How Do Students Understand the Concept of “Curiosity?” Critical Implications for Research From a Multiple Methods Investigation

Although “validity” is not a concern of many qualitative approaches to inquiry outside of the post-positivist tradition (see Schwandt et al., 2007), systematicity, rigor, and transparency are key quality indicators for these projects. Merriam (1998) writes that an assumption of non-positivistic work using qualitative methods “is that reality is holistic, multidimensional, and ever-changing; it is not a single, fixed, objective phenomenon waiting to be discovered, observed, and measured” (p. 202). Merriam offers several strategies for demonstrating credibility of analysis; for example, participatory research and disclosure of researcher biases. This study was conducted at a school with which the second author has a long-standing, mutually respectful and beneficial partnership. The larger project and other collaborative activities are informed (and, in many cases, directed) by the school’s administrators and participating teachers (e.g., Louick & Emery, 2024; Louick & Wang, 2021; Wang & Louick, 2020).

Merriam also suggests that researchers account for the lenses they bring to data collection, analysis, and interpretation, which we (Emery and Louick) will do in several of the following sections via positionality statements and transparent descriptions of the axiology informing our results. Further, Merriam suggests that researchers provide an audit trail. Following below, we present exemplars of our coding techniques in text and tables, and present data from participants in our Results section (additional, anonymized data are available upon request). We triangulated our findings with the use of multiple interviews and the classroom observations we conducted. Interviewing the participants on multiple occasions enabled us to see whether there were consistencies between the participants’ statements at different times; observing in the classrooms provided an opportunity to see if the interview statements were borne out in classroom actions. We also engaged in a round of coding in which we specifically looked for counterexamples or counterstories. Further, members of the school team involved in the larger project have seen versions of these analyses prior to public dissemination, which we present with their approval.

Researchers’ embodied experiences necessarily shape the ways they plan studies, collect data with and from participants, engage in analyses, and interpret data (Boveda & Annamma, 2023; Lauterbach et al., 2022). As such, it is important to note the following: I (Emery) identify as a white-appearing, cisgender, monolingual woman. A former special education teacher, I now work at a public, research-intensive university with a commitment to improving access to resources and opportunities for people in Iowa. I approach research with the belief that the unearned privileges afforded to me—as a function of, for instance, the way I am racialized as white—carry with it responsibilities to (1) name power structures that contribute to the invisibilization of marginalized communities and (2) interrogate how knowledge is processed and produced. I endorse the sociopolitical model of disability, which posits that “disability is not a personal defect or deficiency, but rather primarily the product of a disabling environment” (Hahn, 1991, p. 17). It is my belief that the field of educational psychology has actively harmed the learning experiences of students with disabilities through both a history of eugenics and explicit exclusion in studies that otherwise aim to generate “universal” theories. As such, the focus of my program of research is to consider whether and how current work in the field of educational psychology adequately accounts for the experiences of students with disabilities, and how the scholarly body of empirical evidence related to learning and motivation could be improved with better representation. In this particular study, I operated under the assumption that our participants had unique insights to share, were the experts on their own experiences, and were the population best positioned to help our research team answer the research questions we identified as a gap in the literature.

I (Louick) identify as a White, cisgender and monolingual woman without disabilities, working at a public university, with prior experience as a middle and high school special education teacher. I share in the first author’s beliefs about the ways in which people like myself, who are privileged in terms of race, ethnicity, and disability status, have a critical responsibility to identify power structures and question forms of knowledge. I see merit in both the functional/environmental and the sociopolitical models of disability (Smart & Smart, 2006). I am particularly interested in ways we can improve educational opportunities for adolescents with learning disabilities: both by training teachers to recognize and address motivational challenges, and by developing innovative instructional materials and learning environments.

As an undergraduate student and passionate future educator, I (Stelter) was excited to join the first two authors on this project. As a preservice teacher in the process of earning my special education endorsement, the opportunity to spend time exploring the field of Educational Psychology was a way for me to engage early in professional development, lifelong learning, and my overall commitment to bettering myself as an educator. Both the opportunity to participate in the authorship process and my interest in education more generally were, and will continue to be, influenced by the identities that I hold. As a neurotypical, white, cisgendered male, the rural and suburban schooling systems I attended were, in many ways, designed specifically for someone like me to succeed. I recognize the privilege that a system of that design that affords someone like me, not only to earn a degree in Higher Education, but to navigate the space in a way that allows for the opportunity to participate in a different function of Iowa State University’s purpose.

