Students’ epistemological resources and framing of stoichiometry assessment items across linguistic backgrounds: insights for equitable design

Abstract

Chemistry assessments shape not only what students know, but how they interpret what kinds of reasoning are valued. For multilingual learners (MLs), linguistic complexity in assessment items can constrain access to epistemological resources, masking conceptual understanding. This study examines how undergraduates' epistemological framing and resource activation are influenced by linguistic features of stoichiometry items, with attention to language background. We introduce translinguistic framing, an epistemological frame describing how MLs orient to disciplinary tasks by coordinating linguistic and semiotic resources. We conducted semi-structured think-aloud interviews with 40 undergraduates (23 English monolingual, 17 multilingual) in an introductory chemistry course at a research-intensive U.S. university. Participants examined stoichiometry item pairs combining original and linguistically revised versions. Using Hammer and Elby's (2003) epistemological resources framework, we analyzed how students framed each task and what knowledge they drew upon. Findings reveal three interrelated frames—answer-making, sensemaking, and translinguistic framing—flexibly activated depending on item design, time constraints, and linguistic load. MLs faced barriers when chemistry terminology differed from classroom language, often shifting to symbolic or procedural representations. Emotional and strategic stances, including anxiety and time-aware strategizing, further mediated resource activation, limiting conceptual engagement. Translinguistic framing, in contrast, enabled MLs to sustain disciplinary reasoning by integrating first-language terms, English keywords, and symbolic notation. This study expands models of epistemological framing by highlighting an epistemic dimension of science assessment. Implications for equitable assessment design include minimizing unnecessary linguistic complexity, offering multimodal scaffolds, and recognizing translinguistic framing as a legitimate epistemological stance supporting opportunities to demonstrate conceptual understanding.

---

Introduction

Chemistry assessments do more than evaluate content knowledge—they shape how students interpret what kinds of thinking are valued. Prior research has shown that while many undergraduate students can successfully perform algorithmic procedures such as balancing equations or carrying out stoichiometric calculations, they often struggle to engage in deeper conceptual reasoning about underlying chemical relationships, instead relying on memorized formulas or procedural shortcuts (Nurrenbern and Pickering, 1987; Hunter et al., 2021).

The difficulty lies not only in interpreting the chemical processes represented by the calculations, but also in understanding and interpreting what assessment tasks are actually asking of them. As educators, we aim to help students build meaningful links between algorithmic procedures and conceptual understanding. However, a persistent disconnect remains. This disparity reveals more than just a gap in skills—it highlights deeper challenges in how students make sense of chemistry and navigate the demands of its assessments.

Students’ understandings of the kinds of knowledge, reasoning, and justification are essential issues of epistemological framing (e.g., Hammer et al., 2005; Russ et al., 2009). Epistemological framing extends beyond descriptions of procedural versus conceptual understanding, and rather, refers to what students think they are being asked to do within a given task or context As described by Elby and Hammer (2010), epistemological framing is concerned with students’ understanding of “what is going on here” (p. 414) in a particular problem, and what kinds of ways of knowing and thinking are appropriate. Furthermore, epistemological frames are comprised of smaller-grained epistemological resources which can be activated or suppressed depending on the context. The epistemological resources include resources about knowing, such as the nature of knowledge and learning, and affective orientations (Hammer et al., 2005; Elby and Hammer, 2010).

Examples of epistemological frames in chemistry education include Hunter and colleagues’ (2021) description of the epistemic games of “answer-making” and “sensemaking.” In answer-making, students’ perceive their goal is narrowly focused on obtaining the correct numerical or symbolic response through the application of memorized procedures. In contrast, sensemaking refers to a frame in which students aim to “figure something out”—that is, to ascertain the mechanism underlying a phenomenon, exploring the “why” and “how” questions of reactions and grappling with inconsistencies to refine their conceptual grasp (Odden and Russ, 2019). Sensemaking is deeply interconnected with the authentic practices of science. Even though instructors often intend to support sensemaking, assessment contexts—especially time-pressured, calculation-heavy exams—frequently cue answer-making frames, shaping students’ engagement in ways that may be misaligned with disciplinary goals (Berland et al., 2016; Hunter et al., 2021).

Assessment-driven epistemological cues also have important equity implications in chemistry education. Assessment designs often privilege particular ways of knowing and problem-solving that align with dominant instructional norms, potentially marginalizing students whose epistemological resources and prior experiences do not align seamlessly with these expectations (Grapin et al., 2023). For multilingual learners in particular, chemistry assessment may interact with linguistic resources in ways that shape not only which frames become constrained, but also which become newly available through strategic use of their full linguistic repertoires. This raises a critical yet underexplored question: How do students’ framing resources interact with the epistemic demands of stoichiometry assessment tasks? While prior work has examined linguistic complexity as a barrier in chemistry assessment (Lee and Orgill, 2021; 2025; Siegel, 2007), less attention has been paid to how students’ linguistic assets influence their epistemological framing—such as how drawing on multiple languages influences learners’ interpretations of what kinds of knowledge and reasoning are useful in assessment contexts.

Students approach chemistry problems with a pre-existing toolkit of resources for meaning-making, shaped by their educational and other experiences. Recent scholarship has highlighted ways that multilingual learners (MLs) frequently engage in translanguaging—the actions through which learners strategically integrate multiple languages or semiotic resources to support their reasoning and meaning-making in science (Suárez, 2020; Grapin et al., 2023). Grapin et al. (2023) highlighted translanguaging as a powerful practice allowing multilingual learners to exploit their entire linguistic repertoires when engaging in rigorous science tasks strategically. Similarly, Suárez (2020) underscores translanguaging as multilingual learners’ resources for constructing and communicating sophisticated scientific explanations.

While the research on translanguaging has emphasized MLs’ actions, or “what language users do” (Grapin et al., 2023; p. 1007), translanguaging does not account for the epistemological dimension: how learners interpret the task itself as one that can and should be navigated through multiple languages and representational systems. We distinguish what they do from broader epistemological frames that describe how and when they apply translanguaging resources. We seek to better understand how linguistic assets shape inform epistemological framing, particularly, how MLs' linguistic resources influence their stance on perceptions of what knowledge and practices are valued within disciplinary assessment contexts. In addition to answer-making and sensemaking frames, MLs’ epistemological engagement may involve a distinctive linguistic dimension. A student may translanguage (e.g., read a formula in Portuguese) without necessarily framing the entire assessment encounter as legitimately multilingual. Conversely, a student who adopts translinguistic frame treats the coordination of linguistic resources not merely as a coping strategy but as an epistemological approach to disciplinary reasoning. This distinction matters because it shifts the analytic focus from the linguistic practices learners deploy to the epistemological stances they adopt—that is, how they interpret the nature of the task and what counts as a legitimate resource for engaging with it.

Therefore, our study explores learners’ translinguistic framing, that is, the application of a frame that accounts for an underlying belief or orientation that translanguaging practices are a useful means for problem solving. To do so, we also examine the multiple finer-grained linguistic resources students employ when approaching disciplinary tasks. We focus on understanding (1) how language background influences the resources that learners draw upon, and (2) how students frame their engagement with stoichiometry assessment items. Specifically, we seek to answer two primary research questions:

RQ1: What is the nature of undergraduate students’ epistemological resources when solving stoichiometry assessment items, and how does their language background influence the use of these resources?

RQ2: How do students’ epistemological frames (answer-making, sensemaking, and translinguistic framing) vary according to different design features in chemistry assessment tasks?

