There has been little Southern African research attention on the potentials of m-learning to support quality mathematics learning for young children and their caring adults. This article argues that m-learning research has shifted from claims of being promising to claims of effect in educational settings of both classrooms and homes. This is particularly the case in mathematics, where there is increasing evidence of positive (although modest) improvement in learning outcomes.
This article modifies an analytical framework for initial descriptions of m-learning interventions. Comparison between interventions in the Southern African Development Community (SADC) context is then possible.
Three large-scale m-learning interventions focused on early grade mathematics in the SADC countries.
Targeting the early grades and building on an existing framework for describing m-learning interventions, three large-scale m-learning interventions from within the SADC were purposively selected. The three interventions exemplify a possible way to describe the learning theory and pedagogical emphasis underlying the design of their mathematics programmes.
The cases themselves contribute to understanding the m-learning landscape and approaches to early grade mathematics in the SADC in more detail.
A modified analytical framework is offered as a means of describing m-learning in ways that attend to children’s and caregivers’ use of mobile devices, as well as the underlying learning theories.
A relatively recent (2016) special issue of the
[t]he way forward lies in focusing on teachers and their instructional practices and abilities to deal with contextual realities to forge navigational maps to improve the lives of children in vulnerable circumstances. (p. 2)
However, none of the interventions included in this special issue are related to teachers (caring adults or parents) and their instructional practices for mathematics. Neither did any of the interventions consider the possibility of utilising mobile technologies in this quest. We share Ebrahim and Pascal’s (
This article is based at the confluence of three educational premises relating to early education in the Global South: that improving mathematics and reading outcomes is a global priority; that early interventions are necessary to try and reduce the learning gap evident between children from wealthy and poor backgrounds; and that mobile learning is being looked to as a possible means for children and their caregivers to access better quality educational opportunities. Evidence supporting each premise is presented in brief.
Improving mathematics and reading outcomes is a global priority. The sustainable development goals articulate education as a global priority with ‘Goal 4: Ensure inclusive and equitable quality education and promote lifelong learning opportunities for all’. The recent United Nations report (2018) on progress towards these goals has highlighted the progress with regard to increasing access to education, but laments the poor quality of learning in schools, noting that (Guterres
617 million children and adolescents of primary and lower secondary school age worldwide – 58 per cent of that age group – are not achieving minimum proficiency in reading and mathematics. In 2016, an estimated 85 per cent of primary school teachers worldwide were trained; … [where this is only] 61 per cent for sub-Saharan Africa. (p. 6)
The World Development Bank Development Report has echoed this concern, labelling it a ‘global learning crisis’ and reporting that ‘schooling is not the same as learning’, raising serious concerns about basic levels of literacy and numeracy which are most acute in lower income countries (World Bank
Early interventions are necessary to try and reduce the learning gap evident between children from wealthy and poor backgrounds. The World Bank Development Report (
The policy message is simple and stark: for most children, learning deficits are already so substantial by the middle of primary school that many doors have already closed for them. Whilst efforts to ameliorate these deficits at higher levels are important and must continue for the sake of those who may still benefit from them, the greatest effort is required in the early school years, if not before. (p. 41)
Mobile learning is being looked to as a possible means for children and their caregivers to access better quality educational opportunities. There is a growing body of research directed to considering the educational benefits of m-learning. For example, several studies suggest that m-learning has the potential to extend education resources by opening access to disadvantaged peoples (e.g. women, homeless, offenders, disabled, sick and rural poor) and increase equity of access to education (e.g. Vosloo & Botha
Despite the growing evidence of efficacy of m-learning intervention in general, and in mathematics in particular, one has to question this evidence in terms of at least four aspects: (1) the limited scale of the studies on which the efficacy claims are made; (2) the extent to which they are representative of learning across the globe; (3) the absence of analytical frameworks that allow for comparison between different m-learning interventions; and (4) the lack of detail on their underlying learning theories and pedagogic practices.
