**1. Introduction**

Globalization, which involves a fierce competition among companies to retain consumers and simultaneously to find new ones, is the driving force that keeps food and other industries innovating to make well-informed decisions in the marketplace. In the course of innovation, many concepts are brainstormed, and only a few selective ideas are taken forward for bench- or pilot-scale testing. However, the success of these prototypes in the marketplace is not guaranteed. Indeed, 50–70% of newly launched food products do not last long in the market [1], despite their intensive market research. Sensory and consumer sciences provide a few tools in this context [2–5] to better understand products [6–9] and population categories [10], and to minimize the risk of failure [10,11]. One of the most popular and extensively used methods to quantify affective responses in sensory science is the acceptability test using the nine-point hedonic scale [12]. However, product sensory liking does not necessarily always convert into a purchase [13], and formulators should look beyond the liking scores to make a product successful [14]. Thus, many other methods [3,13,15] are being explored to provide insights into food-choice behaviours. A novel sensory and emotional approach is the use of non-verbal cues, such as facial expressions, which can communicate highly detailed information about the individual's experiences [16]

**Citation:** Mehta, A.; Sharma, C.; Kanala, M.; Thakur, M.; Harrison, R.; Torrico, D.D. Self-Reported Emotions and Facial Expressions on Consumer Acceptability: A Study Using Energy Drinks. *Foods* **2021**, *10*, 330. https://doi.org/10.3390/foods 10020330

Academic Editor: Derek V. Byrne Received: 2 October 2020 Accepted: 29 January 2021 Published: 4 February 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

and helps in understanding product liking and different emotions that influence the purchase intention. Emotions play a significant role in the comprehension of food preferences and consumers' likings. Moreover, emotions are decisive factors in our food choices since the consumed foods can evoke certain emotions [14,17,18]. Several factors affect emotions, such as age, satiety, health, economic condition, and expectations. In addition to these, two types of emotional responses, conscious and unconscious, can be elicited by consumers when exposed to different products [19,20]. However, most market research is based on conscious arousal and measured with self-reporting scales [21,22], such as EsSense Profile® [23] and PrEmo® [24].

Recent studies using the EsSense profile® questionnaire have validated its discriminating power within and between food product categories [25]. The EsSense Profile® questionnaire is cost-effective, easy to use and interpret, covers a wide range of emotions and has provided rich insights into the consumers' perceptions as well as the liking of products such as beer [26], wine [27,28] and coffee [29,30]. However, the explicit method of emotional measurement requires cognitive thinking to convert the experiences into expressions, which sometimes lose the actual meaning of the emotions felt.

Automatic facial expression recognition (hereafter AFER) is one of the important novel methods to study emotional responses and human behaviours. AFER is a non-verbal and arguably universal language [31] that helps communicate countless emotions, such as happiness, sadness, anger, fear, surprise, and others among humans. Recent studies have been conducted to find the correlation between the AFER and emotions, and how both are comparatively linked with the self-reported likings [32]. Infants and children have been using facial expressions to communicate their emotions and feelings [33], especially in the case of sweet foods, when they smack their lips, protrude the tongue and smile; while, for bitter food, they wrinkle their nose and turn their heads [34]. Researchers have found that facial expressions can recognize and differentiate the basic tastes and odours [35] and found that consumers elicit negative expressions (dislike) more accurately than positive expressions [36]. AFER and autonomous nervous system responses provide insights into food preference in relation to food properties for different kind of foods [37–39] and its influence on the purchase behaviours of consumers. The facial expressions can elucidate the consumer acceptability of products based on emotional responses and familiarity. In the present study, energy drinks were selected as the beverage model for evaluating the hedonic and emotional responses of consumers. Energy drinks are soft beverages that contain several stimulants such as caffeine and glucose and have a wide range of flavours and mouthfeels [40], which can affect mood and mental energy [41]. Many studies have proved the effect of caffeine or the combination of caffeine with other ingredients on the mood and cognitive performance [41]. Caffeine in doses of 75 and 150 mg enhances positive emotions (such as happiness and calmness), while it reduces tenseness [42,43] and caffeinated taurine drinks improve alertness [44] and attention [45]. The overall liking and explicit emotion measurements were investigated using a self-reported questionnaire, and implicit emotion measurements were observed from the facial expressions using the Affectiva® software. The study aimed to investigate whether self-reporting liking and explicit emotion measurements provide similar differentiation among the samples. It also evaluated whether the positive/negative emotions elicited during the implicit emotion measurements were correlated to the sensory attributes of the products. The explicit and implicit measurements of emotions on the differentiation of samples were also investigated. Finally, self-reported liking, and explicit and implicit measurements of emotions were studied on the samples' differentiation. Therefore, the following hypotheses were proposed for this research: H1—The self-reported liking and positive explicit emotion measurement are positively correlated; H2—The positive facial expression emotions and liking will depict similar product differentiation; and H3—The self-reported liking, explicit and implicit emotions, exhibit similar behaviours.