Context and methods

This study draws on semi-structured interviews with 40 undergraduate students enrolled in multiple sections of the same introductory chemistry course at a large, research-intensive university in the northeastern United States. The (blinded) Institutional Review Board approved this study and approved that it complied with ethical practices for human subjects research. One graduate student, unaffiliated with the course, obtained informed consent. The researchers conducting the interviews were one graduate student and one undergraduate student, both of whom were not instructors or teaching assistants for the class. Participants were informed that they could withdraw from the study at any time and their identity was protected following approved methods of data storage and anonymization.

Participants

We interviewed two categories of undergraduates enrolled in the first semester of general chemistry for this study: (1) English monolingual students (EMs) and (2) multilingual students (MLs). We defined English monolingual students as those who were born in the United States and grew up speaking English as their first and only language. Multilingual students were those who met the following criteria:

1. Born outside of the USA and/or other countries where English is the primary language

2. Have a non-English first language

3. Able to read, write, and understand basic conversational English

4. Have lived in the USA for ten years or less

These criteria were informed by previous literature on language acquisition, which suggests that students learning English typically develop basic conversational proficiency within the first two years of immersion in an English-speaking environment, but may require eight or more years to acquire cognitive academic language proficiency (Cummins, 1980, 1981, 1984, 2000; Collier, 1987, Hakuta et al., 2000). We acknowledge that the term multilingual encompasses students who can also be bilingual yet still in the process of developing academic English language proficiency.

Among the interviewed students, 30 students are from a large campus (approximately 40 000 undergraduates), and ten students are from a smaller campus (approximately 550 undergraduates). All students attended an in-person section of the course at their respective campus. All participants completed a one-hour interview focused on how they interpreted and solved a series of stoichiometry assessment items. Among the 40 students, there were 23 English monolingual speakers (57.5%) and 17 multilingual learners (42.5%), representing language groups of Arabic, Chinese, Korean, Spanish, and Thai. The intended purpose of the interviews was to investigate how chemistry assessment items can become more accessible for students, especially for multilingual students. In this study, ten participants are selected for closer analysis to capture a range of perspectives and reasoning: five English monolingual students (Participant 7, 12, 22, 23, and 26) and five multilingual learners (Participant 5, 6, 9, 21, and 40). These participants were chosen for the richness and clarify with which they articulated their reasoning during interviews. All were first-year undergraduates enrolled in the same general chemistry course, and the selected excerpts reflect diversity in gender, campus affiliation, and language experiences. Table 1 summarizes the demographic background and language histories of the ten students, including their gender, first language(s), duration of residence in the United States, and the language in which they received prior chemistry instruction.

Table 1 Demographic background of the ten focal participants

Participant	L1	Previous chemistry experience	Duration of residence in the United States	Gender
5	French/Creole	Not explicitly mentioned	9 years	Female
6	Korean	Took AP Chemistry in high school (in English)	5 years	Female
7	English	Not explicitly mentioned	U.S. only	Male
9	Portuguese	Took chemistry in Brazil in Portuguese	3–4 months	Female
12	English	Implied high school chemistry experience	U.S. only	Female
21	Chinese	Took AP Chemistry in high school (in English)	7–8 years	Female
22	English	Chemistry class in high school	U.S. only	Male
23	English	Basic chemistry in 10th grade, but it was interrupted by COVID	U.S. only	Female
26	English	Implied high school chemistry experience	U.S. only	Male
40	Spanish	Took general science in Spanish in Dominican Republic	1.5 years	Female

---

Interview protocol & conceptual framework

During the interview, students were presented with pairs of stoichiometry assessment items in a random order, as shown in Fig. 1. The pairs of items include the original version of item (indicated with an O) and the revised version of item (R) modified by the Equitable Framework for Classroom Assessment (Siegel, 2007; EFCA), reducing unnecessary linguistic complexity without modifying the difficulty level of the assessment item. The EFCA is grounded in sociocultural perspectives of learning and aims to make assessment items more linguistically accessible for multilingual learners while maintaining the same chemical content and difficulty level (Lee and Orgill, 2021). The framework emphasizes creating assessments that align with pedagogical objectives, enhance linguistic and cultural accessibility, maintain cognitive rigor, effectively reveal student comprehension, and provide appropriate language support during the learning process (Lee and Orgill, 2021). While EFCA successfully addresses linguistic barriers through language simplification, it does not explicitly leverage or validate the rich linguistic resources that multilingual students bring to their learning. Building upon EFCA's equity-focused ground, our study extends these principles by investigating how assessment design can actively incorporate multilingual learners’ diverse linguistic practices and examining the potential impact of such inclusive approaches on students’ epistemological framing. Students were asked to explain how they approached and solved a total of two or three pairs of stoichiometry assessment items by “thinking aloud” about their reasoning process. Then, the students reflected on how they interpreted the problems. The students then gave feedback to the interviewer on which parts of the items were helpful and challenging and what they liked or disliked about the problems.

Fig. 1 Revised version of item 7.

Analytic approach

We employed reflexive thematic analysis (RTA; Braun and Clarke, 2019) to examine participants' epistemological framing as they engaged with chemistry assessment items. Reflexive thematic analysis was selected as an appropriate method because it emphasizes the researcher's active role in making interpretive theory-informed judgments about patterns in qualitative data (Braun and Clarke, 2019). Consistent with Braun and Clarke's recommendation for theory-informed but inductively driven analysis, we began the analysis by coding grounded in participants’ own words, reasoning, and contextual explanations. Analysis focused on how students made sense of and responded to different item designs, particularly in relation to their language background. Initial inductive coding was conducted using a subset of five representative transcripts (three from English monolingual speakers and two from multilingual students), with codes generated to capture patterns in reasoning strategies, language processing, and affective responses. MAXQDA was used to support data organization and the coding process, allowing for systematic annotation and comparison across participant responses.

Codes were iteratively refined through continued engagement with the data and the first authors’ analytic memos. Although the theoretical construct of epistemological framing (Hammer et al., 2005; Russ et al., 2009) informed later stages of interpretation for RQ1, it did not constrain initial coding. Analytic decisions were explicitly guided by the research questions. For RQ1, we focused on identifying epistemological resources evident in students’ talk—such as the source of knowledge, representational preferences, and affective or strategic stances—drawing inductively from the data and later interpreted using Hammer and Elby's (2003) epistemological resources framework. For RQ2, we focused on how these epistemological resources were coordinated in response to assessment design features to constitute broader epistemological frames, including answer-making, sensemaking, and translinguistic framing.

A reflexive codebook was developed to document the evolving codes and their meanings (see Table 2). These included codes such as time-aware strategizing, conceptual prioritization, language load sensitivity, and affective filtering, which were used to characterize epistemological resources students activated during problem solving (RQ1). In subsequent analyses, patterns in how these resources were coordinated were examined to interpret broader epistemological frames, such as answer-making, sensemaking, and translinguistic frames (RQ2).

