View in EDS HTML Full Text PDF Full Text

Temporal Change of Emotions: Identifying Academic Emotion Trajectories and Profiles in Problem-Solving

Saved in:
Bibliographic Details
Title: Temporal Change of Emotions: Identifying Academic Emotion Trajectories and Profiles in Problem-Solving
Language: English
Authors: Zheng, Juan (ORCID 0000-0002-4896-8198), Lajoie, Susanne P., Li, Shan, Wu, Hongbin
Source: Metacognition and Learning. Aug 2023 18(2):315-345.
Availability: Springer. Available from: Springer Nature. One New York Plaza, Suite 4600, New York, NY 10004. Tel: 800-777-4643; Tel: 212-460-1500; Fax: 212-460-1700; e-mail: customerservice@springernature.com; Web site: https://link.springer.com/
Peer Reviewed: Y
Page Count: 31
Publication Date: 2023
Document Type: Journal Articles
Reports - Research
Education Level: Higher Education
Postsecondary Education
Descriptors: Independent Study, Psychological Patterns, Problem Solving, Learning Processes, Reflection, Performance, Simulated Environment, Medical Students, Clinical Diagnosis, Personality Traits, Student Interests, Time Factors (Learning), Self Control
DOI: 10.1007/s11409-022-09330-x
ISSN: 1556-1623
1556-1631
Abstract: Academic emotions play an important and complex role in self-regulated learning (SRL). However, few studies have examined how academic emotions unfold in different phases of SRL and how the changes in these emotions influence learning performance. The current study examines 98 students' academic emotion trajectories and profiles across the three phases of SRL (i.e., forethought, performance, and self-reflection) as they solve a clinical problem in BioWorld. Specifically, BioWorld is a simulated learning environment where medical students are tasked with diagnosing virtual patient diseases. We identified the three phases of SRL based on students' problem-solving behaviors and we asked students to self-report their achievement and epistemic emotions at the end of each phase of SRL. The growth curve model results showed that curiosity and confusion declined across the three phases of SRL, whereas boredom increased in the self-reflection phase of SRL. The initial levels of curiosity and enjoyment positively predicted students' performance. Latent transition analysis revealed three emotion profiles: curious-positive, confused-negative, and medium-low. Curious-positive students maintained a relatively stable profile through the SRL phases, whereas students in the confused-negative and medium-low groups exhibited specific transition patterns in their emotions. This study makes theoretical contributions by highlighting the temporal and dynamic nature of emotions in problem-solving. Findings from this study have educational implications regarding the role of specific emotions in learning, the development of one's awareness of their emotions, and emotion regulation.
Abstractor: As Provided
Entry Date: 2023
Accession Number: EJ1386273
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwFheTs3HcsUBahJGtDn2VEbAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDBlAPo3Z4rDPbteC6AIBEICBmzFDUF7sR6gTfJtuxsUTL7qyRA4Bsbtb4_q4X9L0AIR1rHLSuZaaIvPebX_jcZXqpA3t2_rqJJfbPbOpSU982ZWpKMF4x_QY91hiPvMCfqIni6_Vddv4spnc-mjNOAdLY037y0u46_bXDZW2Eh2IyzVQvSW7GNlxr-35edZHc9RwbphZ7NsFtD02vlRmw7wNOVTi9XZqwq2e4K7T
Text:
  Availability: 1
  Value: <anid>AN0168593744;[3d0h]01aug.23;2023Aug01.07:02;v2.2.500</anid> <title id="AN0168593744-1">Temporal change of emotions: Identifying academic emotion trajectories and profiles in problem-solving </title> <p>Academic emotions play an important and complex role in self-regulated learning (SRL). However, few studies have examined how academic emotions unfold in different phases of SRL and how the changes in these emotions influence learning performance. The current study examines 98 students' academic emotion trajectories and profiles across the three phases of SRL (i.e., forethought, performance, and self-reflection) as they solve a clinical problem in BioWorld. Specifically, BioWorld is a simulated learning environment where medical students are tasked with diagnosing virtual patient diseases. We identified the three phases of SRL based on students' problem-solving behaviors and we asked students to self-report their achievement and epistemic emotions at the end of each phase of SRL. The growth curve model results showed that curiosity and confusion declined across the three phases of SRL, whereas boredom increased in the self-reflection phase of SRL. The initial levels of curiosity and enjoyment positively predicted students' performance. Latent transition analysis revealed three emotion profiles: curious-positive, confused-negative, and medium–low. Curious-positive students maintained a relatively stable profile through the SRL phases, whereas students in the confused-negative and medium–low groups exhibited specific transition patterns in their emotions. This study makes theoretical contributions by highlighting the temporal and dynamic nature of emotions in problem-solving. Findings from this study have educational implications regarding the role of specific emotions in learning, the development of one's awareness of their emotions, and emotion regulation.</p> <p>Keywords: Academic emotion; Self-regulated learning; Emotion trajectory; Emotion profile; Clinical problem-solving</p> <p>Copyright comment Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.</p> <p>Patient safety depends on medical practitioners' effective clinical problem-solving skills, which determine accurate and efficient diagnoses. Clinical problem-solving is a dynamic and recursive thinking process wherein medical practitioners collect evidence, propose hypotheses, plan or implement treatment, evaluate outcomes, and reflect on the clinical reasoning process (Kuiper, [<reflink idref="bib35" id="ref1">35</reflink>]). In clinical problem-solving, medical practitioners face the challenges of building correct mental representations of clinical problems, striking a balance between efficiency and accuracy, and making decisions under uncertainty (Isen et al., [<reflink idref="bib28" id="ref2">28</reflink>]). These challenges present clear opportunities for knowledge conflict and for success or failure, which have the potential to elicit various and intense emotions.</p> <p>Emotions play an important role in clinical problem-solving. As an illustration, whether positive or negative, mild emotions can impair clinical problem-solving performance (McConnell et al., [<reflink idref="bib45" id="ref3">45</reflink>]). Studies of medical students found that low-performers experienced more emotions than high-performers, especially negative emotions such as anxiety, frustration, and boredom (Artino et al., [<reflink idref="bib6" id="ref4">6</reflink>]; Lajoie et al., [<reflink idref="bib39" id="ref5">39</reflink>]). Understanding the functioning of emotions in clinical problem-solving can help instructors improve their approaches and interventions to better support the training of medical practitioners.</p> <p>In an attempt to understand how emotions affect achievement and performance, researchers have used self-regulated learning (SRL) theory as a framework to examine the role of emotions in clinical problem-solving and other contexts. SRL theory provides a solid framework for understanding the problem-solving process and its underlying cognitive, metacognitive, and affective mechanisms (Panadero, [<reflink idref="bib57" id="ref6">57</reflink>]). For instance, in multiple cross-sectional studies, Pekrun et al. ([<reflink idref="bib59" id="ref7">59</reflink>]) revealed that positive emotions were positively related to SRL strategies and negative emotions were negatively associated with SRL strategies. Mega et al. ([<reflink idref="bib46" id="ref8">46</reflink>]) also found the positive effects of positive emotions on students' SRL strategies in preparation for an exam. Nevertheless, Taub et al. ([<reflink idref="bib74" id="ref9">74</reflink>]) found that surprise negatively predicted students' metacognitive strategies, whereas frustration positively predicted cognitive strategies. In general, the current literature about emotions in SRL emphasizes how certain types of academic emotions interact with self-regulated learning strategies. In addition, researchers have focused on summary statistics of emotion-related variables, with their temporal structure and patterns traditionally being overlooked (Artino et al., [<reflink idref="bib6" id="ref10">6</reflink>]; Mega et al., [<reflink idref="bib46" id="ref11">46</reflink>]; Pekrun et al., [<reflink idref="bib59" id="ref12">59</reflink>]; Taub et al., [<reflink idref="bib74" id="ref13">74</reflink>]).</p> <p>Emotions are likely to fluctuate during SRL processes in response to ever-changing appraisals related to unfolding learning processes (Li et al., [<reflink idref="bib42" id="ref14">42</reflink>]; Taub et al., [<reflink idref="bib74" id="ref15">74</reflink>]). Although SRL and emotion-related models (e.g., Efklides, [<reflink idref="bib17" id="ref16">17</reflink>]; Muis et al., [<reflink idref="bib49" id="ref17">49</reflink>]; Pekrun, [<reflink idref="bib58" id="ref18">58</reflink>]) provide helpful theoretical assumptions about possible emotion changes in SRL, how emotion changes in SRL remain poorly understood from an empirical viewpoint, especially in the context of clinical problem-solving. We are interested in understanding the dynamic attributes of emotions, how emotions change throughout the clinical problem-solving process, the consequences of the dynamic changes in emotions, and the resultant emotion profiles that emerge. Knowing how emotions evolve, interact, and influence learning and engagement is important in understanding how emotions can influence education (D'Mello, [<reflink idref="bib12" id="ref19">12</reflink>]). Moreover, what is currently missing from the vast literature on SRL is the inclusion of diverse academic emotions, such as achievement emotions and epistemic emotions. With these insights in mind, we aim to explore students' academic emotion trajectories and profiles across SRL processes when they solve clinical problems.</p> <hd id="AN0168593744-2">Academic emotions</hd> <p>Although there are different definitions of emotions, there is a consensus that emotions consist of five general components: subjective feelings, motor, appraisal, action, and physiological tendencies (e.g., Plutchik, [<reflink idref="bib64" id="ref20">64</reflink>]). Learners experience these five emotional components across different learning contexts. For example, a student who experiences anxiety during an exam may have unpleasant and nervous feelings, cry as a way to communicate their anxiety (motor), think about the importance of the exam all the time (appraisal), jitter their hands (action tendency), or have an increased heart rate or blood pressure (physiological). In alignment with the five broad components of emotions, educational researchers define emotions as coordinated affective, expressive, motivational, cognitive, and physiological processes (Pekrun, [<reflink idref="bib58" id="ref21">58</reflink>]; Scherer, [<reflink idref="bib69" id="ref22">69</reflink>]). Academic emotions are tied directly to academic settings, including attending classes, studying, and taking exams (Pekrun et al., [<reflink idref="bib59" id="ref23">59</reflink>]). Several vital issues regarding the conceptualization of emotions are discussed below to benefit the understanding of academic emotions.</p> <p>Emotions can be distinguished as traits and states. Trait emotions are so semantic, conceptual, and decontextualized that they reflect a relatively stable way of responding to the world. In contrast, state emotions are episodic, experiential, and contextual and can be influenced by situational cues (Goetz et al., [<reflink idref="bib23" id="ref24">23</reflink>]). Similarly, trait-like academic emotions are typical course-related emotional experiences pertaining to a specific course or an exam, whereas state-like academic emotions are momentary emotional experiences within a single episode of academic life (Ahmed et al., [<reflink idref="bib3" id="ref25">3</reflink>]). The differences between trait and state emotions can trace back to the factors that may affect emotions. Trait emotions are derived from memory and are influenced by students' subjective beliefs and semantic knowledge (Robinson & Clore, [<reflink idref="bib67" id="ref26">67</reflink>]). In contrast, the variance of state emotions is affected by the students' interactions between learning contents and learning environments in a single learning episode.</p> <p>Based on the focus of objects (stimulus of emotions), academic emotions can be grouped into <emph>achievement emotions, epistemic emotions, topic emotions,</emph> and <emph>social emotions</emph> based on the focus of objects (Pekrun & Stephens, [<reflink idref="bib61" id="ref27">61</reflink>]). <emph>Achievement</emph> and <emph>epistemic emotions</emph> are the most commonly occurring emotions in clinical problem-solving, where individuals are required to solve medical problems independently and efficiently. Thus, we only studied achievement and epistemic emotions in the research context of clinical problem-solving.</p> <p>Achievement emotions are the emotions pertaining to achievement activities or outcomes that are typically judged by competency-based standards (Pekrun, [<reflink idref="bib58" id="ref28">58</reflink>]). According to the control-value theory, achievement emotions are induced by a joint effect of control and value appraisals. Subjective control refers to individual evaluations of agency over activities and outcomes (i.e., controllability). Subjective value is the personal evaluation of the degree of importance for oneself and the direction of importance. The combination of control and value appraisals induces a variety of achievement emotions. This study focuses on boredom, pride, anxiety, and enjoyment, which were identified as the more intensive emotions in clinical problem-solving (Jarrell, [<reflink idref="bib29" id="ref29">29</reflink>]). Boredom is the product of value appraisals, whereas pride is assumed to be control-dependent emotion. For example, a student may experience boredom if she considers clinical problem-solving skills unimportant for their medical career. In contrast, pride is assumed to be instigated if success is judged to be caused by controllable internal causes. Similarly, the enjoyment of achievement activities is induced when students experience both the control and value of these activities. For example, a student is expected to enjoy the clinical reasoning task when she feels competent to solve the task and perceives the task as important. This implies that the student may feel anxious if she is unable to solve the highly valued clinical task.</p> <p>In addition to achievement emotions, students may experience epistemic-related emotions triggered by knowledge and knowledge-generating qualities in cognitive tasks and activities (Pekrun et al., [<reflink idref="bib62" id="ref30">62</reflink>]). When personal knowledge conflicts with external knowledge, namely cognitive incongruity, epistemic emotions may be activated by the epistemic nature of the task (Muis et al., [<reflink idref="bib48" id="ref31">48</reflink>], [<reflink idref="bib50" id="ref32">50</reflink>]). A typical situation for the arousal of epistemic emotions is when students process new and discrepant information that can trigger surprise and curiosity (D'Mello & Graesser, [<reflink idref="bib14" id="ref33">14</reflink>]). Confusion will be induced if students cannot integrate the new information and address the cognitive incongruity properly (D'Mello et al., [<reflink idref="bib15" id="ref34">15</reflink>]). The arousal of different epistemic emotions depends on whether students process the incoming information successfully in problem-solving. Confusion, surprise, and curiosity are natural epistemic emotions since knowledge and knowledge generation generally trigger these three emotions (Pekrun et al., [<reflink idref="bib62" id="ref35">62</reflink>]).</p> <p>Along with the theoretical development of academic emotion and the prevalence of computer-supported learning (CSL), researchers have shown considerable interest in the broad array of emotions that occur during CSL. In general, CSL is likely to trigger more viable emotions than just enjoyment and anxiety since CSL can provide learners with different learning techniques and authentic learning experiences (Loderer et al., [<reflink idref="bib43" id="ref36">43</reflink>]). Identifying the occurrences of emotional states in CSL and their relationships to learning outcomes is becoming a fundamental question in the field (D'Mello, [<reflink idref="bib13" id="ref37">13</reflink>]). For example, D'Mello ([<reflink idref="bib13" id="ref38">13</reflink>]) conducted a meta-analysis of 24 studies examining the role of emotions in different CSL environments (e.g., intelligent tutoring systems, game-based learning environments, and simulation environments). They found that the occurrence and effects of emotions on task performance varied substantially across studies, especially the emotions of boredom, confusion, curiosity, and happiness. As claimed by D'Mello ([<reflink idref="bib13" id="ref39">13</reflink>]), the authenticity of learning contexts accounted for greater heterogeneity of emotions than the use of learning technologies. In sum, despite the incidence of multifaceted and variable emotions in CSL, researchers generally agree on the occurrence, antecedents, and outcomes of emotions in similar learning contexts.</p> <p>For this study, we provided students with an authentic learning and problem-solving experience by leveraging a computer-simulated environment designed for medical students to deliberately practice clinical reasoning skills. In clinical problem-solving, researchers found that students experienced seven state-like academic emotions (i.e., boredom, pride, anxiety, enjoyment, confusion, surprise, and curiosity) with a higher frequency (Duffy et al., [<reflink idref="bib16" id="ref40">16</reflink>]). Self-regulated learning theory explains what, when, and how these state-like academic emotions may occur and how they may change in clinical problem-solving.