Notes
-
[1]
The design of our experiment does not include the text with a sad emotional content as it appeared ethically questionable to induce sadness in children through recall of a sad memory, especially with regards to emotion regulation skills still developing in children that age.
-
[2]
Degrees of freedom for the reported t-tests were rounded to the nearest whole number when Levene’s test of equality of variance was not significant.
1. – Introduction
1 A recent study by Fartoukh, Chanquoy and Piolat (2014) replicated results from Cuisinier, Sanguin-Bruckert, Bruckert and Clavel’s (2010) study on the influence of the emotional content of a text on children’s orthographic performance in a dictation task. Results from both studies showed children made more spelling errors when the text had a pleasant (joyful) or unpleasant (sad) emotional valence compared to a neutral valence. Cuisinier et al.’s (2010) study tested two hypotheses: that of a direct effect of the emotional content, and that of an indirect effect through the influence of the emotional content on children’s emotional state. In both studies (Cuisinier et al., 2010; and Fartoukh et al., 2014) the design involved reading of a text with an emotional content (joyful, sad, neutral-three independent groups) before dictation of part of the text (of equal length and complexity in the three groups). In Cuisinier et al.’s (2010) study children’s emotional experience was not modified by the emotional content of the joyful and sad texts so the authors concluded in favour of a direct effect of the emotional content on performance. However, children´s emotional state was measured retrospectively, after the dictation task (i.e., “How did you feel during the reading of the text”) and therefore did not really allow an examination of an emotion induction congruent with the emotional content of the text. To clarify the question of induction with the emotional content, Fartoukh et al. (2014) measured children’s emotional state right after reading the text and read the text twice instead of once. As reading of the texts (by an experimenter) induced an emotional state in children congruent with their respective emotional valence, they concluded in favour of an indirect influence supporting the prediction of a detrimental effect of emotion on performance (Resource Allocation Model, Ellis & Moore, 1999). However, as emotion was induced with the emotional content of a text, it is difficult to conclude regarding the validity of the indirect effect hypothesis. Indeed, we cannot exclude that children’s performance was influenced by their processing of the emotional content of the text, especially if we take into account the lack of stability of induction evidenced by a decrease in children’s experience of pleasant emotions after the dictation task. Therefore, a question remains unresolved: whether it is the emotional state induced by the text which affects performance (indirect effect hypothesis) or whether the effect is due to processing of the emotional content itself (direct effect hypothesis) or both. The experiment reported in this paper aimed at clarifying this issue. On a theoretical level this clarification is necessary to better understand the interplay between emotion (i.e., child’s emotional experience, emotion content of the task) and cognition (i.e., performance) in children during a complex school task (dictation); on a pragmatic level, this clarification is important to inform educational practices (for a review see for example Pons, Giménez-Dasi, Sala, Molina, Tornare & Andersen, 2015).
1.1. – Influence of emotion in literacy activities
2 A large amount of research has contributed to our understanding of the cognitive processes that underlie literacy activities, but few studies have integrated emotion in this endeavor. A growing amount of research has nonetheless revealed that both one’s emotional state as well as the emotional content of text may influence the processes involved in literacy activities.
1.1.1. – Influence of mood and emotional content
3 Pioneering work by Bower, Gilligan and Monteiro (1981) showed that adults identified with a protagonist in the same emotional state as had been induced in them through hypnosis (happy or sad), and recalled more details about this protagonist (see also Laird, Wagener, Halal & Szegda, 1982). In short, textual information congruent with one’s emotional state may be better recalled. In a study examining the effect on textual comprehension of adult readers’ emotional state induced with emotionally laden photographs, Blanc and Tapiero (2002) found a similar congruency effect on the generation and judgment of inferences.
1.1.2. – Influence of mood
4 Induced mood also seems to influence students’ strategy use while reading expository texts; happy-induced readers are more likely to make causal inferences than sad-induced readers, but both happy and sad mood induction seems to foster memory for the text (Bohn-Gettler & Rapp, 2011). Conversely, sad mood as opposed to neutral mood has been found to lead to processing decrements in a textual comprehension tasks (Ellis, Ottoway, Varner, Becker, & Moore, 1997; Ellis, Varner, Becker, & Ottoway, 1995).
1.1.3. – Influence of emotional content
5 Other researches interested in the influence of the emotional content of a text on memory have shown that adult readers remember a dramatic version of a text (journalistic account) better than a neutral version (Legros, 1988). Regarding memory for narratives (constructed for the purpose of the study), Davidson, Luo and Burden (2001) observed that children (mean age 6 to 11 years of age) remember emotional information better (whatever their valence). When emotion was explicitly mentioned, children’s memory for the text as a whole was increased. This suggests that emotion facilitates selection of important ideas in the text by providing retrieval indices (see also Liwag & Stein, 1995). More recently, Vinson, Ponari and Vigliocco (2014) showed emotional words were processed faster than neutral words, independently of arousal. One explanation of this facilitating effect of emotional content on processing may be, as suggested by the authors, that the meaning of words are embedded with emotional experience. Children aged 7 and 9 years, were also found to better recall pleasant (“positive”) emotional words than neutral words when controlling for non-emotional characteristics such as word length, familiarity, imageability (Monnier & Syssau, 2008; Syssau & Monnier, 2012).
6 The emotional valence of a text has also been shown to influence children’s reading comprehension (Clavel, 2007; Clavel & Cuisinier, 2008): eleven years old children were found to have trouble understanding a text conveying pleasant (e.g., joy, pride…) emotions, but to process more deeply and better understand a text conveying unpleasant emotions. This echoes the negativity bias hypothesis, whereby readers’ tend to process negative information more carefully regardless of their own emotional state (Egidi & Gerrig, 2009). Other researches regarding children’s understanding of narratives revealed that children’s (from 5 to 10 years of age) comprehension builds, in part, on emotional information, especially when emotion is related to the character’s behavioral expression of emotion (Blanc, 2010), and is explicitly mentioned (Creissen & Blanc, 2014). In a study examining the effect of an instruction to write an autobiographical text with a positive, negative or neutral emotional content on writing processes, Fartoukh, Chanquoy and Piolat (2012) found the number of grammatical spelling errors varied as a function of children’s working memory capacity when children were instructed to write about a neutral memory. This effect of working memory capacity was not observed when children were instructed to write about a happy or sad memory. Emotional vocabulary matching instructions was also found to increase writing fluidity when children wrote about a happy autographical memory (see also Lynton & Salovey, 1997).
7 In summary, effects of emotion in literacy activities (i.e., emotional states and content) are not always consistent, suggesting that emotion effects on cognitive processes may vary as a function of the task (e.g. memory for textual information, text comprehension, orthographic and vocabulary production) and emotion valence (pleasant-unpleasant). Nonetheless, these findings contribute to demonstrate the entanglement of cognitive and emotional processes in both adults and children. The theoretical models proposed to account for this entanglement (for a review see for example Blanchette & Richards, 2010; Corson, 2002; Pons, de Rosnay, & Cuisinier, 2010) posit either a facilitating influence of positive mood linked to a greater activation of concept in memory and increased flexible and creative thinking (Semantic Network Theories; Bower, 1981; Isen, 2008) as well as a greater influence of mood (either positive or negative) on complex processes (Affect Infusion Model; Forgas, 1995), or a detrimental influence of mood on performance (Resource Allocation Model; Ellis & Ashbrook, 1988; Ellis & Moore, 1999). According to the Resource Allocation Model (Ellis & Ashbrook, 1988) emotion place a burden on attention and working memory resources by producing task-irrelevant thoughts, which in turn create interference and lead to performance decrements in complex problem solving tasks such as textual comprehension (Ellis et al., 1995; 1997). In other words, the cognitive resources allocated toward processing emotion may no longer be available to successfully engage in complex problem solving (Ellis, Seibert, & Varner, 1995). Moreover, it seems both positive and negative mood can reduce cognitive resources and produce irrelevant thinking leading to poorer performance in problem solving (Oaskford, Morris, Grainger, & Williams, 1996). In the last version of this model, Ellis & Moore (1999) consider the interfering and detrimental influence of emotion as independent of valence.
