Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


osu1282868174.pdf (529.43 KB)

ETD Abstract Container

Abstract Header

Multivariate Approaches for Relating Consumer Preference to Sensory Characteristics

Liggett, Rachel Esther
http://rave.ohiolink.edu/etdc/view?acc_num=osu1282868174

Abstract Details

2010, Doctor of Philosophy, Ohio State University, Food Science and Nutrition.
Preference mapping refers to a category of statistical methods used to relate consumer acceptance to a product’s characteristics (often measured by sensory descriptive analysis). Several techniques for creating preference maps exist, and they vary in the manner by which these two types of data are related. Techniques are generally classified into three main categories: external, internal, and hybrid. For external preference mapping, the sensory perceptual space is set by descriptive sensory data and consumer preference information is subsequently overlaid onto this sensory space. For internal preference mapping, the consumers’ preference for the products (usually overall liking ratings) is used to create a liking space upon which sensory descriptive ratings are subsequently mapped. Hybrid techniques locate products on the map using both consumer and sensory descriptive data simultaneously onto a restricted liking space. Preference mapping studies consume a considerable amount of time and resources. When such investigations are unsuccessful, it is highly desirable to determine the root cause and avoid the same issue in the future. Preference mapping studies are necessarily complex and error can be introduced at any of several steps along the way, such as product selection, descriptive analysis, consumer testing, data analysis and interpretation of the outputs. The purpose of this investigation was to examine these issues and develop best practices for conducting preference mapping studies. One popular technique from each category was selected for investigation using a common dataset. Fifteen commercially available Swiss-type cheeses (10 domestic Swiss cheeses, 4 Baby Swiss cheeses, and 1 imported Swiss Emmenthal) were evaluated by twelve trained panelists using the Spectrum method. Significant differences between the cheeses were exhibited for 15 flavor attributes. The same 15 cheeses were also evaluated for overall liking by 101 untrained consumers (53 female; ages 18-65). Significant differences in liking of the cheeses were also found. External preference mapping (EPM) was able to fit some consumers but generally performed poorly. The resulting preference map explained only 42% of variability. The region of highest appreciation for products was quite broad and determination of an optimum profile was impossible. Landscape Segmentation Analysis® (LSA), a version of internal preference mapping, explained 90% of variability once two outlying products were removed. An optimal profile and 7 liking drivers were obtained. Partial Least Squares (PLS) Regression, a hybrid preference mapping technique, explained 93% of variability after 1 outlying product was removed and an optimal profile and 7 liking drivers were obtained similar to LSA. PLS and LSA offered advantages over EPM in understanding consumer liking, seemingly because the product space was set at least in part using consumer information. Optimal profiles and liking drivers for PLS and LSA were strikingly similar, suggesting both are successful and valid preference mapping techniques. Even though PLS requires pre-treatment of the data to identify possible consumer segments, LSA does not. However, PLS offers a slight advantage over LSA because its output is easier to interpret. Research was funded through financial and material gifts from the Swiss Cheese Consortium, Ohio Agricultural Research & Development Center, and USDA Cooperative State Research, Education, and Extension Service.
Michael E. Mangino, PhD (Advisor)
Jeannine F. Delwiche, PhD (Advisor)
W. James Harper, PhD (Committee Member)
M. Monica Giusti, PhD (Committee Member)
94 p.
Food Science; Psychology; Statistics
Preference Mapping; Sensory; Modeling; Swiss Cheese; Consumer Assessment; Descriptive Analysis; Response Surface External Preference Mapping; Partial Least Squares Regression; Landscape Segmentation Analysis

Recommended Citations

Citations

  • Liggett, R. E. (2010). Multivariate Approaches for Relating Consumer Preference to Sensory Characteristics [Doctoral dissertation, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1282868174

    APA Style (7th edition)

  • Liggett, Rachel. Multivariate Approaches for Relating Consumer Preference to Sensory Characteristics. 2010. Ohio State University, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu1282868174.

    MLA Style (8th edition)

  • Liggett, Rachel. "Multivariate Approaches for Relating Consumer Preference to Sensory Characteristics." Doctoral dissertation, Ohio State University, 2010. http://rave.ohiolink.edu/etdc/view?acc_num=osu1282868174

    Chicago Manual of Style (17th edition)

osu1282868174
5,538
© 2010, all rights reserved.
This open access ETD is published by The Ohio State University and OhioLINK.