</p> <hd id="AN0168593744-3">The trajectories of academic emotions in self-regulated learning</hd> <p>SRL was initially defined as the degree to which students deliberately use cognitive and metacognitive processes to attain their planned goals (Zimmerman, [<reflink idref="bib81" id="ref41">81</reflink>]). Current researchers conceptualize SRL as a multi-dimensional construct that includes not only cognitive and metacognitive components but also emotional, motivational, and contextual components (Greene & Schunk, [<reflink idref="bib24" id="ref42">24</reflink>]). In a review of the six predominant SRL models, Panadero ([<reflink idref="bib57" id="ref43">57</reflink>]) found that all the models view SRL as a dynamic and cyclical process consisting of several phases and sub-processes to form a complete learning circle. In particular, Zimmerman's ([<reflink idref="bib82" id="ref44">82</reflink>]) social cognitive model describes that SRL consists of three phases: forethought, performance, and self-reflection. In the <emph>forethought</emph> phase, students focus on analyzing and making plans for the task. They execute the task by monitoring and controlling learning strategies in the <emph>performance</emph> phase. Students finally evaluate success or failure attributed to the task in the <emph>self-reflection</emph> phase.</p> <p>In summary, SRL has evolved as a general umbrella term and widely adopted theoretical framework to describe students' learning; however, researchers use various terms to describe SRL components, including terms referring to the emotional aspects of SRL. In summary, SRL has evolved as a general umbrella and widely adopted theoretical framework to describe students' learning using various terms, including motivation component (e.g., self-efficacy, task value, and achievement goal orientation), learning strategies (e.g., rehearsal, elaboration, organization, and critical thinking strategies), metacognition component (e.g., metacognitive self-regulation, control, monitoring), and emotion component (e.g., test anxiety, achievement emotions, academic emotions). In line with the theoretical framing of SRL, we consider SRL a multi-dimensional process whereby students cognitively, metacognitively, emotionally, and motivationally engage in their learning. For this study, we focus on the cognitive and emotional aspects of SRL.</p> <p>Although a growing body of literature exists on the role of academic emotions in SRL, theoretical and empirical evidence on the trajectory of state-like emotions in SRL processes is largely lacking. The existing theoretical models typically emphasize how academic emotions influence SRL strategies. For example, Pekrun's ([<reflink idref="bib58" id="ref45">58</reflink>]) control-value theory postulates that positive achievement emotions such as enjoyment and pride promote the use of elaborative and metacognitive strategies, and negative achievement emotions such as anxiety and boredom undermine the use of all SRL strategies. Control-value theory integrates control and value appraisals of achievement-related activities as the antecedents of achievement emotions. With regard to epistemic emotions, Muis et al. ([<reflink idref="bib49" id="ref46">49</reflink>]) developed a model that identifies information-related appraisals (e.g., novelty and complexity of the task) as the antecedents of epistemic emotions. They also postulated that epistemic emotions could influence how students plan, set goals, and use cognitive and metacognitive strategies. Efklides's ([<reflink idref="bib17" id="ref47">17</reflink>]) model describes how emotions could be triggered in different phases of SRL by considering the state-like attribute of emotions and event-like attributes of SRL processes. In the <emph>forethought</emph> phase of SRL, where students analyze the task and make plans, students either have an effortless and fluent analysis of familiar tasks or undergo cognitive incongruity due to task novelty and complexity. These findings imply that a mismatch between task features and the student's prior knowledge could trigger the occurrence of epistemic emotions (Muis et al., [<reflink idref="bib49" id="ref48">49</reflink>]). Students may experience emotions like surprise, curiosity, and confusion when analyzing the task (Efklides, [<reflink idref="bib17" id="ref49">17</reflink>]). In the <emph>performance</emph> phase of SRL, where students execute the task by employing learning strategies, emotions could be positive or neutral in the case of effortless task processing. In the case of effortful processes, the rate of progress and fluency of the process determines the positive or negative emotions students may experience (Ainley et al., [<reflink idref="bib1" id="ref50">1</reflink>]). In other words, achievement-related activities, including progress rate and process fluency, trigger emotions at this phase. Students experience either positive or negative achievement-related emotions in the <emph>self-reflection</emph> phase, depending on the evaluation of learning outcomes (Zimmerman, [<reflink idref="bib83" id="ref51">83</reflink>]).</p> <p>The models described above lend support to the theoretical explications of the trajectory of academic emotions in SRL. Unstable epistemic emotions and achievement emotions occur throughout the phases of SRL. However, the intensity of epistemic and achievement emotions differs in SRL phases. Epistemic emotions are assumed to play a major role in the <emph>forethought</emph> phase of SRL since new information in task analysis is very likely to instigate students' curiosity, surprise, or even confusion (Efklides, [<reflink idref="bib17" id="ref52">17</reflink>]; Muis et al., [<reflink idref="bib49" id="ref53">49</reflink>]). Thus, it is likely to expect epistemic emotions with a higher intensity than achievement emotions in the <emph>forethought</emph> phase if the task is new and challenging. In the <emph>performance</emph> phase, more achievement emotions start to occur in accordance with student learning progress and fluency, whereas epistemic emotions continue to play by cognitive incongruity (Ainley et al., [<reflink idref="bib1" id="ref54">1</reflink>]; Muis et al., [<reflink idref="bib49" id="ref55">49</reflink>]). Finally, it is tempting to assume that more achievement-related emotions and less epistemic emotions may occur at this phase of SRL following the control-value theory. According to the control-value theory, students' control appraisal of the task and how students identify themselves in the process rate determines the occurrence of achievement emotions (Pekrun, [<reflink idref="bib58" id="ref56">58</reflink>]). This kind of achievement emotion may even be boosted by students' evaluation of their achievement in the <emph>self-reflection</emph> phase.</p> <p>Although academic trajectories are not frequently studied in the literature, the available empirical evidence shows that academic emotions are not static in SRL. Taub et al. ([<reflink idref="bib74" id="ref57">74</reflink>]) investigated students' confusion, frustration, and enjoyment in the process of scientific reasoning. Different phases of science problem-solving were divided by identifying the key actions students performed in the game-based learning environment (i.e., read more content, perform laboratory tests, submit final diagnosis). They found that students were happy reading relevant content and confused when conducting laboratory tests. Students' academic emotions are also not static in exams and courses. Peterson et al. ([<reflink idref="bib63" id="ref58">63</reflink>]) examined how achievement emotions change across an assessment event. They employed repeated measures over three weeks (9 times): the study week before the exam, the exam week, and the week of receiving exam feedback. The descriptive statistics show that students experienced a dynamic trajectory in positive emotions (e.g., enjoyment): a decrease before the exam, an increase after the exam, and a slight decrease after receiving feedback. As for negative emotions (e.g., anxiety), Peterson et al. ([<reflink idref="bib63" id="ref59">63</reflink>]) found that students had a relatively stable level of anxiety before the exam, followed by a significant decrease after the exam and a slight increase after receiving feedback. In contrast, Niculescu et al. ([<reflink idref="bib55" id="ref60">55</reflink>]) found that negative achievement emotions (i.e., anxiety, boredom, and hopelessness) were relatively stable throughout a mathematics and statistic course. Different from Peterson et al.'s ([<reflink idref="bib63" id="ref61">63</reflink>]) study, Niculescu et al. ([<reflink idref="bib55" id="ref62">55</reflink>]) measured students' achievement emotions two times: one time halfway through the course and the second time at the end of the course before the exam. Based on these different findings, we argue that the trajectories of academic emotions can differ depending on the learning context and the measurement of emotions. The academic emotion trajectories that occur in a problem-solving event, an assessment event, or a course could be different. In clinical problem-solving, Lajoie et al. ([<reflink idref="bib38" id="ref63">38</reflink>]) explored the interaction between SRL processes and emotions and found that engagement in the <emph>forethought</emph> phase positively predicted negative deactivating emotions (e.g., boredom). The occurrence of deactivating emotions aligns with Efklides's ([<reflink idref="bib17" id="ref64">17</reflink>]) assumption that some emotions may occur due to effortless analysis of familiar tasks in the <emph>forethought</emph> phase. Engagement in the <emph>performance</emph> phase positively predicted negative activating emotions (e.g., anxiety), whereas engagement in the <emph>self-reflection</emph> phase positively predicted the level of positive activating emotions. Different academic emotions play different roles in the three phases of SRL. These findings lend support to the existence of an academic emotion trajectory in the context of problem-solving. The field needs more research to extensively examine students' academic emotion trajectories in problem-solving and how emotion trajectories affect students' learning outcomes.</p> <hd id="AN0168593744-4">Academic emotions and learning outcomes</hd> <p>There are several theoretical assumptions on the effects of academic emotions on learning outcomes. Traditional emotion theory assumes that positive emotions are maladaptive by making "lazy thinkers" reduce their motivation to pursue challenging goals (Aspinwall, [<reflink idref="bib7" id="ref65">7</reflink>], p. 7). However, in the perspective of control-value theory, positive achievement emotions generally exert positive effects on learning outcomes, whereas the effects of negative emotions can be variable depending on the arousal level of the negative emotions (Pekrun & Perry, [<reflink idref="bib60" id="ref66">60</reflink>]). Control value theory also acknowledges the adverse effects of excessive positive emotions, such as pride. Overall, academic emotions can be either adaptive or maladaptive to learning outcomes.</p> <p>Researchers also found that the effects of academic emotions on learning outcomes were complex. Among epistemic emotions, the effects of confusion on learning outcomes are less stable, as evidenced by three studies conducted on undergraduate students by Vogl et al. ([<reflink idref="bib78" id="ref67">78</reflink>]). Vogl et al. ([<reflink idref="bib78" id="ref68">78</reflink>]) found significant positive relationships between confusion and learning outcomes in two studies but not in the third study. D'Mello et al. ([<reflink idref="bib15" id="ref69">15</reflink>]) also confirmed that confusion benefits learning by promoting the use of appropriate strategies. However, a contradictory finding was found in Muis et al.'s ([<reflink idref="bib48" id="ref70">48</reflink>]) study. They found that elementary students' confusion in a complex mathematical problem-solving task negatively predicted their learning outcomes. As for other epistemic emotions, curiosity positively influences learning outcomes as a matter of direct or indirect effects (Muis et al., [<reflink idref="bib48" id="ref71">48</reflink>]; Munzar et al., [<reflink idref="bib51" id="ref72">51</reflink>]; Vogl et al., [<reflink idref="bib77" id="ref73">77</reflink>]). In terms of direct effects, surprise positively influenced learning outcomes (Vogl et al., [<reflink idref="bib78" id="ref74">78</reflink>]). But the negative indirect effects of surprise on learning outcomes were also found in empirical studies (Muis et al., [<reflink idref="bib48" id="ref75">48</reflink>], [<reflink idref="bib50" id="ref76">50</reflink>]). In contrast, the relationship between achievement emotions and effects was more consistent in empirical findings. Extensive studies have verified the negative effects of test anxiety on learning outcomes irrespective of the population (Ahmed et al., [<reflink idref="bib3" id="ref77">3</reflink>]). Boredom was found to negatively predict learning outcomes (Daniels et al., [<reflink idref="bib11" id="ref78">11</reflink>]; Putwain et al., [<reflink idref="bib65" id="ref79">65</reflink>]).</p> <p>As discussed above, the majority of studies have examined how academic emotions, at a single point in time, affect subsequent or concurrent learning outcomes. There are a few studies focusing on the changes in emotions in relation to learning outcomes. For example, Ahmed et al. ([<reflink idref="bib3" id="ref80">3</reflink>]) examined the emotional changes of Grade 7 students in mathematics throughout a school year. They found that changes in anxiety and boredom negatively predicted the changes in learning outcomes, but changes in enjoyment and pride were positively associated with changes in learning outcomes. The initial level of anxiety, boredom, and pride is negatively associated with learning outcomes, whereas the initial level of enjoyment is positively associated with learning outcomes. However, Peterson et al. ([<reflink idref="bib63" id="ref81">63</reflink>]) did not find a significant relationship between the changes in emotions and learning outcomes in an intensive longitudinal diary study with tertiary university students, as their emotions were measured in a three-week assessment period of a general education course. As this empirical evidence suggests, dynamic academic emotions have various effects on learning outcomes. Therefore, it is essential to take account of both initial levels of emotions and how these emotions change over time in considering their relationships to learning outcomes.</p> <hd id="AN0168593744-5">Emotion profiles</hd> <p>An alternative way to identify the relationship between academic emotions and learning outcomes is to utilize a person-oriented approach by examining how particular combinations of emotion trajectories are associated with learning outcomes (Ganotice et al., [<reflink idref="bib22" id="ref82">22</reflink>]). A person-oriented approach accounts for individual heterogeneity in the levels and types of emotions students experience. Given that a single emotion is likely to operate in junction with other emotions, a person-oriented approach allows researchers to investigate complex emotional experiences by identifying each individual's emotion profile and then grouping individuals with a similar profile (Fernando et al., [<reflink idref="bib19" id="ref83">19</reflink>]; Martinent et al., [<reflink idref="bib44" id="ref84">44</reflink>]). A person-oriented approach is well-suit to investigate multiple emotions without assuming linear relations since some research indicates that the relationship between emotions and learning outcomes may be curvilinear (Keeley et al., [<reflink idref="bib32" id="ref85">32</reflink>]; Robinson et al., [<reflink idref="bib66" id="ref86">66</reflink>]).</p> <p>Generating emotion profiles using a person-oriented approach is not new. Ganotice et al. ([<reflink idref="bib22" id="ref87">22</reflink>]) used cluster analysis to identify students' emotion profiles based on trait-like achievement emotions (i.e., enjoyment, hope, pride, anger, anxiety, shame, hopelessness, and boredom). They found four emotion profiles in a domain-general context and in the context of studying mathematics: high positive and low negative emotions, high positive and moderate negative emotions, low positive and high negative emotions, and low emotions. Students in these profiles demonstrated different learning outcomes regarding math achievement. Robinson et al. ([<reflink idref="bib66" id="ref88">66</reflink>]) also identified four emotion profiles (i.e., positive, deactivated, negative, and moderate-low) from students' four emotion dimensions (i.e., positive activating, positive deactivating, negative activating, and negative deactivating emotions) in the context of a science course. They found that students with positive and deactivated profiles had better learning outcomes. In the context of solving a clinical problem, Jarrell et al., ([<reflink idref="bib30" id="ref89">30</reflink>], [<reflink idref="bib31" id="ref90">31</reflink>]) examined state-like achievement emotions of medical students in two studies. Both studies found three profiles (i.e., positive, negative, and low emotions) and the association between these profiles and concurrent learning outcomes. Emotion profiles are not always static. Martinent et al. ([<reflink idref="bib44" id="ref91">44</reflink>]) followed students' emotional profiles across intensive training (i.e., before, middle, and end of the training). They found four profiles: high positive and low negative, moderate-high positive and low negative, moderate-high positive and negative, and moderated positive and negative. More interestingly, they found that individuals transit membership of emotion profiles in the training process.</p> <p>The studies reviewed above provide initial evidence that a person-oriented approach effectively distinguishes subgroups of students who experience similar emotion patterns and that these profiles are associated with academic outcomes. However, the majority of the prior studies focus on trait-like achievement emotions across a course. Jarrell et al., ([<reflink idref="bib30" id="ref92">30</reflink>], [<reflink idref="bib31" id="ref93">31</reflink>]) examined state-like emotions in the clinical problem-solving process, including an array of achievement emotions. Nevertheless, including epistemic emotions in students' emotion profiles may be important since many knowledge-generation activities could be involved in clinical problem-solving. In addition, the emotion profile transitions found by Martinent et al. ([<reflink idref="bib44" id="ref94">44</reflink>]) also imply the necessity to explore temporal changes in emotion profiles. As can be seen, the academic benefits and detriments of various emotion patterns can be more accurately assessed when both state-like achievement emotions and epistemic emotions are measured in reference to concurrent learning outcomes.