1.2. – Learning to spell
8 Learning to spell, according to Fayol (2010), implies learning of three complementary skills: 1. the alphabetic principle and phonological spelling; 2. frequent letter associations also called graph-tactic regularities (general orthographic knowledge; Jaffré & Fayol, 2005) and the lexical form of words that constitute the orthographic lexicon (specific orthographic knowledge); 3. morphology: inflection which concerns number and genre agreement of nouns, adjectives and verbs as well as form and tense of verbs and derivation which is a mean to form words from words already known within the same morphological family (for example in French inferring the noun plant ends with a silent “t” because the associated verb is planter).
9 The French language, as other alphabetical languages, offers correspondences between graphemes and phonemes (Fayol & Morais, 2004) one needs to learn to be able to read and write. However several specificities of the French language (like other “opaque” languages) such as the higher number of phonemes than graphemes (36 and 130 respectively), as well as mute letters, final letters, morphological marks of plural and genre which have no phonological correspondence make inferring the correct written form of a words from its oral form difficult (Lété, Peereman, & Fayol, 2008). Therefore, learning the alphabetical principle and phonological spelling is not enough to be able to spell correctly. When writing one may also directly retrieve the lexical form of words stored in memory or infer spelling of a new word through analogy with similar known words or one’s knowledge of regularities. From their first year of primary school children have already memorized several lexical forms (Martinet, Valdois, & Fayol, 2004) and learned regular and frequent letter associations (Pacton, Perruchet, Fayol, & Cleeremans, 2001; Pacton, Fayol, & Perruchet, 2005). Other spelling strategies consist in generalizing knowledge about the lexical form of words to write new words, for example by adding a prefix or suffix (derivational morphology) or to use knowledge about grammar and conjugation such as marks of number and genre agreements and tense (inflectional morphology). Research has shown that children soon use derivational morphology to spell words (Sénéchal, 2000; Sénéchal, Basque, & Leclaire, 2006; Sanguin-Bruckert & Bruckert, 2004) by relying on graph-tactic regularities rather than on the application of rules (Pacton et al., 2005). Concerning inflectional morphology, results from developmental research suggest it is acquired progressively throughout primary school, first for nouns, then adjectives and finally verbs (Fayol, Totereau, & Barouillet, 2006). Learning of inflectional morphology is assumed to start with explicit teaching of rules (i.e. declarative knowledge) whose use becomes more and more automatized with experience and thus requires less and less attention (Anderson, 1983). Explicit teaching of agreement rules has indeed been shown to increase first and third graders’ performance within a three month period although stability of learning was not assessed (Thévenin, Totereau, Fayol, & Jarousse, 1999). The use of agreement rules may be overridden by the direct retrieval of inflected forms in memory (Fayol, Largy, & Lemaire, 1994; Cousin, Largy, & Fayol, 2002), a faster and less cognitively demanding strategy although sometimes leading to spelling mistakes (Largy, Fayol, & Lemaire, 1996; Largy, Cousin, Bryant, & Fayol, 2007). In summary, learning to spell is a long and difficult process which is sensible to cognitive overload in adults (Fayol et al., 1994; Largy et al., 1996) and even more so in children (Chanquoy & Negro, 1996; Fayol, Hupet, & Largy, 1999; Negro & Chanquoy, 2000) if we consider that the burden placed on cognitive resources varies as a function of the automation of the various strategies used to spell. In that respect, we may hypothesize that morphological strategies are automatized later than lexical strategies in that they imply learning of rules. Moreover, as stressed by Cuisinier et al. (2010) dictation in itself is a very complex and cognitively demanding activity in that it requires a variety of processes—perceptual, phonological, orthographic and graph-motor—which greatly solicit both working and long term memory. Results from Cuisinier et al. (2010) and Fartoukh et al. (2014) pointed the emotional content of a text used in a dictation task impairs children’s performance, maybe because of its impact on cognitive resources (attention, inhibition) although the underlying mechanisms of this detrimental effect still need to be specified.
2. – The current study
10 The aim of the present article was to clarify the influence of the emotional content of a text in a dictation task. The replication of Cuisinier et al.’s (2010) findings by Fartoukh et al.’s (2014) pointed toward a detrimental effect of emotional valence on spelling performance, however there is still debate over the explanation of this effect: is it due to a direct influence of the emotional content (direct effect hypothesis), or to an indirect influence of the emotional content through its effect on children emotional experience (indirect effect hypothesis).
11 In order to be able to contrast these hypotheses, we used part of the method used in Cuisinier et al. (2010) and Fartoukh et al. (2014) and introduced another experimental manipulation. We compared children’s performance when the emotional content of the text was neutral (control group), joyful, and when children’s experience of pleasant emotions was modified with an induction technique independent of textual emotional content (vivid recall of a joyful experience in school) and the content of the text was neutral. Recall of an emotion-laden autobiographical memory has been shown to be one of the most efficient ways to induce an emotion (Westermann, Spies, Stahl, & Hesse, 1996) and is one of the most used induction method with children (Brenner, 2000). This kind of induction is also both ethically acceptable and ecologically valid in that it may trigger representations related to children’s academic-self and could easily be implemented in the classroom. In order to distinguish the influence of children emotion experience and that of the emotional content on performance in the dictation task, it was necessary to induce emotion through other means than the emotional content of the text and to read and dictate a text with a neutral valence [1]. Indeed, an effect of induced pleasant emotion on children’s spelling performance when the text is neutral would support the indirect effect hypothesis (i.e. the emotional content influences children’s performance through induction of an emotional state). In contrast, if the joyful emotional content of the text influences children’s performance and induced pleasant emotion does not, it would provide evidence for the direct effect hypothesis (i.e. children’s performance is influenced directly by the emotional content). If both the joyful emotional content and induced pleasant emotion affect children’s performance, than we may infer that these two hypotheses are not mutually exclusive, especially if the joyful text does not induce an emotion congruent with its emotional valence. Moreover, as in Fartoukh et al. (2014) we measured emotion three times, a first time before reading of the text by the experimenter; a second time right after reading of the text by the experimenter in the joyful text and control conditions, and in the induction condition, right after reading of the text by the experimenter and induction; and a third time right after the dictation task in the three conditions. This temporality in the measurement of emotion experience allowed us to assess a potential effect of the text on emotional experience as well as the effectiveness of induction. This procedure also enabled us to test the effect of the dictation task on children’s emotion experience and to gain insight into the temporal continuity of induction. In order to investigate children’s processing of textual information and especially of the joyful emotional content we also used a free-recall task: children were asked to write down all the words that came to their mind when thinking about the text read before dictation after a break following completion of the dictation task in the three experimental conditions. Thus, this investigation aimed at gaining insight into the possible explanations of a detrimental effect of emotional content and/or emotional state on performance. We might expect, based on previous findings (e.g., Davidson et al., 2006; Monnier & Syssau, 2008; Syssau & Monnier, 2012) children would recall more information from the joyful text, because of its emotional content, than from the neutral text.