The data for this study were collected at Alcott School (a pseudonym), an independent, therapeutic day school for students with emotional disabilities and other learning support needs. Alcott boasts an extraordinary 100% graduation rate and a 100% college placement rate, compared to national rates for students with disabilities of 40% and 37.5% respectively. Located in one of the largest cities in the United States, Alcott provides students with small class sizes, and a daily “student prep” period during which they can go to any of their teachers for academic support. So as to facilitate a feeling of community, students refer to their teachers (and all other staff) on a first-name basis. Students also have access to a counseling center available to them on-demand. It is common for students to come and go from the counseling center as needed; teachers prepare for the daily possibility that at least some of their students will need to leave for at least part of the class period. Alcott offers the same slate of courses as might be expected at any other high school, including Advanced Placement courses, but also has the flexibility to offer courses another school cannot. For example, three of the five students who took part in this study were taking a course called “Scientific Literacy,” in which they studied how to distinguish between evidence-based science and misinformation.

In Merriam’s (1998) approach to case studies, participants are purposively sampled, driven by the framing literature and research question. We recruited one math and one science teacher, Matthew and Harrison (pseudonyms), who shared students to participate in a larger project exploring student affect and motivation in math and science learning. The teachers had classes ranging in size from five to 10 students, depending on the level of academic support students required, as well as who was present at school and/or in the counseling center. For this analysis, we purposively sampled students shared by both teachers. Five of their students agreed with parental consent to participate; we call them Irving, Justin, Katherine, Lydia, and Mark in this study (see Table 2 for student demographics). As required for enrollment at Alcott, participating students all had emotional or behavioral disorder diagnoses thus qualifying them for the therapeutic services at Alcott, indicating they have practice talking about their affective experiences. Individual students also disclosed dyscalculia, dyslexia, and ADHD.

Across 1 month (May 9–June 8, 2022), students participated in three interviews lasting between 30 and 60 minutes, conducted during their prep period in a private space that the Alcott administration arranged for the project. All interviews were semi-structured (Miles et al., 2014) and conducted via Zoom. Interview questions were refined during a pilot study conducted at Alcott the prior year, following Merriam’s (1998) guidelines for interview protocols. One member of the research team met with each student individually for each interview, and took an audio recording of the conversation. In our first interview, we asked students questions related to their emotions and motivation during science and math learning (across their lives, but also in Matthew’s and Harrison’s classes, specifically). In the second interview, we engaged the participants in a “think aloud” about common math and science learning activities they experienced in their classrooms, and asked them to describe how they might think, feel, and act during those scenarios. The third interview involved responding to the items from the Learning Climate Questionnaire (Williams & Deci, 1996). We encouraged students to elaborate verbally on the answer options they chose with examples first from their math class with Matthew, and then their science class with Harrison. This interview technique, which can be described as cognitive interview with emergent probes (Beatty & Willis, 2007), is necessary when researchers want to determine participants’ conceptions of survey items (e.g., Karabenick et al., 2007). The students completed the interview protocols in the same order.

To address our third research question (“Are the behaviors students report engaging in when feeling curious observable in the classroom?”), we collected video recordings of students’ math and science classes. The inclusion of classroom observations was particularly important, as it enabled us to determine if students feel curious in compulsory learning situations. Five sessions were audio- and video-recorded for each of the three classes (two sections of Matthew’s math classes that included student study participants, and Harrison’s science class that included those same students) for a total of 15 class videos. The lessons took place between May 24 and June 2, 2022, within the same month that the interviews took place. In all classes, students spent most of the class time working on their final projects. Harrison’s students were engaged in research projects about the facts that could be used to debunk unsubstantiated theories (“Flat Earth,” extrasensory perception, etc.); over this set of 5 days, the students were taking turns presenting their preliminary findings to their classmates and receiving feedback meant to prepare them for their final project. Matthew’s students were working on a final project related to architectural drawing; they used mathematical concepts they had learned throughout the year to design either a dream room (students with intermediate math proficiency) or a dream floor of a building (students with advanced math proficiency). The students in the advanced class spent 2 days finishing up a trigonometry unit, and then moved on to their final project for the remaining three classes.

