Consumer Panels

Consumer panels are conducted in our consumer-lab area located in historic Babcock Hall on the UW-Madison campus. Consumers are generally selected based on the demographics you specify and data is collected with the end goal of measuring preferences or degrees of liking (hedonics). Consumer testing is also known as affective testing.

Preference testing

Preference testing is useful to assess the appeal of certain products over others. In the simplest scenario, consumers are presented with two different products (i.e. paired preference testing) and asked which they prefer or if they have no preference. These responses are recorded and statistically analyzed to determine if there is a product that is truly preferred. It is important to remember paired preference testing sheds light on which product was preferred, not how much each product was liked.

Table 1. Consumer preference results between two products. Difference column assumes 95% confidence level (α<0.05). Product 1 is preferred over Product 2.

SAMPLE PREFERRED DIFFERENT?
Product 1 83% Yes
Product 2 17 %

Acceptance testing

Acceptance testing is used to establish how much a product is liked (hedonics) by consumers. Consumers sample various products and use structured scales to rate how much they like the product. Commonly a nine-point scale is used ranging from dislike extremely to like extremely, with neither like nor dislike in the middle. Hedonic responses are generally collected for: overall liking, appearance liking, texture liking, flavor/aroma liking and any other attributes of interest. Hedonic scales can be assigned numbers to generate average liking scores for each product tests. Hedonic responses are usually combined with “Just About Right” (JAR) scales in order to obtain a measure of adequacy of various attributes.

Table 2. Overall liking (1-9 scale) summary table. Samples with different letters are statistically different from each other (α<0.05). Product 3 is liked the most.

PRODUCT OVERALL LIKING STATISTICAL EQUIVALENCY
Product 3 7.2 A
Product 2 5.7 B
Product 1 5.4 B
Product 4 4.8 B

Figure 1. Results from amount of seasoning JAR scale question. Percentage of responders shown.


Difference testing

Also known as discrimination testing, difference testing is used to identify if products are perceptibly different from one another. In many cases, difference testing is used to access whether or not consumers can tell the difference when a product is reformulated or is compared against a competitor. Examples of differences tests include: duo-trio, triangle, and paired comparison, and ranking tests. Please contact us to determine which would suit your needs. Trained panelists can be used in cases where differences are subtle and/or too complex to explain to consumers.

Table 3. Table outlining Product 2 as being more bitter than Product 1. (α<0.01) Semi-trained panelists (n=40) used in evaluation.

SAMPLE CHOSEN AS MORE BITTER DIFFERENT?
Product 1 18 Yes
Product 2 22

Ranking

Ranking tests are used in order to generate insight into three or more products. Consumers are asked to rank products in order of preference. Ranking tests with large number of samples can be quite fatiguing to consumers. Care must be taken to limit number of products to be ranked. Ranking tests cannot give indication of the extent of the difference between products.

Table 4. Summary of ranking test. (α<0.05)

PRODUCT RANK SUM
Product 3a 30
Product 2ab 24
Product 1b 15
Product 4b 11