Affective Simon Effect: A study in children and adolescents

Efecto Simon afectivo: un estudio en niños y adolescentes

Zamora, Eliana V. eliana.zamora@conicet.gov.ar

Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Argentina

 https://ror.org/03cqe8w59

Introzzi, Isabel

Consejo Nacional de Investigaciones Científicas y Técnicas, Argentina

 https://ror.org/03cqe8w59

del Valle, Macarena

Consejo Nacional de Investigaciones Científicas y Técnicas, Argentina

 https://ror.org/03cqe8w59

Aydmune, Yesica

Consejo Nacional de Investigaciones Científicas y Técnicas, Argentina

 https://ror.org/03cqe8w59

Richard´s, María M

Consejo Nacional de Investigaciones Científicas y Técnicas, Argentina

 https://ror.org/03cqe8w59

The Simon effect, and the tasks which make use of it, take as a premise the prepotent tendency of human beings to respond ipsilaterally (i.e., on the same side where the stimulus appears; Hommel et al., 2004; Lu & Proctor, 1995). According to De Houwer and Eelen (1998), three elements define the Simon paradigm: (1) a relevant feature of the stimulus which determines the correct response, (2) an irrelevant feature of the stimulus which must be ignored, and (3) a response which is related solely to the irrelevant feature. In this type of task, participants must press a corresponding key at the right or left side of the keyboard when the target stimulus appears on the screen. Whether they have to press the right or left key depends on a relevant feature of the stimulus (e.g., identity, color, or form), regardless of its location.

Experimentally, the effect is manifested in (a) a decrease of Reaction Times (RT) and in (b) an increase of accuracy when the correct response is ipsilateral to the stimulus, compared whit when the correct response is contralateral (Kornblum & Lee, 1995). In the latter case, the participant must inhibit the prepotent tendency to respond ipsilaterally. As the irrelevant feature (location) is processed jointly with the relevant feature, and both generate contradictory responses, a conflict appears (Fuentes Melero & García Sevilla, 2008). The stimulus location automatically activates the tendency to respond accordingly to that location. But this response competes with the one required by task instructions. Although location is irrelevant, people respond more slowly (additional delay 20-30ms approx.) when the stimulus appears on the opposite side of its associated response (Lu & Proctor, 1995; Pellicano et al., 2010;Proctor & Vu, 2006). In these conflicting tasks, the retention of an impulsive response is not the object of interest. Instead, what is assessed is the capacity to provide a new response which is incompatible with the prepotent response (Carlson & Moses, 2001;Cragg, 2016). This makes this type of task suitable for assessing complex, high-level processes such as response inhibition.

In the last few years, many classic paradigms used to assess stimuli characteristics (this is, to assess stimuli features such as color, or shape) have been adapted to explore the automatic processing of these characteristics and their subsequent control. Among them is the emotional Stroop Paradigm (Williams et al., 1996). In the emotional Stroop task, participants are asked to name the word color as fast as they can, ignoring their content (neutral or emotional). Emotional words show longer RTs. Pratto and John (1991) and Schimmack and Derryberry (2005) suggested the existence of automatic processing of negative emotional stimuli, which appeared as an evolutionary and critical for survival trait of danger awareness. Other paradigms adapted to explore this topic were the priming paradigm (also known as affective priming paradigm), the rapid serial visual presentation paradigm (i.e., attentional blink), and the Simon paradigm. Regarding the latter, in the last years, interest in the Simon effect has increased (Wentura, 2019). Findings allowed researchers to establish different approaches, like the Affective Simon effect (i.e., the Simon effect in emotionally salient contexts). The task is based on this paradigm, and is called the Affective Simon Task (AST). De Houwer and Eelen (1998) and De Houwer et al. (2001) agree that there are three elements of the Simon paradigm that can be found in the AST: (1) a relevant feature that determines the correct response (e.g., grammatical category or color); (2) an irrelevant feature that should be ignored (i.e., the stimulus valence); and (3) responses that are significantly related solely to the irrelevant feature.