</p> <hd id="AN0168593744-6">Current study</hd> <p>The current study aimed to contribute to the growing knowledge of academic emotion trajectories and emotion profiles by addressing the following three research questions: (<reflink idref="bib1" id="ref95">1</reflink>) Do academic emotions (i.e., achievement emotions and epistemic emotions) change over the three phases (i.e., forethought, performance, and self-reflection) of SRL?; (<reflink idref="bib2" id="ref96">2</reflink>) Does each of the academic emotions play a unique role in influencing learning performance?; and (<reflink idref="bib3" id="ref97">3</reflink>) what are students' temporal changes of emotion profiles across the SRL phases? For the first research question, we examined the nature of change in seven academic emotions (i.e., confusion, curiosity, surprise, boredom, pride, anxiety, and enjoyment) across the self-regulated learning phases in clinical problem-solving using unconditional growth curve modeling. Based on the theoretical development in the extant literature, we expected a decline in epistemic emotions and an increase in achievement emotions across the three phases of SRL. Regarding our second research question, we performed a series of conditional growth models with the trajectories of all seven emotions as predictors of students' performance in the clinical reasoning task. We will describe students' task performance in detail in the methods section. Given that mixed results exist in the literature (Ahmed et al., [<reflink idref="bib3" id="ref98">3</reflink>]; Peterson et al., [<reflink idref="bib63" id="ref99">63</reflink>]), and no research has examined the trajectories of epistemic and achievement emotions in clinical reasoning, we cannot propose specific directional hypotheses. We anticipate that both the initial levels of the seven emotions and their growth trajectories affect students' task performance. In line with our first research hypothesis, we further anticipate that the roles the growth trajectories of achievement emotions play in determining students' task performance can be opposite to that of epistemic emotions. To answer our third research question, we applied latent cluster analysis and latent transition analysis to investigate students' temporal changes in emotion profiles in solving clinical problems. We acknowledge that this study is exploratory in nature and there is no direct evidence to support our research hypothesis related to this research question. We expect that students demonstrate specific patterns of emotions and emotional changes in clinical reasoning. The reason is that students' behavioral patterns in problem-solving were well-studied in the literature and emotional changes were found to coincide with behavioral changes (Timmermans et al., [<reflink idref="bib75" id="ref100">75</reflink>]).</p> <hd id="AN0168593744-7">Methods</hd> <p></p> <hd id="AN0168593744-8">Participants</hd> <p>The data in this study was part of a larger project. This study was conducted with the approval of the institutional board of two universities (i.e., the affiliations of the researchers and the participants) to protect the rights of human research subjects. Following the ethics guideline, we allowed participants to skip any part of the research or withdraw at any time. One hundred twenty-nine medical students from a leading university in China consented to participate in this study. However, thirty-one students skipped at least one self-report of their state emotions. For this study, we only used the data of the participants who had fully completed the task for analyses, leaving a sample of 98. Little's ([<reflink idref="bib40" id="ref101">40</reflink>]) test of Missing Completely at Random (MCAR) was not significant, χ<sups>2</sups>(<reflink idref="bib35" id="ref102">35</reflink>, _I_N_i_ = 129) = 32.8, <emph>p</emph> = 0.57, indicating that the missing data was random. A Chi-square test of independence was performed to reveal that there was no significant relationship between student study participation (whether the student skipped at least one time of reporting state emotions) and problem-solving accuracy (whether the student diagnosed the patient correctly) (χ<sups>2</sups> = 1.44, <emph>p</emph> = 0.23). Among the remaining 98 students, 30 of them did not consent to reveal their demographic information. Out of the 98 students, 40 (58.8%) of them were male students. The mean age of the participants was 20.98, with <emph>SD</emph> = 0.72. The participants were all in their third year of medical school and were enrolled in the Pathophysiology course. The course was about the study of abnormal changes in body functions that are the causes, consequences, or concomitants of disease processes (Turner & Gellman, [<reflink idref="bib76" id="ref103">76</reflink>]). Therefore, the participants gained the domain knowledge to diagnose patients in BioWorld.</p> <hd id="AN0168593744-9">The learning environment and learning task</hd> <p>BioWorld is a simulated environment for clinical problem-solving, and it supports analyzing, testing, and diagnosing a virtual patient (Lajoie, [<reflink idref="bib37" id="ref104">37</reflink>]). Bioworld (see Fig. 1) provides students with tools such as evidence panels, medical lab tests, and medical library databases that enable students to control and monitor their thinking and progress when diagnosing a patient. These tools scaffold students to successfully complete the self-regulated learning circle by analyzing the patient, performing the diagnosis, and reflecting on their diagnosis. More specifically, students start the task by reading a case description, which includes the symptoms and medical records that a doctor usually obtains from a patient. At this stage, students can familiarize themselves with the patient by identifying the critical observations, which will be saved in the evidence panel for later argumentation (i.e., collecting observations). For example, a student may collect symptoms such as diarrhea, and weight loss, as important observations from the case description. Afterward, students may have a hypothesis in mind or may need some lab tests to confirm or disconfirm their hypotheses (i.e., ordering lab tests). BioWorld includes the most current medical lab test results that help students evaluate their original hypotheses and make relevant adjustments to the final decision-making. For example, the anti-endomysial antibody was positively found in the patient's blood test, indicating that the patient is more likely to be diagnosed with <emph>Celiac disease</emph>. In addition to lab tests, students can search the library to obtain related knowledge about the case (i.e., searching the library). For example, a student may search with the keyword <emph>Celiac disease</emph> to refresh their memory about this disease. More importantly, students may organize relevant evidence in favor of, or against, a specific hypothesis, during which a final hypothesis will be reached. After submitting a final diagnosis, students can reflect on and evaluate their diagnosis by reorganizing the selected evidence and writing a summary to reproduce their argumentation process and provide patient treatment solutions. Finally, the self-reflection process was extended to the reading of feedback, where students compared their evidence with an expert's evidence list.</p> <p>Graph: Fig. 1Screen capture of BioWorld</p> <p>Participants in this study were asked to diagnose a patient with Diabetes Mellitus Type 1. Diabetes Mellitus Type 1 is most commonly diagnosed in adolescents affecting millions of people worldwide (Menart-Houtermans et al., [<reflink idref="bib47" id="ref105">47</reflink>]). This disease has symptoms including thirst, excessive hunger, fatigue, weight loss, and nausea (Ciechanowski et al., [<reflink idref="bib9" id="ref106">9</reflink>]). Within the optimal model to identify diabetes, the Fasting Blood Glucose Level test and Glycated Hemoglobin (HbA1c) test are the two critical clinical tests to screen for diabetes (American Diabetes Association, [<reflink idref="bib5" id="ref107">5</reflink>]). Urinalysis of Glucose, a test of white blood cells, and Serum Ketones are the additional tests to differentiate between Diabetes Mellitus Type 1 and Type 2 (Laffel, [<reflink idref="bib36" id="ref108">36</reflink>]; Menart-Houtermans et al., [<reflink idref="bib47" id="ref109">47</reflink>]). An experienced physician and medical and educational scholar in Canada created the case based on an authentic clinical patient. A panel of medical experts and learning scientists in China reviewed the case and provided an expert solution. The expert solution is a specific learning trajectory for diagnostic reasoning, which will be used for dynamic assessment of students' clinical reasoning competence (Lajoie, [<reflink idref="bib37" id="ref110">37</reflink>]). In comparison with an expert solution, students' performance was assessed based on the same relevant observations and relevant medical lab tests they collected and linked in problem-solving. The evidence and lab tests were weighted according to their importance to the final diagnosis following the scientific diagnosis of Diabetes (American Diabetes Association, [<reflink idref="bib5" id="ref111">5</reflink>]).</p> <hd id="AN0168593744-10">Measures</hd> <p></p> <hd id="AN0168593744-11">Task performance</hd> <p>As described above, we evaluate students' performance based on their efficacy in solving the case. More specifically, students' efficacy was assessed based on the relevant observations and relevant medical lab tests they collected and linked with a hypothesis during their learning trajectory. In other words, students need to identify the correct evidence to support their final hypothesis. The evidence and lab tests were weighted according to their importance to the final diagnosis (see Table 1). In particular, we used the following formula to calculate the student's percentage of the match in learning trajectory compared with the expert solution.</p> <p> <ephtml> <math display="block" xmlns="http://www.w3.org/1998/Math/MathML"><mrow><mi mathvariant="normal">Performance</mi><mo>=</mo><msubsup><mo>∑</mo><mrow><mi>j</mi><mo>=</mo><mn>1</mn></mrow><mi>n</mi></msubsup><msub><mi>W</mi><mi>j</mi></msub><msub><mi>I</mi><mi>j</mi></msub><mo stretchy="false">/</mo><msubsup><mo>∑</mo><mrow><mi>j</mi><mo>=</mo><mn>1</mn></mrow><mi>n</mi></msubsup><msub><mi>W</mi><mi>j</mi></msub></mrow></math> </ephtml> </p> <p>Graph</p> <p>where <emph>n</emph> = the total number of lab tests and evidence items, <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>W</mi><mi>j</mi></msub></math> </ephtml> refers to the weight of item <emph>j</emph>, <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>I</mi><mi>j</mi></msub></math> </ephtml> is equal to one if an individual linked item <emph>j</emph> to a diagnostic hypothesis and otherwise equals zero.</p> <p>Table 1 The weight of lab tests and evidence</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left" /><th align="left"><p>Item</p></th><th align="left"><p>Note</p></th><th align="left"><p>Weight</p></th></tr></thead><tbody><tr><td align="left" rowspan="5"><p>Lab Test</p><p>(5)</p></td><td align="left"><p>Fasting Blood Glucose Level</p></td><td align="left" rowspan="2"><p>Primary screening tests for diabetes</p></td><td align="left" rowspan="2"><p>3</p></td></tr><tr><td align="left"><p>Glycated Hemoglobin (HbA1c)</p></td></tr><tr><td align="left"><p>Urinalysis/Glucose</p></td><td align="left" rowspan="3"><p>Additional tests to distinguish between type 1 and type 2 diabetes</p></td><td align="left" rowspan="3"><p>2</p></td></tr><tr><td align="left"><p>White blood cells (WBC)</p></td></tr><tr><td align="left"><p>Serum Ketones</p></td></tr><tr><td align="left"><p>Evidence</p><p>(8)</p></td><td align="left"><p>Eight relevant symptoms such as increased thirst, frequent urination, and blurred vision</p></td><td align="left"><p>Symptoms of Type 1 diabetes</p></td><td align="left"><p>1</p></td></tr></tbody></table> </ephtml> </p> <hd id="AN0168593744-12">Achievement emotions</hd> <p>To assess the state achievement emotions students experienced as they solved the task, we used a modified version of the Achievement Emotion Adjective List (AEAL; Raccanello et al., [<reflink idref="bib68" id="ref112">68</reflink>]). The original AEAL was developed in Italian (Brondino et al., [<reflink idref="bib8" id="ref113">8</reflink>]) and was subsequently translated into English using a back-translation procedure (Raccanello et al., [<reflink idref="bib68" id="ref114">68</reflink>]). Considering the modifications made to the English version of the AEAL, the content validity of the modified AEAL scale was examined using a structured sorting task (Agarwal, [<reflink idref="bib2" id="ref115">2</reflink>]). Two researchers with expertise in the literature on achievement emotions examined the modified AEAL scale, and the researchers classified the category of each emotion with 100% accuracy. Specifically, AEAL is a self-report questionnaire of achievement emotions in which each emotion is measured with three single-word adjectives. Items with single-word adjectives are simple and easy to administer, which allow for repetitive measures without annoying participants. The multi-item measures of achievement emotions also maintain psychometrically sound emotion scores (Pekrun et al., [<reflink idref="bib62" id="ref116">62</reflink>]). In this study, we used a modified and English-translated version of AEAL to measure enjoyment, pride, boredom, and anxiety. Some of the items were modified to accurately reflect the definition of achievement emotions in the literature (Pekrun, [<reflink idref="bib58" id="ref117">58</reflink>]) and to align with other English multi-item emotion measures (Jarrell, [<reflink idref="bib29" id="ref118">29</reflink>]). These four achievement emotions were included, as suggested by the low intensity of anger and shame in prior studies (Jarrell et al., [<reflink idref="bib31" id="ref119">31</reflink>]; Naismith & Lajoie, [<reflink idref="bib53" id="ref120">53</reflink>]). All items were measured on a 5-point Likert scale from 1 (strongly disagree) to 5 (strongly agree) to indicate the intensity of each emotion. Cronbach's reliability coefficients were 0.88 for enjoyment, 0.90 for pride, 0.94 for boredom, and 0.82 for anxiety at Time 1. At Time 2, Cronbach's reliability coefficients were 0.92 for enjoyment, 0.93 for pride, 0.87 for boredom, and 0.88 for anxiety. At Time 3, Cronbach's reliability coefficients were 0.94 for enjoyment, 0.95 for pride, 0.84 for boredom, and 0.80 for anxiety.</p> <hd id="AN0168593744-13">Epistemic emotions</hd> <p>To capture students' epistemic emotions experienced as they solved the task, we used Pekrun and his colleagues' ([<reflink idref="bib62" id="ref121">62</reflink>]) Epistemically-Related Emotion Scales (EES). Similar to the achievement emotions measures, each item of EES consists of a single word describing one emotion. The emotions relevant to this study included curiosity, surprise, and confusion. There are three items to measure each emotion, such as curious, inquisitive, and interested are the adjectives used to measure curiosity. Participants were asked to rate along a 5-point Likert scale how strongly they felt about each of the three emotions. Cronbach's reliability coefficients were 0.80 for curiosity, 0.84 for surprise, and 0.71 for confusion at Time 1. At Time 2, Cronbach's reliability coefficients were 0.87 for curiosity, 0.86 for surprise, and 0.81 for confusion. At Time 3, Cronbach's reliability coefficients were 0.89 for curiosity, 0.79 for surprise, and 0.67 for confusion.</p> <hd id="AN0168593744-14">Procedures</hd> <p>In April of the academic year 2019–2020, approximately two weeks before data collection with students, instructors assigned students a self-training task to get themselves familiar with BioWorld. Students practiced on the sample case (i.e., diagnose Celiac disease) following step-by-step instructions. Students could work on the sample case repeatedly, during which they were asked to report their demographic information. Subsequent data collection took place in two consecutive classes (i.e., 2 h) during regular classroom hours. Figure 2 displays the classroom settings for the study and the data collection. The participants independently solved the patient case in the same classroom as they logged in to the BioWorld system with their laptops. Four instructors spent the first 30 min providing technical support and training on BioWorld to ensure that all students had access to the platform and were familiar with the learning environment and data collection procedures. In the process of solving the task, they were directed to fill in the state academic emotion questionnaire three times within the system. To capture participants' state emotions across the three phases of SRL, we asked students to report their emotions at the end of each phase of SRL using the following instruction "Using the scale below, indicate how you currently feel, keeping in mind the activity you are currently working on. For each emotion, please indicate the strength of that feeling by selecting the number that best describes the intensity of your emotion." The state academic emotion questionnaire was triggered at three-time points based on the problem-solving behaviors of students. The first time is when students finish reading the patient description and are about to solve the task by switching to other pages (i.e., at the end of the forethought phase). The second time is when students submit their final diagnosis (i.e., at the end of the performance phase). The third time is when students finish reading their feedback and are about to exit from the task (i.e., at the end of the self-reflection phase). The three actions were chosen based on Zimmerman's ([<reflink idref="bib82" id="ref122">82</reflink>]) three phases of SRL model since BioWorld was designed to support the processes of SRL. During the entire process, the four instructors were available to provide technical support. The task took approximately 40 min for the students to complete.