Firstly, the very limited scale of studies included in m-learning meta-analyses is a concern. By way of example, in a recent systematic review of mobile and ubiquitous learning practices (drawing from 50 studies), Wong (
Secondly, in relation to representation across the world, most studies of m-learning are based on research in developed-world contexts, specifically in countries classed by the World Bank (
Thirdly, perhaps because m-learning is a relatively new research domain, there are very few common data collection tools and analytical frameworks for describing such educational interventions. Researchers tend to frame their enquiry in relation to experimental designs where m-learning interventions are contrasted to ‘traditional’ teaching interventions. These studies are then included in meta-analysis studies or reviews and judgements are made on whether or not m-learning for mathematics ‘works’. For example, Cheung and Slavin (
Finally, related to the above, studies on m-learning interventions may include a simple identification of the school subject which is in focus but seem to pay very little attention to orienting theories and underlying conceptualisations of how children learn in relation to that subject. The m-learning research does not sufficiently attend to what is (in our view) really of interest: the pedagogy underlying the intervention, how these m-learning tools have been utilised and what this reveals about the learning of young children.
This article makes a small contribution to starting to fill the gaps identified in m-learning research, particularly the paucity of evidence relating to m-learning in developing country contexts. It focuses on three large-scale (tens of thousands of learners), early grade (first 4 years of schooling) m-learning projects in the Southern African Development Community (SADC) countries
This article uses a simple definition of m-learning, as ‘learning through mobile devices (such as smart mobile phones and tablet PCs)’ (Chee et al.
The purpose of this study is to refine a modified analytical framework that could be used to offer initial descriptions of m-learning interventions in similar contexts where the descriptions all attended to the same features.
This article answers the following research questions:
How can a modified analytical framework for m-learning configurations be easily and cheaply used to develop common descriptions of m-learning interventions in SADC countries which utilise m-learning to focus on mathematics in the early grades?
How can additional detail about the mathematical pedagogy underlying the interventions be easily and cheaply obtained from project coordinators?
In answering these questions, it was hoped that an analytical framework for describing m-learning interventions using a common set of spectra (defined in terms of m-learning configurations) would emerge.
The analytical framework adopted for this study considers the m-learning configurations: the potential ways in which m-learning services are designed or intended for use by learners. We briefly describe each configuration in turn with reference to the literature informing them.
Strigel and Pouezevara (
This [Nokia Mobile mathematics] mobile mathematics service was informal (used out-of-school) but supported formal learning (school mathematics) in terms of learning spectrum … The service was towards the mobile end of the kinetic spectrum as the service could be used while the learners … [were] moving …, although this movement was not a requirement for engaging with the service. Finally, in terms of the collaborative spectrum, the service was nearer to the individual end of the spectrum … [in that] individual learners typically worked independently on the service. However, the service included a limited collaborative aspect in that the learners’ points (attainment and activity levels) were visible to each other in a community of mathematics learners, and learners could send messages to other learners from within the service. (Roberts et al.
Roberts et al. (
In the resource-constrained context of South Africa, where consideration of m-learning interventions should focus on redress and equity; we consider this spectrum to be a fundamental consideration. We think that this ranges from free public access to suitable devices and free broadband data on one end, to Bring Your Own Device (BYOD) access models and private individual data contracts for broadband data on the other. Subsidised data (by government or operators) and public investments into improved access to mobile devices fall somewhere on this spectrum. (Roberts et al.
In putting forward the
These arguments are made with specific reference to South Africa, but the concerns raised in this context are relevant to the SADC community, where other member countries are also resource constrained when compared to the Global North.
The modified analytical framework put forward in this article builds on the work of Strigel and Pouezevara (
In terms of the
These strands (or slight variations on them) have been adopted in several international curricula – including Australia, the United States, Singapore and Malaysia (Groves
The five strands are mutually supportive and interconnected (hence Kilpatrick et al.
Analytical framework for m-learning configurations.
Consideration is also given to the relative importance placed on the five strands of mathematical proficiency.
These strands are not on a continuum as the categories are not in a sequence.
To select the cases for inclusion in this article, the authors drew on, and then extended, empirical research work initially collected for the Spencer-Smith and Roberts (
The three projects were selected from the landscape review based on there being sufficient grey literature of their project documentation and support from the project coordinators to participate in the new research study. The three focal projects are: Mwabu in Zambia, Mathematics Curriculum Online in South Africa and Unlocking Talent in Malawi.