2.1. – Hypotheses
2.1.1. – Indirect effect hypothesis: Influence of induced pleasant emotions
12 In order to test the indirect effect hypothesis, we examined the effect of pleasant emotion induced through recall of a joyful experience on children’s spelling performance.
13 Hypothesis 1: Based on the Resource Allocation Model (Ellis & Ashbrook, 1988; Ellis & Moore, 1999) and on results from Fartoukh et al. (2014) we expected that induced pleasant emotions would interfere with the processes at stake in the dictation task and impair children’s performance. Also, as the model stipulates greater performance decrements of complex processes, we expected it would have a greater detrimental effect on morph-syntactic than on lexical processes and even more so in children with lower spelling ability level.
2.1.2. – Direct effect hypothesis: Influence of the joyful emotional content
14 Hypothesis 2: Based on Cuisinier et al.’s (2010) findings we expected children to make more mistakes in the joyful text condition than in the neutral condition. Indeed, if attention is engaged towards the emotional content, less cognitive resources will be available for orthographic production. This hypothesis echoes that of a detrimental effect of induced emotion predicted by the Resource Allocation Model (Ellis & Ashbrook, 1988; Ellis & Moore, 1999). However, they differ in that we expect the emotional content to have a direct effect on performance, meaning without it modifying children’s emotional state. We assumed, as in hypothesis 1, performance decrements would vary as a function of the complexity of the processes involved and be moderated by ability level.
3. – Method
3.1. – Participants
15 The sample consisted of 134 fifth grade pupils from five Parisian schools, comparable in terms of socioeconomic background of the families (lower to middle-class). All children were allowed to participate by their parents and took part to the study on a voluntary basis. None of the children had identified personality, linguistic or behavioral problems. Five children with very low spelling ability scores (below -2 standard deviations on the spelling ability test) were excluded from the sample. None of the children scored above +2 standard deviations. The final sample comprised 129 fifth grade pupils (64 girls and 65 boys; Mage = 11 years 1 month; SD = 4 months).
3.2. – Tasks and Measures
3.2.1. – Dictation task
16 The joyful and neutral texts from Cuisinier et al. (2010) were used. The joyful text comes from Sempé’s Les vacances du Petit Nicolas (see appendix B.1), and the neutral text from La joie des yeux (see appendix B.2). Both of these books were recommended reading from the French Ministry of Education for children in fifth grade. One text was read to children in each experimental group and then a part of the text was dictated. Cuisinier et al. (2010) have constructed the dictated parts to ensure their equivalence in terms of length as well as lexical and syntactic complexity and content. The dictated parts were almost identical in both texts. Spelling errors were marked on twenty-four words, common to both texts. These words presented either a lexical or syntactic difficulty. The lexical difficulties were 1. double consonant (3 words: vallon, maisonnette, traditionnelle); 2. grapheme/phoneme correspondence (4 words: chaume, logis, âtre, barreau); 3. silent letter (5 words: aboiement, envahie, bergerie, vingtaine, fond). The syntactic difficulties encompassed 1. five number agreement marks: nominal (moutons), adjectival (profonds), verbal (découvrirent, pressait), from past participle (bordé) and one genre agreement mark (éclairée); 2. a distinction between infinitive tense and past participle (entrer); 3. five choices between grammatical homophones (à, ses, ce, c’était, où).
17 As in Cuisinier et al. (2010), only the difficulties shown in italic and related errors were considered for scoring. Each error was granted 1 point. The maximum possible error-score was 12 for each error type (lexical and morph-syntactic).
3.2.2. – Spelling ability
18 We used the Spelling Collective Screening (Repérage Orthographe Collectif, 2006) to evaluate children’s spelling ability. The test consists of three measures: 1. a measure of spelling judgement designed to tap children’s ability to identify misspelt words within a text and to correct them (14 misspelt words: 1 point granted when correctly identified, 1 point granted when corrected properly, 1 point subtracted when a word written properly was identified as misspelt; maximum possible score was 28); 2. a dictation task measuring both lexical and syntactic spelling ability (maximum score for each category was 10). A total score was calculated as follows: Spelling judgment /3 + Lexical spelling + Syntactic spelling. The maximum possible score was 29.33. This measure enabled us to create three spelling ability groups: low (mean -0.5 standard deviation), medium (mean ±0.5 standard deviation), high (mean + 0.5 standard deviation).
3.2.3. – Emotion Questionnaire
19 Children’s emotional experience was measured with the same questionnaire as that used by Cuisinier et al. (2010) and Fartoukh et al. (2014). Children rated on a 5 point Lickert scale going from 1 not at all to 5 a lot the intensity to which they felt happy, joyful, proud, content, worried, angry, sad, ill at ease, bored.
3.2.4. – Qualitative evaluation of the text
20 As in Cuisinier at al. (2010) children were asked to indicate on a 5 point Lickert scale going from 1 not at all to 5 a lot if they found the text pleasant, surprising, interesting, boring, joyful, sad.
3.2.5. – Text recall task
21 In order to gain further insight into children’s processing of textual information we asked them to write down words that came to their mind when thinking of the text read before the dictation task (the text included the dictated part). The instruction was formulated as follows: “Before the dictation task, I read you a text. Now I would like to know what you remember from this text. You do not have to write down sentences. You can just write all the words that come to your mind when you think about the text”. The time children could spend on the task was not restricted and they were provided half an A4 page with lines to respond. Based on our analysis of children’s answers we elaborated three categories: recall of dictated text information (e.g., un vallon près d’un bois; une étable avec des moutons; un chien qui aboie; un monsieur), recall of non-dictated text information (e.g., joyful text: les vaches; les chefs; les enfants jouent dans la forêt; ils mangent des tartines; les enfants courent après les poules; neutral text: la montagne; une ferme dans la montagne; une famille en vacances), recall of associated representations, which included inferences children generated about the texts (e.g., joyful text: les enfants sont en colonie de vacances; les enfants sont contents, joyeux; neutral text: la nature; un vieux chalet). The number of items recalled from each category was divided by the total number of items recalled. Three scores (quotients) were thus calculated for each child. Scores ranged from 0 to 1.
Experimental procedure (session 2)
Experimental procedure (session 2)
3.3. – Procedure
22 The study took place in children’s schools in their usual classrooms during spring. Children of a same classroom were randomly assigned to the three experimental groups (2 classrooms per group): in the control group the neutral text was read and part of it dictated; in the induction group the neutral text was read and part of it dictated but before dictation children were asked to recall a joyful experience in school (joy induction – 2 minutes); in the joyful text group, the joyful text was read and part of it dictated. The texts were read aloud by a trained experimenter (first author of this article) who made sure to use consistent pace and prosody in each reading the joyful and neutral text. In the first session, children filled in a questionnaire on their perception of school activities. The aim of this questionnaire was to familiarize them with the functioning of self-report scales. In the second session, children started by filling in the emotion questionnaire (baseline, T1; about 2 minutes) before they were read the text in its entirety (neutral or joyful depending on the group; 4 minutes). They answered again the emotion questionnaire right after the text was read to them (post-reading, T2). In the induction group, children were asked to vividly recall a joyful experience in school before filling out the emotion questionnaire for the second time (T2). Children were not asked to report in writing the experience they recalled. Children were then informed about the dictation task and part of the text was dictated (10 minutes). After dictation, children were invited to answer questions on their qualitative evaluation of the text (about 2 minutes), and filled in the emotion questionnaire for the third time (post-dictation, T3). All the questionnaires and writing sheet for the dictation task were collected in a single booklet. Children were instructed not to turn the pages of the booklet unless instructed to do so. A sheet with the instruction “wait for the signal to turn the page” separated the emotion questionnaires of T1 and T2 and the writing sheet for the dictation task. The experimental procedure lasted for approximately 25–30 minutes in each experimental group. The duration was a bit longer in the induction group because of the induction phase (about 2 minutes). After the scheduled break (20 minutes) children did the text recall task, which lasted approximately 5 minutes (average time observed – the task was not limited in time). In the third session, children completed the spelling ability test (5 minutes for the spelling judgment task and 8 minutes for the dictation task). At the end of the last session the experimenter explained the goal of the study to the children, and answered their questions.
