Keywords

Introduction

According to the OECD (2019), there are good reasons for supporting students to develop agency in their learning: “when students are agents in their learning, they are more likely to have learned how to learn – an invaluable skill that they can use throughout their lives” (p. 2). As understood in the context of the OECD Learning Compass 2030, the concept of agency is based on the conviction that “students have the ability and the will to positively influence their own lives and the world around them” (p. 2)

In this chapter we are reviewing the literature with respect to studies in mathematics education where students (increasingly) develop agency in selecting and using digital resources (DRs) for their own learning, often in innovative mathematics learning environments. To provide an example of students’ selection and use of resources in such innovative learning environments, Pepin and Kock (2021) have described students’ use of (digital) resources in an innovative course at the end of a bachelor program at a technical university. Multidisciplinary student groups were given a real-life challenge formulated by an external stakeholder, typically from a company. The students selected a concrete problem based on the challenge, investigated the problem, and developed a prototype solution. In this process, the students exercised considerable agency, because they had to find and select relevant resources that would help them solve the problem (e.g., scientific papers, software tools). The students were guided by feedback from social resources (e.g., their academic tutors and the stakeholders). In more traditional learning environments students have sometimes been given the choice between different resources (e.g., going to a lecture or watching online videos; Meehan and McCallig 2018), which has allowed them to develop and exercise some agency.

Answering the question why students are, or should be, given the opportunity to select (or even design) materials for their learning is provided in concepts such as self-regulated learning (SRL; e.g., Steffens 2011). There, it is said that this can enhance their feelings of empowerment, ownership, and sense of purpose, and thus contribute to live long learning. By directing their own learning, students can more easily study according to what works best for them given their mathematical proficiency, their interest, their personal situation, and their learning objectives when they work in open-ended projects at their own pace. In addition, by allowing students to direct their learning (e.g., with particular resources), they can tailor their interests more. It is assumed that students will be more engaged in the learning activities, if they can focus on something they have chosen and find relevant according to their interests. With the explosion of information and an increasing availability of different DRs on the internet that students can access, it is becoming even more important for students (and teachers) to find ways of not only coping, but rather managing this influx of information (Steffens 2011).

The literature (e.g., Bron and Veugelers 2014) provides at least five arguments for involving student voice in curriculum design: (1) the normative argument; (2) the developmental argument; (3) the political argument; (4) the educational argument; (5) the relevance argument. While the normative argument sets out that students are entitled to participate in decision making in their education (“young people are entitled to the right to have a voice in matters that affect them,” p. 125), the developmental argument claims that students are developmentally ready to participate and capable of having an influence on their own education, “as they often assume more responsibility and autonomy outside school than allowed within” (p. 125). The political argument rests on the assumption that a general “one size fits all” curriculum is not appropriate, as students have different needs and interests, and they experience the curriculum differently in different contexts (and cultures). Hence, asking that the curriculum should be open to improvement and adaptation to elements related to time and place, it is claimed that involving students in curriculum design is one way of doing so. The educational argument includes the development of knowledge, skills, attitudes, and a sense of belonging. Having students reflect on their learning experiences with respect to their interests and opportunities can be used as an input for the curriculum, in order to connect in- and out-of-school/university learning, for example. The relevance argument (“Involving students in curriculum design improves the relevance of curricula”) rests on the assumption of four quality criteria for the curriculum: relevance, consistency, practicality, and effectiveness (Thijs and van den Akker 2009). To consider providing agency for students by involving them in curriculum design, it is necessary to see the curriculum and its associated resources not as a product or a fixed set of requirements and tools, but as a negotiation process wherein external aims give direction but moreover where (teachers and) students influence what is actually experienced in class. For example, Dewey (1938) opposed the idea that the curriculum is a prescription of what learners have to undergo. He argued that learning cannot happen by the external motivation of a prescribed curriculum and the provided resources, but that learning starts with the experiences and interests of the learner and is built up by negotiation (between teacher and student), toward a more systematic growth of knowledge and insights of the learner.

Leaning on this argumentation, in this chapter we are interested in findings about how, under which circumstances, and in which environments students develop agency in choosing and using DRs for their learning of mathematics.

Main research question:

How do students develop agency in selecting and using digital resources for their learning of mathematics?

Sub-question 1::

What kinds of (digital) resources do students select, and how do they use these resources?

Sub-question 2::

Which factors influence students’ selection and use of resources?

Following this introductory section, we explain the methodology in the second section and the theoretical frames in the third section. In the fourth section we provide our findings answering the first sub-question and in the fifth section our insights answering the second sub-question. The sixth section provides a discussion of the findings and conclusions drawn with respect to answering the overall research question.

Methodology

To identify the relevant literature to answer our research questions, we conducted a literature review. This review mainly relied on the procedures of a thematic synthesis (Xiao and Watson 2019) with the overarching aim to build on the current body of literature, to summarize what is known about students’ selection and use of resources and related influential factors. Since research on students’ use of (digital) resources has only increased over recent years, we reduced our literature review to publications that were published since 2015. We went through all titles and abstracts from the following list of journals, conference proceedings, and books:

  • Journals: Digital Experiences in Mathematics Education, Educational Studies in Mathematics, For the Learning of Mathematics, International Journal of Science and Mathematics Education, Journal of the Learning Sciences, Journal of Mathematics Teacher Education, Journal for Research in Mathematics Education, Mathematical Thinking and Learning, Research in Mathematics Education, ZDM – Mathematics Education

  • Conference proceedings: Congress of the European Society for Research in Mathematics Education (CERME), International Conference on Technology in Mathematics Teaching (ICTMT), International Conference on Mathematics Textbook Research and Development (ICMT), Proceedings of the International Network for Didactic Research in University Mathematics (INDRUM), Conference of the International Group for the Psychology of Mathematics Education (PME)

  • Books: Mathematics Education in the Digital Age: Learning, Practice, and Theory (Clark-Wilson et al., 2021), The Resource Approach to Mathematics Education (Trouche et al. 2019)

The journals were chosen from the top 15 of journals with the highest quality based on the study by Williams and Leatham (2017). The choice was made based on the authors’ shared judgment of relevance for the theme. We also included conference proceedings of large international conferences with peer-reviewed proceedings (CERME, PME) and of specialized international conferences that were thematically related to resources in mathematics education. Books that are relevant to the theme and are published by publishing houses with a reputation in mathematics education were also included.

In the titles and abstracts, we looked for information indicating that the paper comprised students’ own choice of digital curriculum resources and influencing factors. Studies investigating students’ interactions with particular and predetermined digital resources, as well as studies investigating the effect of digital curriculum resources on students’ learning or studies focusing on teachers’ choice or interaction with resources, were not included.