In 2016 and 2017, the project coordinators who participated in the Spencer-Smith and Roberts (
Coding and analysis of the updated project data was then conducted making use of the modified analytical framework. For all the spectra (except the detailed considerations for mathematical pedagogy), the second author classified the project interventions against the analytical spectra. This was based on the project description and grey literature. This coding was blind-checked by the first author, and no changes in classification were made. The project coordinators were then presented with the way in which their intervention had been coded for each spectrum in the modified analytical framework. Once again, the project coordinators were invited to validate or change how their intervention had been coded. Some engagement was necessary over the ‘learning spectrum’, when interventions were in support of the formal school curriculum, but where use of the intervention took place outside of school time. The ‘collaborative spectrum’ also required some discussion, as in some cases the usual way of engaging with the content was individual although collaboration was possible (or vice versa).
For the mathematical pedagogy considerations, a different approach was required, as no descriptions of mathematical pedagogy were available in the grey literature. As such, the project coordinators were invited to respond to the following two questions via email:
Please describe, in 3–4 sentences, the mathematical pedagogy underlying your intervention/service (this aims to capture how you approach the teaching and learning related to the mathematics of the service).
Which two of Kilpatrick et al.’s (
While it was expected that all five strands may feature within a single project, how the project coordinators prioritised these was considered revealing of their emphasis and hence their pedagogic approach.
The purpose of this study was, however, to refine a modified analytical framework which could be used to offer initial descriptions of m-learning interventions in similar contexts where the descriptions all attended to the same features. Its intent was therefore limited to the project description level, and there was no intention to make evaluative comment on efficacy or impact (which would require far more empirical work). The data collection was limited to grey literature and direct engagement with the project coordinators.
This research drew on secondary sources available in the public domain to select possible case study projects. The project coordinators were then contacted via email and asked for their voluntary participation in the research. Upon their agreement, the case descriptions were circulated back to the project coordinators for validation (on three different occasions over time). No empirical data at the level of children were collected. As such, this research applied the ethical principles of voluntary, informed consent for research participants.
In this section, we offer a brief description of each of the three exemplar projects, followed by an application of the modified analytical framework.
For the Mwabu project, learning takes place in school during mathematics lessons and supports the formal school curriculum.
The Mathematics Curriculum Online project learning takes place in school time during mathematics lessons and supports the formal school curriculum. Learners are stationary while engaging with the mathematics content. Most of the interaction with the device is individual, although there is the potential for collaboration. Because of there being a limited number of devices per class, the weekly ‘Brain Quests’ are completed collaboratively (with two learners per device); however, the ‘Formal Assessment Tasks’ are completed individually. Schools use devices supplied by the provincial education department, by a donor or by the schools themselves. Data costs are paid by the school or as part of a Provincial Education Department initiative. There are annual subscription costs for Mathematics Curriculum Online that can be paid by the school or as part of a wider Provincial Education Department or donor programme (most are paid by Education Districts within the provinces as part of their mathematics curriculum programmes). The intervention prioritises procedural fluency and productive disposition.
The
For the Unlocking Talent project, learning takes place during formal schooling while learners are stationary. The learning centre where the devices are accessed is another classroom in the school. Most engagement happens individually, although there is opportunity to collaborate. Learners are provided with both a device and the data required to access the service. There are no subscription costs. The project prioritises conceptual understanding and procedural fluency.
The project descriptions included reference to the reach of the intervention in relation to the number of learners, geographic spread, technology used and implementation partners. Developing the project descriptions was not time-consuming, and there were only minor changes requested from project coordinators or managers. In one case, there was a name change and a new website to consult; in all cases, the reach of the project required updating.