4. – Results
23 Before testing the direct and indirect effect hypothesis and text recall according to experimental group, we conducted a series of preliminary analysis. We checked the three groups were equivalent on spelling ability, as well as variation in emotion experience as a function of group and time of measurement to verify the efficiency of induction, the effect of the emotional content of the joyful text on children’s emotion experience, the stability of induction and variations in emotion experience after completion of the dictation task. We also verified whether emotion experience varied according to children’s ability level and children qualitative evaluation of the text.
4.1. – Preliminary Analysis
4.1.1. – Spelling ability across groups
24 Because spelling ability is likely to influence performance we made sure the three groups were comparable on spelling ability with an ANOVA (3 groups). Results showed spelling ability was equivalent across the three experimental groups F(2,126) = .97, p =.38 ns (see for descriptive statistics). Spelling ability score was categorized into three ability groups at the sample level: low (mean -0.5 standard deviation), medium (mean ± 0.5 standard deviation), high (mean + 0.5 standard deviation). There were no gender differences on spelling ability F(1,127) = .32, p =.57 ns.
Descriptive statistics for spelling ability as a function of group
Descriptive statistics for spelling ability as a function of group
4.1.2 – Effects of group and time of measurement on emotional experience
25 Children emotional state was measured three times, before (T1) and after they were read the text and/or induction (T2) and after the dictation task (T3). In order to verify the efficiency of induction, the effect of emotional content on emotion experience, the stability of induction as well as the effect of completion of the dictation task on emotion experience we conducted a repeated measure MANOVA (3 Groups x 3 Time) on the 9 emotions assessed. Multivariate analysis revealed a significant effect of time Λ = .49, F(18,87) = 5, p = .000, ηp² = .51 and a significant interaction effect between time and group Λ = .48, F(36,174) = 2.18, p = .000, ηp²= .31 but no significant effect of group Λ = .79, F(18,192) = 1.32, p = .18 ns. Univariate tests revealed a significant effect of time only on worry and an interaction between time and group on happiness, joy, pride, boredom and contentment (see for descriptive statistics). To specify these effects we compared change in emotion experience between T1 and T2 as well as between T2 and T3 within each experimental group using paired sample t-test with Bonferroni correction for multiple testing (α = .01 required for significance; t-test results are presented in ). These comparisons are detailed below.
Descriptive statistics for emotional experience as a function of time of measurement and group
Descriptive statistics for emotional experience as a function of time of measurement and group
Results from paired-sample t-tests on emotion items between T1 -T2 and T2 -T3 by group
Results from paired-sample t-tests on emotion items between T1 -T2 and T2 -T3 by group
4.1.2.1. – Effect of induction and of the joyful text on emotion experience
26 The comparisons between T1 and T2 (see ), revealed children experienced less worry at T2 whatever the group, more boredom at T2 in the control group and less boredom at T2 in the induction group. Experiences of happiness, joy and contentment increased at T2 in the induction group. These results suggest induction was efficient at increasing children’s experience of pleasant emotions whereas the joyful and neutral text did not significantly modify children’s emotion experience.
27 To further test the effect of induction on emotion experience we conducted a MANOVA (3 groups) on the 9 emotions experienced at T2. Multivariate tests showed a significant effect of group Λ = .64, F(18,220)= 3.10, p = .000, ηp² = .20. Univariate tests revealed a significant effect of group on happiness F(2,118) = 16.20, p = .000, ηp² = .22, joy F(2,118) = 10.95, p =.000, ηp² = .16, contentment F(2,118) = 11.71, p = .000, ηp² = .17 and boredom F(2,118) = 4.96, p = .01, ηp² = .08. Contrasts with Bonferroni correction (α = .01 required for significance) revealed no differences in emotion experience between children in the control and joyful text groups, whereas children in the induction group experienced more happiness, joy, and contentment than children in the control and joyful text groups (contrasts induction vs. control and joyful text together for happiness t(125) = -5.57, p = .000; joy t(125) = 2.61, p = .01; contentment t(123) = -4.69, p = .000). Children in the induction group also experienced less boredom than children in the control and joyful text group t(125) =3.29, p = .001. To sum up, these results confirm joy induction succeeded in increasing children’s experience of pleasant emotions, which allows testing of the direct effect hypothesis of emotional state on performance.
4.1.2.2. – Stability of induction, effect of the dictation task on emotion experience
28 The comparisons between T2 and T3 (see ) revealed children experienced less happiness and joy after the dictation task whatever the group, less pride and contentment in the induction and joyful text groups and more boredom in the neutral and induction groups. To sum up, involvement in the dictation task modified children’s emotion experience regardless of experimental manipulations (type of text, induction).
4.1.3. – Effects of spelling ability level and time of measurement on emotional experience
29 In order to examine whether emotion experience varied as a function of spelling ability level (low, medium, high) and time we conducted a repeated measure MANOVA (3 Ability levels x 3 Time) on the 9 emotions assessed. Multivariate tests indicated a significant effect of time Λ = .51, F(18,86) = 4.53, p = .00, ηp² = .49 but non-significant effect of ability level Λ = .83, F(18,192)= 1.01, p =.45 ns, or any interaction between time and ability level Λ = .71, F(36,172) = .88, p = .67 ns on emotion experience. Based on these results effects of induction and emotional content on performance in the dictation task may be analyzed without considering potential differences in emotional experience according to spelling ability level.
4.1.4. – Effect of group on qualitative evaluation of the text
30 After the dictation task children indicated to what extent they found the text pleasant, boring, surprising, interesting, joyful, or sad. In order to check whether children’s perception of the text varied across groups we conducted a MANOVA (3 groups) on the 5 items. Children’s perception varied as a function of group F(12,236) = 2.45, p = .01, ηp² = .11 on the items joyful and boring only. The joyful text was perceived as more joyful F(2,123) = 6.24, p = .002 ηp² = .09 than the neutral text (with or without induction). The neutral text was perceived as more boring than the joyful text whether children received a joy induction or not F(2,123) = 4.90, p = .01 ηp² = .07. These results indicate children identified the emotional valence of the joyful text. Descriptive statistics are presented in.
Descriptive statistics for qualitative evaluation of text as a function of group
Descriptive statistics for qualitative evaluation of text as a function of group
4.2. – Analysis regarding testing of the direct and indirect effect hypothesis and text recall
4.2.1. – Effect of group and ability level on the number of spelling errors
31 To examine the effect of group and ability level on the number of spelling errors we conducted a MANOVA (3 Groups x 3 Ability level) on the two error types (lexical and morph-syntactic). Multivariate tests revealed a significant effect of group Λ = .66, F(4,230) = 13.14, p = .00, ηp² = .19, of ability level Λ = .34, F(4,230) = 40.29, p = .00, ηp² = .41 but a non-significant interaction between group and ability level Λ = .89, F(8,230)= 1.73, p = .09 ns, ηp² = .06 on the number of spelling errors (lexical and morph-syntactic errors together).