</p> <p>Graph: Fig. 2The classroom setting of data collection</p> <hd id="AN0168593744-15">Analysis</hd> <p>We performed a series of latent growth curve models to analyze the change of academic emotions (i.e., achievement emotions and epistemic emotions) throughout the process of problem-solving. The latent growth curve model identifies different emotion trajectories by priori-defined linear or quadratic growth patterns (Muthén & Muthén, [<reflink idref="bib52" id="ref123">52</reflink>]). The latent growth model estimates the latent means of all academic emotions with an intercept and one slope (Muthén & Muthén, [<reflink idref="bib52" id="ref124">52</reflink>]). The intercept represents the initial intensity of an emotion, whereas the slope represents linear or quadratic growth in relation to the initial intensity. To examine the change of seven academic emotions in the first research question, we compared the model fit index between the intercept-only and linear unconditional latent growth model. The diagnostic performance was then added to the second conditional growth model to examine how the change could influence performance in emotions. Finally, we adopt latent transition analysis to identify emotion profiles of individuals across three time points and explore the stability and change of emotion profiles at the individual level.</p> <hd id="AN0168593744-16">Results</hd> <p></p> <hd id="AN0168593744-17">Unconditional growth models</hd> <p>To address our first research question regarding the nature of change in academic emotions, we first compared the intercept-only latent growth model (LGM) and unconditional LGM to determine which was optimal to represent the average pattern of change in emotions for the sample. In particular, we compared only intercept-only and linear growth models, considering that the quadratic growth model (a growth trajectory curving either up or down) requires at least four times repeated measurements. The model estimation terminated normally for all intercept-only and linear growth models. However, the fit of the intercept-only model was poorer than a linear model for each emotion since the intercept-only models had larger model chi-square and lower CFI values (see Table). Therefore, linear models were selected. Overall, the linear models across all emotion variables showed excellent fit except for the emotions of boredom and pride (see Table 2). For boredom and pride, although the RMSEA (Root Mean Square Error of Approximation) values were larger than 0.08, the CFI values (0.983 and 0.952, respectively) indicated adequate model fit.</p> <p>Table 2 Fit statistics for unconditional latent growth models</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left"><p>Model</p></th><th align="left"><p>χ<sup>2</sup></p></th><th align="left"><p><italic>p</italic></p></th><th align="left"><p><italic>df</italic></p></th><th align="left"><p>RMSEA</p></th><th align="left"><p>CFI</p></th><th align="left"><p>ΔCFI</p></th></tr></thead><tbody><tr><td align="left"><p>Confusion</p></td><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left"><p> Intercept-only</p></td><td align="left"><p>6.65</p></td><td align="left"><p>0.156</p></td><td align="left"><p>5</p></td><td align="left"><p>0.082</p></td><td align="left"><p>0.944</p></td><td align="left"><p>—</p></td></tr><tr><td align="left"><p> Linear</p></td><td align="left"><p>0.01</p></td><td align="left"><p>0.935</p></td><td align="left"><p>8</p></td><td align="left"><p>0.000</p></td><td align="left"><p>1.000</p></td><td align="left"><p>0.056</p></td></tr><tr><td align="left"><p>Curiosity</p></td><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left"><p> Intercept-only</p></td><td align="left"><p>44.66</p></td><td align="left"><p>0.000</p></td><td align="left"><p>5</p></td><td align="left"><p>0.322</p></td><td align="left"><p>0.654</p></td><td align="left"><p>—</p></td></tr><tr><td align="left"><p> Linear</p></td><td align="left"><p>0.02</p></td><td align="left"><p>0.892</p></td><td align="left"><p>8</p></td><td align="left"><p>0.000</p></td><td align="left"><p>1.000</p></td><td align="left"><p>0.346</p></td></tr><tr><td align="left"><p>Surprise</p></td><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left"><p> Intercept-only</p></td><td align="left"><p>2.96</p></td><td align="left"><p>0.564</p></td><td align="left"><p>5</p></td><td align="left"><p>0.000</p></td><td align="left"><p>1.000</p></td><td align="left"><p>—</p></td></tr><tr><td align="left"><p> Linear</p></td><td align="left"><p>0.29</p></td><td align="left"><p>0.588</p></td><td align="left"><p>8</p></td><td align="left"><p>0.000</p></td><td align="left"><p>1.000</p></td><td align="left"><p>0.000</p></td></tr><tr><td align="left"><p>Boredom</p></td><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left"><p> Intercept-only</p></td><td align="left"><p>9.48</p></td><td align="left"><p>0.050</p></td><td align="left"><p>5</p></td><td align="left"><p>0.118</p></td><td align="left"><p>0.957</p></td><td align="left"><p>—</p></td></tr><tr><td align="left"><p> Linear</p></td><td align="left"><p>3.14</p></td><td align="left"><p>0.076</p></td><td align="left"><p>8</p></td><td align="left"><p>0.148</p></td><td align="left"><p>0.983</p></td><td align="left"><p>0.026</p></td></tr><tr><td align="left"><p>Pride</p></td><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left"><p> Intercept-only</p></td><td align="left"><p>21.71</p></td><td align="left"><p>0.000</p></td><td align="left"><p>5</p></td><td align="left"><p>0.213</p></td><td align="left"><p>0.842</p></td><td align="left"><p>—</p></td></tr><tr><td align="left"><p> Linear</p></td><td align="left"><p>6.42</p></td><td align="left"><p>0.011</p></td><td align="left"><p>8</p></td><td align="left"><p>0.235</p></td><td align="left"><p>0.952</p></td><td align="left"><p>0.110</p></td></tr><tr><td align="left"><p>Anxiety</p></td><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left"><p> Intercept-only</p></td><td align="left"><p>5.50</p></td><td align="left"><p>0.240</p></td><td align="left"><p>5</p></td><td align="left"><p>0.062</p></td><td align="left"><p>0.981</p></td><td align="left"><p>—</p></td></tr><tr><td align="left"><p> Linear</p></td><td align="left"><p>0.03</p></td><td align="left"><p>0.865</p></td><td align="left"><p>8</p></td><td align="left"><p>0.000</p></td><td align="left"><p>1.000</p></td><td align="left"><p>0.019</p></td></tr><tr><td align="left"><p>Enjoyment</p></td><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left"><p> Intercept-only</p></td><td align="left"><p>7.08</p></td><td align="left"><p>0.132</p></td><td align="left"><p>5</p></td><td align="left"><p>0.089</p></td><td align="left"><p>0.977</p></td><td align="left"><p>—</p></td></tr><tr><td align="left"><p> Linear</p></td><td align="left"><p>0.60</p></td><td align="left"><p>0.440</p></td><td align="left"><p>8</p></td><td align="left"><p>0.000</p></td><td align="left"><p>1.000</p></td><td align="left"><p>0.023</p></td></tr></tbody></table> </ephtml> </p> <p> <emph>N</emph> = 98. p-values greater than or equal to 0.05 indicate good fit; RMSEA = Root Mean Square Error of Approximation, CFI = Comparative Fit Index, ΔCFI = The CFI difference between an intercept-only model and a linear model</p> <p>The latent growth model within the framework of structural equation modeling (SEM) is a large-sample approach. Therefore, we further examined whether or not the sample size of this study (<emph>N</emph> = 98) was appropriate to yield valid and reliable parameter estimates. We acknowledged that some researchers considered <emph>N</emph> = 100 to <emph>N</emph> = 150 to be the minimum sample size, whereas attention is increasingly given to the ratio of observations (<emph>N</emph>) to the number of free parameters (<emph>q</emph>) being estimated in a SEM related model (Wang & Wang, [<reflink idref="bib79" id="ref125">79</reflink>]). Specifically, there should be at least five observations per free parameter; however, the ratio (<emph>N: q</emph>) should be ten or greater to maintain precision and statistical power for the data with significant kurtosis (Hoogland & Boomsma, [<reflink idref="bib26" id="ref126">26</reflink>]; Kline, [<reflink idref="bib34" id="ref127">34</reflink>]; Wang & Wang, [<reflink idref="bib79" id="ref128">79</reflink>]). In this study, the ratio of the observations to the number of free parameters (<emph>q</emph> = 8) of linear models was 12.25. Therefore, we deemed the sample size reasonable.</p> <p>Parameter estimates for the linear growth models of the seven emotions are presented in Table 3. As shown in Table 3, the intercept variances in all seven models were significant, suggesting that students varied significantly in their initial levels of epistemic and achievement emotions. On average, students reported significantly higher levels of confusion and curiosity than all achievement emotions (i.e., boredom, pride, anxiety, and enjoyment) at the initial stage of problem-solving (see Table 4). In addition, students experienced a significantly lower level of surprise in this phase than confusion, curiosity, and all achievement emotions except enjoyment.</p> <p>Table 3 Unstandardized parameters for unconditional latent growth models</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left"><p>Model</p></th><th align="left"><p>Intercept Mean</p></th><th align="left"><p>Intercept Variance</p></th><th align="left"><p>Slope Mean</p></th><th align="left"><p>Slope Variance</p></th><th align="left"><p>Intercept & Slope Cov</p></th></tr></thead><tbody><tr><td align="left"><p>Confusion</p></td><td align="left"><p>3.13*** (0.08)</p></td><td align="left"><p>0.43** (0.15)</p></td><td align="left"><p>-0.11* (0.05)</p></td><td align="left"><p>0.05 (0.07)</p></td><td align="left"><p>-0.07 (0.08)</p></td></tr><tr><td align="left"><p>Curiosity</p></td><td align="left"><p>3.13*** (0.09)</p></td><td align="left"><p>0.89*** (0.17)</p></td><td align="left"><p>-0.27*** (0.05)</p></td><td align="left"><p>0.22*** (0.06)</p></td><td align="left"><p>-0.24** (0.08)</p></td></tr><tr><td align="left"><p>Surprise</p></td><td align="left"><p>2.45*** (0.09)</p></td><td align="left"><p>0.45** (0.15)</p></td><td align="left"><p>0.06 (0.04)</p></td><td align="left"><p>0.01 (0.07)</p></td><td align="left"><p>-0.04 (0.08)</p></td></tr><tr><td align="left"><p>Boredom</p></td><td align="left"><p>2.74*** (0.10)</p></td><td align="left"><p>0.61*** (0.15)</p></td><td align="left"><p>0.08 (0.05)</p></td><td align="left"><p>0.10 (0.07)</p></td><td align="left"><p>-0.03 (0.07)</p></td></tr><tr><td align="left"><p>Pride</p></td><td align="left"><p>2.75*** (0.10)</p></td><td align="left"><p>0.78*** (0.16)</p></td><td align="left"><p>-0.06 (0.06)</p></td><td align="left"><p>0.26*** (0.07)</p></td><td align="left"><p>-0.20* (0.08)</p></td></tr><tr><td align="left"><p>Anxiety</p></td><td align="left"><p>2.81*** (0.09)</p></td><td align="left"><p>0.56*** (0.15)</p></td><td align="left"><p>0.02 (0.05)</p></td><td align="left"><p>0.15* (0.07)</p></td><td align="left"><p>-0.11 (0.08)</p></td></tr><tr><td align="left"><p>Enjoyment</p></td><td align="left"><p>2.56*** (0.09)</p></td><td align="left"><p>0.76*** (0.15)</p></td><td align="left"><p>-0.03 (0.05)</p></td><td align="left"><p>0.14* (0.05)</p></td><td align="left"><p>-0.16* (0.07)</p></td></tr></tbody></table> </ephtml> </p> <p>Standard errors are displayed in parentheses. * <emph>p</emph> < 0.05; ** <emph>p</emph> < 0.01; *** <emph>p</emph> < 0.001</p> <p>Table 4 Differences in intercept means of the seven emotions</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="2" colspan="3" /><th align="left" rowspan="2"><p><italic>t</italic></p></th><th align="left" rowspan="2"><p><italic>df</italic></p></th><th align="left" rowspan="2"><p><italic>p</italic></p></th><th align="left" rowspan="2"><p>MD</p></th><th align="left" rowspan="2"><p>SED</p></th><th align="left" colspan="2"><p>95% CI for MD</p></th><th align="left" rowspan="2"><p>Cohen's d</p></th></tr><tr><th align="left"><p>Lower</p></th><th align="left"><p>Upper</p></th></tr></thead><tbody><tr><td align="left"><p>Confusion</p></td><td align="left"><p>-</p></td><td align="left"><p>Curiosity</p></td><td align="left"><p>0.028</p></td><td align="left"><p>97</p></td><td align="left"><p>0.978</p></td><td align="left"><p>0.003</p></td><td align="left"><p>0.121</p></td><td align="left"><p>-0.237</p></td><td align="left"><p>0.244</p></td><td align="left"><p>0.003</p></td></tr><tr><td align="left"><p>Confusion</p></td><td align="left"><p>-</p></td><td align="left"><p>Surprise</p></td><td align="left"><p>7.064</p></td><td align="left"><p>97</p></td><td align="left"><p> < 0.001***</p></td><td align="left"><p>0.667</p></td><td align="left"><p>0.094</p></td><td align="left"><p>0.479</p></td><td align="left"><p>0.854</p></td><td align="left"><p>0.714</p></td></tr><tr><td align="left"><p>Confusion</p></td><td align="left"><p>-</p></td><td align="left"><p>Boredom</p></td><td align="left"><p>2.703</p></td><td align="left"><p>97</p></td><td align="left"><p>0.008**</p></td><td align="left"><p>0.323</p></td><td align="left"><p>0.120</p></td><td align="left"><p>0.086</p></td><td align="left"><p>0.560</p></td><td align="left"><p>0.273</p></td></tr><tr><td align="left"><p>Confusion</p></td><td align="left"><p>-</p></td><td align="left"><p>Pride</p></td><td align="left"><p>2.765</p></td><td align="left"><p>97</p></td><td align="left"><p>0.007**</p></td><td align="left"><p>0.391</p></td><td align="left"><p>0.141</p></td><td align="left"><p>0.110</p></td><td align="left"><p>0.672</p></td><td align="left"><p>0.279</p></td></tr><tr><td align="left"><p>Confusion</p></td><td align="left"><p>-</p></td><td align="left"><p>Anxiety</p></td><td align="left"><p>4.441</p></td><td align="left"><p>97</p></td><td align="left"><p> < 0.001***</p></td><td align="left"><p>0.323</p></td><td align="left"><p>0.073</p></td><td align="left"><p>0.179</p></td><td align="left"><p>0.468</p></td><td align="left"><p>0.449</p></td></tr><tr><td align="left"><p>Confusion</p></td><td align="left"><p>-</p></td><td align="left"><p>Enjoyment</p></td><td align="left"><p>4.735</p></td><td align="left"><p>97</p></td><td align="left"><p> < 0.001***</p></td><td align="left"><p>0.571</p></td><td align="left"><p>0.121</p></td><td align="left"><p>0.332</p></td><td align="left"><p>0.811</p></td><td align="left"><p>0.478</p></td></tr><tr><td align="left"><p>Curiosity</p></td><td align="left"><p>-</p></td><td align="left"><p>Surprise</p></td><td align="left"><p>5.774</p></td><td align="left"><p>97</p></td><td align="left"><p> < 0.001***</p></td><td align="left"><p>0.663</p></td><td align="left"><p>0.115</p></td><td align="left"><p>0.435</p></td><td align="left"><p>0.891</p></td><td align="left"><p>0.583</p></td></tr><tr><td align="left"><p>Curiosity</p></td><td align="left"><p>-</p></td><td align="left"><p>Boredom</p></td><td align="left"><p>1.916</p></td><td align="left"><p>97</p></td><td align="left"><p>0.058</p></td><td align="left"><p>0.320</p></td><td align="left"><p>0.167</p></td><td align="left"><p>-0.011</p></td><td align="left"><p>0.651</p></td><td align="left"><p>0.194</p></td></tr><tr><td align="left"><p>Curiosity</p></td><td align="left"><p>-</p></td><td align="left"><p>Pride</p></td><td align="left"><p>4.672</p></td><td align="left"><p>97</p></td><td align="left"><p> < 0.001***</p></td><td align="left"><p>0.388</p></td><td align="left"><p>0.083</p></td><td align="left"><p>0.223</p></td><td align="left"><p>0.552</p></td><td align="left"><p>0.472</p></td></tr><tr><td align="left"><p>Curiosity</p></td><td align="left"><p>-</p></td><td align="left"><p>Anxiety</p></td><td align="left"><p>2.324</p></td><td align="left"><p>97</p></td><td align="left"><p>0.022*</p></td><td align="left"><p>0.320</p></td><td align="left"><p>0.138</p></td><td align="left"><p>0.047</p></td><td align="left"><p>0.593</p></td><td align="left"><p>0.235</p></td></tr><tr><td align="left"><p>Curiosity</p></td><td align="left"><p>-</p></td><td align="left"><p>Enjoyment</p></td><td align="left"><p>7.846</p></td><td align="left"><p>97</p></td><td align="left"><p> < 0.001***</p></td><td align="left"><p>0.568</p></td><td align="left"><p>0.072</p></td><td align="left"><p>0.424</p></td><td align="left"><p>0.712</p></td><td align="left"><p>0.793</p></td></tr><tr><td align="left"><p>Surprise</p></td><td align="left"><p>-</p></td><td align="left"><p>Boredom</p></td><td align="left"><p>-2.912</p></td><td align="left"><p>97</p></td><td align="left"><p>0.004**</p></td><td align="left"><p>-0.344</p></td><td align="left"><p>0.118</p></td><td align="left"><p>-0.578</p></td><td align="left"><p>-0.109</p></td><td align="left"><p>-0.294</p></td></tr><tr><td align="left"><p>Surprise</p></td><td align="left"><p>-</p></td><td align="left"><p>Pride</p></td><td align="left"><p>-2.354</p></td><td align="left"><p>97</p></td><td align="left"><p>0.021*</p></td><td align="left"><p>-0.276</p></td><td align="left"><p>0.117</p></td><td align="left"><p>-0.508</p></td><td align="left"><p>-0.043</p></td><td align="left"><p>-0.238</p></td></tr><tr><td align="left"><p>Surprise</p></td><td align="left"><p>-</p></td><td align="left"><p>Anxiety</p></td><td align="left"><p>-3.823</p></td><td align="left"><p>97</p></td><td align="left"><p> < 0.001***</p></td><td align="left"><p>-0.344</p></td><td align="left"><p>0.090</p></td><td align="left"><p>-0.522</p></td><td align="left"><p>-0.165</p></td><td align="left"><p>-0.386</p></td></tr><tr><td align="left"><p>Surprise</p></td><td align="left"><p>-</p></td><td align="left"><p>Enjoyment</p></td><td align="left"><p>-1.031</p></td><td align="left"><p>97</p></td><td align="left"><p>0.305</p></td><td align="left"><p>-0.095</p></td><td align="left"><p>0.092</p></td><td align="left"><p>-0.279</p></td><td align="left"><p>0.088</p></td><td align="left"><p>-0.104</p></td></tr><tr><td align="left"><p>Boredom</p></td><td align="left"><p>-</p></td><td align="left"><p>Pride</p></td><td align="left"><p>0.423</p></td><td align="left"><p>97</p></td><td align="left"><p>0.673</p></td><td align="left"><p>0.068</p></td><td align="left"><p>0.161</p></td><td align="left"><p>-0.251</p></td><td align="left"><p>0.387</p></td><td align="left"><p>0.043</p></td></tr><tr><td align="left"><p>Boredom</p></td><td align="left"><p>-</p></td><td align="left"><p>Anxiety</p></td><td align="left"><p>0.000</p></td><td align="left"><p>97</p></td><td align="left"><p>1.000</p></td><td align="left"><p>0.000</p></td><td align="left"><p>0.116</p></td><td align="left"><p>-0.231</p></td><td align="left"><p>0.231</p></td><td align="left"><p>0.000</p></td></tr><tr><td