In the AST, participants are asked to give an emotionally positive or negative response to the target stimuli based on a stimulus feature, while ignoring its emotional salience. For example, participants should say “good” whenever a word referred to the “people” category is presented, and say “bad” when a word referred to the “animal” category is presented. Results show that even though the target stimulus valence is irrelevant for the task –and must be ignored–, responses are faster when they are congruent with the target stimulus valence (e.g., responding “good” to the word “friend” because it belongs to the “people” category). On the contrary, responses are slower and less accurate in incongruent conditions (e.g., responding “good” to the word “enemy” because belongs to the “people” category; De Houwer et al., 2001, Experiment 1). Also, in addition to emotional words, figures with emotional valence have been used as stimuli (De Houwer et al., 2001, Experiment 3).

Different studies have explored the Affective Simon paradigm (De Houwer, 2003a, 2003b;De Houwer et al., 2001;De Houwer et al., 1998; De Houwer & Eelen, 1998; de Jong et al., 2003; Duscherer et al., 2008; Tipples, 2001;Voß et al., 2003). However, these studies have focused on the adult population. Although both, the classic Simon paradigm and the affective Simon paradigm, are formally equivalent in adult population, Duscherer et al. (2008) suggest that accurate interpretations of this effect require further analyses. Specifically, the analysis of the relationship between stimuli and responses, as well as of the population under study.

In this regard, the Simon effect in children and adolescents has only been studied using the classic paradigm. But the Affective Simon variant has never been analyzed in these populations. Regarding the classic Simon, it was found that, as children improve their fine motor skills, they become faster and more accurate (Amso & Casey, 2009). The fact that Simon effect size decreases with age suggests that as children grow up, they take less time to inhibit incongruent prepotent responses (Amso & Johnson, 2005; Davidson et al., 2006). According to Casey (2005), performance in this type of task continues improving during infancy until approximately twelve years of age. After this age and until approximately twenty years of age, performance in these tasks is better adjusted to a non-linear developmental trajectory, characterized by advances and setbacks, and even by stability periods (Casey, 2005; Constantinidis & Luna, 2019; Pauls et al., 2013). In contrast with adults, during childhood and adolescence cognitive processes develop at different speeds. Moreover, it seems that pubertal arousal and motivation changes precede the development of regulatory competence, creating a disjunction between the adolescent’s affective experience and his/her ability to self-regulate (Dahl, 2001;Steinberg, 2005). Also, according to Casey (2015), adolescents and children show an imbalance between the limbic regions -functionally fully developed in adolescents- and the prefrontal cortex –not yet fully developed in children-. Therefore, it could be interesting to compare cognitive process development between children and adolescents.

De Houwer and Eelen (1998) argue that the AST is a flexible tool for studying the automatic vs. controlled affective processing because it allows the analysis of a biologically-determined response and its subsequent control. Hence, the AST might become a valid tool to assess automatic vs. controlled processes, and even further, to measure inhibitory control. Traditionally, response inhibition has been assessed using relatively abstract decontextualized tasks that lack a significant affective or motivational component. More recent characterizations of response inhibition have suggested that tasks vary in motivational significance, with emotionally significant tasks described as ‘‘hot’’ and more neutral tasks described as ‘‘cool’’ (Zelazo & Carlson, 2012; Zelazo & Cunningham, 2007). Hot tasks involve stimuli, decisions, and outcomes that are emotionally salient.

Also, in the AST, stimuli of any modality and complexity can be used. Moreover, different relevant features for each type of stimulus can be chosen. Despite this fact, in the classic AST, words were mostly used as stimuli, and their grammatical categories were chosen as the relevant features. And despite the extensive research in this field, literature shows a limited comprehension of this paradigm in children and adolescents. This issue is especially relevant because children’s and adolescents’ skills to assess information in emotionally significant contexts are different from adults’ (Silvers et al., 2012). Since the similarity of this effect in children and adolescents has not been studied, this study aims to provide evidence about the Affective Simon effect on these populations. To that end, pictures of objects and animals were used, varying their affective valence and their relevant features. Likewise, the Affective Simon effect was analyzed in different children and adolescent’s age groups aiming at studying age-related differences.