32 For lexical errors univariate tests revealed a significant effect of group F(2,116) = 16.19, p =.000, ηp² = .22 and ability level F(2,116) = 58.11, p = .000, ηp² = .50 and a significant interaction F(4,116) = 2.92, p = .02, ηp² = .09 (see figure 1). Contrasts indicated children in the control and induction groups made the same amount of lexical errors t(123) = -.09 ns, and made fewer lexical errors than children in the joyful text group t(123) = -2.46, p = .02. Children with low ability made more lexical errors than children with medium ability t(122) = 4.92, p = .000, who themselves made more errors than children with high ability t(122) = 4.74, p = .000. Children with low and medium ability made more lexical errors in the joyful text group compared to children in the control and induction groups (contrasts joyful text vs control and induction together for low t(23) = -4.30, p = .000 and medium t(26 [2]) = -3.73, p = .001). The number of lexical errors made by children with high ability did not significantly vary across groups (contrast control vs. induction group t(15) = -.93, p = .37 ns; contrast joyful text group vs control and induction groups together t(27) = -1.24, p =.23 ns).
33 For morph-syntactic errors univariate tests also showed a significant effect of group F(2,116)= 22.91, p = .000, ηp² = .28, and ability level F(2,116) = 86.19, p = .000, ηp² = .60, but a non-significant interaction F(4,116) = .23, p = .92 ns (see figure 2). As for lexical errors, contrasts revealed children in the control and induction groups made the same amount of morph-syntactic errors t(75) = -.40, p = .69 ns and made less errors than children in the joyful text group t(97) = -2.80, p = .01. Children with low ability made more morph-syntactic errors than children with medium ability t(84) = 5.36, p = .000, who themselves made more errors than children with high ability t(78) = 5.74, p = .000. Descriptive statistics for both types of errors are presented in appendix A.
4.2.2. – Effect of group on text recall after the dictation task
34 Recall of the text was split in three categories: recall of dictated text information, recall of non-dictated text information, recall of associated representations. In order to test an effect of group on recall of the text we conducted a MANOVA (3 groups) on the 3 recall dimensions. Results indicated a significant effect of group Λ = .83, F(6,228) = 3.64, p = .002, ηp² = .09. Univariate tests revealed a significant effect of group on recall of dictated text information F(2,116) = 7.21, p = .001, ηp² = .11 and recall of non-dictated text information F(2,116) = 7.17, p = .001, ηp² = .11 but no significant effect on associated representations F(2,116) = 0.46, p = .63 ns. Contrasts showed the percentage of recalled information from the dictated and non-dictated parts of the text did not differ significantly between the control and induction group (t(116)= .84, p = .40 ns; t(65) = -1.21, p = .23 ns, for dictated and non-dictated information respectfully). Children in the joyful text group recalled more non-dictated text information than children in the control and induction group t(116) = 3.69, p = .000. Conversely, children in the control and induction groups recalled more dictated text information than children in the joyful text group t(91) = -3.50, p =.001. Descriptive statistics are presented in.
Descriptive statistics for text recall (three categories) as a function of group.
Descriptive statistics for text recall (three categories) as a function of group.
5. – Discussion
35 The present research aimed at clarifying the influence of the emotional content of a text on 5th graders performance in a dictation task by distinguishing two hypotheses put forward by Cuisinier et al. (2010): that of an indirect effect of the emotional content and that of a direct effect. The indirect effect hypothesis posits that the emotional content influences performance through its effect on the emotional state (i.e. the emotional content performance through induction of an emotional state congruent with the emotional valence of the text); the direct effect hypothesis predicts it is the emotional content itself, which influences performance. In order to contrast these two hypotheses we compared children’s performance in a dictation task in three conditions: 1. joy induction independent of text valence (through recall of a joyful experience in school) with reading by an experimenter and dictation of a neutral text (induction group); 2. reading by an experimenter and dictation of a text with a joyful emotional content (joyful text group); 3. reading by an experimenter and dictation of a neutral text (control group). We formulated hypothesis regarding the influence of the joy induction (hypothesis 1) and of the joyful emotional content (hypothesis 2).
5.1. – Effect of induction on children’s performance
36 Contrary to the predictions we made based on the Resource Allocation Model (Ellis & Ashbrook, 1988; Ellis & Moore, 1999) and results from Fartoukh et al. (2014) induction did not significantly influence children’s performance although it increased children’s experience of pleasant emotions (happiness, joy, contentment). A significant influence would have given support to the indirect effect hypothesis, however it appears induction does not influence children’s performance when the dictated text is neutral. Based on this finding, we may infer the detrimental influence of emotion found in Fartoukh et al. (2014) is probably not due to the induction of an emotional state congruent with text valence.
37 Interestingly, the non-significant influence of induced pleasant emotion on 5th graders performance in a dictation task is congruent with previous findings in math word problem solving (Efklides & Petkaki, 2005). However, a facilitating influence of induced pleasant (positive) emotion/mood was found on other types of tasks such as Duncker problem and verbal fluency (Greene & Noice, 1988), Kohs Block Design (Rader & Hughes, 2005), simple math problems (Bryan & Bryan, 1991). This contrasted evidence suggests induced pleasant emotions only facilitate children’s performance in certain types of tasks. When the task is open and/or complex and greatly mobilizes cognitive resources induced pleasant emotions do not seem to influence (i.e. they neither enhance or impair) children’s performance. Further research is needed to entangle the differential effect of induction as a function of the nature and complexity of the task. The stability of induction also appears as another critical factor to consider (Brenner, 2000).
38 Our analysis of emotion variations indicated that the dictation task modified children’s emotion experience. As in Cuisinier et al. (2010) and Fartoukh et al. (2014) the dictation task modified children’s emotion experience. Pleasant emotions (happiness, joy, pride and contentment) decreased after the dictation task and boredom increased in the induction and control groups. These results also echo variations observed in a math word problem-solving task (Efklides & Petkaki, 2005). These changes in emotion experience suggest that our attempts to modify children’s emotion experience with induction possibly interfere with their on-going appraisals of the problem solving situation (appraisal theories of emotion, Frijda, 1986; Lazarus, 1991; Scherer, 2009). Considering emotion experience results from the appraisal one makes of a situation warrants the need to investigate what shapes children’s emotion experience in problem solving situations and how this experience is regulated. The question of the determinants of children emotion experience after problem solving addressed by Tornare, Czajkowski & Pons (2015) revealed children’s metacognitive experiences of difficulty and success (i.e. feeling of difficulty and of success) are important predictors of their emotions. Feeling of difficulty increases the intensity of children’s experience of hopelessness and decreases that of joy and contentment. Feeling of success increases the intensity of children’s experience of pride and joy and decreases that of shame; it also mediates the influence of self-concept (self-perception of competence and control) on these three emotions (i.e. pride, joy and shame). In order to broaden our understanding of the interplay between emotion and learning future research should continue examining the question of the determinants of children’s emotion through a situated approach and during problem solving. The dynamic and labile nature of emotional experience (Scherer, 2009) makes the selection of appropriate time intervals between multiple measures difficult (Sansone & Thoman, 2005), nevertheless, recently developed moment-to-moment measurements methods (D’Mello & Graesser, 2012) offer an interesting paradigm to investigate more finely the interplay between emotions, self-regulation processes and performance in problem solving task. Moreover, considering emotion is the result of the synchronization of five components (i.e. appraisal, arousal, expression, action tendencies, and feeling) warrants the need to supplement self-report measures with physiological measures of emotion experience (such as skin conductance, heart rate) in order to approach the question of the interaction between emotion and cognition more finely.
