Psychological biases affect hedonic ratings
Loading...
Date
2013-03-29
DOI
Open Access Location
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Rights
Abstract
Psychological biases in consumer testing may lead to misinterpretation of results and lower experimental power. Reports on various hedonic scales associated with psychological biases induced by sample presentation are limited in the literature. An appropriate experimental protocol could enable sensory scientists to accurately determine if a product is more or less liked. Overall, in this study some drawbacks of hedonic scales were revealed and some recommendations were made under specific circumstances. A more powerful design (SPRCBD) helped minimize positional and First Serving Order (FSO) biases in consumer tests by extracting more explained variances, resulting in decreased Type-II error in the model. Logistic regression analysis was proven to be an alternative methodology to quantify sensory contrast effects. For sensory testing, a multidimensional attribute tended to be more affected by the contrast effects than a simpler attribute. Several scales have been used for assessing the degree of food liking/disliking. This study provided a good practice protocol, suggesting use of a regular scale length (100 mm.) for assessing a degree of food liking/disliking while Labeled Affective Magnitude (LAM) would be an alternative choice where the scale length effects may be a critical issue. Depending on the type of scale and its polarity, a negative attribute (e.g., bitterness) was more affected than was a positive attribute. When testing extremely liked product, one should be aware of contrast biases that affected more toward positive attributes than negative attributes. This study demonstrated some psychological biases that affected the hedonic ratings. There are many more factors that could sway sensory responses and prevent experimenters from getting accurate, valid and actionable outcome. Understanding of psychological biases, proper product selection, and proper data analysis should be further studied to minimize misinterpretation of hedonic ratings.
Description
Keywords
Citation
2013