In all of these classes, final projects were multi-step, ongoing endeavors that had been broken down into concrete steps. Opportunities for feedback from peers and teacher were built into the projects. In Harrison’s class, students were using a rubric to “grade” their classmates’ presentations, and a specific question-and-answer period was built into the end of the presentation time. Opportunities for student choice were also built into all settings. Harrison allowed the students to choose their unsubstantiated theory of choice, and to decide the manner in which they wanted to present their final project (traditional paper, slides, etc.). Matthew allowed the students to choose the space they wanted for their architectural drawing, as well as what pieces of furniture they wanted in their space.

Because of the second author’s long-standing research relationship with the school, the students were accustomed to video-recording taking place in school classrooms; in addition, student assent and parent consent forms assured participants and families that they could opt out of recording at any time. Although some may worry that students changed their behavior due to the presence of video recording equipment, we captured one instance of a student asking which project the camera was for and several instances of students making routine requests of their teachers to access normal school supports and resources, suggesting the students did not feel inhibited by the video recording. Initial transcripts were completed by a transcription service; Louick then went through the transcripts and compared them with the videos themselves, making corrections for accuracy. We did not include exchanges between any non-consented students and consented/assented participants or the teachers in our final transcriptions.

The first and second author began by reading through all of the students’ transcripts by round (e.g., the first interviews we conducted with students, then the second, and then the third). Having familiarized ourselves with the entire data corpus, we focused our analyses on the units of meaning within the students’ interview responses that were related to curiosity and/or interest. These included any time the students used terms such as “curious,” “curiosity,” “interest,” “interested,” or “interesting,” in relation to their current science and math classes, as well as times when they otherwise described instances when they felt curious about something in math or science learning more broadly. As Merriam (1998) describes data analysis as a recursive and interactive process, we met regularly to compare our selections, resolving any discrepancies by consensus to ensure we were conducting subsequent rounds of analyses on exactly the same units of meaning. We then completed this same descriptive coding on the transcripts from our teacher interviews.

In our second round of data analysis, we conducted simultaneous rounds of pattern and process coding. Emery led the pattern coding, pulling together related material into “more meaningful and parsimonious unit[s] of analysis;” (Saldaña, 2013, p. 210). Louick led the process coding (identifying actions with gerunds, or “-ing words,” that students said they engaged in; Charmaz, 2002; Saldaña, 2013). Although we each took the lead on one type of coding, we took supporting roles in the other coding protocol to determine reliability of the codes developed. We met regularly to review and refine codes, repeating our coding protocols until our separate coding efforts resulted in concurrence for both sets of codes. Exemplars for each are presented in Tables 3 (process codes) and 4 (pattern codes).

Although most typically associated with grounded theory, we chose to employ axial coding (Charmaz, 2006) to organize and draw connections between the process and pattern codes that we developed in the previous round of coding. This resulted in five categories. The first included codes related to the role of the teacher in students’ descriptions of feeling curious. The remaining categories were each related to the student experience of feeling curious, and included students describing where knowledge is held, the importance of multiple sources or viewpoints, personal reasons for pursuing resolution to a knowledge gap, and actions taken to pursue deeper or additional knowledge. An exemplar is presented in Table 5.

In our final round of coding interview transcripts, we reviewed the entire data corpus of transcripts explicitly looking for counterexamples or alternative explanations. We also specifically looked for instances in which the students described a lack of curiosity and/or interest.

In the process of analyzing the interview transcripts, we identified actions that individual students claimed to engage in when they were feeling curious about a math- or science-related learning opportunity. Although some actions could not be tracked from our classroom data (e.g., doing independent research at home about a topic that sparked curiosity), others, such as the action of asking questions (which Mark, Justin, and Katherine all said they did when they were curious) were observable in the course of daily classroom activity. Thus, we reviewed the class meeting transcripts to capture whether (and when) students engaged in the observable behaviors they described in their interviews. This additional analysis allowed for possible triangulation of findings from the interview portion of the study, and also highlighted and clarified the difficulty of tracking curiosity-related behaviors in real time. A list of the behaviors we aimed to observe appears in Table 6.