In the Simon paradigm, participants must respond to a relevant feature while they ignore an irrelevant feature related to the response (Kornblum, 1992). While in the classic version of the paradigm, the irrelevant feature and the response are spatially related, in this study we introduced a variant. In the task participants had to respond by choosing a happy or sad face depending on whether the picture presented was an object or an animal, ignoring its affective valence. In some trials participants must respond congruently (a pleasant object, of positive valence), and in other trails they must respond incongruently (an unpleasant object, of negative valence). Results of Affective Simon in adult population show that, although participants were instructed to ignore the affective meaning, RTs are increased in incongruent conditions compared with congruent conditions (De Houwer et al., 1998; De Houwer et al., 2001; De Houwer, 2003a, 2003b;Duscherer et al., 2008; Tipples, 2001;Voß et al., 2003). However, until now, this effect had not been analyzed in children and adolescent, and thus neither was evidence about the generalization of the affective Simon effect to these populations.

Table 4
Post hoc comparisons and Bayes factors by age-group for each condition

Note. BF: Bayes' Factor analyses, G1: Group 1, G2: Group 2, G3: Group 3, G4: Group 4.

The results showed that children (G1 and G2) presented lower performance at the incongruent condition than at the congruent condition, which indicates the presence of the Affective Simon effect. This is evidenced by higher LISAS scores (PO and PE). Although the results for G1 are not conclusive (see Table 4), the strong difference observed in G2 in LISAS PO could be taken as evidence that the effect exists in G1. In contrast, in the adolescent group (G3 and G4) evidence of invariance between both conditions was found for both variables (LISAS PO and LISAS PE). RT interference scores also show significant differences for the G1 and G2 groups, and revealed an absence of the effect on G3. In the case of G4, the results of this index were not conclusive.

Therefore, these results provide some evidence that suggests that the Affective Simon effect only appear in children. These findings can be explained from different perspectives. Some developmental changes could explain the performance differences found between children and adolescents. For example, Davidson et al. (2006) point out that the motor interference generated by the Simon effect can be solved efficiently during this developmental stage. Authors also point out that it is to be expected that as long as there are age-related performance improvements, both conditions become increasingly similar. However, the Affective Simon effect has been reported in the adult population. This discrepancy can arise from differences between this study and previous ones. Specifically, in the AST designed by De Houwer et al. (2001) the affective pictures were associated with negative and positive valence scores (Experiment 3). According to Schimmack and Derryberry (2005), the picture’s arousal could have explained the Affective Simon effect found, but it was not controlled in the stimuli selection process. This can be supported by the fact that, generally, extreme valence scores are also associated with high arousal scores. In contrast, in our study, arousal was controlled to be medium across all conditions and age groups, with valence as the only parameter determining whether the stimulus was negative or positive. Thus, our results show a valence-related Affective Simon effect only in children but not in adolescents. This implies that the Affective Simon effect reported in the adult population (e.g., De Houwer et al., 1998;De Houwer et al., 2001;Duscherer et al., 2008; Tipples, 2001;Voß et al. 2003) could be due to the arousal but not the valence scores of the pictures. Therefore, a possible arousal-related Affective Simon effect can be explored in future studies.

Also, results showed that performance in both conditions improved with age, in line with previous studies (Amso & Johnson, 2005; Casey, 2005). This could be attributed to the operative improvement during these developmental periods in diverse processes, such as processing speed or inhibition (Introzzi et al., 2019; Kail & Salthouse, 1994). Inhibition implies the capacity of resisting automatic and/or prepotent responses, in pursuit of an objective. Therefore, it is logical to expect that experience and maturation linked to development could affect the intensity of inhibition. The more intense the inhibition is, the faster and more accurate the responses are. And processing speed is the maximum execution speed of cognitive operations. Thus, the higher the processing speed, the better the performance in cognitive tasks that do not necessarily imply speed as the main component. Studies analyzing development through the classic Simon paradigm have found a significant improvement of the processes involved in all groups with differences between 7 and 11 years of age (e.g., Cragg, 2016;Dahl, 2001;Luna et al., 2004;Richardson et al., 2018). In our study, there were significant differences between almost all age groups, except between G1 and G2 (in which no age-associated differences were found in the congruent condition, and even a weak invariance was detected for the incongruent condition). However, a general age-related performance improvement can be seen, in line with the literature. Thus, adolescence could also be considered as a sensitive developmental period, due to the occurrence of significant cognitive and emotional changes (Casey, 2005; Constantinidis & Luna, 2019; Steinberg, 2005).