In order to not miss any important papers published in journals other than the ones we looked through systematically, we further conducted a literature search of titles, abstracts and keywords using Scopus with the search terms: mathematics AND students AND digital AND (resource OR material OR textbook OR tool OR technology) AND (usage OR use OR selection OR choice). This yielded 114 results. We scanned titles and abstracts according to the same rules that we used when we systematically went through the journals. Finally, this yielded seven additional papers that were relevant and not captured by our previous search.

After identifying a corpus of the relevant literature, we went through the literature identifying results on students’ selection and use of resources at primary, secondary, and tertiary level, and identified influencing factors. The influential factors that were found in the literature were grouped inductively and yielded the following four categories: (1) Availability and nature of resources; (2) Nature and structure of the course/s and pedagogical approaches; (3) Institutional frameworks; (4) Student beliefs and goals.

Theoretical Frames

In this section we explain and conceptualize the following terms: “lens of resources”; and “student agency.” The term “digital resources” (DR) has already been defined in the introductory chapter of this handbook, and we have picked it up under the section “Lens of Resources.”

Lens of Resources

We draw here on the “lens of resources,” as outlined by Trouche, Gueudet, and Pepin (2019) in connection with the Documentational Approach of Didactics (DAD), and more particularly on the Instrumental Approach (Rabardel and Bourmaud 2003; Trouche 2004), to address the question of students exercising and developing agency in their selection and use of DRs. For example, for engineering students in applied mathematics, we assume that the development of competencies is related to students’ strategies when orchestrating and integrating different types of resources. This integration involves the process of instrumentation, where the affordances of resources influence student practice and knowledge, and the process of instrumentalization, where students adapt the resources to their own needs.

Related to the definition outlined in the introductory chapter of this handbook, in terms of “resources” we draw on the categories of resources as outlined by Pepin and Gueudet (2018), who define mathematics curriculum resources as “all the material resources that are developed and used by teachers and students in their interaction with mathematics in/for teaching and learning, inside and outside the classroom” (p. 172/173). According to this, curriculum resources would include the following:

  • Text resources (e.g., textbooks, teacher curricular guidelines, websites, worksheets, syllabi, tests)

  • Other material resources (e.g., manipulatives, calculators)

  • Digital-/ICT-based curriculum resources (e.g., interactive e-textbooks)

Moreover, we distinguish digital curriculum resources (e.g., e-textbooks), from instructional technology (e.g., digital geometry software), by leaning on work by Pepin et al. (2017a):

It is the attention to sequencing – of grade-, or age- level learning topics, or of content associated with a particular course of study (e.g., algebra) – so as to cover (all or part of) a curriculum specification, which differentiates Digital Curriculum Resources from other types of digital instructional tools or educational software programs. … Of course, Digital Curriculum Resources make use of these other types of tool and software: indeed, what differentiates them from pre-digital curriculum programs is that they are made accessible on electronic devices and that they often incorporate the dynamic features of digital technologies. (p. 647)

We acknowledge that there are other “nonmaterial” resources used by students: for example, social and cultural resources (e.g., conversations with tutors, peers, and friends), and cognitive resources (e.g., concepts and techniques), and general resources (e.g., software, internet, and other digital resources). Digital resources help to dissolve the boundaries between these different types of resources. For example, a student using a messenger app to communicate about the solution to a problem with his/her peers uses social, cognitive, and general resources. Therefore, we do not strictly distinguish between the different types, but our main focus lies in the selection of the use of digital resources.

We use the notion of “resource/s” that students have access to and interact with in and for their learning and studying, assuming that the ways students learn mathematics are influenced/shaped by their use of the various resources at their disposal. By “use of resources” we denote, for example, the resources students choose (among the ones on offer) and for what purpose (e.g., revision); the ways they align and orchestrate them (e.g., first lecture then checking the textbook); the ones that seem central to achieve particular learning goals (e.g., for weekly course work, for examinations, for their mathematical topic area).

Student Agency

Student agency and voice as a means to incorporate the ideas of young people in education have (re)emerged over the last 25 years (e.g., McIntyre et al. 2005). Some authors have compared it to a movement. For example, Sinnema and Ludlow (2013) compared the ideas surrounding educational reform (e.g., in Australia, England, New Zealand, Northern Ireland, Scotland, and Wales), finding that student agency and voice are one of the central aspects of a new policy framework in all of these countries. Despite the apparent enthusiasm, examples of students’ involvement in curriculum development and choice of their own curriculum resources are rare.

According to the OECD report (2019) on “Student agency for 2030,” there is no global consensus on the definition of “student agency.” To unpack the term, student agency refers to the level of autonomy and power that a student experiences in their learning environment. It represents the ability of students to play a critical role in their own development (what they want to learn), practice (how they are learning what they want to learn), and reflection on what and how they learnt (or contemplation on what they wanted to learn and how they wanted to learn it) (OECD 2019). Research has shown that students who have agency in their learning are more motivated, experience greater satisfaction in their learning, and, consequently, are more likely to achieve academic success (Lin-Siegler et al. 2016, p. 297).

As developed in the context of the OECD Learning Compass 2030, student agency implies “a sense of responsibility as students participate in society and aim to influence people, events and circumstances for the better” (p. 4). Agency requires the ability to frame a guiding purpose and identify actions to achieve a goal. It is said to be about acting rather than being acted upon; shaping rather than being shaped; and making responsible decisions and choices rather than accepting those determined by others (OECD 2018). Often, it is said, the term “student agency” is mistakenly used as a synonym for “student autonomy”, “student voice” and “student choice.” The authors of the OECD Learning Compass explain that “acting autonomously does not mean functioning in social isolation, nor does it mean acting solely in self-interest. Similarly, student agency does not mean that students can voice whatever they want or can choose whatever subjects they wish to learn” (p. 4, OECD 2019).

Student agency refers to (mathematics) learning through activities that are meaningful and relevant to students, driven by their interests, and often self-initiated with appropriate guidance from teachers (OECD 2019, p. 5). It gives students voice and often choice in how they learn and which resources they use for their learning. This includes the choice of resources that they regard as meaningful for them, and how they use them for their mathematics learning and study (e.g., Pepin and Kock 2019). Beside “choice,” another term commonly associated with student agency is “voice, ”which conveys a situation where students are more active stakeholders in their own learning (of mathematics) and educational journey (study) (e.g., McIntyre et al. 2005). Student voice is not simply about giving students the opportunity to communicate ideas and opinions; it is about students having the power to influence and change their learning and study.