Table 2 Codebook with descriptions, examples, and associated epistemological frame

Code	Description	Example quote	Framing
Time-aware strategy	Strategic decisions made in response to time pressure or test constraints	Participant 12 (EM): I probably wouldn't [write an explanation] if I was like on a time crunch… I would definitely make something up or at least try to go for like half credit. (item 8.R., Fig. 6)	Answer-making
Conceptual prioritization	Emphasis on sensemaking phenomena over merely making correct answers	Participant 5 (ML): I go like in order because I feel like each of the questions, like, help you answer the next… Like, if you know how to define a limiting reagent, then you'd know what the significance of it is in the product of the reaction. (item 4.O., Fig. 5)	Sensemaking
Language load sensitivity	Awareness of the cognitive load associated with language-dominant question formats	Participant 21 (ML): Sometimes the the bunch of words and only few informations are benefit, but and others are… It's not beneficial for us to solve the question… Less wording is… that's what I'm saying. Like OK, wording is a big problem for me.	Answer-making
Preference for symbols/numerical values	Preference for numerical/symbolic representations over verbal formats (to use their L1 knowledge)	Participant 6 (ML): I am more confident in looking at those equation or numbers rather than like reading English words and stuff… So just by even though it's the same question, I would rather prefer to have a less English and then it feels like it's more straightforward.	Translinguistic framing (content-dependent)
Affective filtering	Emotional responses shaping engagement with the item (stress, relief, anxious, threw me off)	Participant 21 (ML): As a student, we get nervous because time's passing really quickly… I think it will affect me to do the other question because… It's emotion, stuff like, can affect how a student do the question.	Often Answer-making
Framing shifts	Moments where students switch or reflect on different ways of approaching the item	Participant 9 (ML): If I'd never seen either of these before I would probably pick 2O, just because I'm comfortable with that but… I definitely would be interested in seeing a question like 2R because I feel like it'd be a good way to try to display concepts without taking up too much time. (comparing items 2.O. and 2.R., Fig. 8)	Transitional
Translinguistic resource activation	Strategically draw on their first language (L1) to interpret chemistry content, treating multilingual repertoires as resources	Participant 9 (ML): There are certain words that do go to Portuguese. Reagent is one of them because it's very close to what it is in Portuguese… mostly it's names of compounds, so I read the molecular formula in Portuguese in my head. I won't read the name in English… numbers as well, I'll think in Portuguese when it comes to math questions. Some organic functions, I identify in Portuguese… I don’t always know how to pronounce them, but I don’t really need to on exams.	Translinguistic
Translinguistic bridging	Rely on multiple representations (e.g., name + symbol) to connect L1 knowledge with English terminology	Participant 40 (ML): I think for me it was like more easy to read… because sometimes I like when they like, pull like the names and then they just put like… for example hydrogen gas. H2. Yeah, I like that because… I get confused in the names. That's one of my problems… like in chemistry, because I know the name in Spanish, but I don’t know in English some of them… So I just like… when they put like silver or lead and then they pull like the symbol… because we get sometimes if you just put like H2, I don’t know what you’re talking about.	Translinguistic

---

In line with the assumptions of reflexive thematic analysis (Braun and Clarke, 2019), this study does not employ traditional reliability markers such as inter-rater agreement. Instead, analytic rigor was supported through reflexive memoing, iterative engagement with the data, and transparency in the development of themes. Trustworthiness was supported by analytic dialogue among the authors and validity in this approach is grounded in the clarity, coherence, and depth of interpretation, rather than in reproducibility or coding consensus.

The thematic findings are organized around recurring reasoning patterns and their associated epistemological framings, with a particular emphasis on how multilingual students navigated linguistic and conceptual dimensions of chemistry assessments.

Findings

The analysis revealed distinct patterns in how students approach stoichiometry assessment items, particularly regarding their epistemological framing and resource utilization. The findings are organized around two research questions, beginning with an examination of epistemological resources (RQ1) and then followed by an exploration of epistemological frames (RQ2).

RQ1. Epistemological resources and the influence of language background

To address the first research question, we examined the kinds of epistemological resources students use while solving stoichiometry items and how language background mediates their accessibility and use. To do so, we drew on Hammer and Elby's (2003) framework of epistemological resources. Hammer and Elby (2003) conceptualize epistemological resources as context-sensitive elements about knowledge and learning, such as sources of knowledge, the representations of ideas (e.g., verbal, visual) through which knowledge is expressed, and stances towards knowledge.

Our findings show that students’ resource activation varies by the design of a task design and by language background. In particular, multilingual learners (MLs) frequently described the importance of specific terminology and viewed knowledge about word-level meanings as something that is propagated (Hammer and Elby, 2003) from instructor to learner. MLs had more difficulty when terminology appeared in unfamiliar forms or dense English prose. To manage this, they often relied on symbolic representations or procedural scaffolds. Students also expressed affective and strategic stances, such as simplifying to save time or avoiding verbal explanations, that influenced which resources they engaged.

The themes that follow are organized around three dimensions of epistemological resource use: Theme 1A, source of knowledge with attention to the limits of using propagated stuff in unfamiliar language; Theme 1B, representations of ideas comparing symbolic and verbal formats and their epistemic roles; Theme 1C, emotional and strategic stances shaped by time, language, and assessment demands.

Theme 1A. The relationship between language and the reproduction of received knowledge. Overall, the multilingual students demonstrated a stronger reliance on the precision of terminology, with greater emphasis on memorizing exact terminology from lectures compared to English monolingual speakers. They perceived that variation in wording could disrupt their understanding. The excerpt below is illustrative because it shows how unfamiliar terminology constrained the student's ability to access chemistry knowledge as it had been previously learned, highlighting the limits of treating knowledge as something to be reproduced from the instruction. For example, Participant 6 expressed uncertainty about an unfamiliar term, “initial amount”, when presented with the revised version of the following item (Fig. 1).

Participant 6 (ML): I never learned like the word “initial amount” when we like learned this uh problem like with our professor. So like it really got me at first, so I was like. What does initial amount mol mean? (discussing item 7.R., Fig. 1).

When students' prior chemistry knowledge was acquired in their first language, activating that prior knowledge in English-language assessment contexts became even more complex. The following excerpt is illustrative because it shows how unfamiliar chemistry terminology led the student to skip contextual text entirely to locate the calculation demand. The student demonstrates a strategic approach to overcome contextual information they perceived as extraneous. Participant 40 explicitly described how terminology confusion affected problem-solving:

Participant 40 (ML): Sometimes when I, like, read this question, I just go to the to the, like question at at the beginning… at the end because it's like sometimes I don't understand like the chemistry part. Because if I mean it's English, so I just like go to the end if they say like how many grams of this? (discussing item 1.O., Fig. 2).

Fig. 2 Original version of item 1.

Similarly, the following excerpt further illustrates how unfamiliar terminology creates a barrier to accessing their knowledge:

Participant 40 (ML): I don't know what that means. That word like in English so… for example, if the question is gonna ask something about that word I don't know how to begin to to like answer that question. (referring to “stoichiometric” in item 1.O., Fig. 2).

Challenges with terminology when solving chemistry exam questions reflect the students’ epistemological resource that views knowledge as something that reproduced from authorities, such as teachers, textbooks, or memorized formulas. Hammer and Elby (2003) aptly referred to this as a view of “knowledge-as-propagated stuff.” From this perspective of knowledge, responding to an assessment item heavily depends on the students’ ability to recognize and apply vocabulary and phrasing exactly as it was presented in lectures or textbooks. For multilingual learners, their recognition of terms appears to be more constrained, and participants’ ability to recognize words in new contexts is limited to the specific wording used by instructors before the assessment.

In response to these constraints, students described their strategic workarounds. For instance, Participant 40 explained their strategy skipping over textual descriptions to focus only on the calculation demand. Participant 6 noted that unfamiliar terms disrupted their understanding unless they matched the exact phrasing they had memorized from class, as with the term, initial amount.