5.2. – Effect of the emotional content on children’s performance
39 Unlike induction, the joyful emotional content of the text, as we expected (hypothesis 2) had a detrimental influence on children’s performance. Children in the joyful text group made significantly more lexical and morph-syntactic errors than children in the induction and control groups. Moreover, as in Cuisinier et al. (2010) children ability level was found to moderate the detrimental effect of the joyful emotional content of the text on lexical errors. Children with low and medium ability level made more lexical errors in the joyful text group than in the induction and control groups whereas the number of lexical mistakes made by children’s with higher spelling ability level did not significantly differ across the three conditions. This interaction pattern on lexical errors is slightly different from that in Cuisinier et al. (2010), maybe because spelling ability was not assessed with the same tool. In addition no interaction was found between ability level and condition for morph-syntactic errors. The detrimental effect of the joyful emotional content then appears to be compensated by ability level for lexical processes but not morph-syntactic processes which take longer to be automatized through learning.
40 It is also important to note, the text read to the children did not induce an emotional state congruent with text valence, nor did it significantly modify children’s emotion experience. As in Fartoukh et al. (2014) we measured emotion before and after reading of the text, but the text was only read to the children once; this may explain why the text did not induce an emotion congruent with text valence in our study. Indeed, reading the text twice might allow an induction of emotion congruent with that expressed in the text by making more salient the emotional content to children. However, a closer look at the effect of induction (with reading of the text) in Fartoukh et al. (2014) indicated variation in children’s emotion experience after they were read the joyful text was only significant for happiness with a small effect size. Although children in the present study did not feel pleasant emotions with more intensity after they were read the joyful text, they identified its valence.
41 Moreover, as shown by our investigation of children’s free-recall of the text after the dictation task and a 20 minute break they seem to have processed the joyful text differently: children in the joyful text group remembered more information from the text than children in the other conditions whereas children in the induction and control groups remembered more information from the dictated part of the text. The fact children in the joyful text group recalled more information from the text read to them before dictation than from the dictated part of the text suggests their attention to the story was greater than in the two other groups who were read a neutral text. This finding echoes previous empirical evidence of children’s greater memory for emotional information within a text (Davidson et al., 2001) and pleasant emotional words (Monnier & Syssau, 2008; Syssau & Monnier, 2012), which suggest children primarily process emotional information. It is also worth noting that children perceived the joyful text—which tells the story of a group of children having fun during an expedition in the country side, as less boring than the neutral text. Children may therefore have recalled more information and more associated representations of the joyful text because they especially enjoyed it. In other words, children recall of the joyful text, may be explained by their attention to the emotional content but also by their liking of the text. The detrimental effect of the joyful emotional content on children’s performance and the mobilization of attention it appears to have raised contrasted with the non-significant influence of induction on children’s performance when the content of the text is neutral give support to the direct effect hypothesis. In that respect, this finding helps to clarify the underlying mechanisms of the detrimental effect of the emotional content on children’s performance in a dictation task; this direct influence nonetheless needs to be explained. As suggested by Cuisinier et al. (2010) we may hypothesize processing of the emotional content by involving children attention reduces cognitive resources available to resolve spelling difficulties. In other words, the emotional content appears to create interference that adds to the cognitive load caused by task processing. This interpretation echoes that of the Resource Allocation Model regarding the influence of one’s emotional state on performance, however findings from this study indicate processing of the emotional content may similarly use up the cognitive resources available to successfully complete the task (in dictation but also in textual comprehension Clavel (2007)) without modifying children’s emotional experience. Interestingly, a study on textual comprehension showed orienting children’s attention with an instruction to read to memorize or to understand the text annulled the detrimental effect of the emotional content (Tornare, 2014). Such a finding warrants further investigation of the interplay between processing of the emotional content and of meaningful aspects of the tasks (i.e. necessary for successful completion). Future research should indeed examine more precisely processing of the emotional content in relation to cognitive resources such as working memory and inhibitory control. In an applied perspective congruent results of a detrimental effect of textual emotional content on children’s performance in dictation suggest teachers should take into account this influence of the emotional content when choosing or designing activities.
42 Received September 21, 2015.
43 Revision accepted April 14, 2016.
Appendix A. – Descriptive statistics for lexical and morph-syntactic errors as a function of experimental group and spelling ability level.
Appendix B. – Joyful and neutral texts read before dictation
Appendix B.1. – Joyful text read before dictation (dictated abstract is in italic)
45 Nous sommes sortis du camp, avec M. Rateau devant nous et tous les autres rangés par équipes avec nos chefs. C’était comme un défilé. Ce qui est dommage, c’est que comme c’était tôt, il n’y avait personne pour nous voir, surtout quand on est passé devant les hôtels où les autres gens sont en vacances. Nous avons traversé un champ, et beaucoup ne voulaient pas y aller parce qu’il y avait trois vaches; mais nous on a dit que nous étions des hommes et que c’était que les bébés qui étaient des chiques molles ! Après nous sommes sortis du champ pour enter dans les bois. C’est chouette les bois. Avec des tas d’arbres, on peut jouer à cache-cache, faire peur aux autres, pousser des cris bizarres. Il fallait voir Paulin, il réclamait toujours son papa et sa maman. M. Rateau, le chef des chefs nous a dit qu’on allait faire un truc mais que bon, il ne pouvait rien dire car même les « petits chefs » étaient pas au courant. Il fallait quand même choisir d’envoyer des éclaireurs et les suivre à la queue leu leu. Nous, comme on est petits, on a tous décidé que ce serait les chefs d’équipe qui iraient en premier.
46 Au fond d’un vallon bordé de bois profonds, nos chefs découvrirent une maisonnette à toit de chaume, une vraie chaumière! Alors nous, on a couru comme des fous mais les aboiements d’un chien et ses grognements ont signalé notre présence. Le maitre de ce logis nous attendait. Il nous fit entrer dans une pièce. C’était une cuisine traditionnelle éclairée par un feu brillant dans l’âtre et envahie par l’odeur de l’étable. Pouah! Il y avait même une bergerie où se pressait une vingtaine de moutons avec tout au fond, sur des barreaux, des poules.
47 Mais c’était vraiment chouette, avec les copains on a joué à saute-mouton mais le plus drôle c’est de poursuivre les poules, c’est vraiment bête une poule! Après le monsieur nous a raconté des histoires de quand il était petit et on a mangé des grosses tartines. Même le chien qui s’appelait Tartarin a eu droit à une tartine. C’était cool! Après, on est rentré en chantant tellement on était content.