While agency has historically been concerned with the individual’s free will (or lack of it), it is important to acknowledge that we are mostly not free to do whatever we please, and that our actions are constrained by various forces. Studies of agency examine how resources (mostly artefacts) and people shape our actions and decision making, and it can be claimed that agency is a central construct in educational studies that consider teachers’ and students’ actions: these are “channeled by culture, other people and the material world” (including resources) (p. 4, Carlsen et al. 2016). To human and materials agency, Pickering (1995) has added disciplinary agency (of mathematics): “It is, I shall say, the agency of a discipline—elementary algebra, for example—that leads us through a series of manipulations within an established conceptual system” (p. 115). He coins the term “dance of agency” in performance where agencies “emerge in the temporality of practice and are definitional of and sustain one another” (p. 21). This metaphorical dance can be seen in mathematics lessons/projects with DRs: the teacher may take the lead at the start of the lesson/project, but as the lesson/project proceeds, other agents come into play: students, DRs, mathematics (Carlsen et al. 2016). Grootenboer and Zevenbergen (2007) note that mathematics teachers have to engage in a “dance of agency” when to decide to encourage students’ own agency as mathematicians. Boaler (2003) uses the “dance of agency” metaphor when illustrating the importance for mathematics learners to have an empowering identity in relation to school mathematics. To know when to draw on mathematical ideas and to be able to solve mathematical problems is a critical part of the dance of agency according to Boaler.

The OECD (2019) provides key constructs related to “student agency”:

Student agency relates to the development of an identity and a sense of belonging. When students develop agency they rely on motivation, hope, self-efficacy and a growth mindset (the understanding that abilities and intelligence can be developed) to navigate towards well-being. This enables them to act with a sense of purpose, which guides them to flourish and thrive in society. (p. 5)

The notion of identity is often linked to student agency and references to the notion of identity are abundant in the mathematics education research literature (e.g., Boaler and Greeno 2000; Cobb et al. 2009; Sfard and Prusak 2005). Cobb et al. (2009), for example, propose an interpretative scheme for analyzing the identities that students develop in mathematics classrooms; a scheme that can inform instructional design and teaching. They introduce the key concepts of normative and personal identity. Advocates of the construct of identity contend that it enables researchers to broaden the scope of their analyses beyond an exclusive focus on the nature of students’ mathematical reasoning by also considering the ways that students think about themselves in relation to the mathematics and the extent to which they have developed a commitment to, and have come to see value in, mathematics as it is realized in the classroom (Cobb et al. 2009).

It is important to note that agency is not just individual; it can also be exercised within social practices: “Agency lies in the improvisations that people create in response to particular situations” (p. 279, Holland et al. 2003). Another cautionary note relates to the following: promoting student agency should not be confused with providing unlimited autonomy without structure and guidance. A lack of structure and guidance may jeopardize student opportunities to learn successfully, in particular if students are relative beginners in a domain (Kirschner et al. 2006). This may also go at the cost of student motivation. A productive balance between autonomy and structure/guidance needs to be found, which may be a challenging task for teachers and educational designers (Kock et al. 2013).

In terms of student agency and the learning of mathematics, the education research literature has shown that students who have (or develop) agency in their learning are more motivated, experience greater satisfaction in their learning, and, consequently, are more likely to achieve academic success (Lin-Siegler et al. 2016, p. 297; Cobb et al. 2009). It has been said that student agency is one of the most transformative ingredients for learning, also of mathematics (Brown 2009), and that learning mathematics with understanding is best achieved when it is driven by the student/learner (Hiebert et al. 1997). Hence, it can be said that students developing agency in and for their learning may lead to more motivated, more satisfying, and more successful learning.

The definition of agency provided by the OECD (2019) requires students to become self-navigating learners who are capable of making choices, who purposefully make choices about the directions that their mathematical inquiries will take and about the resources (e.g., material, digital, social/human, or cognitive) they may need to undertake such inquiries. Hence, one of the main ingredients of agency development relates to the purposeful choice and meaningful use of resources by students.

The selection and use of DRs in different educational contexts have often been linked with more student-centered approaches to learning in more open learning environments. This includes providing students with possibilities to choose among different opportunities to learn and to build their own learning trajectories, to increase their self-regulated learning (Beishuizen 2011).

Selection/Use of Resources and Influencing Factors

In this section we focus on students’ preferences when selecting among different resources, on the choices they make when using the resources, and on influential factors and support mechanisms that foster students’ agency in using the resources.

Students’ Selection and Use of Resources for Their Study and Learning of Mathematics

In this section we answer the first research sub-question (What kinds of (digital) resources do students select, and how do they use these resources?), first at primary and secondary (school) level and second at tertiary/university level in mathematics education.

Primary and Secondary Level

The textbook used to be the main learning resource for primary and secondary students. It was widely agreed that teachers mediate students’ textbook use and thus students were not provided agency to choose their own learning resources or even select content within the resource (e.g., Pepin and Haggarty 2001). Although Rezat (2013) has already challenged this view with regard to textbooks, this situation has drastically changed in the past years due to the increased availability of DRs. Especially through the internet, students have access to a plethora of different DRs that are intended to support them in their learning of mathematics or can be used as resources to get assistance when learning mathematics. Among these are online dictionaries such as Wikipedia, YouTube videos, learning platforms, chats, discussion forums. and social media. In this situation, primary and secondary students have the opportunity to choose among the different resources and instrumentalize and orchestrate them for their learning of mathematics. In this context, achieving curricular coherence when building their own curriculum that relies on and combines different resources becomes a major issue not only for teachers but also for students (Confrey et al. 2017).

Nevertheless, it is probably still common in primary and secondary education that teachers mediate students’ use of resources in the mathematics lessons, and students have varying degrees of freedom to select content and learning opportunities from the resources. As Rezat (2013) has already shown for students’ use of traditional mathematics textbooks, even a single resource is not necessarily used as a whole in a predictable way, but lower and upper secondary students make choices of opportunities to learn within it. These choices depend on different factors such as the design of the resource, students’ goals and beliefs, their previous knowledge about using this or a similar resource, and the pedagogical approaches that provide the context for students’ learning activities.

Although students have increasing access to a wide range of DRs for supporting their learning, there is relatively little research about how students at primary and secondary level make use of these possibilities for learning mathematics. Barry et al. (2019) surveyed 1212 U.S. students who just left high-school, about their access to, use of, and perceived benefits of instructional materials. Their results indicate that students preferred printed textbooks over digital or electronic texts because print was more dependable, tactile, lacked distractions, and was easier to navigate and notate. Students also appreciated to watch instructional videos when learning mathematics.

Muir (2014) surveyed 153 students in Grades 5–9 (in the Australasian region) about their use of online resources. Forty two of these students were also interviewed. The results showed that students at all grade levels used online resources to help with learning mathematics at home and mostly not directed by the teacher. YouTube was used by 33–68% of the students across the different grade levels on a daily basis. In the case of Google, the range was even between 48% and 86%. Learning platforms such as Mathletics were mostly used at school when directed by the teacher. Reasons for accessing online resources were, among others, boredom and the need to know or understand. Enjoyment was mentioned to a lesser extent.