align="left"><p>Boredom</p></td><td align="left"><p>-</p></td><td align="left"><p>Enjoyment</p></td><td align="left"><p>1.573</p></td><td align="left"><p>97</p></td><td align="left"><p>0.119</p></td><td align="left"><p>0.248</p></td><td align="left"><p>0.158</p></td><td align="left"><p>-0.065</p></td><td align="left"><p>0.562</p></td><td align="left"><p>0.159</p></td></tr><tr><td align="left"><p>Pride</p></td><td align="left"><p>-</p></td><td align="left"><p>Anxiety</p></td><td align="left"><p>-0.470</p></td><td align="left"><p>97</p></td><td align="left"><p>0.640</p></td><td align="left"><p>-0.068</p></td><td align="left"><p>0.145</p></td><td align="left"><p>-0.355</p></td><td align="left"><p>0.219</p></td><td align="left"><p>-0.047</p></td></tr><tr><td align="left"><p>Pride</p></td><td align="left"><p>-</p></td><td align="left"><p>Enjoyment</p></td><td align="left"><p>2.519</p></td><td align="left"><p>97</p></td><td align="left"><p>0.013*</p></td><td align="left"><p>0.180</p></td><td align="left"><p>0.072</p></td><td align="left"><p>0.038</p></td><td align="left"><p>0.322</p></td><td align="left"><p>0.254</p></td></tr><tr><td align="left"><p>Anxiety</p></td><td align="left"><p>-</p></td><td align="left"><p>Enjoyment</p></td><td align="left"><p>1.957</p></td><td align="left"><p>97</p></td><td align="left"><p>0.053</p></td><td align="left"><p>0.248</p></td><td align="left"><p>0.127</p></td><td align="left"><p>-0.003</p></td><td align="left"><p>0.500</p></td><td align="left"><p>0.198</p></td></tr></tbody></table> </ephtml> </p> <p>MD = Mean Difference, SED = Standard Error Difference</p> <p>The epistemic emotions of confusion and curiosity decreased over the problem-solving process, as indicated by the significant negative slopes in the confusion and curiosity models, respectively. The rates of decline in confusion and curiosity did not significantly differ since the confidence intervals for the slope of confusion, 95% CI [-0.197, -0.035], overlapped the slope of curiosity, 95% CI [-0.365, -0.177]. Furthermore, nonsignificant variance in the slope of confusion suggested that students were not significantly different from one another in the rate of decline of confusion. However, the slope variance for curiosity was significant, indicating that students followed significantly different growth trajectories. For curiosity, the covariance of the intercept and the slope was negative and significant (<emph>p</emph> < 0.01), meaning that students who reported higher initial levels of curiosity tended to decrease more rapidly over time. With regard to the epistemic emotion of surprise and all four achievement emotions, the slopes were not significantly different from zero, which meant that there was no development over time on average for these emotions. Nevertheless, the variances of slopes for pride, anxiety, and enjoyment were significant, indicating that there was significant variability in their growth rates. As displayed in Fig. 3, students maintained a medium level of anxiety across the phases of problem-solving, whereas the mean of pride and boredom changed from the <emph>performance</emph> phase to the <emph>self-reflection</emph> phase. As three-time points were not sufficient to perform a curvilinear analysis (Hesser, [<reflink idref="bib25" id="ref129">25</reflink>]), we performed paired t-tests with Bonferroni correction to confirm that students experienced a significant increase of boredom (<emph>t</emph> = 2.44, <emph>p</emph> = 0.022 < 0.025) but a decrease of pride at the final stage of problem-solving (t = -2.62, <emph>p</emph> = 0.01 < 0.025).</p> <p>Graph: Fig. 3Model-implied trajectories for unconditional models of seven academic emotions</p> <hd id="AN0168593744-18">Conditional growth models with a distal outcome of performance</hd> <p>For our second research question regarding the relationship between emotions and learning performance, we added students' diagnostic performance into each emotion model as a distal outcome of the intercept and slope (see Fig. 4). Although the variances in slopes for confusion, surprise, and boredom were not significantly different from zero, we decided to keep them to test these constructs in growth models with a performance outcome. We were interested in the relationships between the variances in slopes and performance, as well as how the variance in initial levels of emotions influenced performance, as the variance in intercepts for all emotions was significant. Results in Table 5 showed that the variances in slopes could not predict students' performance for the seven emotions, whereas both initial levels of curiosity and enjoyment positively predicted students' performance, with <emph>b</emph> = 0.061, <emph>p</emph> = 0.045, and <emph>b</emph> = 0.080, <emph>p</emph> = 0.016, respectively.</p> <p>Graph: Fig. 4A latent linear growth model with a distal outcome of performance, modeled separately for confusion, curiosity, surprise, boredom, pride, anxiety, and enjoyment</p> <p>Table 5 Latent growth models with a distal outcome of performance</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="2"><p>Model</p></th><th align="left" colspan="3"><p>Performance</p></th></tr><tr><th align="left"><p><italic>b</italic></p></th><th align="left"><p>S.E</p></th><th align="left"><p><italic>p</italic></p></th></tr></thead><tbody><tr><td align="left"><p>Confusion</p></td><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left"><p> Intercept</p></td><td align="left"><p>-0.007</p></td><td align="left"><p>0.051</p></td><td align="left"><p>0.895</p></td></tr><tr><td align="left"><p> Slope</p></td><td align="left"><p>0.018</p></td><td align="left"><p>0.249</p></td><td align="left"><p>0.944</p></td></tr><tr><td align="left"><p>Curiosity</p></td><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left"><p> Intercept</p></td><td align="left"><p>0.061</p></td><td align="left"><p>0.031</p></td><td align="left"><p>0.045*</p></td></tr><tr><td align="left"><p> Slope</p></td><td align="left"><p>0.075</p></td><td align="left"><p>0.063</p></td><td align="left"><p>0.233</p></td></tr><tr><td align="left"><p>Surprise</p></td><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left"><p> Intercept</p></td><td align="left"><p>0.001</p></td><td align="left"><p>0.521</p></td><td align="left"><p>0.998</p></td></tr><tr><td align="left"><p> Slope</p></td><td align="left"><p>-0.500</p></td><td align="left"><p>6.121</p></td><td align="left"><p>0.935</p></td></tr><tr><td align="left"><p>Boredom</p></td><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left"><p> Intercept</p></td><td align="left"><p>-0.068</p></td><td align="left"><p>0.037</p></td><td align="left"><p>0.066</p></td></tr><tr><td align="left"><p> Slope</p></td><td align="left"><p>-0.043</p></td><td align="left"><p>0.106</p></td><td align="left"><p>0.682</p></td></tr><tr><td align="left"><p>Pride</p></td><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left"><p> Intercept</p></td><td align="left"><p>0.044</p></td><td align="left"><p>0.032</p></td><td align="left"><p>0.163</p></td></tr><tr><td align="left"><p> Slope</p></td><td align="left"><p>0.051</p></td><td align="left"><p>0.055</p></td><td align="left"><p>0.357</p></td></tr><tr><td align="left"><p>Anxiety</p></td><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left"><p> Intercept</p></td><td align="left"><p>0.015</p></td><td align="left"><p>0.040</p></td><td align="left"><p>0.713</p></td></tr><tr><td align="left"><p> Slope</p></td><td align="left"><p>-0.028</p></td><td align="left"><p>0.086</p></td><td align="left"><p>0.747</p></td></tr><tr><td align="left"><p>Enjoyment</p></td><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left"><p> Intercept</p></td><td align="left"><p>0.080</p></td><td align="left"><p>0.033</p></td><td align="left"><p>0.016*</p></td></tr><tr><td align="left"><p> Slope</p></td><td align="left"><p>0.099</p></td><td align="left"><p>0.084</p></td><td align="left"><p>0.238</p></td></tr></tbody></table> </ephtml> </p> <p>S.E. = Standard Error; * <emph>p</emph> < 0.05</p> <hd id="AN0168593744-19">Latent transition analysis</hd> <p>Regarding our third research question, we applied latent cluster analysis (LCA) and latent transition analysis (LTA) to identify students' emotion profiles at each phase of SRL. A series of latent cluster analyses were applied to the data to determine the optimal number of latent emotion classes at each time point. We relied on the model fit indices, including the Bayesian Information Criterion (BIC), adjusted BIC, the Bootstrapped Likelihood Ratio (BLR) test, and the Lo-Mendell-Rubin (LMR) adjusted likelihood ratio test. The BIC and adjusted BIC are descriptive goodness-of-fit indices wherein smaller values indicate better model fit. A significant <emph>p-value</emph> generated for either BLR or LMR suggests a model improvement when comparing a <emph>k</emph> class (e.g., 3-class) to a <emph>k-1</emph> class solution (e.g., 2-class). According to (Nylund et al., [<reflink idref="bib56" id="ref130">56</reflink>]), the <emph>p-value</emph> for the BLR test performed the best among all the fit indices when determining an optimal number of classes in latent class analysis, followed by the BIC and adjusted BIC indices. In addition, we checked the entropy value and the number of individuals in each group to determine the quality of class membership classification. In particular, an entropy value larger than 0.80 suggests latent class membership is clearly classified (Wang & Wang, [<reflink idref="bib79" id="ref131">79</reflink>]). The three-class solution emerged as the best fit for the data at Time 2 and Time 3, as indicated by the lowest AIC, lowest BIC, lowest BIC, and highest entropy values in Table 6. Although lower AIC and BIC values are associated with the 4-cluster solution at Time 1, we deemed that the 3-cluster solution was optimal considering the insignificant <emph>p-value</emph> (<emph>p</emph> = 0.32) for BLR and the smallest class size for the 4-cluster solution (<emph>N</emph> = 5).</p> <p>Table 6 Fit indices for the latent profile analysis models with 1–5 classes for three time points</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left" /><th align="left"><p>Model</p></th><th align="left"><p>AIC</p></th><th align="left"><p>BIC</p></th><th align="left"><p>ABIC</p></th><th align="left"><p><italic>p BLR</italic></p></th><th align="left"><p><italic>p LMR</italic></p></th><th align="left"><p>Entropy</p></th><th align="left"><p>No</p></th></tr></thead><tbody><tr><td align="left"><p>Time 1</p></td><td align="left"><p>1-class</p></td><td char="." align="char"><p>1845.99</p></td><td char="." align="char"><p>1882.18</p></td><td char="." align="char"><p>1837.97</p></td><td align="left"><p>-</p></td><td align="left"><p>-</p></td><td align="left"><p>-</p></td><td align="left" /></tr><tr><td align="left" /><td align="left"><p>2-class</p></td><td char="." align="char"><p>1733.39</p></td><td char="." align="char"><p>1790.26</p></td><td char="." align="char"><p>1720.79</p></td><td align="left"><p>0.000</p></td><td align="left"><p>0.0013</p></td><td align="left"><p>0.931</p></td><td align="left"><p>44(0.45)</p></td></tr><tr><td align="left" /><td align="left"><p>3-class</p></td><td char="." align="char"><p>1710.75</p></td><td char="." align="char"><p>1788.30</p></td><td char="." align="char"><p>1693.56</p></td><td align="left"><p>0.000</p></td><td align="left"><p>0.58</p></td><td align="left"><p>0.911</p></td><td align="left"><p>9(0.097)</p></td></tr><tr><td align="left" /><td align="left"><p>4-class</p></td><td char="." align="char"><p>1662.32</p></td><td char="." align="char"><p>1760.55</p></td><td char="." align="char"><p>1640.55</p></td><td align="left"><p>0.000</p></td><td align="left"><p>0.32</p></td><td align="left"><p>0.941</p></td><td align="left"><p>5(0.051)</p></td></tr><tr><td align="left" /><td align="left"><p>5-class</p></td><td char="." align="char"><p>1701.48</p></td><td char="." align="char"><p>1820.39</p></td><td char="." align="char"><p>1675.13</p></td><td align="left"><p>0.000</p></td><td align="left"><p>0.38</p></td><td align="left"><p>0.900</p></td><td align="left"><p>1(0.01)</p></td></tr><tr><td align="left"><p>Time 2</p></td><td align="left"><p>1-class</p></td><td char="." align="char"><p>1893.81</p></td><td char="." align="char"><p>1930.00</p></td><td char="." align="char"><p>1885.79</p></td><td align="left"><p>-</p></td><td align="left"><p>-</p></td><td align="left"><p>-</p></td><td align="left" /></tr><tr><td align="left" /><td align="left"><p>2-class</p></td><td char="." align="char"><p>1757.75</p></td><td char="." align="char"><p>1841.62</p></td><td char="." align="char"><p>1745.15</p></td><td align="left"><p>0.000</p></td><td align="left"><p>0.10</p></td><td align="left"><p>0.901</p></td><td align="left"><p>32(0.33)</p></td></tr><tr><td align="left" /><td align="left"><p>3-class</p></td><td char="." align="char"><p>1699.55</p></td><td char="." align="char"><p>1777.10</p></td><td char="." align="char"><p>1862.36</p></td><td align="left"><p>0.000</p></td><td align="left"><p>0.22</p></td><td align="left"><p>0.932</p></td><td align="left"><p>11(0.11)</p></td></tr><tr><td align="left" /><td align="left"><p>4-class</p></td><td char="." align="char"><p>1715.55</p></td><td char="." align="char"><p>1813.78</p></td><td char="." align="char"><p>1693.78</p></td><td align="left"><p>0.000</p></td><td align="left"><p>0.18</p></td><td align="left"><p>0.591</p></td><td align="left"><p>13(0.13)</p></td></tr><tr><td align="left"><p>Time 3</p></td><td align="left"><p>1-class</p></td><td char="." align="char"><p>1859.80</p></td><td char="." align="char"><p>1895.99</p></td><td char="." align="char"><p>1851.78</p></td><td align="left"><p>-</p></td><td align="left"><p>-</p></td><td align="left"><p>-</p></td><td align="left" /></tr><tr><td align="left" /><td align="left"><p>2-class</p></td><td char="." align="char"><p>1695.39</p></td><td char="." align="char"><p>1752.26</p></td><td char="." align="char"><p>1682.78</p></td><td align="left"><p>0.000</p></td><td align="left"><p>0.006</p></td><td align="left"><p>0.910</p></td><td align="left"><p>27(0.28)</p></td></tr><tr><td align="left" /><td align="left"><p>3-class</p></td><td char="." align="char"><p>1648.17</p></td><td char="." align="char"><p>1725.72</p></td><td char="." align="char"><p>1630.99</p></td><td align="left"><p>0.000</p></td><td align="left"><p>0.009</p></td><td align="left"><p>0.933</p></td><td align="left"><p>15(0.15)</p></td></tr><tr><td align="left" /><td align="left"><p>4-class</p></td><td char="." align="char"><p>1664.17</p></td><td char="." align="char"><p>1762.40</p></td><td char="." align="char"><p>1642.40</p></td><td align="left"><p>0.000</p></td><td align="left"><p>1.00</p></td><td align="left"><p>0.636</p></td><td align="left"><p>16(0.16)</p></td></tr></tbody></table> </ephtml> </p> <p>No. = Smallest class count and proportion<emph>, p</emph> BLR = <emph>p</emph> values for the Bootstrapped Likelihood Ratio test, <emph>p</emph> LMR = <emph>p</emph> values for the Lo-Mendell-Rubin adjusted likelihood ratio test</p> <p>We successively performed latent transition analysis (LTA) with parameter restrictions that constrained emotion profile parameters to be equal at T1, T2, and T3 for three. The final LTA model identified emotion profiles of individuals across three-time points. The composition of three emotion profiles at three-time points is displayed in Figs. 5, 6, 7. The curious-positive profile consists of students who demonstrated high levels of curiosity and positive emotions. Students in this profile had the highest level of curiosity, pride, and enjoyment at the three-time points. The confused-negative profile exhibited high levels of confusion, boredom, and anxiety. The two negative emotions, boredom and anxiety, co-occur across the three phases of SRL. More interestingly, the epistemic emotion of confusion co-occurs with anxiety and boredom. The third profile is moderate-low because of its low to moderate levels of all emotions.</p> <p>Graph: Fig. 5Composition of three latent classes at time 1</p> <p>Graph: Fig. 6Composition of three latent classes at time 2</p> <p>Graph: Fig. 7Composition of three latent classes at time 3</p> <p>The final LTA model also examined the stability and change of emotion profiles at the individual level. Table 7 displays the probability of remaining in the same class or moving to another class across the three-time points. Individuals were more likely to stay in the same profile even though some individuals transitioned between profiles. Individuals from the T1 confused-negative profile either moved to the T2 curious-positive profile (20%) or the moderate-low profile (24%). Some individuals from the T1 moderate-low profile moved to the T2 curious-positive profile (18%). Individuals from the T2 moderate-low profile also moved to the T3 curious-positive profile (35%). Some individuals moved from the T2 confused-negative profile to the T3 moderate-low profile (16%).