Comparison of Methods for Evaluating Drivers of Liking for Yakju: Ideal Napping versus the Check-All-That-Apply Method

Abstract

Check-all-that-apply (CATA) and napping are cost- and time-efficient alternatives to conventional descriptive analysis for evaluating product sensory characteristics. This study introduces ideal napping as a novel sensory evaluation method that integrates the napping procedure with the concept of constructing an ideal product profile. Ideal napping was used to identify drivers of liking for 12 commercial yakjus (Korean rice wines). When comparing the product configurations generated by the multiple factor analysis between the two methods, a similar representation of product locations, including the ideal product, was observed. The ideal product was characterized by statements such as "fruit-related," "green plum-related," and "sweetness," while the "yeast-related" attributes were far from ideal according to the results of both the methods. Drivers of liking or disliking were identified by partial least squares regression for these two methods and yielded similar results, revealing the main drivers of liking to be "sweetness," "sweet aroma," "fruit flavor and aroma," "green plum flavor," "floral aroma," and "carbonic mouthfeel," while the main drivers of disliking were "acridity," "bitterness," "yeast flavor and aroma," and "mushroom flavor and aroma." Ideal napping was able to provide additional insights into the consumer perception of ideal products with performance similar to CATA. The findings offer practical insights for liquor product development and reformulation. By identifying sensory attributes that influence consumer preferences, manufacturers can better align products with market expectations. Additionally, comparing ideal napping with CATA provides valuable guidance for improving consumer satisfaction with traditional beverages such as yakju.

Keywords: check-all-that-apply; consumer perception; ideal napping; sensory attributes; yakju.

Fig. 1

Multiple factor analysis biplot based on (A) ideal napping and (B) check-all-that-apply. The first two dimensions explain 48.54% and 72.76% of total variance, respectively. Blue squares indicate samples; red squares indicate sensory attributes. Samples near the “ideal” point indicate higher consumer preference.

Fig. 2

Comparative multiple factor analysis of individual configurations of 12 yakju samples based on ideal napping (IN) and check-all-that-apply (CATA). The first two dimensions explain 63.90% of the total variance (F1: 44.52%, F2: 19.38%). Red triangles indicate samples (A－L); colored squares indicate individual configurations from IN and CATA, color-coded by product.

Fig. 3

Standardized regression coefficients from partial least squares models predicting overall liking of 12 yakju samples based on sensory attributes from (A) ideal napping and (B) check-all-that-apply. The y axis shows the standardized coefficients. Positive values indicate drivers of liking; negative values indicate drivers of disliking.