Additional information about the interviews and coding processes not included in this manuscript or accompanying tables are available upon request. This includes de-identified data and the research team’s full codebook. This study was not preregistered.

In our original interview protocol, we were explicit about using the word “curiosity” instead of interest. However, when we asked participants something specific about experiences of curiosity, the students often referenced “interest” or “interesting” in their response. An example of this appears in the following exchange between Justin and a researcher:

Here, we see Justin was asked about times he felt curious, and he responded describing something he thought was “interesting” in science class. In other words, Justin described feelings and experiences that are consistent with an educational psychologist’s conception of “curious,” even though he labeled it as relating to the “interest” construct. In our analyses, we counted 35 instances of students being asked explicitly about curiosity, and responding using the words “interest” or “interesting.”

Given our focus on curiosity at school throughout the interviews, it made sense that students often discussed teachers and their roles in exposing students to content and facilitating their curiosity; as Irving said, “if I wasn’t forced to do it, I wouldn’t get interested in it.” As we looked across the data, we saw three primary ways that students connected their teachers to their experiences of curiosity, all of which are consistent with other findings in the extant literature. The first was connecting material to students’ own experiences; Justin said he experienced this frequently during scientific literacy class, as Harrison ensured that everyone in the class had “a base understanding, and then using that [for students] to know what is and isn’t misinformation.” In other words, Justin felt that Harrison established a common set of knowledge to which they could compare things they had learned or believed previously. The second was focusing on how the content could be applied and used in real life: Lydia said specifically that if the material at hand “is something that I’m able to apply . . . that kind of piques my interest more;” she went on to explain that during trigonometry, “my teacher had brought up how you can apply what we’re learning to . . . find out where people are through cell phone calls.” The third was designing appropriately challenging tasks. For example, when asked how he would feel if he was asked to build a model of a rainforest in class, Mark said, “I’d probably not want to build the model, ’cause that sounds difficult. But I’ll do it because it sounds like an interesting challenge.”

Although Pekrun (2006) does not make specific ontological claims about whether knowledge is pre-existing or is generated, CVT helps explain the dynamic processes in which students engage to assess the perceived control over and valuing for their learning. Our study participants described curiosity with an underlying assumption that if they encountered a gap in their knowledge, they could fill it with knowledge that already existed. In other words, they described curiosity as accessing extant knowledge, rather than generating something entirely new. This reflects the control they felt (e.g., they were capable of accessing knowledge, or they had access to knowledge) and valuing (if the knowledge was something they wanted). One clear example of this came when students spoke of accessing knowledge of a field new to them. Justin talked about how curious he was when Harrison mentioned subjects that he (Justin) had never heard of before. Similarly, Mark and Lydia both talked about the experience of learning that multiple fields existed within psychology, and how learning about these fields made them curious enough that they ended up doing their own research outside of traditional class time. Students also described curiosity to learn how someone else thinks, especially when presented with a point of view they had not considered before. For example, Mark said: “I’d be like ‘Whoa, no way,’ and then I look more into it, and then make my decision on what I think of it later.”

Additionally, students described drawing connections between classes; for example, Justin described how fundamental concepts he learned about scientific literacy applied to other academic subjects. They also described understanding contextual differences in word meaning, as when Irving reported seeking out new definitions for words he heard in unfamiliar contexts. Further, students aimed to situate knowledge within its cultural context, as reflected in Lydia’s curiosity about the history of math and where fundamental math concepts came from.

In line with Pekrun’s (2006) CVT, in which epistemic emotions are thought to arise in situations related to knowledge acquisition, our participants distinguished between pursuing knowledge to fulfill an obligation (e.g., to complete an assigned task) and pursuing knowledge for other reasons. It was the latter that participants described as related to curiosity. Additionally, the students were able to make distinctions between intensities of curiosity, describing being mildly curious or deeply curious for varying reasons. For example, they described being curious about learning for the sake of having fun (e.g., “[Listening to a guest speaker] would be fun . . . because they would have a lot of knowledge about the topic and I would be able to learn from them”), but also for purposes such as knowing more about topics related to personal or professional goals (e.g., “Engineering is something I wanna do eventually, so I love hands-on projects. I would 110% do that [build a model of the rainforest]”).