</p> <p>Table 7 Emotion profile transition probabilities</p> <p> <ephtml> <table frame="hsides" rules="groups"><tbody><tr><td align="left" /><td align="left" colspan="2"><p><italic>Time 2</italic> profiles</p></td><td align="left" /></tr><tr><td align="left"><p><italic>Time 1</italic> profiles</p></td><td align="left"><p>Curious-Positive (<italic>N</italic> = 59)</p></td><td align="left"><p>Confused-Negative</p><p>(<italic>N</italic> = 13)</p></td><td align="left"><p>Moderate-Low (<italic>N</italic> = 26)</p></td></tr><tr><td align="left"><p>Curious-Positive (<italic>N</italic> = 52)</p></td><td align="left"><p>0.98</p></td><td align="left"><p>0.021</p></td><td align="left"><p>0</p></td></tr><tr><td align="left"><p>Confused-Negative</p><p>(<italic>N</italic> = 20)</p></td><td align="left"><p>0.20</p></td><td align="left"><p>0.56</p></td><td align="left"><p>0.24</p></td></tr><tr><td align="left"><p>Moderate-Low</p><p>(<italic>N</italic> = 26)</p></td><td align="left"><p>0.18</p></td><td align="left"><p>0.063</p></td><td align="left"><p>0.76</p></td></tr><tr><td align="left" /><td align="left" colspan="2"><p><italic>Time 3</italic> profiles</p></td><td align="left" /></tr><tr><td align="left"><p><italic>Time 2</italic> profiles</p></td><td align="left"><p>Curious-Positive (<italic>N</italic> = 62)</p></td><td align="left"><p>Confused-Negative</p><p>(<italic>N</italic> = 12)</p></td><td align="left"><p>Moderate-Low (<italic>N</italic> = 24)</p></td></tr><tr><td align="left"><p>Curious-Positive (<italic>N</italic> = 59)</p></td><td align="left"><p>0.89</p></td><td align="left"><p>0.016</p></td><td align="left"><p>0.098</p></td></tr><tr><td align="left"><p>Confused-Negative</p><p>(<italic>N</italic> = 13)</p></td><td align="left"><p>0.00</p></td><td align="left"><p>0.84</p></td><td align="left"><p>0.16</p></td></tr><tr><td align="left"><p>Moderate-Low</p><p>(<italic>N</italic> = 26)</p></td><td align="left"><p>0.35</p></td><td align="left"><p>0.00</p></td><td align="left"><p>0.65</p></td></tr></tbody></table> </ephtml> </p> <hd id="AN0168593744-20">Discussion</hd> <p>The fundamental premise underlying research on state-like emotions in education is that emotions could change across a learning task, and multiple emotions are likely to occur together. Nevertheless, this dynamic nature of academic emotions remains relatively unexamined. This current study attempted to fill in this gap by investigating the trajectories of academic emotions (i.e., achievement emotions and epistemic emotions) across the course of clinical problem-solving, by examining the associations between academic emotions and learning performance, and by identifying students' temporal changes of emotion profiles in the process of solving clinical problems. Regarding academic emotion trajectories, we found that confusion and curiosity decreased significantly throughout the problem-solving process. Students in this study reported significantly higher levels of epistemic emotions of confusion and curiosity than achievement emotions (i.e., boredom, pride, anxiety, and enjoyment) at the forethought phase of clinical problem-solving. In terms of the relationship between academic emotions and learning performance, we found that both initial levels of curiosity and enjoyment positively predicted students' performance. However, the change in academic emotions did not significantly predict students' performance. With regard to students' emotion profiles, we identified three emotion profiles: positive-curious, negative-boredom, and moderate-low. These three emotion profiles were relatively stable over time.</p> <hd id="AN0168593744-21">Emotion trajectories</hd> <p>The decrease in epistemic emotions was confirmed for curiosity and confusion but not for surprise. The intensity of surprise was significantly lower than curiosity and confusion, even though the growth rate coefficient for surprise was not statistically significant. These findings were partly consistent with prior studies that reported a low frequency of surprise (D'Mello & Graesser, [<reflink idref="bib14" id="ref132">14</reflink>]) and a decrease in epistemic emotions (Nerantzaki & Efklides, [<reflink idref="bib54" id="ref133">54</reflink>]). It is very likely that the mechanism underlying the occurrence and transition of surprise leads to the low intensity of surprise and the insignificant growth rate of surprise. According to Foster and Keane ([<reflink idref="bib20" id="ref134">20</reflink>]), the level of surprise is a function of the level of novelty of the information, and learners can continue to engage in learning even with a low level of surprise. Students in the current study have received pre-training a week before the data collection to get familiar with the learning environment. They may have mentally prepared for what to expect even in the <emph>forethought</emph> phase of problem-solving. There was little new information that would give rise to surprise. Another possible reason for the insignificant growth rate of surprise may have to do with the transition from surprise to curiosity and confusion in the <emph>forethought</emph> phase of problem-solving. Surprise is an emotion that may result in curiosity or confusion depending on the complexity of the task (Foster & Keane, [<reflink idref="bib20" id="ref135">20</reflink>]; Silvia, [<reflink idref="bib71" id="ref136">71</reflink>]). According to Silvia ([<reflink idref="bib71" id="ref137">71</reflink>]), curiosity is more likely to follow surprise if the complexity of the task is low, whereas confusion is more likely to follow surprise if the complexity of the task is high. As the antecedents of curiosity and confusion, surprise will transit to curiosity and confusion in the task analysis process at the <emph>forethought</emph> phase of SRL. Curiosity and confusion can decline when there is an incidence of impasse resolution and perceived success in the <emph>performance</emph> and <emph>self-reflection</emph> phase of problem-solving (D'Mello & Graesser, [<reflink idref="bib14" id="ref138">14</reflink>]). Students do not randomly shift between emotions. This finding also supports D'Mello and Graesser's ([<reflink idref="bib14" id="ref139">14</reflink>]) dynamic affect model that posits the interplay between epistemic emotion trajectory and external events that trigger impasses and obstacles to learning.</p> <p>Although the growth rate coefficients were positive for negative achievement emotions (i.e., anxiety and boredom) and negative for positive emotions (i.e., pride and enjoyment), they were not statistically significant. Contrary to our expectations, achievement emotions did not show a significant increasing or declining trend. More specifically, anxiety appeared to be relatively stable across the three phases of problem-solving, which aligns with the within-course and within-year change of anxiety in prior studies conducted by Niculescu et al. ([<reflink idref="bib55" id="ref140">55</reflink>]) and Ahmed et al. ([<reflink idref="bib3" id="ref141">3</reflink>]). Regarding other achievement emotions, the increase of boredom and the decrease of enjoyment and pride only occur at the <emph>self-reflection</emph> stage of problem-solving. As of yet, specific studies on the trajectory of achievement emotions in problem-solving are currently unavailable, and direct comparison of any sort is not possible. One possible reason is that students' achievement emotions were more closely attached to learning outcomes than learning activities. Achievement emotions are directly linked to achievement activities or achievement outcomes (Pekrun et al., [<reflink idref="bib62" id="ref142">62</reflink>]). It is likely that achievement outcomes at the <emph>self-reflection</emph> phase of SRL are a more robust indicator of achievement emotions in problem-solving. Students' intensity of emotions differs depending on whether there is a positive or negative learning outcome related to problem-solving (Taub et al., [<reflink idref="bib74" id="ref143">74</reflink>]). Thus, an increase in boredom and a decrease in enjoyment and pride at the end of problem-solving may have to do with problem-solving outcomes and feedback.</p> <p>In addition to the trajectory of epistemic emotions and achievement emotions, we find students varied significantly in epistemic and achievement emotions at the <emph>forethought</emph> phase of SRL, as indicated by the significant variances of intercepts in all emotion models. More specifically, students experienced a higher level of curiosity and confusion from the beginning of problem-solving. This finding confirms our assumption that higher intensity of epistemic emotions occur at the <emph>forethought</emph> phase of SRL since epistemic emotions are likely to be triggered by new information in the task analysis process. Furthermore, the intercept and slope of curiosity were negatively correlated, suggesting that higher levels of curiosity at the <emph>forethought</emph> phase were associated with more dramatic declines in curiosity over time. These are promising findings with regard to implications for early intervention, as it suggests that epistemic emotions should be the primary focus of early intervention in problem-solving since the high initial level of curiosity may serve as a risk factor for declines in the following phase of problem-solving.</p> <hd id="AN0168593744-22">Emotions and their relationship to performance</hd> <p>The change of emotions did not significantly predict students' performance. However, we found that both initial levels of curiosity and enjoyment positively predicted students' performance. The positive relationship between initial levels of enjoyment and performance corroborates previous cross-sectional studies (Ahmed et al., [<reflink idref="bib3" id="ref144">3</reflink>]; Frenzel et al., [<reflink idref="bib21" id="ref145">21</reflink>]; Pekrun et al., [<reflink idref="bib59" id="ref146">59</reflink>]). Individuals who start with a more positive state tend to be more proficient at elaborating, organizing, and categorizing incoming information, and consequently have the potential to produce positive learning outcomes (Isen, [<reflink idref="bib27" id="ref147">27</reflink>]). The role of curiosity in learning performance showed a similar pattern to that of enjoyment. Consistent with the prior study findings reported by Munzar et al. ([<reflink idref="bib51" id="ref148">51</reflink>]), we found that the initial level of curiosity positively predicted students' performance. Curiosity is defined as students' desire to acquire new knowledge and new experience (Efklides, [<reflink idref="bib18" id="ref149">18</reflink>]). Curiosity is the result of impasses or gaps in knowledge. Students who are able to resolve impasses are confident in their ability to complete the task, meaning that a higher level of curiosity should be accompanied by a higher perception of self-control (Munzar et al., [<reflink idref="bib51" id="ref150">51</reflink>]). Thus, students who start with a higher level of curiosity should demonstrate better performance. The current study findings further support the importance of early intervention in enjoyment and curiosity. The feelings of enjoyment and curiosity may best be kept at the early stage of problem-solving, although curiosity may decline throughout the problem-solving process.</p> <hd id="AN0168593744-23">Emotion profiles</hd> <p>This study identifies three emotion profiles across problem-solving processes: curious-positive, confused-negative, and moderate-low. These findings confirm the existence of three emotion profiles in problem-solving identified by prior researchers (Jarrell et al., [<reflink idref="bib30" id="ref151">30</reflink>], [<reflink idref="bib31" id="ref152">31</reflink>]; Sinclair et al., [<reflink idref="bib72" id="ref153">72</reflink>]). The identification of three emotion profiles provides a complement to Jarrell's et al. ([<reflink idref="bib30" id="ref154">30</reflink>], [<reflink idref="bib31" id="ref155">31</reflink>]) research findings wherein a profile is only characterized by achievement emotions. More specifically, the inclusion of epistemic emotions further extends our understanding of academic emotions, providing empirical support for the unique role of curiosity in problem-solving. Furthermore, this study contributes to the literature by considering the stability and change of emotion profiles over time. We found that most students exhibited stable emotion profiles across the problem-solving processes. Students who belonged to the curious-positive profile at the <emph>forethought</emph> phase were more likely to stay in this emotion profile during both the <emph>performance</emph> and <emph>self-reflection</emph> phases. Conversely, students who were in the confused-negative profile at <emph>the forethought</emph> phase had the possibility to transit to either the curious-positive profile or moderate-low profile at the <emph>performance</emph> phase. More interestingly, students in the moderate-low profile at the <emph>forethought</emph> and <emph>performance</emph> phase either stayed in the same profile or transited to the curious-positive profile. These findings suggest the urgency of providing early support for students in the confused-negative profile. Early interventions may be best taken on students in the boredom-negative profile to help them transit to the curious-positive profile instead of the medium–low profile.</p> <hd id="AN0168593744-24">Educational implications</hd> <p>Findings from this study have implications for instruction. The emotion trajectories and their relationship to learning performance illustrate the critical role emotions play in specific SRL phases, providing instructors with insights regarding what emotions to expect at each phase of SRL. More importantly, understanding the temporal emotion profile of students can lead to early instructional interventions in the learning process. Instructors or learning systems can obtain information to decide when it is optimal to provide students with emotional support.</p> <p>First, our results suggest that instructors should expect epistemic emotions rather than achievement emotions at the <emph>forethought</emph> phase of self-regulated learning. Intensive knowledge and knowledge-generation activities can occur at this phase to trigger students' epistemic emotions. Instructors should be prepared to address students' curiosity and confusion when they are analyzing the task and obtaining new information. Moreover, instructors should maintain students' high level of curiosity and enjoyment at this phase since the initial level of these two emotions positively predicts students' final performance. Some positive and inspiring feedback can keep students happy, and certain enlightening questions may make students curious about the following steps of problem-solving.</p> <p>Second, our findings suggest that the support of achievement emotions should focus on the <emph>self-reflection</emph> phase of SRL<emph>. Self-reflection</emph> leads to students' emotional reactions to the quality and quantity of learning outcomes (Schmitz et al., [<reflink idref="bib70" id="ref156">70</reflink>]), which is evidenced by the significant decrease in positive emotions (i.e., enjoyment and pride) and an increase in boredom at the <emph>self-reflection</emph> phase of the current study. As such, instructors should provide the necessary support to avoid negative consequences associated with a decrease in positive emotions and an increase in boredom. For example, a debrief at the end of problem-solving may help students reaccumulate their interests and positive feelings towards similar tasks in the future. Simultaneously, instructors must teach students correspondent emotion regulation strategies to tackle unfavorable emotion changes to minimize the detrimental effects of these changes on their learning processes and future learning plans.</p> <p>In alignment with prior research findings (Jarrell et al., [<reflink idref="bib31" id="ref157">31</reflink>]), our study recognizes that students in the confused-negative group are the ones who need more external support from instructors. About half of the students in the confused-negative group move to either the curious-positive group or the moderate-low group in the <emph>performance</emph> phase of SRL, suggesting that the <emph>performance</emph> phase is the right time for instructors to provide emotional support for students in the negative-boredom group. More importantly, our results highlight the possibility of changing students from the confused-negative profile into the curious-positive profile during problem-solving. Although the majority of students maintain a stable emotion profile, instructors should not despair when students display a confused-negative emotion profile. Instead, instructors should engage in early interventions to help students transit into a curious-positive profile. It is possible that students may temporally feel bored by the impasse of the challenging task. With certain external support from the instructor, students may feel curious and positive once they work through the impasse of the task.</p> <p>Finally, our findings also offer insight into the design of technology-rich learning environments to support problem-solving. Students' academic emotions (i.e., achievement emotions and epistemic emotions) are not static across the phases of self-regulated learning (Li et al., [<reflink idref="bib41" id="ref158">41</reflink>]; Zheng et al., [<reflink idref="bib80" id="ref159">80</reflink>]). Technology-rich learning environments that can automatically and dynamically detect academic emotions can provide learners with personalized feedback and timely support to help them resolve the underlying impasse and regulate emotions. For example, researchers are developing multimodal user emotion detection systems to identify real-time emotions from speech and facial expressions (Alonso-Martin et al., [<reflink idref="bib4" id="ref160">4</reflink>]; Kim et al., [<reflink idref="bib33" id="ref161">33</reflink>]). Integrating these systems into a learning environment can benefit learners by unobtrusively detecting an individualized emotion trajectory and profile that can lead learners to perform in-depth self-reflection. Advanced technologies make it possible to display students' emotion trajectory and profile in the form of visual elements. This study provides evidence for incorporating educational technologies into the emotional aspect of learning in the future.</p> <hd id="AN0168593744-25">Limitations and future direction</hd> <p>Although growth models have been fitted into samples as small as 22 (Curran et al., [<reflink idref="bib10" id="ref162">10</reflink>]), our study is limited to a small sample size and low variability of the sample population. The participants in this study were medical students from a single university in China, which limits the generalizability of the results. Future studies should attempt to replicate this study with large sample sizes, taking cultural backgrounds, gender, ages, socioeconomic backgrounds, and other demographic factors into consideration. Moreover, a total of 129 students participated in this study. However, 31 did not fully report their emotional states in problem-solving and thus were not included in data analyses. Therefore, readers should be aware of the representativeness of the remaining students when interpreting the research findings, although we have checked the selection bias. Furthermore, the generalization of our findings is also limited to the context of clinical problem-solving. Future research should examine the emotion trajectory and their functional roles in other learning subjects and contexts, such as an exam.