Serious games represent a growing area of computer applications used to improve or evaluate different skills. They are appealing, interactive, enhance ecological validity, and allow players to take on realistic roles to cope with problems and to make decisions [1,2]. Games are entertaining, but they can also be educational [3].

The use of computer software has several advantages: the environment is predictable, consistent, free from social demands, and users can work alone. Furthermore, lessons can be repeated, and motivation can be maintained through rewards and feedback [4]. Virtual and mixed realities present the possibility of creating new, immersive, and motivational places where patients can be evaluated and trained while playing [5].

There are various serious games available, such as some for training skills [6], for prevention [7], for psychological therapy [8], or for cognitive training. Other types of games help users to deal with those with special needs, such as the elderly [9], people with physical disabilities, or blind children [10]. An example of a serious game is Happy Farm [11], a software for young people designed to increase their awareness of the risks related to psychoactive substances. Another program, VEPSY (updated telemedicine and portable virtual environments for clinical psychology) [3], was created to investigate the effects of virtual reality systems aimed at dealing with several clinical disorders, such as social phobia, obesity, bulimia, or male impotence. The project combines treatments and assessments with virtual reality. Similar games have been developed to induce mood enhancement on both clinical and nonclinical samples [12]. The EMMA project (Engaging Media for Mental Health Applications) provides innovative ways of coping with distressful emotions for users who suffer psychological problems, users with restricted mobility, or the general population [13]. Another group of serious games has been created to assess and train the components of the theory of mind.

The theory of mind covers mental skills related to understanding, explaining, and predicting the psychological states of oneself and others [14]. The theory of mind was first established in animal studies with chimpanzees [15] and later in infant developmental psychology and autism [16]. The theory of mind permits typically functioning individuals to infer the mental and emotional states of others as a means of engaging reciprocal communication and maintaining relationships. Recognition of emotional facial expressions is an essential part of the theory of mind. The face is the way that emotions can be exteriorized and expressed in a nonverbal way, something essential for a person to adapt to the social environment around them, as shown in “Mind Reading: The Interactive Guide to Emotions” (Jessica Kingsley Publishers, London, United Kingdom). This is a multimedia computer program that has been used to address emotion recognition.

Attempts to teach components of the theory of mind to people with autism spectrum conditions have used computer-based training [17-19] or virtual environments [20]. Golan and Baron Cohen used “Mind Reading: The Interactive Guide to Emotions,” during a study on adults with Asperger syndrome or high-functioning autism [21]. They used it as an interactive guide to teach emotions in a systematic and comprehensive format, as it includes an emotion library, a learning center, and a game zone. The results of this study revealed that the use of the program significantly improves emotional recognition skills in adults with autistic spectrum conditions.

In 2002, Bölte [18] developed a computer-based program to teach and test the ability to identify facial emotions, known as the “Frankfurt Test and Training of Social Affect” (FEFA). The training was conducted for five weeks, for two hours a week, and the participants improved significantly on the facial recognition task. The Motion Picture Mind Reading test is a naturalistic mind-reading test designed to measure individual differences among a young adult population watching TV films showing characters in various social situations [22]. There are several collections of material and databases with facial emotion information and photographs, pictures [23,24], or virtual faces [25]. Different questionnaires have been developed to assess facial recognition ability, including the Florida Affect Battery test [26] or the Feel test [27]. However, a few types of software evaluate facial emotion recognition and empathy in children through serious virtual games.

The present study evaluated a new program, Deusto-e-motion1.0, which was developed to assess and train components of the theory of mind in children between the ages of 8-11 years old. Specifically, this paper presents the development and preliminary evaluation of the Spanish version of Deusto-e-motion1.0 to test the recognition of facial emotions in a sample of 1236 children.

According to Salovey [32], the skills associated with emotional intelligence include the assessment, expression, and regulation of one’s own emotions as well as those of others, and the understanding of emotions and their use in an adaptive way to perform other activities, such as cognitive or behavioral tasks. In this context, the face is the way that emotions can be exteriorized and expressed in a nonverbal way, something essential for children to adapt to their social environment [33].

Deusto-e-motion1.0 is a serious game designed to evaluate components of the theory of mind, specifically facial recognition and empathy in 8-11-year-old children. This study presents the design and validation of the Spanish version of the Deusto-e-motion1.0 serious game.

There is no doubt that a test with suitable psychometric properties contrasted with a representative sample of participants would be beneficial for both the evaluation of the ability to recognize emotional facial expressions and for planning individual or group intervention programs in the area of children’s interpersonal relationships. Moreover, the importance of developing such instruments with demonstrated validity and reliability to have appropriate protocols and paradigms to be applied in basic research should be highlighted, as in the case of neuroimaging studies [34].

This article explores content validity and piloting, internal validity, concurrent validity, and discriminant validity. The test shows moderate concurrent validity by correlating its scores with the Feel test that assesses similar capacity for emotion recognition. These results may be mediated by the characteristics of each test, in the sense that the Deusto-e-motion1.0 test includes, on the one hand, fewer items, and on the other, items with a dynamic and static nature. When working with variables of a very diverse nature, like different types of facial expression stimuli (cultural precedence or faces, photographs versus drawings), a lower correlation will support a high relation. It is suggested that in future studies, the rate of concurrent validity is calculated with a questionnaire with a higher number of static items [35].

There were significant differences in each gender’s scores for the static, dynamic, sadness, and disgust emotions. Results showed that the younger the age of the participants, the slower the reaction time. Overall, there were no significant differences in each gender’s scores in the virtual scenes section. However, age is an important variable to compare both the total answer scores and reaction time scores. With increasing age, facial expression recognition becomes faster and more accurate, possibly due to increased efficiency in understanding faces [36,37]. It is generally accepted that children’s ability to recognize the emotions of unfamiliar faces improves between the ages of 5-10 years old [38].

It should be noted that this study is not without its limitations, and results should be considered with caution. First, the results only indicate the comparability of the classic basic emotions, as described by Ekman. However, daily, pure basic emotions are encountered only rarely. Future research should primarily focus on investigating more ambiguous and nuanced emotional expressions. Second, the Feel test presents only static pictures of adults of Asian and European descent, whereas Deusto-e-motion1.0 presents static and dynamic virtual faces of a boy. Third, the test shows a higher percentage of masculine faces. In the development of a new version of the instrument, stimuli of female faces will be included.