Keywords

1 Introduction

Learning analytics (LA) refers to “the measurement, collection, analysis, and reporting of data about learners and their contexts, for the purposes of understanding and optimizing learning and the environments in which it occurs” [1] (p. 1382). Its implementation covers various levels—from the micro- and meso-levels, which target respectively individual learners and the institution as a whole, to the macro-level which focuses on a cross-institutional perspective [2]. By bringing together the advances in technological, pedagogical and social development, LA has been regarded as having high potential for enhancing learning and addressing the diverse needs of the stakeholders in higher education institutions [3].

LA has had an impact on changing educational practices and improving learning experiences. For example, in reviewing the case studies on the implementation of LA between 2008 and 2013, Avella, Kebritchi, Nunn, and Kanai [4] found that LA was beneficial to students’ learning behaviours and outcomes, instructors’ performance, curriculum development, personalised learning, and research in the field. Also, Sclater, Peasgood and Mullan [5] revealed that LA enables higher education institutions to enhance teaching quality and improve student retention, and helps students to take control of their own learning.

Despite the growing popularity of LA in higher education, its development in Asia is relatively slow. The existing work has been mostly done in the USA, Australia, and the UK [5]. From a survey of academics’ views on the adoption of LA, Drachsler and Greller [6] found that the responses from Romance and Latin American countries were rather limited when compared with those from Anglo-Saxon countries. There was even a lack of response from Russia, China and India. One possibility is a general lack of awareness of LA in these countries. According to a survey targeting instructors in higher education institutions in China, Xiong and Zhang [7] found that more than half of the respondents regarded LA as their most unfamiliar teaching technology. In analysing the authorship of the Third Conference on Learning Analytics and Knowledge, Ochoa, Suthers, Verbert and Duval [8] revealed that, while Europe and North America contributed a total of 134 authors, there was only one from Asia. The notably lower proportion of Asian researchers in this field may have also constrained the development of LA in the region.

This paper seeks to unveil the development of LA in higher education institutions in Asia and explore the future directions in the field. It presents findings from interviewing academics and senior managers from various higher education institutions in China, Japan, India, the Philippines and Malaysia. They shared their views on and experience of LA, covering their institutions’ position, the progress in implementation, factors contributing to the effective implementation, and challenges encountered.

2 Literature Review

Despite a few papers illustrating LA’s overall development across the globe, the existing literature has yet to provide a thorough coverage of the situation in Asia. For example, Sclater et al. [5] examined the case studies from the USA, the UK and Australia. From the LACE Evidence Hub [9], which collects and summarises the features of LA cases, these three counties together contributed more than half of the cases in its latest collection.Footnote 1 Other literature reviews of LA also have not put regional speciality as a focus. For instance, Arroway, Morgan, O’Keefe and Yanosky [10] covered in general the driving factors, uses, institutional readiness, and strategies to guide LA implementation. Also, Leitner, Khalil and Ebner [11] analysed more than a hundred papers on LA and identified the research strands, LA techniques, limitations of research studies, and stakeholders.

In contrast, the related literature mainly presents the situation or reports particular projects in some Asian countries. The following summarised examples of the literature for several of these countries.

2.1 China

In China, the literature focuses more on the conceptual aspect. Gao and Fu [12] reported a systematic review of the LA literature in China between 2012 and 2014, and showed that they were mostly concerned with the value of analytical data, the importance of LA in constructing a smart learning environment, and the research and application of LA in the West. Jiang, Zhao, Li and Li [13] conducted a detailed analysis of LA dashboard applications to examine their potential benefits and limitations. Meng et al. [14] categorised different LA tools and compared their functionalities, providing a guideline on the selection of learning tools in learning and teaching. Also, Xiong and Zhang [7] collected instructors’ views on LA and identified the major obstacles that restrained its implementation in Chinese higher education, viz. limited sources of data due to the insufficient popularity of e-learning, the difficulty of developing an analytical model, and the limited level of instructors’ digital literacy.

There is a limited number of case studies presenting the practices of LA in China. For example, Zhang et al. [15] collected and analysed students’ data in a college, such as library records, Internet access, and course performance, in order to improve the effectiveness of a programme. Ma et al. [16] assessed the role of instructors for students’ engagement in an online learning environment by tracking the weblog data related to the activities of instructors and students in a university.

2.2 South Korea

In South Korea, Jo [17] developed a dashboard, called the “Learning Analytics for Prediction and Action” (LAPA), which was implemented in a private university with the aims of visualising students’ online learning behaviours, and promoting their development of a smart and personalised learning style. Jo also identified the difficulties of interpreting the visualised LA data, and recognised the importance of LA in providing personalised and timely educational opportunities and feedback to learners on their needs and ability. In addition, Jo et al. [18] suggested more LA components for tracking the weblog data, and discovered factors which were significant in predicting students’ performance, such as the login frequency and regularity of the learning interval in a learning management system (LMS).