To address the first research question—What is the nature of undergraduate students’ epistemological resources when solving stoichiometry assessment items and how does their language background influence the use of these resources? —we examined whether English monolingual students described similar difficulties. Some English monolingual students did note moments of confusion related to wording:

Participant 26 (EM): Sometimes the questions on there are worded weird… I know how to do something and then I like psych myself out for the exam and then I forget the word. Like, you know, the wording is a little funky.

This excerpt shows how the nature of the difficulty of English monolingual students differed from that described by multilingual learners. English monolingual students more commonly attributed wording-related challenges to assessment design or test-taking anxiety. This distinction suggests that although complex vocabulary can pose challenges across student populations, for multilingual students, the difficulty operates at the level of linguistic recognition as an epistemological resource, shaping whether knowledge-as-propagated-stuff is accessible in the first place.

Theme 1B. Representations of ideas and procedural scaffolds. An epistemological resource (Hammer and Elby, 2003) view also posits that learners activate different epistemological resources in response to context. Learners drew on both the representations of ideas available in the item and the procedural scaffolds they could enact. By representations of ideas, we mean the expressive formats through which chemistry is articulated; symbolic/numeric (e.g., formulas, variables, coefficients) and verbal/terminological (e.g., compound names, technical verbs). In this sense, the different representations of ideas can influence how students can engage available epistemological resources. While not directly aligned with Hammer and Elby's (2003) taxonomy of “forms” (e.g., story, rule, game), this theme interprets students’ preferences for different representations of ideas as expressions of resource activation in situated assessment contexts. We selected Participant 6's response as the opening excerpt because it shows how symbolic and numeric representations functioned as epistemological resources that reduced linguistic load and provided a sense of control for multilingual learners. Participant 6 expressed a preference for equations and numbers over verbally dense fomats, even when the underlying chemistry content was identical:

Participant 6 (ML): I am fluent in English, but…I still have like language barrier and it's my second language and also reason that I choose that STEM major, like I am more confident in looking at those equation or numbers rather than reading English words and stuff… even though it's the same question, I would rather prefer to have a less English and then it feels like it's more straightforward.

For Participant 6, interpreting the chemistry context through mathematical and symbolic communication was more readily accessible than relying on terminology, “English words and stuff.” Mathematics and the symbolic language of chemistry allows students, including multilingual learners, to sidestep dense verbal explanations.

Conversely, linguistic representations could both support and hinder engagement, depending on the student's language background and disciplinary experience. For some multilingual students, disciplinary terms overlapped with their first-language knowledge or shared recognizable roots of terminology, which allowed them to treat familiar terms as reliable resources for understanding. Participant 9, for example, described how familiar terms helped them interpret technical content:

Participant 9 (ML): I guess I’m just in a unique position that ‘reagent’ in Portuguese is written almost exactly the same, just with an e at the end. So that's a word that I can take out really well… In my specific position, I would prefer ‘reagent’… because I’m familiar with it.

These connections became valuable linguistic resources that helped bridge understanding: noted that “reagent” in English is “almost exactly the same” as in Portuguese (reagente), describing chemical terms as feeling universal. Such cognates appeared to serve as valuable bridges between languages, facilitating conceptual understanding.

The following excerpts are illustrative because they shows how step-by-step structure enabled students to rely on procedural resources rather than interpretive linguistic work.

Participant 2 explicitly described a procedural scaffold approach in their interpretation of a revised assessment item:

Participant 2 (ML): [The revised item] just walks through the entire question and just give you the numbers and you can pretty much just like you know, like what equation to plug it in… you can just write it down, just plug it in and get it immediately.

Likewise, Participant 5 showed a strategic approach to sequential question structure by building off each part:

Participant 5 (ML): I go like in order because I feel like each of the questions, like, help you answer the next…Like, if you know how to define a limiting reagent, then you’d know the significance of it is in the product of the reaction.

Together, these students show how symbolic and numeric representations, particularly when structured through procedural scaffolds, can provide a more accessible and confidence-building path to problem-solving, particularly under language or time constraints.

These findings suggest ML students view symbolic and numeric representations, particularly when structured through procedural scaffolds as less hindered by language barriers, whereas linguistic representations can be either an obstacle or an asset, depending on how closely it aligns with their first language.

English monolingual students also expressed preferences for symbolic and equation-based formats, though for different reasons. When discussing whether compound-names or chemical formulas were more helpful, Participant 22 (EM) noted: “Switching it to a formula would make it quicker to read and more universally understandable.”

Unlike multilingual students who described symbolic representations as a way to leverage alternative semiotic systems when navigating dense English text (e.g., Participant 6 (ML): “I am more confident in looking at those equation or numbers rather than reading English words”), English monolingual students framed their preference in terms of efficiency and accessibility. This comparison suggests that while both groups may favor symbolic formats, the primary rationale differs for multilingual learners, symbolic forms offer an alternative pathway for meaning-making, whereas for English monolinguals, they offer a faster way to locate and process information.

Theme 1C: stances toward knowledge – emotional stance and strategic stance. While Hammer and Elby (2003) describe epistemological stances such as acceptance, understanding, and puzzlement, this theme extends the concept to include affective and strategic responses, such as self-blame, anxiety, or confidence that shape students’ engagement with knowledge during assessment. These stances shape not only what students do but how they engage cognitively and emotionally with knowledge in real time. In many cases, affective reactions such as stress or anxiety acted as a filter, influencing how students activated their epistemological resources under pressure.

We selected Participant 21's responses because they are especially illustrative of how emotional reactions entered into the student's reasoning about how to approach the problem. In reflecting on their choice between two item formats, Participant 21 described emotion as a factor that shaped how they engaged with the question. Fig. 3 shows both the original and revised versions of Item 5. Compared to the revised version, the original version more often made students feel overwhelmed by terminology and complex sentence structures.

Fig. 3 Original and revised versions of item 5.

Participant 21 (ML): It's emotion, stuff like, can affect how a student does the question (choosing 5.R. over 5.O.)

This comment followed a reflection on being overwhelmed by extraneous or unclear information, especially under time constraints:

Participant 21 (ML): As a student, we get nervous because time's passing really quickly…I think it will affect me to do the other question.

Together, these excerpts illustrate how affective states—such as nervousness under time pressure—shaped how students oriented to assessment tasks and anticipated their own performance. Participant 6's response further illustrates how affective reactions could arise immediately upon encountering an item, even before engaging in substantive problem-solving. Participant 6 described an immediate stress response triggered by verbal overload in a complex appearance of assessment item:

Participant 6 (ML): Like the one on the left, 7.O. Part B, it's like asking one more question and it's just… it really stressed me just first looking at it (item 7.O, Fig. 4).

Fig. 4 Original version of item 7.

Participant 6's emotional overwhelm demonstrates how affective factors can directly influence students’ engagement, prompting shifts away from sensemaking and toward quicker, surface-level strategies to “cope” with problem solving.

In addition to emotional filtering, students also demonstrated internalized, self-blaming stances. For example, Participant 5 repeatedly framed performance difficulties as personal failings, not as a function of assessment design:

Participant 5 (ML): “I was quite worried about that because I was like, oh, it's not as hard as I heard it was going to be, but I still ended up feeling… and that's completely my fault because I was so excited, and I wasn't reading the questions right… it was my fault. Overall, I can say that they're fair.” (talking about exam experience in general chemistry course)

“Do we really need the first sentence? Because it kind of confused me… I don’t know what the first sentence is for.” (item 1.O., Fig. 2)

“For a non-native English speaker, not knowing what these words specifically mean would make me frustrated and having to write an explanation about [them]. I wouldn't like it.” (item 4.O., Fig. 5)

Fig. 5 Original version of item 4.