</p> <p>Another limitation is the inclusion of just three-time points, which may restrict the forms of growth model that can be studied. The three-time points used in the current study are meaningful time points chosen to avoid participants being overwhelmed by too frequent requests for answering questionnaires. However, more time points could increase the power and accuracy of detecting more patterns (Singer & Willett, [<reflink idref="bib73" id="ref163">73</reflink>]). Future research should collect additional time points by including other streams of data to model complex growth patterns. As discussed earlier, facial expressions and body gestures are behavioral channels that are relatively inconspicuous, which allows learners to complete learning without too much interruption from researchers. Similarly, physiological measures, such as electrodermal activities (EDA), are also promising for real-time measurement of emotions with the development of technology. Compared to self-report measures, behavioral and physiological measures capture the onset and transition of SRL as they occur in real-time and in the format of concurrent trace data that yield very reliable findings.</p> <p>Finally, future directions can aim to develop standardized tests to measure students' content knowledge related to the task. Although we used indices generated during students' learning trajectory as indicators of performance to model growth, these indices may overlook students' knowledge change over time. In future research, standardized tests may be used to model the longitudinal associations between emotions and SRL, considering the complexity of tasks. These standardized tests can also ensure that students are uniformly assessed with pre-test and post-test to monitor their learning gains after the task.</p> <hd id="AN0168593744-26">Conclusion</hd> <p>Despite the limitations discussed above, the present findings contribute to the advancement of scientific knowledge on academic emotion and SRL in several ways. First, the study is one of the first to explore academic emotion trajectories during the phases of SRL and problem-solving. Achievement emotions are more widely examined in education. The research on epistemic emotions, especially the temporal and dynamic nature of epistemic emotions, is very limited. Therefore, there have been calls for research on an academic emotion trajectory (D'Mello, [<reflink idref="bib12" id="ref164">12</reflink>]); we believe that we have attempted to answer such calls. More importantly, the current study results have practical implications regarding when to attend to students' emotions and what emotions need more attention from instructors. These practical implications provide instructors with straightforward clues for early interventions that aim at cultivating pleasant emotions to enhance students' SRL and performance. Finally, the person-oriented approach used in the study provides a parsimonious and meaningful way to summarize the multifaceted emotions students experienced in clinical problem-solving and explore the stability and change of emotion profiles. More specifically, the current study identified three emotion profiles: positive-curious, confused-negative, and moderate-low. Although emotion profiles are relatively stable, the confused-negative and moderate-low students may change their emotion profile in the learning processes. These results provide evidence for the importance of individualized emotional support.</p> <hd id="AN0168593744-27">Author contribution</hd> <p>Juan Zheng: Conceptualization, Methodology, Investigation, Data Curation, Formal analysis, Writing – Original Draft.</p> <p>Susanne P. Lajoie: Writing- Reviewing and Editing, Software, Supervision, Funding acquisition.</p> <p>Shan Li: Investigation, Resources, Formal analysis, Writing- Reviewing and Editing.</p> <p>Hongbin Wu: Investigation, Resources, Writing- Reviewing and Editing.</p> <hd id="AN0168593744-28">Funding</hd> <p>The research was supported by the Fonds de Recherche du Québec-Société et Culture (FRQSC) awarded to the first author, as well as in part by the Social Sciences and Humanities Research Council of Canada (SSHRC) awarded to the second author.</p> <hd id="AN0168593744-29">Data Availability</hd> <p>The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.</p> <hd id="AN0168593744-30">Declarations</hd> <p>The research was conducted in accordance with the journal's ethical guidelines. The authors declare that the work described was original research that has not been published previously, and not under considerations for publication elsewhere.</p> <hd id="AN0168593744-31">Conflict of interests</hd> <p>The authors declare that they have no conflict of interest.</p> <hd id="AN0168593744-32">Publisher's note</hd> <p>Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.</p> <ref id="AN0168593744-33"> <title> References </title> <blist> <bibl id="bib1" idref="ref50" type="bt">1</bibl> <bibtext> Ainley, M, Corrigan, M, Richardson, N. (2005). Students tasks and emotions: Identifying the contribution of emotions to students' reading of popular culture and popular science texts. Learning and Instruction, 15(5), 433–447.</bibtext> </blist> <blist> <bibl id="bib2" idref="ref96" type="bt">2</bibl> <bibtext> Agarwal NK. Verifying survey items for construct validity: A two-stage sorting procedure for questionnaire design in information behavior research. Proceedings of the American Society for Information Science and Technology. 2011; 48; 1: 1-8. 10.1002/meet.2011.14504801166</bibtext> </blist> <blist> <bibl id="bib3" idref="ref25" type="bt">3</bibl> <bibtext> Ahmed W, van der Werf G, Kuyper H, Minnaert A. Emotions, self-regulated learning, and achievement in mathematics: A growth curve analysis. Journal of Educational Psychology. 2013; 105; 1: 150-161. 10.1037/a0030160</bibtext> </blist> <blist> <bibl id="bib4" idref="ref160" type="bt">4</bibl> <bibtext> Alonso-Martin F, Malfaz M, Sequeira J, Gorostiza JF, Salichs MA. A multimodal emotion detection system during human–robot interaction. Sensors. 2013; 13; 11: 15549-15581. 10.3390/s131115549</bibtext> </blist> <blist> <bibl id="bib5" idref="ref107" type="bt">5</bibl> <bibtext> American Diabetes Association. Screening for diabetes. Diabetes Care. 2002; 25; 1: 21-24. 10.2337/diacare.25.2007.S21</bibtext> </blist> <blist> <bibl id="bib6" idref="ref4" type="bt">6</bibl> <bibtext> Artino ARJ, Hemmer PA, Durning SJ. Using self-regulated learning theory to understand the beliefs, emotions, and behaviors of struggling medical students. Academic Medicine. 2011; 86; 10: 35-38. 10.1097/ACM.0b013e31822a603d</bibtext> </blist> <blist> <bibl id="bib7" idref="ref65" type="bt">7</bibl> <bibtext> Aspinwall LG. Rethinking the role of positive affect in self-regulation. Motivation and Emotion. 1998; 22; 1: 1-32. 10.1023/A:1023080224401</bibtext> </blist> <blist> <bibl id="bib8" idref="ref113" type="bt">8</bibl> <bibtext> Brondino M, Raccanello D, Pasini MMascio T, Gennari R, Vitorini P, Vicari R, de la Prieta F. Achievement goals as antecedents of achievement emotions: The 3 X 2 achievement goal model as a framework for learning environments design. Methodologies and intelligent systems for technology enhanced learning. 2014; Springer: 53-60. 10.1007/978-3-319-07698-0_7</bibtext> </blist> <blist> <bibl id="bib9" idref="ref106" type="bt">9</bibl> <bibtext> Ciechanowski PS, Katon WJ, Russo JE, Hirsch IB. The relationship of depressive symptoms to symptom reporting, self-care and glucose control in diabetes. General Hospital Psychiatry. 2003; 25; 4: 246-252. 10.1016/S0163-8343(03)00055-0</bibtext> </blist> <blist> <bibtext> Curran PJ, Obeidat K, Losardo D. Twelve frequently asked questions about growth curve modeling. Journal of Cognition and Development. 2010; 11; 2: 121-136. 10.1080/15248371003699969</bibtext> </blist> <blist> <bibtext> Daniels LM, Stupnisky RH, Pekrun R, Haynes TL, Perry RP, Newall NE. A longitudinal analysis of achievement goals: From affective antecedents to emotional effects and achievement outcomes. Journal of Educational Psychology. 2009; 101; 4: 948-963. 10.1037/a0016096</bibtext> </blist> <blist> <bibtext> D'Mello SSeel NM. Monitoring affective trajectories during complex learning. Encyclopedia of the Sciences of Learning. 2012; Springer: 2325-2328. 10.1007/978-1-4419-1428-6_849</bibtext> </blist> <blist> <bibtext> D'Mello S. A selective meta-analysis on the relative incidence of discrete affective states during learning with technology. Journal of Educational Psychology. 2013; 105; 4: 1082. 10.1037/a0032674</bibtext> </blist> <blist> <bibtext> D'Mello S, Graesser A. Dynamics of affective states during complex learning. Learning and Instruction. 2012; 22; 2: 145-157. 10.1016/j.learninstruc.2011.10.001</bibtext> </blist> <blist> <bibtext> D'Mello S, Lehman B, Pekrun R, Graesser A. Confusion can be beneficial for learning. Learning and Instruction. 2014; 29: 153-170. 10.1016/j.learninstruc.2012.05.003</bibtext> </blist> <blist> <bibtext> Duffy MC, Lajoie SP, Pekrun R, Lachapelle K. Emotions in medical education: Examining the validity of the Medical Emotion Scale (MES) across authentic medical learning environments. Learning and Instruction. 2018; 70. 10.1016/j.learninstruc.2018.07.001</bibtext> </blist> <blist> <bibtext> Efklides A. Interactions of metacognition with motivation and affect in self-regulated learning: The MASRL Model. Educational Psychologist. 2011; 46; 1: 6-25. 10.1080/00461520.2011.538645</bibtext> </blist> <blist> <bibtext> Efklides A. Affect, epistemic emotions, metacognition, and self-Regulated learning. Teachers College Record. 2017; 119; 13: 1-13. 10.1177/016146811711901302</bibtext> </blist> <blist> <bibtext> Fernando JW, Kashima Y, Laham SM. Multiple emotions: A person-centered approach to the relationship between intergroup emotion and action orientation. Emotion. 2014; 14; 4: 722-732. 10.1037/a0036103</bibtext> </blist> <blist> <bibtext> Foster MI, Keane MT. Why some surprises are more surprising than others: Surprise as a metacognitive sense of explanatory difficulty. Cognitive Psychology. 2015; 81: 74-116. 10.1016/j.cogpsych.2015.08.004</bibtext> </blist> <blist> <bibtext> Frenzel AC, Pekrun R, Goetz T. Girls and mathematics—A "hopeless" issue? A control-value approach to gender differences in emotions towards mathematics. European Journal of Psychology of Education. 2007; 22; 4: 497. 10.1007/BF03173468</bibtext> </blist> <blist> <bibtext> Ganotice FA, Datu JAD, King RB. Which emotional profiles exhibit the best learning outcomes? A person-centered analysis of students' academic emotions. School Psychology International. 2016; 37; 5: 498-518. 10.1177/0143034316660147</bibtext> </blist> <blist> <bibtext> Goetz T, Becker ES, Bieg M, Keller MM, Frenzel AC, Hall NC. The glass half empty: How emotional exhaustion affects the state-trait discrepancy in self-reports of teaching emotions. PLoS ONE. 2015; 10; 9: 1-14. 10.1371/journal.pone.0137441</bibtext> </blist> <blist> <bibtext> Greene, J. A, & Schunk, D. H. (2018). Historical, contemporary, and future perspectives on self-regulated learning and performance. In D. Schunk & J.A. Greene (Eds.), Handbook of self-regulation of learning and performance (pp. 18–32). Routledge.</bibtext> </blist> <blist> <bibtext> Hesser H. Modeling individual differences in randomized experiments using growth models: Recommendations for design, statistical analysis and reporting of results of internet interventions. Internet Interventions. 2015; 2; 2: 110-120. 10.1016/j.invent.2015.02.003</bibtext> </blist> <blist> <bibtext> Hoogland JJ, Boomsma A. Robustness studies in covariance structure modeling: An overview and a meta-analysis. Sociological Methods & Research. 1998; 26; 3: 329-367. 10.1177/0049124198026003003</bibtext> </blist> <blist> <bibtext> Isen AMManstead ASR, Frijda N, Fischer A. Some perspectives on positive feelings and emotions: Positive affect facilitates thinking and problem-solving. Feelings and emotions: The Amsterdam symposium. 2004; Cambridge University Press: 263-281. 10.1017/CBO9780511806582.016</bibtext> </blist> <blist> <bibtext> Isen AM, Rosenzweig AS, Young MJ. The influence of positive affect on cinical problem solving. Medical Decision Making. 1991; 11; 3: 221-227. 10.1177/0272989X9101100313</bibtext> </blist> <blist> <bibtext> Jarrell, A. (2018). The regulation of achievement emotions in higher education: An examination of emotion (Publication No. 10169573). [Doctoral dissertation, McGill University]. ProQuest Dissertations & Theses Global.</bibtext> </blist> <blist> <bibtext> Jarrell A, Harley JM, Lajoie SP. The link between achievement emotions, appraisals, and task performance: Pedagogical considerations for emotions in CBLEs. Journal of Computers in Education. 2016; 3; 3: 289-307. 10.1007/s40692-016-0064-3</bibtext> </blist> <blist> <bibtext> Jarrell A, Harley JM, Lajoie S, Naismith L. Success, failure and emotions: Examining the relationship between performance feedback and emotions in diagnostic reasoning. Educational Technology Research and Development. 2017; 65; 5: 1263-1284. 10.1007/s11423-017-9521-6</bibtext> </blist> <blist> <bibtext> Keeley, J, Zayac, R, & Correia, C. (2008). Curvilinear relationships between statistics anxiety and performance among undergraduate students: Evidence for optimal anxiety. Statistics Education Research Journal, 7(1), 4–15. Scopus.</bibtext> </blist> <blist> <bibtext> Kim, S, Georgiou, P. G, Lee, S, & Narayanan, S. (2007). Real-time emotion detection system using speech: Multi-modal fusion of different timescale features. In 2007 IEEE 9th Workshop on Multimedia Signal Processing (pp. 48–51). IEEE.</bibtext> </blist> <blist> <bibtext> Kline RB. Principles and practice of structural equation modeling. 20154; Guilford publications</bibtext> </blist> <blist> <bibtext> Kuiper, R. A. (2013). Integration of innovative clinical reasoning pedagogies into a baccalaureate nursing curriculum. Creative Nursing, 19(3), 128–139.</bibtext> </blist> <blist> <bibtext> Laffel L. Ketone bodies: A review of physiology, pathophysiology and application of monitoring to diabetes. Diabetes/metabolism Research and Reviews. 1999; 15; 6: 412-426. 10.1002/(SICI)1520-7560(199911/12)15:6<412:AID-DMRR72>3.0.CO;2-8</bibtext> </blist> <blist> <bibtext> Lajoie SPEricsson KA. Developing professional expertise with a cognitive apprenticeship model: Examples from avionics and medicine. Development of professional expertise: Toward measurement of expert performance and design of optimal learning environments. 2009; Cambridge University Press: 61-83. 10.1017/CBO9780511609817.004</bibtext> </blist> <blist> <bibtext> Lajoie SP, Zheng J, Li S. Examining the role of self-regulation and emotion in clinical reasoning: Implications for developing expertise. Medical Teacher. 2018; 40; 8: 842-844. 10.1080/0142159X.2018.1484084</bibtext> </blist> <blist> <bibtext> Lajoie SP, Zheng J, Li S, Jarrell A, Gube M. Examining the interplay of affect and self regulation in the context of clinical reasoning. Learning and Instruction. 2019; 72. 10.1016/j.learninstruc.2019.101219</bibtext> </blist> <blist> <bibtext> Little RJ. A test of missing completely at random for multivariate data with missing values. Journal of the American Statistical Association. 1988; 83; 404: 1198-1202. 10.1080/01621459.1988.10478722</bibtext> </blist> <blist> <bibtext> Li, S, Zheng, J, & Lajoie, S. P. (2021a). The frequency of emotions and emotion variability in self-regulated learning: What matters to task performance?. Frontline Learning Research, 9(4), 76–91.</bibtext> </blist> <blist> <bibtext> Li Shan, Zheng Juan, Lajoie Susanne P, Wiseman Jeffrey. Examining the relationship between emotion variability, self-regulated learning, and task performance in an intelligent tutoring system. Educational Technology Research and Development. 2021; 69; 2: 673-692. 10.1007/s11423-021-09980-9</bibtext> </blist> <blist> <bibtext> Loderer K, Pekrun R, Lester JC. Beyond cold technology: A systematic review and meta-analysis on emotions in technology-based learning environments. Learning and Instruction. 2020; 70. 10.1016/j.learninstruc.2018.08.002</bibtext> </blist> <blist> <bibtext> Martinent G, Gareau A, Lienhart N, Nicaise V, Guillet-Descas E. Emotion profiles and their motivational antecedents among adolescent athletes in intensive training settings. Psychology of Sport and Exercise. 2018; 35: 198-206. 10.1016/j.psychsport.2018.01.001</bibtext> </blist> <blist> <bibtext> McConnell MM, Monteiro S, Pottruff MM, Neville A, Norman GR, Eva KW, Kulasegaram K. The impact of emotion on learners' application of basic science principles to novel problems. Academic Medicine. 2016; 91; 11: 58-63. 10.1097/ACM.0000000000001360</bibtext> </blist> <blist> <bibtext> Mega C, Ronconi L, De Beni R. What makes a good student? How emotions, self-regulated learning, and motivation contribute to academic achievement. Journal of Educational Psychology. 2014; 106; 1: 121-131. 10.1037/a0033546</bibtext> </blist> <blist> <bibtext> Menart-Houtermans B, Rütter R, Nowotny B, Rosenbauer J, Koliaki C, Kahl S, Simon M-C, Szendroedi J, Schloot NC, Roden M. Leukocyte profiles differ between Type 1 and Type 2 diabetes and are associated with metabolic phenotypes: Results from the German Diabetes study (GDS). Diabetes Care. 2014; 37; 8: 2326-2333. 10.2337/dc14-0316</bibtext> </blist> <blist> <bibtext> Muis KR, Ranellucci J, Trevors G, Duffy MC. The effects of technology-mediated immediate feedback on kindergarten students' attitudes, emotions, engagement and learning outcomes during literacy skills development. Learning and Instruction. 2015; 38: 1-13. 10.1016/j.learninstruc.2015.02.001</bibtext> </blist> <blist> <bibtext> Muis KR, Chevrier M, Singh CA. The role of epistemic emotions in personal epistemology and self-regulated learning. Educational Psychologist. 2018; 53; 3: 165-184. 10.1080/00461520.2017.1421465</bibtext> </blist> <blist> <bibtext> Muis KR, Pekrun R, Sinatra GM, Azevedo R, Trevors G, Meier E, Heddy BC. The curious case of climate change: Testing a theoretical model of epistemic beliefs, epistemic emotions, and complex learning. Learning and Instruction. 2015; 39: 168-183. 10.1016/j.learninstruc.2015.06.003</bibtext> </blist> <blist> <bibtext> Munzar B, Muis KR, Denton CA, Losenno K. Elementary students' cognitive and affective responses to impasses during mathematics problem solving. Journal of Educational Psychology. 2020; 113; 1: 104-124. 10.1037/edu0000460</bibtext> </blist> <blist> <bibtext> Muthén, L. K, & Muthén, B. O. (2012). MPlus: Statistical analysis with latent variables—User's guide.</bibtext> </blist> <blist> <bibtext> Naismith LM, Lajoie SP. Motivation and emotion predict medical students' attention to computer-based feedback. Advances in Health Sciences Education. 2018; 23; 3: 465-485. 10.1007/s10459-017-9806-x</bibtext> </blist> <blist> <bibtext> Nerantzaki K, Efklides A. Epistemic emotions: Interrelationships and changes during task processing. Hellenic Journal of Psychology. 