The original coding of interventions to particular spectra was largely uncontested; however, the spectra were not binary, and finding middle ground to reflect combinations and their relative weighting required some further engagement. This was anticipated and was the motivation for referring to the dimensions under consideration as ‘spectra’: these were not expected to be binary ‘either or’ categories, but the contrasting poles were designed to solicit engagement about particular aspects of the mobile learning configuration. For example, the
This is not necessarily the case as many sessions require the children to do active things when working in their maths work (for example, making a market stall and ‘playing markets’ with pretend money). This is described as an activity to the teacher who then enables this to take place in the classroom. There are, however, also static learning elements on the tablet where a pair of children are working on a learning content with one tablet and are interacting with the learning through questions on screen. The children need to work to answer these quiz questions sometimes on the tablet and sometimes in their books. (C. Stead, Mwabu, pers. comm., 06 October 2016)
With this feedback, we noted that there were ‘mobile moments’ within the service design and requested an estimate of what percentage of the time spent on the service involved learners moving (and not just being at their desk). It was then explained that:
[T]he children will be counting using beans, or stones; so active but at their desks! This is encouraged a lot, so this level of doing is encouraged about 50% – 60% of tasks. However, the big active sessions are probably a lot less frequent. (C. Stead, Mwabu, pers. comm., 11 October 2016)
So, ‘big active sessions’ involving, for example, the market stall with teacher-initiated activity involving the children moving around were estimated at ‘say, 10% of activities’ (C. Stead, Mwabu, pers. comm., 11 October 2016).
In relation to the second research question about the mathematical pedagogy underlying the interventions, we provide the project coordinator responses to the two questions posed, in full.
For Mwabu, the project coordinator responded to the request for 3–4 sentences describing the mathematical pedagogy underlying their intervention or service as follows:
The mathematical pedagogy in our product is based on active, enquiry-based learning, where children are taught maths both as a set of skills, knowledge and understanding, but also as part of their wider learning. For example, in Year 4, when learning about ‘the Island’ as a topic, the children are asked to prepare plans to help the builder build a new school, they need to use length and measure to complete the tasks as well as area and perimeter. In year two, the children learn about the market, and during that topic, make a market stall and apply their newfound knowledge in money to running their pretend shop. (C. Stead, Mwabu, pers. comm., 04 October 2016)
The Mwabu project coordinator prioritised conceptual understanding and procedural fluency, although recognised that the other strands feature to some extent as well. This prioritisation was explained as follows:
The MCO Director responded to the request for 3–4 sentences describing the mathematical pedagogy underlying their intervention or service as follows:
Maths Curriculum Online provides weekly consolidation exercises ‘Brain Quests’ that are exactly mapped to curriculum content for that week. The Brain Quests provide examples of multiple questions styles of increasing difficulty. This structure supports teachers to cover the entire curriculum at the correct pace and level and to the depth required. Assessment for learning is enabled through the real-time per learner, per question feedback summaries. Teachers are able to address barriers to learning or identify learners with specific issues within the same lesson or amend subsequent lessons. Learners can immediately assess their progress. The feedback provides learners a sense of achievement and this improved confidence is transferred back to paper-based Maths activities/assessments. The termly online assessment summary data encourages collaboration within a grade or a phase, as teachers work together to tackle common issues or share specific interventions that they have trialled. (J. Besford, MCO, pers. comm., 03 October 2016)
Here, tight alignment for curriculum structure that follows a tightly defined weekly ‘curriculum pacing’ is the main focus of attention, as well as assessment for learning that aligns to this policy framework. Of the five strands of mathematical proficiency, the Mathematics Curriculum Online coordinator prioritised procedural fluency and productive disposition. In relation to the former, they elaborated that:
The real-time, auto-marking allows learners to practice, check answers and the[n] immediately review their thinking if incorrect. Learners are exposed to many question styles that allow them to develop processes to apply their understanding to a range of situations. Auto-marking requires accuracy of answers (J. Besford, MCO, pers. comm., 03 October 2016).