Tertiary/University Level

There is an increasing number of studies focusing on university students’ selection of (digital) resources to study mathematics from the plethora of resources available (curriculum resources, social resources, cognitive resources, and general resources). A list of resources used by engineering students in introductory calculus courses at three Norwegian universities was compiled by Hillesund (2020) in his dissertation, based on interviews with nine students who were asked to draw Schematic Representations of their Resource Systems (SRRS, see Pepin et al. 2017b) and interviews with university teachers. The list contained more than 50 different resources recommended to or used by the students in the calculus courses. Students tended to select mainly the resources emphasized in the course, but variations were found (e.g., students who focused more on video lectures, while others focused on digital exercises).

Other studies followed a quantitative or mixed methodology, and selected studies used cluster analysis of survey data to look for patterns in the selection of resources and their relations with other student characteristics. In earlier studies in the UK and Norway (e.g., Pampaka et al. 2016) students selected mainly traditional resources, and this was an issue for the transition from school to university mathematics. In terms of availability and combination of resources, the study by Inglis et al. (2011) aimed at identifying whether (and how) a cohort of 534 mathematics and engineering undergraduates from a UK-based university combined the resources provided to them (online lectures, live lectures, support center). By performing a cluster analysis, the authors identified four distinct profiles: students only choosing online lectures; students exclusively attending lectures; students solely visiting the support center; and finally, students using none of the provided resources. Similarly, Howard et al.’s (2018) team examined 522 first year mathematics undergraduates’ choices about attending live lectures and/or watching short online videos for a course offered at a university in Ireland. Students were found to belong to one of four clusters: those predominantly using live lectures (lecture users); those predominately using videos (video users); those switching between both resources but rarely covering the same content with both (switchers); and those using both live lectures and online videos to cover the same learning objectives (dual users). Academic results were the highest for the students who chose to attend lectures to a high level. Meehan and McCallig (2018) analyzed data of another year in the same course estimating also the time spent by students attending lectures and watching videos. They identified six clusters in the data: video preference, lecture preference, high overlap (similar to dual users in Howard’s study), balanced, and two clusters with a low use of resources. A positive correlation was found between the (estimated) mean time spent on lectures and videos and the mean academic result for the course. Student selection of other resources by the students than watching the videos or attending lectures was not considered in these studies (although other resources were offered, e.g., weekly quizzes).

Anastasakis and Lerman (2021) explored the range of resources mathematics and engineering students used when studying mathematics and the ways undergraduates combine these tools. Results from this survey (N = 628) showed that students in their sample used mostly their notes, the university’s virtual learning environment (VLE), other students, and textbooks. Mathematics students were found using more online encyclopedias, the university’s VLE, instant messaging and other students, whereas engineering students reported using more textbooks. By performing a cluster analysis, undergraduates were assigned to five distinct tool-use profiles: the peer-learning group (above average use of other students, instant messaging apps and social media); the online-learning group (above average use of online videos, online encyclopedias, Wolfram Alpha, and the university’s VLE); the blended-learning group (above average use of all the tools at their disposal); the no-users group (below average use of all tools except textbooks); and the selective-learning group (above average use of textbooks, online videos, and lecturers). The majority of mathematics students were distributed across the peer-learning, online-learning, and blended-learning groups, whereas most of the engineering students were found in the blended-learning, no-users, and selective-learning groups.

In the Netherlands, a survey (N = 403) on the range of resources selected and used by students in first year calculus and linear algebra courses was administered to engineering students (Kock et al. 2021). Survey items involved some student background characteristics, the relative frequency of the use of resources, and the five most important resources to study the mathematics course, both at high school and at university. Cluster analysis based on a social network approach showed three clusters (“communities”) of students regarding the relative importance of the resources: (1) students who considered lecturer explanations of content and of problem solving as the most important resources; (2) students who considered the textbook as the most important resource; (3) students who considered other resources, such as materials created by the teacher (e.g., lecture notes), worked solutions, past examinations, and the university’s DLE as the most important. Cluster membership was distributed across the courses. However, between courses, the distribution of cluster membership varied, and within clusters, some differences between the courses existed, which could be partly explained by the course curricula and organization.

The data for these two studies (Anastasakis and Lerman 2021; Kock et al. 2021) had been collected before the COVID-19 pandemic broke out. During the COVID-19 pandemic, Kempen and Liebendörfer (2021) collected survey data (N = 89) on a fully digitalized linear algebra course for mathematics, physics, and computer science students, as well as for students in a program for a teaching degree. They asked how useful different resources were to study linear algebra. Moreover, survey items assessed psychological variables (e.g., mathematics self-efficacy, identification with university mathematics, learning behavior). The authors found that students rated (digitalized) traditional parts of mathematics teaching (lecture, full-class tutorial, small-group tutorial, and lecture notes) as highly useful. DRs (videos, webpages, and chat) were also generally rated as useful, as well as external videos and videos from the teaching team. Communication with other students was rated as the most useful resource. The cluster analysis yielded three clusters: students in the first cluster were very positive about the asynchronous DRs and less positive about the traditional mathematics resources. Students in the second cluster were very positive about the traditional resources and less positive about the supplementary DRs. Students in cluster 3 (the largest group of about 50% of the students) were positive about all resources. There were some relations between the clusters and student characteristics. For example, students in cluster 1 were studying for a teaching degree, were mostly women, and reported relatively low mathematics self-efficacy.

The studies described so far all took place in a context with a clearly prescribed curriculum, where a mathematics textbook or reader, offline or online lectures, and exercise/tutorial sessions were available. Kanwal (2018) observed three engineering students’ selection and use of resources in an online calculus course created for students to work independently and in a self-regulated way. This course did not make use of mandatory lectures. Instead, DRs were provided, such as a digital homework environment (MyMathLab), which also provided help and feedback, tutorial videos, access to a lecturer, and a textbook. A compulsory task for the students was a group project that involved learning and using a programming language. The final assessment consisted of a digital examination, in which the students could use the digital tools. Kanwal found that students, pragmatically, selected those resources that provided the results they needed for the tasks at hand, also with a view on what they could use during the examination. This involved among others the use of electronic calculators (e.g., Wolfram Alpha) and the help provided in the MyMathLab environment. Videos were considered time consuming and only used if necessary. The students also used the programming language to easily produce results of exercises.

In another innovative context, Pepin and Kock (2021) studied the selection of mathematics resources in a challenge-based course at the end of a bachelor program in engineering, using group interviews with six students and their SRRSs. In this course, multidisciplinary groups of students worked on authentic challenges set by external stakeholders (e.g., business or industry) and were supported by process coaches and academic coaches. The students had to define their own project within a challenge. They mostly used resources outside the realm of curriculum resources offered to them in traditional courses. These included “pieces of knowledge” obtained from various sources, scientific papers, websites, mathematics software, peers, and experts in the field. Social resources took on a special role, and the most prominent was the role of the academic coaches: they gave discipline-specific feedback and helped students to re-focus on the project aims, when the students’ ideas went into different directions. Some students mentioned the importance of the basic mathematics courses as cognitive resources, and they valued the knowledge they had obtained now that they were using mathematics as a tool in their projects.