These responses reveal a dual burden for multilingual learners: not only must they interpret linguistically dense questions, but they often bear the emotional weight of misunderstanding as a personal flaw. These stances—emotional filtering and self-blame—can shift a student's epistemological frame from active sensemaking toward performance-focused answer-making, reducing opportunities for deeper conceptual engagement.

RQ2. Epistemological frames for stoichiometry assessment items

Students’ responses to the three contrasting stoichiometry items revealed dynamic shifts in epistemological framing, shaped by their interpretation of time constraints, contextual features, the nature of the knowledge types (procedural vs. conceptual), and, for multilingual learners, the linguistic demands of the task. While some students demonstrated consistent reasoning frames, others transitioned between sensemaking and answer-making framings, often in response to cognitive demands or prior classroom experiences. In addition, analysis of multilingual learners’ responses revealed a third epistemological frame, translinguistic framing, in which students strategically coordinated their linguistic repertoires to sustain disciplinary engagement. Together, these findings illustrate that students’ framing was fluid and context-sensitive, varying not only with assessment design and time pressure but also with the resources available to them across languages and representational systems. These findings are organized into three key themes.

Theme 2A. Accounting for the time constraints of the assessment environment. When faced with time pressure during assessments, students strategically simplify their approach by adopting answer-making frames that prioritize quick execution over conceptual engagement, fundamentally altering how they value and access knowledge. These shifts were not simply pragmatic reactions to time constraints. Rather, they were epistemological choices about what type of knowledge was most appropriate or accessible under those conditions. Across the full sample, 21 out of 40 students (7 MLs and 14 EMs) explicitly described how time limitations shaped their reasoning strategies during assessments. This theme is organized into two subthemes.

Subtheme 2A.1: Strategic simplification under time pressure. Nine students (4 MLs and 5 EMs) students expressed a preference for strategies that minimized effort or complexity during timed assessments. These moments reflected an answer-making frame in which knowledge was treated as a tool for quick execution rather than conceptual engagement. For instance, when discussing written explanation requirements on exams, students explicitly acknowledged how time pressure influenced their approach:

Participant 12 (EM): I probably wouldn’t [write an explanation] if I was like on a time crunch… I would definitely make something up or at least try to go for like half credit. (item 8.R., Fig. 6)

Fig. 6 Original and revised versions of item 8.

Similarly, multilingual students demonstrated acute awareness of how assessment design could either support or hinder their performance under time constraints:

Participant 40 (ML): So [the first part of 3.O] is more for like lecture I don’t have to know that part to do in the test, because the test, it's limited time. So just like go just like direct, just ask why I need to. What I need to do. (item 3.O. and 3.R., Fig. 7)

Fig. 7 Original and revised versions of item 3.

For multilingual students, time pressure compounded linguistic processing demands, shifting their epistemological frame to an answer-making frame:

Participant 6 (ML): This one looks like there's more English that I need to read about so that really like make me sigh. So maybe we just look at it and then like try to like solve this problem… and at the time I'm gonna be like freaking out like how to state like my explanation …so that my like, somebody who's grading, probably like TA or like the professor whoever grades can understand what I'm talking about…. (preferred 8.O over 8.R., Fig. 6)

These responses align with the Time-aware Strategy code, in which students prioritize speed and clarity over depth due to perceived assessment pressures. In this case, Participant 6 notes that having “more English… make[s] me sigh” and described deciding to “just look at it and… try to… solve this problem” rather than spend time crafting a detailed explanation. This explains that, under time pressure, they would be “freaking out” about “how to state my explanation so that somebody who's grading can understand what I’m talking about.” This illustrates how the student consciously streamline their approach to ensure timely completion over an elaborated justification.

Subtheme 2A.2: Epistemological prioritization under cognitive constraints. Despite the tendency toward strategic simplification, students also revealed more nuanced epistemological reasoning when comparing assessment formats. When presented with different versions of stoichiometry problems, students expressed interest in formats that could support both efficiency and conceptual understanding. Students were presented to two versions of a problem concerning the reaction of hydrogen gas with chlorine gas. As shown in Fig. 8, 2.O is the original version of the problem from the Top Hat question bank. Item 2.R represents the revised version. The interviewer asked, “What kind of item would be better if you have one of these items in your exam?”

Fig. 8 Original and revised versions of item 2.

When students engage in solving assessment items, it is often regarded as the context of answer-making (Chen et al., 2013; Hunter et al., 2021). Even in confronting general chemistry assessment items to solve and evaluate, Participant 23 expressed familiarity with traditional formats (2.O) but also showed interest in 2.R, highlighting it as a more efficient way to demonstrate conceptual understanding without extensive calculations:

Participant 23 (EM): If I'd never seen either of these before I would probably pick 2O, just because I'm comfortable with that but…I definitely would be interested in seeing a question like 2R because I feel like it'd be a good way to try to display concepts without taking up too much time of that sometimes we have really long questions on an exam and it's trying to figure out how much time we're going to spend on each question…And this way with with 2.R you can still show that you can learn the concept that you know what you're doing well…also not having to pull a calculator out and periodic tables and all that stuff, yeah.

Participant 23's response indicates a transition in preference from answer-making to sensemaking, motivated by the constraints of exam conditions. Although they initially mention they would choose 2.O “just because I’m comfortable with that,” they go on to express interest in 2.R as “a good way to… display concepts without taking up too much time.” This shift shows a move away from relying solely on a familiar procedural format (answer-making) toward valuing a format that allows them to demonstrate conceptual understanding more efficiently (sensemaking), particularly under time-limited conditions. In such environments, the perceived need to optimize time leads students to selectively prioritize conceptually efficient strategies. Participant 26 echoed this, noting a preference for symbolic representations (like “X” and “Y”) over numerical values in the problem:

Participant 26 (EM): Yeah, what is that… the fact that there were no numbers given here, it was just variables…I honestly I like that a lot that there are no variables cause I just kind of think, specifically on the H2 and the Cl2, and I know obviously you need to have numbers in some of them cause you know that's chemistry, but I did like that a lot just because of how I work. It's just not as many factors going into my thought process, especially in a class where I have 50 minutes like…And I think I've told you this before, but I you know I always like… I feel like I do better when I have more time to do things and like with this one I could figure this one out in a pretty quick amount of time.

These findings suggest that while time constraints typically push students toward answer-making frames, thoughtful assessment design can create opportunities for sensemaking even within time-limited environments. Students demonstrated sophisticated epistemological reasoning about when and how to engage with different types of knowledge based on contextual demands. Seven students (3 MLs and 4 EMs) across the data explicitly showed this tendency.

Theme 2B. Accounting for supplemental, contextual information. Students demonstrated responses to supplemental and contextual information in assessment items, recognizing its value for learning while simultaneously viewing it as potentially distracting under assessment conditions. In our data, 14 out of 40 students (5 MLs and 9 EMs) commented on the role of supplemental or contextual information. While some valued background detail in practice contexts, some described it as distracting or unnecessary in exam settings. For example, when solving item 3.O. and 3.R., shown in Fig. 7, Participant 7 found contextual and background information in assessment items intriguing and worthwhile for practice sessions. This interest reflects an appreciation for the broader context and real-world application of chemistry concepts, aligning with sensemaking approaches. These connections help students understand the relevance and application of chemical concepts in everyday life and historical contexts, fostering deeper engagement with the material during practice sessions.

