Effects of Classroom Activities on Affective State — A Comparison of Third and Fifth Graders

We analysed the influence of classroom activities on children’s affective states. Children perform many different activities in the course of an ordinary school day, some of which may trigger changes in their affective state and thus in the availability of their cognitive resources and their degree of motivation. To observe the effects of two such activities (listening to a text and performing a dictation) on affective state, according to grade, we asked 39 third graders and 40 fifth graders to specify their affective state at several points in the day. Results showed that this state varied from one activity to another, and was also dependent on grade level. Third graders differed from fifth graders in the feelings elicited by the activities. The possible implications of these findings for the field of educational psychology and children’s academic performance are discussed.


Introduction
The likelihood of a close relationship between affective states and cognitive processes is a major point of agreement among many of the researchers who study the genesis and role of emotions (Barrett, 2009;Frijda, 2009;Izard, 2007;Russell, 2003;Scherer, 2009).Learning is one of the most important cognitive activities we perform, and can be influenced by the nature of the individual, the task or the context.Furthermore, emotions and feelings are present in everyday life events and, therefore, in all classroom activities.As our affective state can sometimes influence how we process information, reason, or interact with others (Blanchette & Richards, 2010), it is essential to understand the multitude and variability of the affective states and emotions that occur in the course of classroom activities, as they could have important implications for teaching practice.
For Pekrun (1988), affect is one of the most crucial components of human activities.It has a lasting effect and plays a central role in learning situations.Again according to Pekrun (2006), there are two explanations for the crucial role played by emotion in the educational context: (1) emotions can affect the interest, commitment, success and personal development of students; and (2) emotions make an important contribution to the healthy psychological development and well-being of children.Consequently, it is important to clarify the influence of students' affective state in the context of different school experiences, as well as the determinants of these feelings (Järvenoja & Järvelä, 2005;Wosnitza & Volet, 2005).
1.1 Motivation, Emotion and Classroom Activities Ainley, Corrigan, and Richardson (2005) analysed the emergence of emotion and interest in 13-year-old students during a text reading exercise.First, participants were asked to choose the text they wanted to read from four texts displayed on a computer screen.Each of these texts came in four parts, and before reading each part of the text, students were asked to assess their feelings and whether they wished to continue reading.Results showed a strong link between interest in a portion of text and the desire to continue reading.Conversely, a lack of interest in a portion of text provoked boredom among students, who frequently interrupted their reading.In the same line of research on interest, motivation and classroom activities, Ainley and colleagues (Andrews, Ainley, & Frydenberg, 2004;cited by Ainley, 2006) postulated that a lack of interest could be responsible for a decrease not just in students' motivation, but also in their performance on the educational activities they were asked to undertake.They therefore asked students in Grades 7-10 to imagine either that they were going to do a maths test or else that they had won a fortnight's holiday to the destination of their choice.The students then had to rate the intensity of their interest in each of these scenarios on a scale ranging from 1 (Bored) to 5 (Interested) and express the type and intensity of the affect they experienced while imagining the scenarios by choosing specific emotion icons (angry, anxious, relieved, hopeful, proud, happy, hopeless and shameful) and indicating the degree to which they felt each emotion, on a scale ranging from 1 (Not at all) to 5 (A lot).Results showed that the classroom activity attracted more boredom than interest, and a considerably more negative emotional experience compared with winning a two-week holiday.The affects most frequently expressed in connection with the maths test were the negative ones of anxiety, anger and hopelessness, while joy was the affect most frequently expressed in relation to the holiday.Nonetheless, the students also reported several positive affects for the maths test, such as hopefulness and happiness, indicating a certain ambivalence.Thus, the holiday scenario induced relatively limited feelings, restricted to positive affect, and a great deal of interest, while the maths test scenario attracted a wider range of feelings but far less interest.
The emotions experienced by students can also modify their access to the attentional resources needed to perform classroom activities.When Sarason (1984) studied the impact of anxiety on schoolwork, he showed that this negative emotional state influences students' concentration and attention.More specifically, he claimed that experiencing anxiety in an assessment situation can trigger intrusive, interfering thoughts that diminish attention and the efficiency with which the task is performed.For their part, Alexander, Jetton, and Kulikowich (1995), and Lepper and Henderlong (2000) found that by eliciting positive emotions, which promote interest, they were able to increase the attention displayed by students and their capacity for memorisation.
Other research has shown that negative emotions can increase cognitive load, thus bringing about variations in cognitive resources and academic performances.More precisely, negative emotions such as sadness, hopelessness and sorrow are responsible for mobilising or increasing the demand for cognitive resources (Fartoukh, Chanquoy, & Piolat, 2014a).This resource mobilisation could decrease students' performances during classroom activities that already place high demands on attentional resources (Meinhardt & Pekrun, 2003).For example, in the case of a written composition exercise, Fartoukh, Chanquoy, and Piolat (2012) demonstrated the effects of emotions induced by the composition theme on the length of the texts produced by fourth and fifth graders.Students produced shorter texts when the emotional content was negative than when it was neutral or positive.In addition, when the emotional content was positive or negative, their orthographic (grammatical) performances were less predictable.A similar effect of affective state on performance has been observed in a dictation exercise.The induction of a negative (or positive) affective state in fifth graders caused a higher number of spelling errors, compared with the neutral condition (Cuisinier, Sanguin-Bruckert, Bruckert, & Clavel, 2010;Fartoukh, Chanquoy, & Piolat, 2014b).Thus, the positive or negative emotions that are engendered by the composition theme or have been previously induced can have an impact on students' performances on the traditional classroom exercises of writing and dictation.All these studies therefore suggest that the relationship between emotion, information processing and the attentional resources required for cognitive functioning is a complex one.According to Panksepp, "emotional values interpenetrate with cognitive activities" (2000, p. 253).
These complex interactions are even more difficult to study in schoolchildren who, in addition to improving their performance on classroom activities, also have to learn to identify and harness their emotions.According to Pons, Harris, and de Rosnay ( 2004), children's emotional development, which has been widely studied (for reviews, see Harris, 2000;Manstead, 1994;Pons, Harris, & de Rosnay, 2000;Saarni, Mumme, & Campos, 1998), takes place in stages between the ages of 3 and 11 years.More specifically, between 7 and 11 years, changes in their understanding of the role and the effects of emotions help children to realise that the same event can trigger very different emotions, and that these emotions can be regulated.Most studies have examined how emotions and motivations influence students' cognitive functioning, but it is also necessary to control how different classroom activities can generate different emotional experiences, while looking at how these activities are themselves regulated or modified by these emotional experiences (Sansone & Thoman, 2005).