In summary, it can be said that it is difficult to compare the results from the different studies, partly because of the different university contexts, the nature of the courses which were investigated, and the study level (e.g., bachelor, master, year level) of the students. However, the studies indicate that the selection and use of resources by students are complex processes with many factors playing a role. These we investigate in the next section.

Factors Influencing Students’ Selection and Use of Resources

In this section, we answer the second research sub-question (Which factors influence students’ selection and use of resources?). Leaning on the previous section where studies on student selection and use of resources were identified and discussed, we now investigate the factors that influence and shape the selection and use of resources. From the review of the literature, we found that student selection and use of resources were influenced, and often shaped, by the following four factors: (1) availability and nature of resources; (2) nature and structure of the course/s and pedagogical approaches; (3) institutional framework; (4) and student beliefs and goals.

Availability and Nature of Resources

Concerning primary and secondary education, Rezat (2009) has shown how the structure and the visual appearance of the traditional mathematics textbook influence students’ choices of contents from the book. Especially the visual separation of different blocks appears to influence students’ selection of contents.

The advocacy for more open learning environments that support students to build and follow more individualized learning trajectories raises the issue how students can be possibly supported on their individual learning trajectories. This is directly linked with the nature of resources as an influential factor. While students are required to be more autonomous in the building of their learning trajectory, the resources need to offer features that support them in developing this autonomy.

A small number of studies have developed and investigated tools to support students in developing curricular coherence and to support them on their individual learning trajectories. Confrey et al. (2017) developed a tool that provides “scaffolding to visualize and select content, evaluate student progress, and support classroom discourse around student thinking, all bundled around a means to link to outside resources” (p. 719). Surveying 6–8 grade students’ use of the reports they found that some learners are able to use the feedback to improve their learning. In particular, learners reported specific observations in their feedback related to their learning progress and that they went back to the questions they had missed.

Edson (2017) analyzed U.S. high-school students’ use of learner-controlled scaffolding in a digital learning environment with open-ended problems. He found that the scaffolding was used differently by different students and with different problems. However, it was mostly used after exhausting other provided resources and before asking the teacher. Students used the scaffolding mainly for three different reasons: (a) to make progress in the problem exploration; (b) to move group discussions forward; and (c) to reflect on their own solution of the problem.

In the case of GeoGebra, Olsson (2019) showed that the design of the tasks also influenced how students utilized the DR. Tasks that included guidance of how to engage with the tasks, but not how to find the solution, seemed to foster students’ utilization of the DGS (GeoGebra) and engaged them in creative mathematical reasoning.

In tertiary/university mathematics education, the picture is more complex. In particular in their first year at university, students are offered a plethora of resources, which they can choose (or not) for their learning (Kock and Pepin 2018). While the abundance of resources is generally appreciated, students prefer “focused” resources, that is resources that home in on their particular mathematical problems and the course in general. Textbooks that are too “difficult” make students look for other, more easily understandable, resources (Kock and Pepin 2018); students who get stuck in solving homework problems consult internet resources or fellow students through social media. This brings up the issue of the “quality” of resources for student learning: here quality is related to their relevance (for the particular topic or problem), and practicality (in terms of easy to access and use), while quality in terms of effectiveness is not always guaranteed. Puga and Aguilar (2021) found in an interview study among students in Mexico that one of the reasons to use internet resources is that they can be consulted in private, so that students do not have to reveal their questions, doubts, or perceived lack of knowledge to others.

It is relatively common that video recorded lectures are provided for students. As we have seen earlier, students make their own choices whether to use these or not. It can generally be said that short knowledge clips are preferred over recorded lectures (Guo et al. 2014). Trenholm et al. (2012) examined the use of recorded lecture videos (RLVs) in mathematics instruction, through the lens of cognitive engagement. Cognitive engagement was operationalized using the Revised Two-Factor Study Process Questionnaire (R-SPQ-2F), which measures learning approaches on two major scales: surface and deep. In two mathematics courses at two universities, in Australia and the UK, participants administered the questionnaire near the course start and finish. Overall findings were similar in both contexts: a reduction in live lecture attendance coupled with a dependence on RLVs was associated with an increase in surface approaches to learning. Similarly, in their review study, Lindsay and Evans (2021) claim that the literature is consistent in the opinion that students and administrators positively view Lecture Capture (LC) for its utility and flexibility despite the moderately convincing evidence that most institutions face attendance drops. However, most students, they claim, do tend to see attending lectures/watching recordings as an “either-or.” At the same time, the literature reports a negative association between attainment and the use of LC as a substitute to live lectures. The proportion of students who choose to skip live lectures has steadily increased over the last decade as the student campus culture adjusts to LC. It is claimed that when LC is used imperfectly, it apparently provides false benefits and promotes surface learning strategies. These results are clearly a consequence of students’ selection and use (and not a factor that influences students’ selection and use of resources). However, it is likely to have implications for decisions on which kinds of resources should be made available by the university.

Nature and Structure of the Course/s and Pedagogical Approaches

Several studies have shown that students’ selection and use of resources depended on the nature of the courses and related pedagogical approaches as well as on the teachers’ and students’ roles within these courses. In the case of the use of traditional mathematics textbooks in primary and secondary education, Rezat (2009) has already unveiled different ways how teachers’ and secondary students’ use of resources interact depending on the different ways teachers mediate textbooks use. Aldon (2010) also differentiated between collective and private use of handheld calculators by teachers and students and analyzed how this was related to the construction of knowledge. These studies show that at least in secondary contexts it is important for the construction of knowledge that teachers and students share their ways of using the resources.

Especially in the context of game-based learning, studies have shown how the frame of gaming and the nature of the game-based learning environments can influence secondary students’ learning behavior. Wisittanawat and Gresalfi (2021) showed that students’ framing of playing an educational game affected their approach to the game. If the game was perceived as part of the school culture, the opportunities to learn were likely to be perceived and used differently than if the game was framed in the activity of game playing. This is also supported by Ke et al. (2019) who found that students in grades 7 and 8 initially used domain-generic, trial-and-error gaming strategies as sense-making and problem-solving strategies, and rarely carefully read the narratives. However, with continuing involvement in the game-based activity participants showed a steadily rising trend of in-game problem investigation and systematic solution refinement.