Interviewer 1: What about [the introduction]? And then it says found in ancient Egyptian cosmetics?

Participant 7 (EM): I feel like that would be nice to know in like a practice format because I do think that's kind of cool how they tell us… like what it is.

However, Participant 7 also noted that such information, while enlightening, could become a source of distraction during exams. This feeling was particularly empathetic towards peers for whom English is not the first language, highlighting that the additional information could add to the complexity of understanding the question, impeding their performance under time constraints.

Participant 7 (EM): But like during a test… I feel like I would just be distracting. And if I was like someone who English was a second language…then like if it already takes me like, or it… it's so I'm like a little slower with comprehending the words at all that fluff is just like unnecessary.

Five multilingual students confirmed Participant 7's concern this concern, expressing how extraneous information created cognitive burden during time-pressured assessments. When discussing problematic features of assessment items, one student explained:

Participant 21 (ML): Sometimes the bunch of words and only few informations are benefit, but and others are… It's not beneficial for us to solve the question… Less wording is… that's what I'm saying. Like OK, wording is a big problem for me.

Another multilingual student described the strategic reading approach this required:

Participant 40 (ML): So [the first part of 3.O] is more for like lecture I don't have to know that part to do in the test, because the test, it's limited time. So just like go just like direct, just ask why I need to. What I need to do.

These data illustrate that supplemental and contextual information in chemistry assessments can support different epistemological frames depending on the environment: fostering sensemaking during learning contexts while potentially creating barriers during assessment contexts, particularly for linguistically diverse students who must navigate both conceptual and linguistic demands simultaneously.

Theme 2C: acknowledging differences between procedural and conceptual engagement. Students demonstrated awareness of the distinction between procedural knowledge (knowing how to perform calculations) and conceptual understanding (knowing why procedures work), with expressing a preference for assessment formats that support conceptual engagement over rote procedural execution. When discussing contextual assessment items, students revealed sophisticated thinking about the purpose of different question types in evaluating their understanding. A total of 14 students (6 MLs and 8 EMs) explicitly reflected on the contrast between procedural calculation and conceptual reasoning. For example, Participant 22 acknowledged the usefulness of contextual problems in assessment during solving item 8.O. and item 8.R. (Fig. 6).

Participant 22 (EM): …I think those are good questions for the conceptualization, because just memorizing how to do the math is different than actually understanding like…What the math is telling you basically, and so having like C&D there I think really help check for like actual understanding of the material.

This student's recognition of the value of contextualizing chemistry concepts aligns with sensemaking approaches. Also, Participant 22 emphasized the importance of understanding the conceptual basis of mathematical problems in chemistry rather than just memorizing the procedures. Students often recognize that calculation and mathematical components are included in chemistry items. However, students often see those parts as rote memorization or even obstacles in the answer-making perspective. This student's distinction points to a deeper engagement with the material, indicative of a sensemaking approach. Furthermore, the student highlighted that certain types of questions (“like C&D”) are effective in assessing actual understanding of the material rather than just the ability to perform calculations. This preference for deeper conceptual questions over procedural ones further supports sensemaking in chemistry education.

Multilingual students particularly appreciated assessment designs that scaffolded their thinking through sequential, conceptually connected questions. When discussing multi-part problems, one student explained:

Participant 5 (ML): I go like in order because I feel like each of the questions, like, help you answer the next… Like, if you know how to define a limiting reagent, then you'd know what the significance of it is in the product of the reaction. (item 4.O., Fig. 5)

This preference for scaffolded progression reveals how students can engage in deeper conceptual thinking when assessment design supports their reasoning process. Similarly, another multilingual student valued structured approaches that guided their thinking:

Participant 9 (ML): Because even though they're like multi-step, the steps walk you through the thought process and that's helpful. (item 1.R., Fig. 9)

Fig. 9 Revised version of item 1.

These data suggest that while students generally value conceptual over procedural engagement, assessment design plays a crucial role in enabling sensemaking approaches. Multilingual learners, in particular, benefit from structured formats that scaffold conceptual understanding while reducing unnecessary linguistic complexity, allowing them to demonstrate their chemical reasoning without being hindered by language barriers.

Theme 2D: translinguistic framing. In addition to epistemological frames of sensemaking and answer-making, analysis of multilingual learners’ responses revealed a distinct epistemological frame that we term the translinguistic frame. Epistemological framing refers to how students interpret “what is going on here” (Elby and Hammer, 2010; p. 414) in terms of knowledge and learning: whether the task is about figuring something out, making sense of phenomena, producing a correct answer, or recalling authoritative knowledge (Hammer et al., 2005; Hutchison and Hammer, 2010). Students’ framings shape the knowledge they activate, the strategies they choose, and how they evaluate the relevance or legitimacy of different resources they use.

Translinguistic framing refers to an interpretive stance in which multilingual learners approach disciplinary tasks by actively drawing on their full linguistic repertoires, including their first languages (L1) or the language they used when they learned the relevant knowledge, English, and symbolic or visual representations, to make sense of science problems. This theme was unique to multilingual learners: 5 out of 17 ML students described approaching tasks through a translinguistic frame. In contrast, no EM students demonstrated this framing. This frame is marked not simply by code-switching or translation, but by the strategic and context-sensitive coordination of multiple linguistic and semiotic systems for conceptual clarity, problem-solving efficiency, and epistemological access. In this frame, language becomes not just a medium of communication but a central component of meaning-making that shapes how the student understands the nature of the task and how knowledge is conceptualized.

While translanguaging has been extensively studied as an instructional practice in bilingual education and STEM education for multilingual learners (e.g., García et al., 2017; Suárez, 2020), translinguistic framing goes a step further by capturing the epistemological orientation that multilingual students adopt when they integrate linguistic and symbolic resources into disciplinary reasoning processes. Rather than assuming the goal is to perform in English or merely translate terms, students in this frame approach science problems as tasks that can and should be navigated using all available language-based and symbolic knowledge systems. This framing thus reinterprets what counts as legitimate epistemic activity and expands models of epistemological framing to better reflect the experiences of linguistically diverse learners.

Participant 9 provides a clear example of this interpretive stance. While solving a chemistry item, the student described dynamically shifting between Portuguese and English depending on disciplinary demands:

Participant 9 (ML): There are certain words that do go to Portuguese. Reagent is one of them because it's very close to what it is in Portuguese… mostly it's names of compounds, so I read the molecular formula in Portuguese in my head. I won't read the name in English… numbers as well, I'll think in Portuguese when it comes to math questions. Some organic functions, I identify in Portuguese… I don’t always know how to pronounce them, but I don’t really need to on exams.

This student is not switching languages to compensate for perceived weaknesses in English. The student is choosing which language, Portuguese or English, helps a particular step make sense, and they lean on symbols to keep ideas straight. In doing so, they treat chemistry as something that can be worked through across multiple linguistic and representational systems, not only in English. We read this as a translinguistic framing that differs from sensemaking or answer-making because coordinating languages is the central strategy for understanding. Similarly, Participant 6 expressed a preference for assessment items with minimal English text:

Participant 6 (ML): Even though it's the same question, I would rather prefer to have less English and then it feels like it's more straightforward.