With regard to prioritising productive disposition, they indicated:
Learners are encouraged to review their scores, look at where they struggle, seek assistance. For many it is the first time a personal, active participation in Maths is promoted. The use of technology to deliver activities coupled with the immediate feedback has encouraged a level of engagement and achievement that was not necessarily generated previously. (J. Besford, MCO, pers. comm., 03 October 2016)
The Unlocking Talent project coordinator responded to the request for 3–4 sentences describing the mathematical pedagogy underlying their intervention or service as follows:
In maths, concepts and skills build on each other, with increasing levels of difficulty. It is harder to count to 20 than to 10. It is harder to calculate 47 + 12 than 5 + 2. It is harder to subtract than to add. So, our approach is three-fold: (1) use familiar objects (fruit, flowers, cups, fish) to introduce concepts where possible; (2) build up the work slowly, at each stage showing the child ‘how to’ and (3) offer different approaches where feasible, for example number lines for addition. We also aim to make the activities engaging and fun. (V. Shimizu, Unlocking Talent, pers. comm., 05 November 2016)
In terms of the five strands of mathematical disposition, in the case of Unlocking Talent, conceptual understanding and procedural fluency were emphasised, with each being explained in detail:
Factors that might interfere with understanding include screen layout and the amount of material on screen. We work to ensure that screens look legible and uncluttered. This is often quite a challenge, for example where we show an array of 100 items.
The technology itself is a wonderful aid to conceptual understanding. By simply touching the screen, the child can add an object to a set. By touching an object, the child can take it away. The child can drag a missing number into a sequence or move an object from one place to another.
We do our best to exploit what the technology offers. By means of the pointing hand – our little teacher’s hand – and highlighting, and other colour changes, we draw attention to what is going on and help the child to focus closely. Other animations, and sound effects, help too.
Overall, the tablet can often offer more help in conceptual understanding than a busy teacher can – and especially a busy teacher with limited resources and a large class. For example, the teacher may have few objects to hand with which to explain subtraction. But on the tablet the child can subtract at the lightest touch, and from a wealth of objects. (V. Shimizu, Unlocking Talent, pers. comm., 05 November 2016)
The emphasis on procedural fluency was also elaborated upon in some detail:
It was of interest that the other three strands – although not the main emphasis of Unlocking Talent – were also commented upon:
We think that the above three descriptions from the project teams offer additional details about the approach to mathematics and the underlying learning theories guiding the content development. While
This article illustrates a data collection technique and applies a modified analytical framework for describing and comparing m-learing inverventions. In this section, we therefore briefly discuss what we observe when reflecting across the three exemplar cases. These observations and conjectures would require additional research to establish the extent to which these reflect more general trends in the m-learning field in the SADC region.
Mapping the case study interventions to the m-learning configurations.
In terms of the
In the case of the
With the
We discuss the
In the case of the
We trust that this article makes a contribution towards filling at least two identified gaps. Firstly, by describing three exemplar examples of m-learning interventions from the SADC region, we contribute to the m-learning literature from this under-researched and under-documented geographical region. It offers examples of how teachers and caring adults can be supported via mobile technologies in ‘their instructional practices [of mathematics] to improve the lives of children in vulnerable circumstances’ (Ebrahim & Pascal
In particular, we have distinguished the
In addition, we have introduced a new spectrum relating to the
We hope that the modified m-learning analytical framework will be used, adapted and improved, so that the m-learning field starts to have some common ways of approaching the descriptive component of m-learning work. We recognise that further research and contributions are required for the ongoing quest to have commonly–agreed metrics and approaches to measuring and reflecting on the efficacy of such interventions. Without broadly comparable descriptive details of exactly what each intervention entailed – how the m-learning was configured, the underlying pedagogy of the mathematics and then how the learners and their carers engaged with the m-learning intervention – lessons on how the promises of m-learning enhance or hinder educational outcomes will remain elusive.
The authors declare that they have no financial or personal relationships that may have inappropriately influenced them in writing this article.
N.R. developed the expanded analytical framework. G.S.-S. did the bulk of the data collection. Both authors contributed equally to the writing of the article.
The earlier research study, Spencer-Smith and Roberts (
The following 15 countries are members of this grouping (in alphabetical order): Angola, Botswana, Democratic Republic of Congo (DRC), Lesotho, Madagascar, Malawi, Mauritius, Mozambique, Namibia, Seychelles, South Africa, Swaziland, United Republic of Tanzania, Zambia and Zimbabwe.
These were provided in full to each respondent as part of the question, but have been removed here as they have been explicated earlier in the article.
However, there is also an informal version of Mwabu for home learning.
Thus, their target learners are only partially in the early grades.
These are mathematical apps provided by ‘onebillion’.