Based on a literature review of self-regulated learning in technology-enhanced learning environments involving various school subjects in the Netherlands, Beishuizen (2011) discussed four key characteristics of the technology-enhanced learning environments that influenced self-regulated learning: (1) complexity, (2) interactivity, (3) articulation, and (4) balance. Complexity relates to the affordances of technology to create authentic learning environments which resemble real-life situations and thus foster students’ intrinsic motivation to learn. Complexity can be both an affordance and constraint for successful learning. Complexity is said to have the potential to increase students’ intrinsic motivation, and at the same time it may create a cognitive overload to students. Interactivity as another characteristic may foster learning processes but does not necessarily improve learning outcomes. Articulation relates to making transparent the structure of the learning task and the learning process by visual means and is seen as an important merit of technology-enhanced learning environments fostering both, learning and self-regulation. Finally, balance relates to the requirement to balance the role of the technology and the teacher in technology-enhanced learning environments, as too much reliance on the technology creates dependent learning behavior.

Regarding tertiary education, Kock and Pepin (2018) examined students’ use of all available resources in two lecture/tutor group-oriented courses, calculus and linear algebra, with the aim of understanding how students coordinate their learning. They found that some students continued to rely on their high school habits concerning the use of resources. Others, who were not satisfied with the textbook or the lectures, made their own selection of resources, part curriculum, part social, part general, often in an idiosyncratic way, with the aim to pass the exam. A segment of the student population felt lost owing to the variety and volume of resources available to them at university. To explain this, Kock and Pepin (2018, p. l) argue that “the course organization and the alignment of curriculum materials with the learning goals had an impact on the students’ choice and use of resources.” They argue that when there is an alignment between these factors, students can direct their learning to appropriate resources for understanding (i.e., optimize the student-artifact-mathematics relationship). In another study in a challenge-based learning environment (Pepin and Kock 2021), the focus was on project aims, rather than on examination-related goals. These aims, together with the feedback students received from their supervisors and the absence of curriculum resources, directed the students to look for resources themselves; this was one of the factors distinguishing the use of resources in this course from that in lecture/tutor group-oriented courses.

In this vein, Kanwal (2018) examined a resource system in an online calculus module which featured a dedicated online interactive system “MyMathLab,” tutorial videos, and a textbook. MyMathLab contained formative assessments and students’ assignments. Kanwal (2018) contends that examination goals drove students’ resource choices and techniques, referring to this as students being pragmatic. An example of this pragmatic approach was that students identified videos as being useful for understanding concepts; however, they were also considered as time-consuming in comparison to the support of MyMathLab. Kanwal (2018) also found that websites were being used for explanation of concepts and to check answers. Whether it is the use of university-provided resources through a Learning Management System or external resources such as websites, students are embracing online resources and research shows that students’ attitudes toward these are positive even when the academic results are mixed (Trenholm et al. 2012).

In a study to analyze students’ resource systems in introductory calculus, Hillesund (2018, 2020) examined students’ Schematic Representation of Resource System (e.g., Pepin et al. 2017b). This consists of a student drawing a representation of the resources they used for their module. Analyzing the representations, he noted that students organized their resource representation through considering either categories such as course organization, situations, the purpose of the resources, or the features of the resources. Hillesund distinguished four phases in students’ study processes: introduction (with an emphasis on curriculum resources); practice and evaluation (e.g., using social resources and digital calculation tools); and explanation (e.g., looking for videos that provide additional explanations). Several factors appeared to influence the selection of resources: particular phases of the course (e.g., exam preparation), the difficulty of exercises, the familiarity of resources for the student and how efficient they could be used. Similarly, Wood et al. (2021) found, based on student interviews regarding the use of DRs in flipped and non-flipped classrooms, that student decisions on selecting DRs were based on what they felt would best support their learning, on the ease of accessibility and the type of information they were looking for, on their personal preferences, and on their beliefs about learning. These ideas seem to align with the notion of Actual Student Study Paths (ASSP; Pepin and Kock 2019) which describe how students show different, and often idiosyncratic, pathways to study with the help of resources.

We note that investigating what resources undergraduates actually use when studying mathematics is important due to the associated costs for acquiring or developing certain technologies and recourses. Moreover, understanding undergraduates’ resource preferences has never been more timely given the rapid rise of online learning that started during the COVID-19 pandemic. At the same time, we understood that student use of resources by university students changes during the course of their study (Stadler 2011).

Institutional Framework

Altogether, there is comparatively little research about the institutional framework as an influential factor on students’ selection and use of resources. Comparing different “institutions” (e.g., school – university, university in different countries), it becomes clear that different institutions have different expectations that are manifested in the notion of “didactic contract” (Brousseau 1997), also in terms of student use of resources (Pepin 2014). In many countries, it is usually the school at primary and secondary level who decides which textbooks are made available to students. The didactic contract used to be that teachers mediate textbook content and decide when and how to use resources (e.g., Pepin and Haggarty 2001) The transition from school to university reflects a transition from one institution to another, and hence the didactic contract. Gueudet and Pepin (2018) investigated the didactic contract at three levels: institutional, mathematical subject, and specific mathematical content level, and link it to the use of resources by students. They have analyzed data stemming from two complementary case studies: one in the UK and one in France. In terms of results, they found discrepancies between students’ and teachers’ expectations: at university, lecturers expect students to use lecture notes to learn and understand mathematical concepts, and as a model for certain mathematical practices, for example, for mathematical proof construction (Gueudet and Pepin 2018). These gaps were said to cause misunderstandings and prevent students from learning mathematics effectively at university level. Hence, course leaders could make their expectations more explicit in terms of resource use and in turn provide more support for an appropriate resource use by students. Furthermore, they also contended that the university institution could contribute to a better alignment of the resources offered with actual student practices. It is clear that institutions have an impact, through their educational visions (e.g., Challenge-Based education) and by means of the pedagogical approaches that are fostered (in those visions), which in turn shapes the available infrastructure for students (e.g., availability of mathematics support centers; DRs and learning environments).

Student Beliefs and Goals

In Schoenfeld’s (2011) theory of goal-oriented decision making, resources are understood as people’s knowledge in the context of available materials. Together with their goals and orientations (beliefs, values, dispositions, etc.) they are expected to explain persons’ actions. In line with Schoenfeld’s (2011) theory, students’ goals are considered a key factor for the selection and use of the respective resources, whereby students’ goals are understood as “the conscious or unconscious aims they are trying to achieve” (Schoenfeld 2011, p. iv).

Regarding primary and secondary education, selected studies report on investigations of how secondary students’ beliefs influence their use of resources. For the case of the traditional mathematics textbook, Rezat (2009) had shown how students’ goals and beliefs about the structure of the textbook and about learning of mathematics guide their choices of contents from the textbooks. In a survey with 442 high school students from Iran, Zeynivandnezhad et al. (2020) investigated how the two constructs (a) students’ conception about mathematics and (b) students’ study approaches influence how they use DRs for learning mathematics. They found significant relationships between each of the two constructs and students’ experiences with DRs. Furthermore, they found a significant relationship between the two constructs (a) and (b).