Aims and Hypotheses
The present study explored changes in the affective state of students in Grades 3 and 5 brought about by two very frequent classroom activities: listening to a text read out to them by an adult, and a dictation.Teachers regularly read texts to children, in order to give them examples to follow, improve their understanding of the texts, or hone their literary sensibilities.This reading is also often used to prepare students for a dictation featuring content taken from the text.The dictation exercise itself is used to improve children's spelling skills, encourage them to think about grammar and spelling, and assess their performance.
We set out to test the following hypotheses: (1) Older students (e.g., fifth graders) express different feelings from younger students (e.g., third graders) because of their greater emotional development.We therefore predicted that the students' feelings would vary in terms of intensity and diversity according to their grade level (Pons et al., 2004); (2) Performing a dictation has a negative impact on affective state (Cuisinier et al., 2010;Fartoukh et al., 2014b).Given that a dictation can be perceived of as an assessment exercise, owing to the difficulty of mastering spelling (Fayol, 2008), we predicted that it would generate a more negative affective state than listening to a text being read out (Ainley, 2006;Efklides & Petkaki, 2005); (3) Finally, in an exploratory way, older students (e.g., fifth graders) should react more strongly, in terms of their negative feelings, than younger students (e.g., third graders) to classroom activities that are negatively perceived.We therefore predicted that the fifth graders' feelings during the two activities would be more intense than those of the third graders.

Participants
The participants were 79 children attending a primary school in Southeast France: 39 third graders (mean age: 8; 11 years, range: 8; 5-9; 6 years, 19 girls and 20 boys) and 40 fifth graders (mean age: 11; 1 years, range: 10; 6-12; 3 years, 22 girls and 18 boys).None of them had learning disabilities or other special needs.All the children took part on a voluntary basis, and written informed consent was obtained from all their parents.