2019; 16; 2: 177-199</bibtext> </blist> <blist> <bibtext> Niculescu AC, Tempelaar DT, Dailey-Hebert A, Segers M, Gijselaers WH. Extending the change–change model of achievement emotions: The inclusion of negative learning emotions. Learning and Individual Differences. 2016; 47: 289-297. 10.1016/j.lindif.2015.12.015</bibtext> </blist> <blist> <bibtext> Nylund KL, Asparouhov T, Muthén BO. Deciding on the number of classes in latent class analysis and growth mixture modeling: A Monte Carlo simulation study. Structural Equation Modeling: A Multidisciplinary Journal. 2007; 14; 4: 535-569. 10.1080/10705510701575396</bibtext> </blist> <blist> <bibtext> Panadero E. A review of self-regulated learning: Six models and four directions for research. Frontiers in Psychology. 2017; 8; 422: 1-28. 10.3389/fpsyg.2017.00422</bibtext> </blist> <blist> <bibtext> Pekrun R. The control-value theory of achievement emotions: Assumptions, corollaries, and implications for educational research and practice. Educational Psychology Review. 2006; 18; 4: 315-341. 10.1007/s10648-006-9029-9</bibtext> </blist> <blist> <bibtext> Pekrun R, Goetz T, Titz W, Perry RP. Academic emotions in students' self-regulated learning and achievement: A program of qualitative and quantitative Research. Educational Psychologist. 2002; 37; 2: 91-105. 10.1207/S15326985EP3702_4</bibtext> </blist> <blist> <bibtext> Pekrun R, Perry RPPekrun R, Linnerbrink-Garcia L. Control-value theory of achievement emotions. International handbook of emotions in education. 2014; Taylor & Francis: 120-141. 10.4324/9780203148211.ch7</bibtext> </blist> <blist> <bibtext> Pekrun, R, & Stephens, E. J. (2012). Academic emotions. In K. Harris, S. Graham, T. Urdan, S. Graham, J. Royer, & M. Zeidner (Eds.), Individual differences and cultural and contextual factors. APA educational psychology handbook, Vol. 2 (pp. 3–31). American Psychological Association. https://doi.org/10.1037/13274-001</bibtext> </blist> <blist> <bibtext> Pekrun R, Vogl E, Muis KR, Sinatra GM. Measuring emotions during epistemic activities: The Epistemically-Related Emotion Scales. Cognition and Emotion. 2017; 31; 6: 1268-1276. 10.1080/02699931.2016.1204989</bibtext> </blist> <blist> <bibtext> Peterson ER, Brown GTL, Jun MC. Achievement emotions in higher education: A diary study exploring emotions across an assessment event. Contemporary Educational Psychology. 2015; 42: 82-96. 10.1016/j.cedpsych.2015.05.002</bibtext> </blist> <blist> <bibtext> Plutchik Robert. The Nature of Emotions. American Scientist. 2001; 89; 4: 344. 10.1511/2001.4.344</bibtext> </blist> <blist> <bibtext> Putwain DW, Becker S, Symes W, Pekrun R. Reciprocal relations between students' academic enjoyment, boredom, and achievement over time. Learning and Instruction. 2018; 54: 73-81. 10.1016/j.learninstruc.2017.08.004</bibtext> </blist> <blist> <bibtext> Robinson KA, Ranellucci J, Lee Y, Wormington SV, Roseth CJ, Linnenbrink-Garcia L. Affective profiles and academic success in a college science course. Contemporary Educational Psychology. 2017; 51: 209-221. 10.1016/j.cedpsych.2017.08.004</bibtext> </blist> <blist> <bibtext> Robinson MD, Clore GL. Belief and feeling: Evidence for an accessibility model of emotional self-report. Psychological Bulletin. 2002; 128; 6: 934-960. 10.1037//0033-2909.128.6.934</bibtext> </blist> <blist> <bibtext> Raccanello D, Brondino M, Crane M, Pasini MCaporuscio M, De la Prieta F, Di Mascio T, Gennari R, Gutierrez Rodriguez J, Vittorini P. Antecedents of achievement emotions: Mixed-device assessment with Italian and Australian university students. Methodologies and intelligent systems for technology enhanced learning. 2016; Springer: 183-191</bibtext> </blist> <blist> <bibtext> Scherer, K. R. (2009). The dynamic architecture of emotion: Evidence for the component process model. Cognition and Emotion, 23(7), 1307–1351.</bibtext> </blist> <blist> <bibtext> Schmitz B, Klug J, Schmidt MZimmerman B, Schunk D. Assessing self-regulated learning using diary measures with university students. Handbook of self-regulation of learning and performance. 2011; Routledge New York: 251-266</bibtext> </blist> <blist> <bibtext> Silvia PJ. Confusion and interest: The role of knowledge emotions in aesthetic experience. Psychology of Aesthetics, Creativity, and the Arts. 2010; 4; 2: 75-80. 10.1037/a0017081</bibtext> </blist> <blist> <bibtext> Sinclair, J, Jang, E. E, Azevedo, R, Lau, C, Taub, M, & Mudrick, N. V. (2018). Changes in emotion and their relationship with learning gains in the context of MetaTutor. In R. Nkambou, R. Azevedo, & J. Vassileva (Eds.), Intelligent tutoring systems (pp. 202–211). Springer International Publishing. https://doi.org/10.1007/978-3-319-91464-0_20</bibtext> </blist> <blist> <bibtext> Singer JD, Willett JB. Applied longitudinal data analysis: Modeling change and event occurrence. 2003; Oxford University Press. 10.1093/acprof:oso/9780195152968.001.0001</bibtext> </blist> <blist> <bibtext> Taub M, Sawyer R, Lester J, Azevedo R. The impact of contextualized emotions on self-regulated learning and scientific reasoning during learning with a game-based learning environment. International Journal of Artificial Intelligence in Education. 2020; 30; 1: 97-120. 10.1007/s40593-019-00191-1</bibtext> </blist> <blist> <bibtext> Timmermans T, Mechelen IV, Kuppens P. The relationship between individual differ- ences in intraindividual variability in core affect and interpersonal behaviour. European Journal of Personality. 2010; 24; 8: 623-638. 10.1002/per.756</bibtext> </blist> <blist> <bibtext> Turner, J. R, & Gellman, M. (2013). Encyclopedia of Behavioral Medicine. Springer Science+Business Media. https://doi.org/10.1007/978-1-4419-1005-9</bibtext> </blist> <blist> <bibtext> Vogl E, Pekrun R, Murayama K, Loderer K, Schubert S. Surprise, curiosity, and confusion promote knowledge exploration: Evidence for robust effects of epistemic emotions. Frontiers in Psychology. 2019; 10: 1-16. 10.3389/fpsyg.2019.02474</bibtext> </blist> <blist> <bibtext> Vogl E, Pekrun R, Murayama K, Loderer K. Surprised–curious–confused: Epistemic emotions and knowledge exploration. Emotion. 2020; 20; 4: 625-641. 10.1037/emo0000578</bibtext> </blist> <blist> <bibtext> Wang J, Wang X. Structural equation modeling: Applications using Mplus. 20202; John Wiley & Sons Ltd</bibtext> </blist> <blist> <bibtext> Zheng, J, Huang, L, Li, S, Lajoie, S. P, Chen, Y, & Hmelo-Silver, C. E. (2021). Self-regulation and emotion matter: A case study of instructor interactions with a learning analytics dashboard. Computers & Education, 161, 104061.</bibtext> </blist> <blist> <bibtext> Zimmerman BJ. Self-regulated learning and academic achievement: An overview. Educational Psychologist. 1990; 25; 1: 3-17. 10.1207/s15326985ep2501_2</bibtext> </blist> <blist> <bibtext> Zimmerman, B. J. (2000). Attaining self-regulation: A social cognitive perspective. In M. Boekaerts, P. R. Pintrich, & M. Zeidner (Eds.), Handbook of self-Regulation (pp. 13–39). Academic Press. https://doi.org/10.1016/B978-012109890-2/50031-7</bibtext> </blist> <blist> <bibtext> Zimmerman BJ. Becoming a self-regulated learner: An overview. Theory into Practice. 2002; 41; 2: 64-70. 10.1207/s15430421tip4102_2</bibtext> </blist> </ref> <aug> <p>By Juan Zheng; Susanne P. Lajoie; Shan Li and Hongbin Wu</p> <p>Reported by Author; Author; Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib35" firstref="ref1"></nolink> <nolink nlid="nl2" bibid="bib28" firstref="ref2"></nolink> <nolink nlid="nl3" bibid="bib45" firstref="ref3"></nolink> <nolink nlid="nl4" bibid="bib39" firstref="ref5"></nolink> <nolink nlid="nl5" bibid="bib57" firstref="ref6"></nolink> <nolink nlid="nl6" bibid="bib59" firstref="ref7"></nolink> <nolink nlid="nl7" bibid="bib46" firstref="ref8"></nolink> <nolink nlid="nl8" bibid="bib74" firstref="ref9"></nolink> <nolink nlid="nl9" bibid="bib42" firstref="ref14"></nolink> <nolink nlid="nl10" bibid="bib17" firstref="ref16"></nolink> <nolink nlid="nl11" bibid="bib49" firstref="ref17"></nolink> <nolink nlid="nl12" bibid="bib58" firstref="ref18"></nolink> <nolink nlid="nl13" bibid="bib12" firstref="ref19"></nolink> <nolink nlid="nl14" bibid="bib64" firstref="ref20"></nolink> <nolink nlid="nl15" bibid="bib69" firstref="ref22"></nolink> <nolink nlid="nl16" bibid="bib23" firstref="ref24"></nolink> <nolink nlid="nl17" bibid="bib67" firstref="ref26"></nolink> <nolink nlid="nl18" bibid="bib61" firstref="ref27"></nolink> <nolink nlid="nl19" bibid="bib29" firstref="ref29"></nolink> <nolink nlid="nl20" bibid="bib62" firstref="ref30"></nolink> <nolink nlid="nl21" bibid="bib48" firstref="ref31"></nolink> <nolink nlid="nl22" bibid="bib50" firstref="ref32"></nolink> <nolink nlid="nl23" bibid="bib14" firstref="ref33"></nolink> <nolink nlid="nl24" bibid="bib15" firstref="ref34"></nolink> <nolink nlid="nl25" bibid="bib43" firstref="ref36"></nolink> <nolink nlid="nl26" bibid="bib13" firstref="ref37"></nolink> <nolink nlid="nl27" bibid="bib16" firstref="ref40"></nolink> <nolink nlid="nl28" bibid="bib81" firstref="ref41"></nolink> <nolink nlid="nl29" bibid="bib24" firstref="ref42"></nolink> <nolink nlid="nl30" bibid="bib82" firstref="ref44"></nolink> <nolink nlid="nl31" bibid="bib83" firstref="ref51"></nolink> <nolink nlid="nl32" bibid="bib63" firstref="ref58"></nolink> <nolink nlid="nl33" bibid="bib55" firstref="ref60"></nolink> <nolink nlid="nl34" bibid="bib38" firstref="ref63"></nolink> <nolink nlid="nl35" bibid="bib60" firstref="ref66"></nolink> <nolink nlid="nl36" bibid="bib78" firstref="ref67"></nolink> <nolink nlid="nl37" bibid="bib51" firstref="ref72"></nolink> <nolink nlid="nl38" bibid="bib77" firstref="ref73"></nolink> <nolink nlid="nl39" bibid="bib11" firstref="ref78"></nolink> <nolink nlid="nl40" bibid="bib65" firstref="ref79"></nolink> <nolink nlid="nl41" bibid="bib22" firstref="ref82"></nolink> <nolink nlid="nl42" bibid="bib19" firstref="ref83"></nolink> <nolink nlid="nl43" bibid="bib44" firstref="ref84"></nolink> <nolink nlid="nl44" bibid="bib32" firstref="ref85"></nolink> <nolink nlid="nl45" bibid="bib66" firstref="ref86"></nolink> <nolink nlid="nl46" bibid="bib30" firstref="ref89"></nolink> <nolink nlid="nl47" bibid="bib31" firstref="ref90"></nolink> <nolink nlid="nl48" bibid="bib75" firstref="ref100"></nolink> <nolink nlid="nl49" bibid="bib40" firstref="ref101"></nolink> <nolink nlid="nl50" bibid="bib76" firstref="ref103"></nolink> <nolink nlid="nl51" bibid="bib37" firstref="ref104"></nolink> <nolink nlid="nl52" bibid="bib47" firstref="ref105"></nolink> <nolink nlid="nl53" bibid="bib36" firstref="ref108"></nolink> <nolink nlid="nl54" bibid="bib68" firstref="ref112"></nolink> <nolink nlid="nl55" bibid="bib53" firstref="ref120"></nolink> <nolink nlid="nl56" bibid="bib52" firstref="ref123"></nolink> <nolink nlid="nl57" bibid="bib79" firstref="ref125"></nolink> <nolink nlid="nl58" bibid="bib26" firstref="ref126"></nolink> <nolink nlid="nl59" bibid="bib34" firstref="ref127"></nolink> <nolink nlid="nl60" bibid="bib25" firstref="ref129"></nolink> <nolink nlid="nl61" bibid="bib56" firstref="ref130"></nolink> <nolink nlid="nl62" bibid="bib54" firstref="ref133"></nolink> <nolink nlid="nl63" bibid="bib20" firstref="ref134"></nolink> <nolink nlid="nl64" bibid="bib71" firstref="ref136"></nolink> <nolink nlid="nl65" bibid="bib21" firstref="ref145"></nolink> <nolink nlid="nl66" bibid="bib27" firstref="ref147"></nolink> <nolink nlid="nl67" bibid="bib18" firstref="ref149"></nolink> <nolink nlid="nl68" bibid="bib72" firstref="ref153"></nolink> <nolink nlid="nl69" bibid="bib70" firstref="ref156"></nolink> <nolink nlid="nl70" bibid="bib41" firstref="ref158"></nolink> <nolink nlid="nl71" bibid="bib80" firstref="ref159"></nolink> <nolink nlid="nl72" bibid="bib33" firstref="ref161"></nolink> <nolink nlid="nl73" bibid="bib10" firstref="ref162"></nolink> <nolink nlid="nl74" bibid="bib73" firstref="ref163"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1386273
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Temporal Change of Emotions: Identifying Academic Emotion Trajectories and Profiles in Problem-Solving
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Zheng%2C+Juan%22">Zheng, Juan</searchLink> (ORCID <externalLink term="http://orcid.org/0000-0002-4896-8198">0000-0002-4896-8198</externalLink>)<br /><searchLink fieldCode="AR" term="%22Lajoie%2C+Susanne+P%2E%22">Lajoie, Susanne P.</searchLink><br /><searchLink fieldCode="AR" term="%22Li%2C+Shan%22">Li, Shan</searchLink><br /><searchLink fieldCode="AR" term="%22Wu%2C+Hongbin%22">Wu, Hongbin</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Metacognition+and+Learning%22"><i>Metacognition and Learning</i></searchLink>. Aug 2023 18(2):315-345.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Springer. Available from: Springer Nature. One New York Plaza, Suite 4600, New York, NY 10004. Tel: 800-777-4643; Tel: 212-460-1500; Fax: 212-460-1700; e-mail: customerservice@springernature.com; Web site: https://link.springer.com/
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 31
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2023
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Audience
  Label: Education Level
  Group: Audnce
  Data: <searchLink fieldCode="EL" term="%22Higher+Education%22">Higher Education</searchLink><br /><searchLink fieldCode="EL" term="%22Postsecondary+Education%22">Postsecondary Education</searchLink>
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Independent+Study%22">Independent Study</searchLink><br /><searchLink fieldCode="DE" term="%22Psychological+Patterns%22">Psychological Patterns</searchLink><br /><searchLink fieldCode="DE" term="%22Problem+Solving%22">Problem Solving</searchLink><br /><searchLink fieldCode="DE" term="%22Learning+Processes%22">Learning Processes</searchLink><br /><searchLink fieldCode="DE" term="%22Reflection%22">Reflection</searchLink><br /><searchLink fieldCode="DE" term="%22Performance%22">Performance</searchLink><br /><searchLink fieldCode="DE" term="%22Simulated+Environment%22">Simulated Environment</searchLink><br /><searchLink fieldCode="DE" term="%22Medical+Students%22">Medical Students</searchLink><br /><searchLink fieldCode="DE" term="%22Clinical+Diagnosis%22">Clinical Diagnosis</searchLink><br /><searchLink fieldCode="DE" term="%22Personality+Traits%22">Personality Traits</searchLink><br /><searchLink fieldCode="DE" term="%22Student+Interests%22">Student Interests</searchLink><br /><searchLink fieldCode="DE" term="%22Time+Factors+%28Learning%29%22">Time Factors (Learning)</searchLink><br /><searchLink fieldCode="DE" term="%22Self+Control%22">Self Control</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1007/s11409-022-09330-x
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 1556-1623<br />1556-1631
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Academic emotions play an important and complex role in self-regulated learning (SRL). However, few studies have examined how academic emotions unfold in different phases of SRL and how the changes in these emotions influence learning performance. The current study examines 98 students' academic emotion trajectories and profiles across the three phases of SRL (i.e., forethought, performance, and self-reflection) as they solve a clinical problem in BioWorld. Specifically, BioWorld is a simulated learning environment where medical students are tasked with diagnosing virtual patient diseases. We identified the three phases of SRL based on students' problem-solving behaviors and we asked students to self-report their achievement and epistemic emotions at the end of each phase of SRL. The growth curve model results showed that curiosity and confusion declined across the three phases of SRL, whereas boredom increased in the self-reflection phase of SRL. The initial levels of curiosity and enjoyment positively predicted students' performance. Latent transition analysis revealed three emotion profiles: curious-positive, confused-negative, and medium-low. Curious-positive students maintained a relatively stable profile through the SRL phases, whereas students in the confused-negative and medium-low groups exhibited specific transition patterns in their emotions. This study makes theoretical contributions by highlighting the temporal and dynamic nature of emotions in problem-solving. Findings from this study have educational implications regarding the role of specific emotions in learning, the development of one's awareness of their emotions, and emotion regulation.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2023
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1386273
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1386273
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1007/s11409-022-09330-x
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 31
        StartPage: 315
    Subjects:
      – SubjectFull: Independent Study
        Type: general
      – SubjectFull: Psychological Patterns
        Type: general
      – SubjectFull: Problem Solving
        Type: general
      – SubjectFull: Learning Processes
        Type: general
      – SubjectFull: Reflection
        Type: general
      – SubjectFull: Performance
        Type: general
      – SubjectFull: Simulated Environment
        Type: general
      – SubjectFull: Medical Students
        Type: general
      – SubjectFull: Clinical Diagnosis
        Type: general
      – SubjectFull: Personality Traits
        Type: general
      – SubjectFull: Student Interests
        Type: general
      – SubjectFull: Time Factors (Learning)
        Type: general
      – SubjectFull: Self Control
        Type: general
    Titles:
      – TitleFull: Temporal Change of Emotions: Identifying Academic Emotion Trajectories and Profiles in Problem-Solving
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Zheng, Juan
      – PersonEntity:
          Name:
            NameFull: Lajoie, Susanne P.
      – PersonEntity:
          Name:
            NameFull: Li, Shan
      – PersonEntity:
          Name:
            NameFull: Wu, Hongbin
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 08
              Type: published
              Y: 2023
          Identifiers:
            – Type: issn-print
              Value: 1556-1623
            – Type: issn-electronic
              Value: 1556-1631
          Numbering:
            – Type: volume
              Value: 18
            – Type: issue
              Value: 2
          Titles:
            – TitleFull: Metacognition and Learning
              Type: main
ResultId 1