Worth to mention are the composite scores used in our results. Generally, accuracy and RTs have been used to analyze emotional interference and/or inhibition development. However, in this approach, the trade-off effect could lead to a wrong interpretation of the findings. This effect is often found during childhood and adolescence, where developmental changes occur fast, and cognitive performance improves. Likewise, using only RTs, without considering errors or omissions, could lead to biased results. Thus, by using a composite score as the LISAS (Draheim et al., 2016; Vandierendonck, 2017), it is possible to provide a more reliable and parsimonious measure of the performance of an executive function as, for example, inhibition (e.g., Nweze & Nwani, 2020). However, it is important that future studies consider both the possibility of analyzing other indexes for understanding subtler developmental changes (e.g.,Cragg & Nation, 2008), and the possibility of working in different levels of analysis (e.g., neural, in addition to behavioral; Vara et al., 2014).

Although the idea of the Affective Simon effect results is attractive, the evidence is yet limited and inconclusive. Therefore, this study needs to be replicated, but adding some methodological modifications to overcome its limitations. One of the limitations of our study is the lack of counterbalancing in the response mapping of the task. This could have affected the results because participants’ laterality was not controlled. Another possible source of error is the number of trials per condition. Although the task is based on previous studies, more trials per condition would improve its sensitivity to fatigue and lapsus (and the reliability of errors as a performance measure). Likewise, another limitation to highlight is that the reliability of PE and PO was low, so future studies could explore this type of effect. Also, stimuli used here must be discussed. While the child-adapted version of NAPS (Zamora et al., 2020) was used in G1 and G2, the adult version of NAPS was used in G3 and G4 (Marchewka et al., 2014). Thus, future studies including stimuli validated in the adolescent population could contribute to the replicability and robustness of the results. Larger and different (e.g., clinical population, different socio-cultural contexts) samples should also be considered. It would also be interesting to consider shorter age intervals, in order to analyze changes more accurately. Finally, we used a non-probabilistic sample, so generalization of results should be made with caution.

Regardless of the limitations, several contributions can be summarized. For instance, we applied a task based on the Affective Simon paradigm, which had not been previously applied in children and adolescents. This study contributes to a scarcely explored topic such as the assumption that an implicit activity (i.e., responding through faces), measures a real activity which underlies a behavior (i.e., association of the pleasant to the happy face and association of the unpleasant to the sad face). More importantly, we found that there is a valence-related affective Simon effect only in children which disappears in adolescents. And also, that the affective Simon effect reported in adults could be due to arousal and not valence of stimulus. This paves the way to a series of new studies that could analyze the arousal-related Affective Simon effect, in both children and adult populations.

The full dataset supporting the results of this study is available upon request to the contact author Eliana Vanesa Zamora.

The full suite of analytical methods supporting the results of this study is available upon request from the contact author Eliana Vanesa Zamora.

The entire set of materials supporting the results of this study is available upon request to the contact author Eliana Vanesa Zamora.

Citar este artículo como : Zamora, E. V., Introzzi, I., del Valle, M., Aydmune, Y. & Richard´s, M. M. (2024). Affective Simon Effect: a study in children and adolescents. Revista Argentina de Ciencias del Comportamiento, 16(2), 37-50.

redalyc-journal-id: 3334

https://revistas.unc.edu.ar/index.php/racc/article/view/37662/44978 (pdf)