Materials
In order to administer activities similar to those that are commonly encountered in the classroom (i.e., listening to a text read out by an adult and performing a dictation featuring content drawn from that text), we chose a text and a dictation that were adapted to primary school children (text was neutral in terms of its emotional content, Cuisinier et al., 2010).The text was about a promenade in the mountain, and a farm nearby.The text reading took about 5 minutes, and the dictation lasted 10 minutes.During the reading, children had to listen quietly, whereas during dictation, children had to listen then to write the dictated text.
To gauge the intensity of the children's affective state, we asked them to rate nine items ("Right now I feel Joyful/Glad/Happy/Proud/Sad/Uncomfortable/Worried/Nervous/I am Bored") on a 5-point scale (see Appendix A) ranging from 1 (Not at all) to 5 (Extremely).This method (Efklides & Petkaki, 2005;Pekrun, 2006) is easy to implement in the classroom.However, it is based on the assumption that participants can accurately verbalise their affective state.It is therefore useful to add a nonverbal scale that is easier for children to understand and which allows the experimenter to obtain comparable and complementary observations (Brenner, 2000;Larsen & Fredrickson, 1999).For this reason, we also asked the children to assess their affect on a smiley face 5-point Likert scale taken from Burkitt and Barnett (2006).This scale features five faces, ranging from 1 = to 5 = (see Appendix A).

Procedure
Data were collected in four classes at the beginning of the school day by the same experimenter.The children were not informed of the succession of tasks, and were simply told that they would have to answer a questionnaire about their affective state several times.After a short presentation of the two self-report measures (9-item intensity scale and smiley face scale), the first measurement of affective state was conducted.Immediately afterwards, the experimenter read the text out to half of the participants, while the other half performed the dictation, in order to counterbalance the order of the activities.A second measurement of affective state was then performed.After that, the groups switched around and either performed the dictation or listened to the text.This was followed by a final measurement of the children's affective state.

Verbal Affective State Intensity Scale
To gauge the effects of activity sequence and grade level on the children's affective state, we performed a 2 (grade level: 3 vs. 5) x 2 (activity sequence: listening then dictation vs. dictation then listening) x 3 (time of measurement: before activities vs. after first activity vs. after second activity) x 9 (Items: Joyful vs. Glad vs. Happy vs. Proud vs. Sad vs. Uncomfortable vs. Worried vs. Nervous vs.I am bored) analysis of variance (ANOVA) with repeated measures on the last two factors.The dependent variable was the rating (1-5) for each of the nine items on the 5-point verbal scale (see Appendix B).
The main effect of activity sequence was not significant, F(1, 75) < 1.The mean intensity ratings were similar regardless of the order in which the children performed the activities (Mlistening then dictation = 2.25 and Mdictation then listening = 2.21).This factor did not interact with the other variables.On the strength of these results, we decided to remove the activity sequence factor from subsequent analyses.
The interaction between grade level and time of measurement was significant for the I am bored item, F(2, 150) = 4.98, MSE = 0.57, p < .01(see Figure 3).Neither listening to the text, F(1, 75) < 1, nor performing the dictation, F(1, 75) < 1, had an effect on the mean intensity ratings provided by the third graders.By contrast, for the fifth graders, listening to the text significantly increased the mean intensity ratings, F(1, 75) = 12.27, MSE = 0.72, p < .01,and performing the dictation had the same effect, F(1, 36) = 14.56,MSE = 0.59, p < .0001. Figure 3. Interaction between grade level and time of measurement for the I am bored item