Appendix B.2. – Neutral text read before dictation (dictated abstract is in italic)
48 C’est dans une ferme montagnarde que la famille Lauris s’était installée pour quelques jours de vacances. Au rez-de-chaussée se trouvait une vaste cuisine garnie de boiseries. Au plafond pendaient des pièces de lard et des jambons. Sous le large manteau de la cheminée régnait une cuisinière à bois, plus commode que l’antique foyer avec ses chenets et sa crémaillère, et qui donne plus de chaleur. Sur un dressoir, orgueil de la fermière, de la vieille vaisselle authentique, aux dessins naïfs, était alignée. Une longue table rectangulaire aux vastes tiroirs servait de garde-manger. Dans sa haute caisse en sapin sculpté, une horloge battait son tic-tac régulier. Derrière se trouvaient les chambres ; dans l’une d’elles, des livres étaient rangés sur une étagère : on aime s’instruire dans ce pays. Et c’est par un vestibule intérieur qu’on accédait à l’étable où voisinaient les vaches et les veaux, les porcs, les chèvres et les poules, principales richesses du paysan. Pendant leur séjour, ils firent de longues promenades. Il fut donné à nos touristes de voir un jour une chose à laquelle il ne s’attendait guère.
49 Dans un vallon bordé de bois profonds, les promeneurs découvrirent une maisonnette à toit de chaume. Les aboiements d’un chien et ses grognements signalèrent leur présence. Le maitre de ce logis les attendait sur le pas de la porte. Il les fit entrer dans une pièce. C’était la cuisine traditionnelle éclairée par le feu brillant dans l’âtre et envahie par l’odeur de l’étable. A côté se trouvait une bergerie où se pressait une vingtaine de moutons et au fond, sur des barreaux, des poules.
References
- Allal I., Cheminal-Lancelot R., Devaux M-F., Divry J., Lequette C., Maitrot C., Milesi J., Pouget G., Savin H., & Zorman M. (2006). Spelling Collective Screening [Repérage Orthographe Collectif; Measurement instrument]. Retrieved from www.cognisciences.com/IMG/Roc_09_2009_code.pdf.
- Anderson J. R. (1983). The architecture of cognition. Cambridge: Harvard University Press.
- Blanc N. (2010). La compréhension de conte entre 5 et 7 ans : Quelles représentations des informations émotionnelles? [Comprehension of tales between 5 and 7 years of age: What representation of emotional information?]. Canadian Journal of Experimental Psychology, 4, 256-265.
- Blanc N., & Tapiero I. (2002). Construire une représentation mentale à partir d’un texte : Le rôle des illustrations et de la connotation des informations [Building a mental representation from text: The role of illustration and valence of information]. Bulletin de psychologie, 461, 525-534.
- Blanchette I., & Richards A. (2010). The influence of affect on higher level cognition: A review of research on interpretation, judgement, decision making and reasoning. Cognition and Emotion, 24, 561-595.
- Bohn-Gettler C. M., & Rapp D. N. (2011). Depending on my mood: Mood-driven influences on text comprehension. Journal of Educational Psychology, 103, 562-577.
- Bower G. H. (1981). Mood and memory. American Psychologist, 36, 129-148.
- Bower G. H., Gilligan S. G., & Monteiro K. P. (1981). Selectivity of learning caused by affective states. Journal of Experimental Psychology: General , 110, 451-474.
- Brenner E. (2000). Mood induction in children: Methodological issues and clinical implications. Review of General Psychology, 4, 264-293.
- Bryan T., & Bryan J. (1991). Positive mood and math performance. Journal of Learning and Disabilities, 24, 490-493.
- Chanquoy L., & Negro I. (1996). Subject-verb agreement errors in written production. Study in French children and adults. Journal of Psycholinguistic Research, 25, 553-570.
- Clavel C. (2007). Construction de sens dans les activités de lecture en contexte scolaire : le rôle de l’expérience émotionnelle et des interactions entre pairs [Building the meaning of text in the school context : the role of emotion experience and peer interactions] (Unpublished doctoral dissertation). Université Paris X, Nanterre.
- Clavel C., & Cuisinier F. (2008). Compréhension de texte en cycle 3 : les compétences scolaires médiatisent-elles l’effet de la tonalité émotionnelle ? [Textual comprehension in the further learning cycle: do academic skills mediatize the effect of the emotional valence?] In Loarer E., Mogenet J-L., Cuisinier F., Gottesdiener H., Mallet P., & Vrignaud P., (Eds.), Perspectives différentielles en psychologie [Differential perspectives in psychology] (pp. 441-445). Rennes: Presse Universitaire de Rennes.
- Corson Y. (2002). Variations émotionnelles et mémoire: principaux modèles explicatifs [Main models of effects of emotional variations on memory]. L’Année psychologique, 102, 109-149.
- Cousin M-P, Largy P., & Fayol M. (2002). Sometimes, early learned instances hinder the implementation of agreement rules. A study in written French. Current Psychology Letters, 8, 51-65.
- Creissen S., & Blanc N. (2015). Quelles représentations des différentes facettes de la dimension émotionnelle d’une histoire entre l’age de 6 et 10 ans ? Apports d’une étude multimedia. [Children’s representation of the emotional dimension of a story: The contribution of a multimedia study]. Psychologie Française, doi :10.1016/j.psfr.2015.07.006.
- Cuisinier F., Sanguin-Bruckert C., Bruckert J. P., & Clavel C. (2010). Les émotions affectent-elles les performances orthographiques en dictée? [Do emotions affect spelling performance in dictation?]. L’Année Psychologique, 110, 3-48.
- D’Mello S. K. & Graesser A. C. (2012). Dynamics of affective states during complex learning. Learning and Instruction, 22, 145-157.
- Davidson D., Luo Z., & Burden M. J. (2001). Children’s recall of emotional behaviours, emotional labels, and non-emotional behaviours: Does emotion enhance memory?Cognition and Emotion, 15, 1-26.
- Efklides A., & Petkaki C. (2005). Effects of mood on student metacognitive experiences. Learning and Instruction, 15, 415-431.
- Egidi G., & Gerrig R. J. (2009). How valence affects language processing: Negativity bias and mood congruence in narrative comprehension. Memory and Cognition, 37, 547-555.
- Ellis H. C., & Ashbrook P. W. (1988). Resource allocation model of the effects of depressed mood states in memory. In Fiedler K. & Forgas J. P. (Eds.), Affect, cognition and social behaviour (pp. 25-43). Toronto: Hogrefe.
- Ellis H. C., & Moore B. A. (1999). Mood and memory. In Dalgleish T., & Power M. J. (Eds.), Handbook of cognition and emotion (pp. 193-210). Chischester: Wiley.
- Ellis H. C., Ottoway L. J., Varner L. J., Becker A. S., & Moore B. A. (1997). Emotion, motivation, and text comprehension: The detection of contradictions in passages. Journal of Experimental Psychology: General, 126, 131-146.
- Ellis H. C., Seibert P. S., & Varner L. J. (1995). Emotion and memory: Effects of mood states on immediate and unexpected delayed recall. Journal of Social Behavior and Personality, 10, 349-362.
- Ellis H. C., Varner L. J, Becker A. S., & Ottoway L. J. (1995). Emotion and prior knowledge in memory and judged comprehension of ambiguous stories. Cognition and Emotion, 9, 363-382.
- Fartoukh M., Chanquoy L., & Piolat A. (2012). Effects of emotion on writing processes in children. Written Communication, 29, 389-409.
- Fartoukh M., Chanquoy L., & Piolat A. (2014). Influence d’une induction émotionnelle sur le ressenti émotionnel et la production orthographique d’enfants de CM1 et de CM2 [Influence of an emotional induction procedure on 4th and 5th graders’ emotional feelings and spelling performance]. L’Année Psychologique, 114, 251-288.
- Fayol M. (2010). L’orthographe du français et son apprentissage [The spelling of French and learning]. In Crahay M., & Dutrévis M. (Eds.), Psychologie des apprentissages scolaires [Psychology of school learning]. Bruxelles: De Boeck Université.
