich, or as a component ingredient, such as ketchup. Great care must be taken to make sure that all samples are presented uniformly and simultaneously to all participants, and that there are no external interferences. Pure spring water heated to 120°F serves as a rinse between samples.

Descriptive sensory panels evaluate food in five categories: aroma, appearance, flavor, texture and tactile modes, using an intensity scale to rate each of the products. FMSS’s seven-member sensory panel individually rates the blind samples. The individual data are electronically transmitted to Carr Consulting for statistical analysis. Statistical findings then are returned to FMSS for interpretation, reporting and technical recommendations.

Well-designed, statistically based sensory research that is meticulously executed and anchored by consumer liking is the cost-effective approach that is used by successful manufacturers of retail and foodservice products. Attention to detail defines the competitive marketplace.

Nancy C. Rodriguez is a sensory specialist and president of Food Marketing Support Services Inc. (www.fmssinc.com), Oak Park, IL, a contract food product design firm. Anne Hunt, FMSS writer-in-residence, contributed to this article.

3400 Dundee Rd. Suite #100
Northbrook, IL 60062
Phone: 847/559-0385
Fax: 847/559-0389
E-mail: contactus@foodproductdesign.com
Website: www.foodproductdesign.com