Nonverbal Scale
To observe the effects of activity and grade level on the children's affective state, we performed a 2 (grade level) x 2 (activity sequence) x 3 (time of measurement) ANOVA with repeated measures on the last factor.The dependent variable was the score (from 1 = to 5 = ) on the nonverbal scale (see Appendix B).
There was a significant effect of grade level, F(1, 75) = 22.43, MSE = 1.45, p < .0001,as mean ratings were higher (i.e., more positive) in third grade than in fifth grade (4.31 vs. 3.56).The effect of time of measurement was also significant, F(2, 150) = 14.28,MSE = 0.34, p < .0001,for although mean ratings did not change significantly after listening to the text, F(1, 75) = 3.87, MSE = 0.38, p = .06(M before activities = 4.17 and M after listening = 3.97), they did change after performing the dictation, F(1, 75) = 12.98, MSE = 0.27, p < .0001(M before activities = 4.17 vs. M after dictation = 3.67).Finally, the interaction between grade level and time of measurement was significant, F(2, 150) = 7.87, MSE = 0.34, p < .0001(see Figure 4).For the third graders, neither listening to the text, F(1, 75) = 1.13,MSE = 0.38, p = .28,nor performing the dictation, F(1, 75) = 1.24,MSE = 0.37, p = .26,had an effect on their mean ratings.By contrast, for the fifth graders, although listening to the text again had no significant effect, F(1, 75) = 2.96, MSE = 0.38, p = .08,performing the dictation had a negative effect on their mean ratings, F(1, 75) = 37.30, MSE = 0.237, p < .0001.-An interaction between grade level and time of measurement for the verbal scale (Happy, Glad, and I am bored items) and for the nonverbal scale.For the Happy and I am bored items, the third graders' mean intensity ratings did not vary significantly during the activities, whereas the fifth graders' did.For the Glad item, the third graders' mean intensity ratings only changed after listening to the text, whereas the fifth graders' ratings only changed after performing the dictation.For the nonverbal scale, the third graders' mean intensity ratings did not vary significantly during both activities while the fifth graders' mean ratings varied negatively when performing the dictation.
-A main effect of grade level on the verbal scale (all the positive items, plus the negative I am bored item) and the nonverbal scale.In the case of the positive items and the nonverbal scale, the third graders' mean intensity ratings were higher than those of the fifth graders, whereas the opposite was observed for the negative item.
-Finally, a main effect of time of measurement on both verbal (all the positive items and on the negative Worried, Nervous and I am bored items) and nonverbal scales.More specifically, performing the dictation decreased the mean intensity ratings of all the positive items, and increased those of the negative Worried, Nervous and I am bored items.Furthermore, after the pupils had listened to the text, their ratings for the Proud and Glad items decreased, whereas their ratings for the I am bored item increased.For the nonverbal scale, children's mean ratings only changed significantly after performing the dictation.