There were also times when students described curiosity to engage with ideas that affected them or society on a deeper level. Students talked about curiosities related to their own identity beliefs, and Mark went so far as to describe curiosity-related learning as a means of maintaining a peaceful society: “I believe knowledge is power. It’s the most powerful thing on the planet. . . . I believe that learning and intelligence is what keeps us from falling into violence.” For Mark, pursuing resolution to knowledge gaps in topics he was curious about was akin to a moral obligation. Other students also described significant reasons for pursuing their curiosity. For example, Justin described feeling curious as prompting him to “ask questions, even if it’s not necessarily relevant to the [classroom task] itself,” indicating that pursuing a resolution to his knowledge gaps helps him make connections across his world. The value of establishing tangential connections across fields of knowledge came up for Katherine, too, who said,

I feel like being able to discuss questions that are maybe—well, I know that schools have discussions, but you know–being able to discuss freely what you are, like, learning about, and also ask questions about maybe something a little unrelated but also connected, is something that I think is a little undervalued.

For Katherine, asking questions to close initial knowledge gaps helped her determine whether she would continue to pursue knowledge of that topic—stated differently, getting answers to her initial questions about topics that made her feel curious helped her determine her level of investment in pursuing a topic further.

It is worth noting that while students spent a great deal of time discussing their curiosity to fill their own knowledge gaps with existing information, the dataset did not contain any examples of students describing a sense of curiosity to generate something new: that is, something that had never been identified before.

Students reported seeking information about their topics of curiosity. Three students (Mark, Irving, and Katherine) described recording information about which they were curious for the purposes of planning (and ultimately completing) additional research after class. Three students (Justin, Katherine, and Mark) said that they asked follow-up questions of others, with Justin and Katherine specifying teachers, peers, and/or parents. The students varied in what they sought. For example, Lydia specified that she wanted connections between what she felt curious about in class and the outside world, echoing Justin’s goal to understand how the topics he felt curious about in school applied in other contexts as we described above. Mark, by contrast, prioritized gathering multiple perspectives on the topics he felt curious about.

Relatedly, Irving and Katherine talked about the actions they took when they felt curious as a means of spending more time with those ideas. For these students, this meant extending their learning through independent inquiry, outside of class time and using their own resources (e.g., internet sources they evaluated and selected, etc.). This implies that the time our participants spend may not be on gathering additional or extended information, but rather continuing to engage with the original topics.

We first asked, “How do these students, who have emotional and behavioral disabilities, describe their experiences of curiosity?” We chose this group purposively because of their significant practice reflecting on and describing their emotional experiences, deliberately positioning them as a group we can learn from. Perhaps most significantly, the students in our study used the terms “curiosity” and “interest” interchangeably, even though our research team explicitly asked about curiosity. In other work led by Aslan et al. (2021) participants distinguished between curiosity and interest, but they were explicitly prompted to do so by the research team. Further, students in our study only swapped “interest” for curiosity; they did not use other synonyms for curiosity. The authors of prior work on epistemic emotions (e.g., Fayn et al., 2019; Noordewier & Gocłowska, 2024; Vogl et al., 2019, ) have suggested that epistemic emotions such as confusion and surprise are often triggered in the same ways that curiosity is triggered, and can function similarly to curiosity (for example, by prompting knowledge-attaining behaviors). Because our students used the terms “curiosity” and “interest” interchangeably, but did not swap in any other terms (such as “surprise” or “confusion”), we conclude that curiosity and other epistemic emotions seem to be sufficiently distinct for our participants in ways that curiosity and interest are not.

A further implication of the current study is that students can (and, in the case of our participants, do) distinguish between passing fancies and connecting to a higher virtue. Many of the curiosity studies to date, especially those conducted in controlled laboratory environments rather than in classrooms, have asked participants to engage with relatively arbitrary information. What remains to be untangled is whether researchers in lab settings are actually measuring a drive to attain information informed solely by cognitive incongruence, or truly the experience of curiosity (and the relationship/distinction between those two constructs). In other words, greater information is needed about whether the participants themselves would define their experience as one of curiosity if the term were not supplied by the research team. A related empirical question is whether lab-based studies hold the same ecological validity as those conducted with students in their actual classrooms, navigating content that they have a vested stake in learning. The next phase of curiosity research should aim to capture student descriptions of their experiences of curiosity while they are in authentic learning situations that require them to engage with information they either must learn or choose to learn.