- Fayol M., Hupet M. & Largy P. (1999). The acquisition of subject-verb agreement in written french. From novices to experts errors. Reading and Writing: An Interdisciplinary Journal, 11, 153-174.
- Fayol M., Largy P., & Lemaire P. (1994). Subject- verb agreeement errors in French. Quarterly Journal of Experimental Psychology, 47A, 437-464.
- Fayol M. & Morais J. (2004). La lecture et son apprentissage. In ONL (Ed.), L’évolution de l’enseignement de la lecture en France depuis dix ans [The evolution of the teaching of reading in France for the past ten years]. Paris: Observatoire National de la Lecture
- Fayol M., Totereau C., & Barrouillet P. (2006). Disentangling the impact of semantic and formal factors in the acquisition of number inflections. Noun, adjective and verb agreement in written French. Reading and Writing, 19, 717-736.
- Forgas J. P. (1995). Mood and judgment: The affect infusion model (AIM). Psychological Bulletin, 117, 39-66.
- Frijda N. (1986). The emotions. Cambridge, UK: Cambridge University Press.
- Greene T. R., & Noice H. (1988). Influence of positive affect upon creative thinking and problem solving in children. Psychological Reports, 63, 895-898.
- Isen A. M. (2008). Some ways in which positive affect influences decision making and problem solving. In Lewis M., Haviland-Jones J. M., & Feldman Barrett L. (Eds.), Handbook of Emotions- Third Edition (pp. 548-573). New York, London: Guilford Press.
- Jaffré J-P., & Fayol M. (2005). Orthography and literacy in French. In Joshi R. M. & Aaron P. G. (Eds.), Handbook of orthography and literacy. Mahwah, NJ: L.E.A.
- Laird J. D., Wagener J. J., Halal M., & Szegda M. (1982). Remembering what you feel: Effects of emotion on memory. Journal of Personality and Social Psychology, 42, 646-657.
- Largy P., Cousin M-P., Bryant P. & Fayol M. (2007). When memorized instances compete with rules: The case of number-noun agreement in written French. Journal of Child Language, 34, 425-437.
- Largy P., Fayol M., & Lemaire P. (1996). The homophone effect in written French: The case of verb-noun inflection errors. Language and Cognitive Processes, 11, 217-255.
- Lazarus R. S. (1991). Emotions and adaptation. Oxford, UK: Oxford University Press.
- Legros D. (1988). Etude de l’effet d’un procédé de dramatisation sur la mémorisation d’un récit journalistique [Role of dramatization procedure on the memorization of a story]. L’Année Psychologique, 88, 197-214.
- Lété B., Peereman R., & Fayol M. (2008). Phoneme-to-Grapheme consistency and word frequency effects on spelling among first- to fifth grade French children: A regression-based study. Journal of Memory and Language, 58, 952-977.
- Liwag M. D., & Stein N. L., (1995). Children’s memory for emotional events: The importance of emotional related retrieval cues. Journal of Experimental Child Psychology, 60, 2-31.
- Lynton H., & Salovey P. (1997). The effects of mood on expository writing. Imagination, Cognition and Personality, 17, 95-110.
- Martinet C., Valdois S., & Fayol M. (2004). Lexical knowledge develops from the beginning of literacy acquisition. Cognition, 91, B11-B22.
- Monnier C., & Syssau A. (2008). Semantic contribution to verbal short-term memory: Are pleasant words easier to remember than neutral words in serial recall and serial recognition?Memory and Cognition, 36, 35-42.
- Negro I., & Chanquoy L. (2000). Subject-verb agreement with present and imperfect tenses: A developmental study from 2nd to 7th grade. European Journal of Psychology of Education, 15, 113-133.
- Oaksford M., Morris F., Grainger B., & Williams J. M. G. (1996). Mood, reasoning, and central executive processes. Journal of Experimental Psychology: Learning, Memory and Cognition, 22, 476-492.
- Pacton S., Fayol M., & Perruchet P. (2005). Children’s implicit learning of graphotactic and morphological regularities. Child Development, 76, 324-339.
- Pacton S., Perruchet P., Fayol M., & Cleeremans A. (2001). Implicit learning out of the lab: The case of orthographic regularities. Journal of Experimental Psychology: General, 130, 401-426.
- Pons F., Giménez-Dasi M., Sala N., Molina P., Tornare E., & Andersen B. (2015). Compréhension et régulation des émotions à l’école [The comprehension and regulation of emotion at school]. In Crahay M. & Dutrévis M. (Eds.) Psychologie des apprentissages scolaires. Bruxelles: De Boeck.
- Pons F., de Rosnay M., & Cuisinier F. (2010). Cognition and emotion. In Aukrust V. (Section Editor), Learning and Cognition – Volume 5; Baker E., Peterson P. & McGaw B. (Eds.) International Encyclopedia of Education – 3rd Edition (pp. 218-224). Oxford: Elsevier.
- Rader N., & Hughes E. (2005). The influence of affective state on the performance of a block design task in 6- and 7-year old children. Cognition and Emotion, 19, 143-150.
- Sanguin-Bruckert Ch., & Bruckert J-P. (2004). Le rôle des connaissances morphologiques dans l’apprentissage de l’orthographe au cycle 2 et 3. Lidil, 30, 147-167.
- Sansone C., & Thoman D. B. (2005). Does what we feel affect what we learn? Some answers and new questions. Learning and Instruction, 15, 507-515.
- Scherer K. R. (2009). The dynamic architecture of emotion: Evidence for the component process model. Cognition and Emotion, 23, 1307-1351.
- Sénéchal M. (2000). Morphological effects in children’s spelling of French words. Canadian Journal of Experimental Psychology, 54, 76-85.
- Sénéchal M., Basque M. T., & Leclaire T. (2006). Morphological knowledge as revealed in children’s spelling accuracy and reports of spelling strategies. Journal of Experimental Child Psychology, 95, 231-254.
- Syssau A. & Monnier C. (2012). L’influence de la valence émotionnelle des mots sur la mémoire des enfants. Psychologie Française, 57, 237-250
- Thévenin M. G., Totereau C., Fayol M., & Jarousse J-P. (1999). L’apprentissage/enseignement de la morphologie écrite du nombre en français [Learning/teaching of number morphology in written French]. Revue Française de Pédagogie, 126, 39-52.
- Tornare E. (2014, June). Compréhension en lecture: Effet différencié de la consigne en fonction des compétences en lecture chez l’enfant de CM2 [Reading comprehension: Differential effect of instruction as a function of reading skills in 5th grade children]. Poster presented at the 21èmes Journées Internationales de la Psychologie Différentielle, Paris, France. .
- Tornare E., Czajkowski N. O., & Pons F. (2015). Children’s emotion in math problem solving situations: Contributions of self-concept, metacognitive experiences, and performance. Learning and Instruction, 39, 88-96.
- Vinson D., Ponari M., & Vigliocco G. (2014). How does emotional content affect lexical processing?Cognition and Emotion, 28, 737-746.
- Westermann R., Spies K., Stahl G., & Hesse F. W. (1996). Relative effectiveness and validity of mood induction procedures: A meta-analysis. European Journal of Social Psychology, 26, 557-580.
Notes
-
[1]
The design of our experiment does not include the text with a sad emotional content as it appeared ethically questionable to induce sadness in children through recall of a sad memory, especially with regards to emotion regulation skills still developing in children that age.
-
[2]
Degrees of freedom for the reported t-tests were rounded to the nearest whole number when Levene’s test of equality of variance was not significant.