Discussion
The role and impact of emotions in the classroom are now central concerns for many researchers in developmental and educational psychology.Although it is generally acknowledged that context (worse or better for learning), task difficulty, engagement or involvement and peer interactions that typically occur in the classroom influence the affective state of learners (Ekflides & Petkaki, 2005;Sansone & Thoman, 2005), little information is yet available as to how this state varies across classroom activities according to grade level.It is therefore essential to attempt to determine the effects that common classroom activities can have on children's affective state.
The present study attempted to analyse changes in affective state in the classroom.To this end, third and fifth graders performed two traditional school curriculum activities: listening to a text read out by an adult and performing a dictation.In parallel, they were asked to rate how they were feeling at three different points (before the two activities, after listening to the text and after the dictation) on a 5-point intensity scale featuring nine items and a nonverbal scale featuring five sad-smiley faces.
Our first hypothesis predicted an effect of grade level on affective state, reflecting the development of emotional understanding in children (Pons et al., 2004).As fifth graders are older than third graders, we expected their expression of affective state to be different.This hypothesis was validated, as results showed a developmental effect where the mean ratings provided by the third graders were higher than those of the fifth graders for the positive items in the intensity scale (Happy, Joyful Proud and Glad) and higher (i.e., more positive) for the nonverbal scale.By contrast, for the negative I am bored item, the mean intensity ratings provided by the fifth graders were higher than those provided by the third graders.In line with the findings reported by Pons et al. (2004), these first results indicate that, between the ages of 7 and 11 years, children achieve a greater awareness of their affective state in a given situation and are able to implement better emotional control.In fifth grade, the last level of primary school in France, pupils are prepared for their forthcoming transition to secondary school.In addition, their classroom activities are geared towards promoting greater autonomy and more critical thinking.Older children may feel less enthusiastic about these two activities, being more accustomed to them, and will inevitably have experienced more disappointments in their academic performance.
Our hypothesis that classroom activities have an effect on children's affective state was also validated.
Compared with the baseline measure, children in both grades expressed changes in their affective state after listening to the text and performing the dictation.More specifically, we observed a significant decrease in the mean ratings of all the positive items, as well as in the nonverbal rating.This decrease sometimes came after listening to the text (Proud and Glad items) and always came after performing the dictation (all positive items and nonverbal rating).For the negative items, the dictation had the effect of increasing the mean ratings of the Worried, Nervous and I am bored items, whereas we observed an increase for the I am bored item after the pupils had listened to the text.Thus, overall, performing the dictation seemed to bring about greater variations in the children's affective state than listening to a text.These results, in line with previous studies (Cuisinier et al., 2010;Efklides & Petkaki, 2005;Fartoukh et al., 2014b), indicate that participating in a common classroom activity (e.g., listening to a text) has a slight effect on pupils' affective state.Moreover, if that activity is performing a dictation -an exercise often used as an assessment tool, it has a particularly negative effect on affective state.This observation is reminiscent of Ainley's (2006) finding that assessment activities can reduce students' positive affect and increase their negative affect.
Our third hypothesis predicted an age-and activity-related change in affective state.Although the interactions were not significant for six of the nine items in the verbal intensity rating scale, they were significant for two positive items (Happy and Glad) and one negative item (I am bored).They were also significant for the nonverbal ratings.These results indicate that there is a change in pupils' perceptions of events in the classroom between Grades 3 and 5 that is reflected in their self-reported affective state.The younger children seemed to be less affectively reactive to the nature of the classroom activities and did not really experience boredom, whereas the older children felt less happy, less glad and more bored during the activities, especially after the dictation.Again, these results could be explained by the fact that fifth graders, because of their age and their development, begin to express their affective state more efficiently.
To our knowledge, this is the first study to demonstrate grade-related interactions between classroom activity and affective state.The affective state of the third and fifth graders varied greatly from one activity to the other when it was assessed in the early morning.The latter point could be regarded as one of the limitations of our study.Future studies could thus assess affective state in relation to school activities performed at different times of the day and/or compare assessments made at different times in the day in order to observe the effect of this factor on student ratings.A second limitation could be the 5-point intensity rating scale used in this study.It would seem that not all the items in this self-report tool are relevant for studying affective state in the classroom, as the mean ratings produced by our participants did not differ according to time of measurement.One possibility would be to use just the items that yielded significant results in this study or else to use other self-report tools such as the Positive and Negative Affect Schedule for Children (PANAS-C; Laurent et al., 1999).In addition, in accordance with the vision of Scherer and Davidson (2003), further research on emotions in the world of education is needed, in order to improve measures of emotional experience to introduce more accurate devices, such as those used in psychophysiology and neuroscience (electro dermal sensor temperature, blood pressure), to probe students' emotions.A third limitation concerns the activities used in this study.More research is needed to find out whether our results can be generalised to other grade levels and, above all, to other activities (arithmetic, reading, writing, tests, etc.).
Despite these limitations, our results confirm the need to take the emotion variable into account in the field of educational psychology.Some activities in the classroom may elicit negative or positive emotional changes, which could hamper any subsequent activities.Indeed, the positive or negative emotions that are engendered by the school activities can have an impact on students' performances on traditional classroom exercises (Cuisinier et al., 2010;Fartoukh et al., 2014b).In parallel, having failed an exercise leads to apprehend it and to have a negative emotional feeling for similar activities.Thus, the study of emotions in the school context can enhance our knowledge not just of the underlying problems, but also of their possible solutions (Schultz et al., 2006).One of them concerns the sequence of activities, which could be designed based on the emotional feelings they provoke.Another one could consist in decreasing the negative emotional feeling generated by some activities (in making them more pleasant, more interesting…).
Finally, the present study highlighted that the affective state of pupils performing certain classroom activities undergoes grade-related changes, emphasising the importance of identifying and understanding students' emotional state during classroom activities.

Appendix B
Mean intensity ratings (standard deviations) on the verbal scale ranging from 1 (Not at all) to 5 (Extremely) according to grade level, time of measurement and item.

Figure 2 .
Figure 2. Interaction between grade level and time of time of measurement for the Glad item

Figure 4 .
Figure 4. Interaction between grade level and time of measurement for the nonverbal scale