Second, we asked, “What do these students describe doing during and after experiencing curiosity?” Some of the current scholarly consensus about curiosity is reflected in our findings regarding students’ self-reported actions during and after experiencing curiosity. For example, students described curiosity as prompting knowledge-seeking behaviors, like doing their own research about topics outside of class. Importantly, students in our study shared that they weigh a drive-to-know against other motivational considerations, such as whether they value the knowledge or it feels worth the time they would need to engage in more research about a topic outside of class. Students discussed valuing knowledge as being clearly connected to out-of-classroom life, practical applications, and their career goals. As such, future work might investigate students’ experiences of curiosity in relation to motivational frameworks like Eccles’s and Wigfield’s situated expectancy-value theory (2020). Although valuing is obviously a central component of Pekrun’s (2006) CVT, the dimensions and types of valuing vary across CVT and expectancy-value frameworks.

Embedded in students’ descriptions of curiosity, we also found an epistemological assumption that the knowledge they sought was “out there” for them to attain. That is, students did not describe feeling curious in terms of needing to generate knowledge. Philosopher Adam Morton’s (2010) suggestions about the nature of curiosity support this claim; he writes, “When you are curious about something, you want to know the truth about it. You don’t just want to have a belief, and perhaps even not just a belief that happens to be true,” (p. 392; emphasis original). This line of thinking suggests that knowledge is an external, objective entity that can be discovered or acquired, rather than something constructed or generated through personal engagement or exploration. Such an argument aligns with the students’ perceptions of curiosity as a process of uncovering pre-existing truths, rather than one of active, creative knowledge generation. According to Morton, the curiosity-driven desire to “know the truth” underscores an assumption that knowledge exists independently of the inquirer, and the pursuit of curiosity is framed as a quest to access or reveal this external truth.

From this, Morton develops the idea of a “professional knower,” describing those who have occupations (e.g., detective, researcher) that require knowledge-seeking, even if the individual may be dispassionate or disinterested about the topic. Morton suggests that such “professional knowers” can engage in knowledge-attainment behaviors without being curious at all, if they have sufficient, non-epistemic reasons to perform them. The fact that the participating students assumed that their role was to learn information that already exists, as opposed to creating new knowledge, suggests that individuals who are students in compulsory schooling may think of themselves as the kind of professional knowers described by Morton. That is, they are “motivated . . . because it is part of the job description,” to attain knowledge, rather than for personal edification or enjoyment (Morton, 2010; p. 13). Professional knowers, Morton argues, are often tasked with finding answers that will please their bosses; in the case of students, perhaps their teachers or caregivers. Curiosity as an activating, intrinsically motivating experience may be tempered for learners required to be both present and learning the content assigned. Researchers should recognize the limitations of data collected in such contexts, regardless of the type of data collection measure used.

Finally, we asked, “Are the behaviors students report engaging in when feeling curious observable in the classroom?” Student interviews revealed that many of the behaviors that students described initiating when they felt curious happened at home. Given that our data source for RQ3 was in-classroom observations, we were unable to track the nature of their outside-of-school responses to curiosity. Further, our classroom observations did not reveal many instances of the other behaviors students described initiating when feeling curious beyond asking questions. This has implications for future research: it will likely require study teams to design protocols for data collection that can happen in students’ homes (e.g., experience sampling methodologies or diary studies). Whereas Hidi and Renninger (2019) claimed that teachers’ attempts to prompt curiosity among their students was not likely to result in “self-initiated information” seeking that leads to “longer term, deep involvement with content,” our findings suggest that instances of curiosity at school do, in fact, lead to self-initiated information seeking, and raise the empirical question of whether and how long students stay engaged with those topics.

Further reflecting on Morton’s (2010) analysis of curiosity, in which he argues that a desire for accuracy is inherent to feeling curious, our results raise concerns about the quality of information and sources students might engage with when exploring ideas at home. Sinatra, Lombardi, and their colleagues (e.g., Herrick et al., 2023; Sinatra & Lombardi, 2020) have detailed both the reach and strength of misinformation posturing as fact, and the skills required of individuals to make plausibility judgments. This also has implications for practice, as supporting and encouraging students’ curiosity will require ensuring that students have the skills needed to critically evaluate the sources they find.