In terms of tertiary education, selected studies (e.g., Anastasakis et al. 2017) indicate that students, in their selection of resources, have been motivated by the goal to be successful in examinations (and to obtain high grades). Göller (2021) made an inventory of the resources used by students when studying for mathematics modules (at a university in Germany), and explicitly related these to their learning goals. The most widely used resources were those that the university provided for the students and their own notes. The use of resources, for example mathematics textbooks, was specifically linked to the study of worked examples, which were said to help students to prepare for examinations; albeit this often led to emphasize the surface aspects of the examples (Biza et al. 2016). Relating the use of resources to examination grades, a study by Inglis et al. (2011) found that students who attended lectures or used the university’s mathematics support centers had higher grades than students who often watched online lectures. In Göller’s study (2021), the institutional framework implied two central performance goals (expressed by all first-year students): passing the exams; and achieving 50% of the points for the weekly exercises to achieve admission to the exam. Selected learning goals were also mentioned, but overall, these two performance goals were so dominant that strategies to achieve them often superseded strategies to achieve learning goals.

However, student goals also depend on the stage of mathematics study at university. For example, first year students are more likely to pursue the goal of “exam passing” than their fourth year (master) peers, who typically envisage particular goals when working toward their master (e.g., often related to the relevance of mathematics in a professional environment; see Gijsbers et al. 2020).

Results

In this section we answer the main research question (How do students develop agency in selecting and using DRs for their learning of mathematics?) and develop insights that are elaborated upon in the conclusions.

As a first step, we map the landscape (“figured world,” see Holland et al. 2003) of (a) primary and secondary and (b) tertiary students’ opportunities for developing agency in selecting and using DRs for their learning of mathematics.

  1. (a)

    Primary and secondary students’ landscape:

    • Institutional affordances/constraints: approved textbooks (with some associated online resources) are provided, which DRs are available (or not) depends on the school environment.

    • Student opportunities for use of DRs for agency development depend on availability of DRs at the school, and teacher mediation of the DRs, that is on teacher pedagogy, or on self-regulated use outside of school, opaque to the teacher’s eyes, without influence, guidance, or support by the teacher

    • Goals lie in nationally agreed curricula, examinations, and teacher beliefs (of student beneficial learning with DRs), rather than student aims and goals and how they can best learn with DRs. Exceptions are emerging in the form of game-based and inquiry-oriented environments, but these are most often encountered in academic research setting and have not yet become widespread in schools.

  2. (b)

    Tertiary/university students’ landscape:

    • Institutions set the framing for opportunities for learning with DRs: for example vision of inquiry- and problem-oriented forms of learning where DRs are essential.

    • Through particular course designs, teachers can provide agency development opportunities: for example they offer open-ended tasks where DRs are a necessity. Students can choose innovative/CB courses (or not) depending on their “comfort zone” and willingness to develop or exercise agency. Students are likely to exercise agency concerning the use of resources when they consider it beneficial to do so: when it helps them master the material easily, when they think it contributes to passing the exam, or to accomplish the aims of the CB project they are working on.

As a second step, we look across these landscapes. Our analysis of the research literature about students’ use of resources and factors influencing this shows that students exercise agency in selecting and using resources, at primary, secondary, and tertiary levels. While at school, students are provided with a limited range of resources, surveys show that students select and use additional and different resources at home. Qualitative studies show that students also make choices according to their aims and goals when using the resources and are able to reflect on their learning accordingly (Confrey et al. 2017). At university level, the institutional framing is often quite different to primary and secondary mathematics education, partly because primary and secondary institutions/schools are bound to national curricula (which offer particular constraints). Here, students get access to and are invited to use a larger number of DRs. Student choice and use of the resources on offer then depend on various factors, among them the nature and structure of the course and its associated pedagogical approaches, and students’ beliefs concerning their learning of mathematics (e.g., “I succeed by doing lots of exercises”) and their actual goals. In traditional courses the goals tend to be examination driven; in project-based or challenge-based courses the goals are defined by the problem the students are trying to solve. Students tend to be pragmatic and look for efficiency in their selection of resources. They use curriculum resources when these are aligned with the course goals and examination (e.g., past exam papers). Under particular conditions (e.g., when looking for help) they select internet resources, which offer an almost unlimited amount of information. As a possible downside, students note that there can be too much information, and one can get stuck in the search process (Puga and Aguilar 2015). Implicitly or explicitly, students use quality criteria in their selection of resources, such as the relevance of the resource and their ease of use. Websites are considered more reliable when they are recommended by peers or lecturers and videos when they have received a high number of “likes” (Puga and Aguilar 2021). Students tend to continue using a particular internet resource if its use worked for them in the past (Puga and Aguilar 2021).

However, the results also show that all these instances of exercising agency are very diverse among students and that they are often not systematically considered or triggered by teachers or particular pedagogical approaches. At school level, students develop their agency mostly out of school and thus also out-of-sight from their teachers. At university level, they often share their resources with peers and friends. Nevertheless, the results indicate that at all levels of education, teachers and their choice of pedagogical approaches influence students’ use of resources directly or indirectly, and that more explicit communication about the mutual use of resources could be supportive.

In our review we identified four factors that are likely to influence students’ choice and use of resources: (1) the availability and nature of resources (which would include, e.g., the structure and visual appearance of resources; the nature of the tasks implemented in digital resources); (2) the nature and structure of the course (including, e.g., nature of problems as part of the course) and its associated pedagogical approaches (e.g., teacher mediation of the use of resources); (3) institutional framing; and (4) student beliefs (e.g., about the nature of mathematics and its learning) and goals. These factors could serve as hypothetical variables to be systematically implemented in learning environments to foster students’ agency in selecting and using resources.

As a third step, by considering the answers to RQ1 and RQ2, we try to develop understandings of the learning environments that may foster the development of student agency. As a general line, students developing agency in selecting and using DRs for their learning of mathematics is likely to be related to an individual and/or social construction of knowledge (“(social-)constructivist approach”), and student-centered approaches have been seen to play an essential role in this process. Over the decades a number of student-centered pedagogies such as Inquiry/Problem/Project/Challenge-Based Learning (I/P/P/C/-BL) methods have been developed and investigated; these approaches are often conducted in teams or small groups of students. Such methods have been commonly used in mathematics education at school (e.g., exploratory activity in mathematics, Schoenfeld 1985) and university level education, in particular in mathematics in engineering education at university (e.g., Pepin et al. 2021). For example, Roddick (2001) reported that students who follow an IBL-based mathematics course tend to follow a conceptual approach in solving problems, while students who follow traditional teaching tend to follow a procedural approach in problem solving. As another example, Mokhtar et al. (2010) found that PBL encourages students to search for information and resources and that this stimulates thinking. Uses of student-centered approaches in mathematics instruction have been reported to result in similar (Gijsbers et al. 2020) or sometimes better exam scores (Jaworski and Matthews 2011).