Here, the student's framing suggests that dense English text is not merely a barrier but a disruption to disciplinary engagement. By reducing the amount of language to process, the student keeps the disciplinary content at the center, using symbols or concise wording as a primary channel for engagement.

Similarly, Participant 40 added another layer to this pattern by expressing a preference for assessment items that include both the chemical name and symbol, indicating that multiple forms of representation help bridge language-specific gaps in terminology across languages:

Participant 40 (ML): I think for me it was like more easy to read… because sometimes I like when they like, pull like the names and then they just put like… for example hydrogen gas. H₂. Yeah, I like that because… I get confused in the names. That's one of my problems… like in chemistry, because I know the name in Spanish, but I don’t know in English some of them… So I just like… when they put like silver or lead and then they pull like the symbol… because we get sometimes if you just put like H₂, I don’t know what you’re talking about.

In this comment, the translinguistic frame is evident that the student wants to align symbolic notation with terms in both languages. This is not simply a matter of translation; it is a way of integrating semantic and symbolic resources so that disciplinary meaning remains accessible even when English terminology is unfamiliar. The framing positions chemistry as a bilingual-symbolic practice, in which toggling between languages and representations is an expected and legitimate part of problem-solving.

Discussion and conclusion

This study demonstrates that students' approaches to stoichiometry assessment items are shaped by complex, dynamic interactions between epistemological frames, epistemological resources, and environmental factors, including time constraints, contextual features, and language background. Through interviews with students comparing original and revised chemistry items, our findings illuminate how students' reasoning is contingent on contextual constraints such as time pressure, linguistic complexity, and assessment format. The analysis reveals the epistemological flexibility and resourcefulness that students demonstrate when engaging with assessment items, especially in linguistically demanding environments, challenging deficit-oriented assumptions about diverse learners' capabilities.

The contextual fluidity of epistemological framing

Students' epistemological framing was not fixed but dynamically responsive to perceived expectations embedded in assessment contexts. Our findings align with the resources model of epistemology (DeGlopper et al., 2023), highlighting how different epistemological stances may be productive in different task contexts. Theme 2A illustrates how time constraints often pushed students toward answer-making frames, either through strategic simplification (Subtheme 2A.1) or through epistemological prioritization under cognitive constraints (Subtheme 2A.2). This extends prior research (Chen et al., 2013; Hunter et al., 2021) by demonstrating that implicit features of the design of traditional assessment items—rather than explicit instructions—often signal a preference for rapid execution of mathematical procedures over conceptual understanding. Theme 2B illustrated that supplemental and contextual information played an ambivalent role. While such information fostered sensemaking during practice, it was frequently perceived as distracting or burdensome under exam conditions, particularly for multilingual learners. Theme 2C showed that students were aware of the difference between procedural engagement (knowing how to carry out calculations) and conceptual engagement (understanding why procedures work), and expressed preferences for assessment formats that supported conceptual reasoning. Lastly, Theme 2D introduced translinguistic framing as a third epistemological frame alongside sensemaking and answer-making. In this frame, multilingual learners strategically leveraged their full linguistic repertoires to sustain disciplinary reasoning. This finding contributes a new dimension to epistemological framing theory by showing how framing can be organized around the deliberate coordination of linguistic and semiotic resources. Students framed chemistry not as a subject to be engaged exclusively in English, but as a multilingual enterprise in which toggling between representational systems is a strategy for disciplinary understanding.

Prior studies of epistemological framing (e.g., Hammer et al., 2005; Shar et al. 2020) have primarily explained how students navigate between sensemaking and answer-making in monolingual, English-dominant contexts. Our findings suggest that for multilingual learners, framing is not limited to this binary. Translinguistic framing appears to function as a distinct orientation in which the coordination of linguistic and semiotic resources is itself the organizing principle of engagement, rather than a secondary feature of sensemaking or a compensatory response to linguistic difficulty.

Epistemological resources and linguistic mediation

Building on Hammer and Elby's (2003) framework, our analysis highlights how epistemological resources were mediated by students' language backgrounds in consequential ways. These resources included —source of knowledge (“knowledge-as-propagated stuff”, Theme 1A), representations of ideas and procedural scaffolds (Theme 1B), and affective and strategic stances about knowledge (Theme 1C). Multilingual students in particular faced challenges activating knowledge when terminology or phrasing are different from what was used in class. This constraint was not a reflection of conceptual gaps, but of fragile linguistic mappings between learned content and assessment language. This highlights how language background shapes resource accessibility rather than capacity.

At the same time, students also demonstrated impressive flexibility in how they approached knowledge. Many turned to symbolic and numeric representations, often supported by procedural scaffolds, as more accessible alternatives to dense verbal formats. These forms functioned as “linguistic safe zones,” where disciplinary content could be processed with less interpretive friction. Students’ preferences for scaffolded formats and representational clarity indicate that the modality of knowledge presentation plays a critical role in shaping equitable access to disciplinary knowledge. When assessments rely solely on dense textual formats, they risk creating what Fricker (2007) calls hermeneutical injustice, depriving multilingual learners of the interpretive tools needed to demonstrate their understanding. This underscores that assessment equity requires attention to representational diversity, ensuring that disciplinary reasoning can be expressed through multiple accessible forms rather than privileging a single linguistic mode.

Finally, Theme 1C also revealed that emotional and strategic stances filtered how resources were activated. Stress, time pressure, or self-blame sometimes pushed students toward superficial answer-making, while confidence and clear scaffolding facilitated sensemaking. These findings highlight how epistemological resources are not only cognitive but also affectively mediated.

Implications for equitable assessment design and epistemic access

This study contributes to understanding how epistemological framing and resources intersect with assessment design, with implications for equity. By incorporating Theme 2D's insight into translinguistic framing, we argue that equitable assessment must recognize not only linguistic barriers but also the epistemological value of multilingual students’ flexible use of languages and representations.

Following the concept of epistemic quality (Hudson, 2018; Hudson et al., 2023)—defined as the quality of subject knowledge based on didactic interactions and what students come to know and can do—we argue that designing for equitable epistemic access requires assessments that maximize meaningful engagement with disciplinary knowledge of high epistemic quality. This involves:

• Minimizing extraneous linguistic complexity without diluting disciplinary rigor;

∘ Use plain, concise phrasing for task instruction, while keeping chemistry-specific terminology intact where it is conceptually necessary. Example: “Determine the quantitative yield, expressed in moles…” Instead use “How many moles of—are produced?”

∘ Eliminate information that is not directly related to solving the problem. Example: “In the combustion of a hydrocarbonaceous species, consider the stoichiometric coefficients of the resultant oxidation products that must be ascertained.” Instead use “A sample of a hydrocarbon is completely burned in excess oxygen.”

• Incorporating scaffolded question structures that support conceptual progression;

∘ Sequence questions from representational to conceptual to application levels. For example, (1) Identify what information is given, (2) Translate information into a balance equation, (3) Explain what the coefficients represent, (4) Calculate required quantity, (5) Interpret what the calculated value means in terms of the chemical process.

∘ Break complex problems into subtasks

∘ Use sentence starters or guided stems such as “To find the limiting reactant, first…”

• Providing multimodal representations (linguistic, symbolic, numeric, visual) that honor diverse representational ideas;

∘ Pair textual description with visual diagrams, particulate-level models, and numeric/symbolic representations.