From our review of the literature, one of the ways to develop and exercise agency must include the active engagement of students, and their interaction with learning resources is likely to include their selection and use of these resources, be they e-based (particular resources) or social/human (e.g., tutor). When searching for, selecting and appropriating these resources, they develop not only agency, but also knowledge and social skills. This is in contrast to non-interactive approaches where the student is the receiver of the delivery of knowledge from the teacher, and in which the student is only minimally the participant. Such approaches often do not address the needs of most students (Abdulwahed et al. 2012).

Another issue emerging from the literature that is likely to influence students’ developing agency is related to the authenticity of the task or problem they engage with. Using authentic and real-world tasks/problems is considered essential in student-centered approaches such as PBL (Mokhtar et al. 2010). Dealing with real-world problems is likely to motivate students, for example, when they regard it as their job as future engineers to solve such authentic problems, or when they see the relevance of the mathematics in the problem. Gijsbers et al. (2020) claim that providing “interesting” (for students) and authentic contexts in mathematics problems/assignments in upper secondary school mathematics instruction increases students’ interest in the subject and their perception of applicability and relevance. In terms of agency development, for authentic problems, students have to identify and appropriate DRs (e.g., at university) in order to be able to provide prototype solutions.

Recent approaches to Challenge-Based Learning (CBL) involve students being offered “grand challenges,” from which they themselves identify a particular problem they will address (Kock and Pepin 2021). Students typically search, identify, and use related digital tools, design and create a prototype solution to the problem in multidisciplinary groups (Gallagher and Savage 2020). CBL is considered a student-centered learning (and teaching) approach, where students are actively involved in choosing and developing their own learning trajectory (Pepin and Kock 2021). In this way, CBL experiences can engage students in ways of thinking and learning authentic to the engineering profession, which is said to contribute to deeper learning and meaning making than traditional lecture-based courses. In the case of mathematics, an abstract and pure science, an authentic challenge-based task means something different than in the case of, for example, engineering (Dahl 2018), and research is still being conducted to establish how students can successfully develop advanced theoretical knowledge and skills through such tasks.

Another issue for agency development relates to the form of working, for example working in groups or teams. Working in groups encourages discourse in the classroom and among students. Encouraging communication among students has been an essential element in a calculus reform course (Roddick 2001). Jaworski and Matthews (2011) report the use of small group discussion of inquiry-based mathematics problems for creating conceptual understanding among engineering students. In CBL-related work forms, students discuss and work in multi-disciplinary groups. During their work on the project and in discussions, students realize the value of different perspectives, critical thinking, and reflection (Pepin and Kock 2021).

Of course, one of the main “enablers” of students developing agency rests on the nature of the digital resource itself. Technology has typically been named as the main driver for innovative approaches in mathematics education. While Matlab has been particularly popular in mathematics courses intended for engineering students, GeoGebra has been used to promote inquiry and facilitate conceptual understanding of students in school and in a first-year university mathematics course for engineering students (Jaworski and Matthews 2011). Advances in digital online tools are said to enable students’ autonomy in the learning process. Specific online learning services provide support for mathematics instruction in higher education, such as MyMathLab. Wikis and online forums have been used to facilitate discourse and collaboration among students learning mathematics. These are some of the many examples of how technology has influenced innovative teaching and learning approaches.

However, it is likely that students’ beliefs concerning mathematics (e.g., the goals for its use) as well as their levels of agency development (to select and use particular resources) are influenced by both the content and its presentation in curriculum resources, and the ways they can potentially be used for the learning of mathematics. Therefore, digital curriculum resources and their use may also have the potential to change students’ beliefs and agency development in mathematics education, as well as their levels of participation/engagement. In terms of content, there are indications that the content of digital (curriculum) resources needs to take more account of students’ interests and career aspirations. As important as the content of digital curriculum resources is the way they are expected to be used in the classroom.

Conclusions

In pedagogical terms, and considering the four factors influencing students’ choice and use of DRs, we claim that learning environments beneficial for students developing and exercising agency can be characterized by, or include, the above outlined characteristics: (1) student-centeredness; (2) active engagement of students; (3) authenticity of tasks/problems; (4) forms of working that foster dialogue/communication among students; and (5) the nature of the DR itself that fosters student agency – we contend that these environments can be called “agentic”: they are environments where students are provided with opportunities to exercise and develop agency.

In theoretical terms, referring to our original understandings of and focus on “student agency” (see section “Student Agency”) we now understand the term “agency” as “distributed agency”: distributed over the different “agents” involved (in the activity), may they be students, teachers, (digital) resources (referred to/involved in the learning environment), or the mathematics. In the different studies, we see these agents interacting in the “dance of agencies” (Pickering 1995), that is each agent contributing to the activity. Hence, in line with Carlsen et al. (2016), we prefer to refer to “distributed agency.”

Moreover, we consider the agency in teacher-student and student-student interactions to be of a different nature than the “resource” agency, which we think is related to the student – environment affordances (see instrumentation-instrumentalization; Trouche 2004). As outlined earlier, the features of the resource may have an (unintended) influence/agency on practice (of the student as well as the teacher). Hence, while acknowledging student (and teacher) agency, there is also the “resource agency” (and its influence on practice) that may often be underestimated. Referring to the Vygotskian mediational triangle (subject/tool/object), Cole (1996, p. 119) claims that “the incorporation of tools into the activity creates a new structural relation in which the cultural (mediated) and the natural (unmediated) routes operate synergistically.” This, we claim, is also a case of distributed agency (and mediation in activity). In other words, we think it is important that more explicit attention is paid to the “distribution of agency” over the different agents, in particular the “resource agency.” How exactly the different agents exercise their agency and how distributions of agency can be optimized to learn mathematics in different educational situations have not been fully explored yet and are topics for further research.

In that line, our contribution can be seen as an extension of the “resource lens” usually focusing on the study of teachers’ work. Indeed, it provides a way to investigate how students appropriate and adapt the resources they use for and in their learning. The focus on the processes of selection and use of digital resources allows to identify which affordances, constraints, and knowledge guide the decisions that students make, individually and collectively, about which resources to use, and when to use them.

Another important contribution presented by the manuscript is the “institutional dimension” that regulates the availability of resources and their social uses and that relates to other aspects such as the beliefs and goals of the students. This organic view of students’ uses of resources highlights the social nature of resources and the mathematical work itself. The various studies in different institutions and courses show similarities, but also differences among groups of students and their preferences and choices regarding resources. Rarely studies relate student preferences and choices to student results in terms of course assessment. Further research in this relation could deepen our insights into more and less successful student strategies of exercising agency.

The notion of “student agency” in the selection and use of digital resources is proposed in terms of the students’ self-regulated learning and exhibits the complexity of mathematical students’ work involving personal orientations, feelings, identity, beliefs and responsibilities. We contend that this is an important and necessary dimension to understand students’ work in mathematics education.

Cross-References