∘ Use consistent iconography (molecule drawings, color coding, arrows for processes) to reduce representational overload.

• Creating affectively supportive environments through clarity and accessibility;

∘ Begin assessments with low-stakes warm up questions to reduce anxiety and activate prior knowledge.

∘ Use predictable formats for questions (preferably as students have seen in homework assignments and during class)

• Recognizing translinguistic framing as a legitimate epistemological stance, not a deficit.

∘ Frame multilingualism as an asset, noting that switching linguistic frames can deepen metacognitive awareness and strengthen opportunities to learn

∘ Acknowledge that scientific sense-making can occur in any language, even if the final output must be in English.

∘ Encourage students to use first-language resources (notes, labels, brainstorming) during early reasoning stages.

Collectively, our study reveals nuanced epistemological resources and frames that students bring to bear in solving stoichiometry problems. We have demonstrated how multilingual students frame problem solving as a translinguistic activity. This view challenges deficit narratives about multilingual students’ problem-solving and assessment. By foregrounding epistemological resources in assessment design, we can move toward more inclusive assessment systems that honor the full range of knowledge resources students bring to chemistry—including their linguistic flexibility, representational fluency, and awareness of when and how different epistemological frames serve disciplinary reasoning. This reframing positions equitable assessment not as a matter of simplification or accommodation, but as a matter of justice: ensuring that what we measure reflects students’ thinking rather than their ability to navigate linguistic and contextual barriers.

Ethical practices

We have included the following details regarding our adherence to ethical practices for human subjects research: “The Penn State Institutional Review Board approved this study and approved that it complied with ethical practices for human subjects research. One graduate student, unaffiliated with the course, obtained informed consent. The researchers conducting the interviews were one graduate student and one undergraduate student, both of whom were not instructors or teaching assistants for the class. Participants were informed that they could withdraw from the study at any time and their identity was protected following approved methods of data storage and anonymization.”

Conflicts of interest

There are no conflicts to declare.

Data availability

The data are not publicly available as approval for this study did not include permission for sharing data publicly (Lee and Orgill, 2025).

Acknowledgements

This material is based upon work supported by the National Science Foundation under Grant Number 2125952. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

References

1. Lee E. N. and Orgill M., (2021), Toward equitable assessment of English language learners in general chemistry: Identifying supportive features in assessment items, J. Chem. Educ., 99(1), 35–48.
2. Lee E. N. and Orgill M., (2025), Enhancing the accessibility of chemistry assessments for multilingual learners: understanding challenging features in assessment items, Chem. Educ. Res. Pract., 26(1), 126–140.
3. Berland L. K., Schwarz C. V., Krist C., Kenyon L., Lo A. S. and Reiser B. J., (2016), Epistemologies in practice: making scientific practices meaningful for students, J. Res. Sci. Teach., 53(7), 1082–1112.
4. Braun V. and Clarke V., (2019), Reflecting on reflexive thematic analysis, Qualitative Res. Sport, Exercise Health, 11(4), 589–597.
5. Chen Y., Irving P. W. and Sayre E. C., (2013), Epistemic game for answer making in learning about hydrostatics, Phys. Rev. Spec. Top.–Accel. Beams, 9(1), 010108.
6. Collier V. P., (1987), Age and rate of acquisition of second language for academic purposes, TESOL Quarterly, 21(4), 617–641.
7. Cummins J., (1980), The cross-lingual dimensions of language proficiency: implications for bilingual education and the optimal age issue, TESOL Quarterly, 175–187.
8. Cummins J., (1981), Empirical and theoretical underpinnings of bilingual education, J. Educ., 163(1), 16–29.
9. Cummins J., (1984), Bilingualism and Special Education: Issues in Assessment and Pedagogy, Clevedon, England: Multilingual Matters.
10. Cummins J., (2000), Language, power, and pedagogy: Bilingual children in the crossfire, vol. 23, Multilingual matters.
11. DeGlopper K. S., Russ R. S., Sutar P. K. and Stowe R. L., (2023), Beliefs versus resources: a tale of two models of epistemology, Chem. Educ. Res. Pract., 24(2), 768–784.
12. Elby A. and Hammer D., (2010), Epistemological resources and framing: a cognitive framework for helping teachers interpret and respond to their students’ epistemologies, in Bendixen L. D. and Feucht F. C. (ed.), Personal epistemology in the classroom: Theory, research, and implications for practice, Cambridge: Cambridge University Press, pp. 409–434. ISBN 978-1107412507.
13. Fricker M., (2007), Epistemic injustice: Power and the ethics of knowing, Oxford University Press.
14. García O., Johnson S. and Seltzer K., (2017), The translanguaging classroom: Leveraging student bilingualism for learning, Philadelphia, PA: Caslon.
15. Grapin S. E., Pierson A., González-Howard M., Ryu M., Fine C. and Vogel S., (2023), Science education with multilingual learners: equity as access and equity as transformation, Sci. Educ., 107(4), 999–1032.
16. Hammer D. and Elby A., (2003), Tapping epistemological resources for learning physics, J. Learn. Sci., 12(1), 53–90 DOI:10.1207/S15327809JLS1201_3.
17. Hammer D., Elby A., Scherr R. and Redish E., (2005), Resources, framing, and transfer, in Mestre J. (Ed.), Transfer of learning from a modern multidisciplinary perspective, Information Age Publishing, pp. 89–120.
18. Hakuta K., Butler Y. G. and Witt D., (2000), How long does it take English learners to attain proficiency? Santa Barbara, CA: University of California Linguistic Minority Research Institute.
19. Hudson B., (2018), Powerful knowledge and epistemic quality in school mathematics, London Rev. Educ., 16(3), 384–397.
20. Hudson B., Gericke N., Olin-Scheller C. and Stolare M., (2023), Trajectories of powerful knowledge and epistemic quality: analysing the transformations from disciplines across school subjects, J. Curriculum Studies, 55(2), 119–137.
21. Hunter K. H., Rodriguez J. M. G. and Becker N. M., (2021), Making sense of sensemaking: using the sensemaking epistemic game to investigate student discourse during a collaborative gas law activity, Chem. Educ. Res. Pract., 22(2), 328–346.
22. Hutchison P. and Hammer D., (2010), Attending to student epistemological framing in a science classroom, Sci. Educ., 94(3), 506–524.
23. Nurrenbern S. C. and Pickering M., (1987), Concept learning versus problem solving: Is there a difference? J. Chem. Educ., 64(6), 508.
24. Odden T. O. B. and Russ R. S., (2019), Defining sensemaking: bringing clarity to a fragmented theoretical construct, Sci. Educ., 103(1), 187–205.
25. Russ R., Coffey J., Hammer D. and Hutchison P., (2009), Making classroom assessment more accountable to scientific reasoning: a case for attending to mechanistic thinking, Sci. Educ., 94(3), 506–524.
26. Siegel M. A., (2007), Striving for equitable classroom assessment: guiding teachers to use students’ cultural, linguistic, and personal experiences, Multicultural Perspectives, 9(3), 54–58 DOI:10.1080/15210960701443511.
27. Shar K., Russ R. S. and Laverty J. T., (2020), Student epistemological framing on paper-based assessments, Phys. Rev. Phys. Educ. Res., 16(2), 020101.
28. Suárez E., (2020), “Estoy Explorando Science”: emergent bilingual students problematizing electrical phenomena through translanguaging, Sci. Educ., 104(5), 791–826.

---