2.3 Japan

In Japan, new LA tools and approaches have been proposed for enhancing learning experience. Ogata et al. [19] proposed a system called “System for Capturing and Reusing of Learning Log” (SCROLL), which aimed to help learners to record, recall, and organise their learning logs. Ogata et al. stressed that the system can be further extended to analyse learning content by accumulating data in learning logs, so as to find learning patterns and supply appropriate learning materials in accordance with learners’ habits. Sorour, Goda and Mine [20] collected comment data from a course in the Kyushu Institute of Information Science and applied text mining techniques in order to predict the grades of students. By applying latent semantic analysis, they found that students’ comment data was influential in predicting their grade. Ogata et al. [21] implemented LA based on e-book data in Kyushu University. They tracked and analysed educational big data from the LMS, the e-portfolio system and the e-book system, and found that the time students spent on reading or viewing e-books had a positive effect on their study results.

2.4 India

In the literature focusing on India, Pratheesh and Devi [22] analysed a collection of students’ opinions and argued for the importance of adopting LA in software engineering education. They discovered that most students preferred a technology-based collaborative learning environment, in which LA would be helpful in detecting learners’ learning styles and preparing learning materials that suited them. Also, Boulanger et al. [23] carried out an experimental study in the Anna University in India, implementing an LA system called “Smart Causal Analytics on Learning”, which aimed at “collecting learning traces from any learning domain and analysing those learning traces to extract the underlying competency levels in the same learning domain” (p. 291). They found that the classes which adopted this system generally outperformed the others.

2.5 Summary

Overall, the literature regarding LA in Asia has been limited. Case studies reporting practices of LA cover only certain countries, such as South Korea and Japan, and most of them were only at an initial stage. Although this literature review only provides a glimpse of the adoption of LA in individual countries, there is a scarcity of studies showing the situation of LA in Asia, such as the progress, the goals, the success factors and challenges. This study aims to address this limitation by collecting the views and experience of academics and managers in higher education institutions in Asia on the implementation of LA in their respective institutions.

3 Methodology

3.1 Participants

A total of eight senior academics and managers were invited to participate in semi-structured interviews. They came from higher education institutions in a total of five Asian countries—China, India, Japan, Malaysia and the Philippines. They were all Professors or unit heads in their institutions.

3.2 Semi-Structured Interviews

The semi-structured interviews were conducted in July 2017, which took about 5 to 20 min. Below are the key interview questions. The researcher also asked other follow-up questions according to the interviewees’ responses.

Interview questions

  1. 1.

    Have you heard of learning analytics?

  2. 2.

    Is your institution developing learning analytics?

  3. 3.

    How long has your institution been developing learning analytics?

  4. 4.

    What are the goals of your institution for developing learning analytics? What is expected to be gain from the learning analytics?

  5. 5.

    What is the progress in the development of learning analytics in your institution?

  6. 6.

    What obstacles, if any, have your institution encounter during the development?

  7. 7.

    What are the future plans on the development of learning analytics in your institution?

  8. 8.

    What do you think of learning analytics?

4 Results and Discussion

The results showed that all the participants had heard of LA. Six of them reported that their institutions were developing LA. Most of the other institutions were at the early stage of the development, e.g. less than five years; and some of them are still in the process of planning. In this section, the findings were categorised into the following themes and discussed along with the results of the past research.

  1. 1.

    The development of learning analytics in higher education institutions mainly aims to enhance student retention, better pedagogy, and improve student learning experiences.

The institutions aimed to develop LA to achieve various goals. During the interviews, goals at different levels were mentioned by the participants. At the institutional level, most of the institutions aimed to maximise the student attendance, improve student-teacher interaction, and enhance retention with the use of LA. This finding is consistent with the research done in the UK and US, which shows that one of the drivers for universities implementing LA is to use it as a presdictive tool to identify students at risk of attrition so as to increase their continuation on the programme [10]. One participant mentioned that his university applied LA in order to discover trends and problems in education that could not be identified by using conventional means. This reflected a positive attitude from the university’s senior management towards investment in LA.

At the teacher level, the participants replied that the application of LA in higher education would offer insights for teachers to improve their teaching. Specifically, it was expected that LA would help to achieve the goal of facilitating policy-making to improve pedagogy, meet the teachers’ needs, monitor students’ learning progress, and gather student feedback.

At the student level, most participants expected that the use of LA would improve students’ learning experiences, making learning and teaching more meaningful.

To sum up, the development of LA was considered to fulfil goals in the areas of university administration, pedagogy, and students’ learning experiences and performance. The most important driver for the senior management for implementing LA was to enhance student retention; and to improve the pedagogy and students’ learning experiences and performance were the main goals of academics.

  1. 2.

    The implementation of learning analytics in higher education institutions needs to listen to students’ voices.

Three participants mentioned that a small working group had been formed for the implementation of, or a pilot study on, LA in their institutions. One of the institutions even provided funds for the preliminary research on this topic. As for the progress, five institutions were planning to collect or had been collecting student data for the LA projects. The main sources of data included student demographic information, the LMS login data, information on enrolment and retention, course performance (e.g. attendance rates, assignments, and exam pass rates), and course evaluation.

  1. 3.

    The difficulties of implementing learning analytics in higher education institutions consist of concerns from teachers and students and technical issues of learning analytics.

The participants identified several difficulties in implementing LA in their institutions. In some institutions, the academics hate changes and so it was difficult to get them on board. In addition, they were not happy that some information is shared with students through student dashboard, which may increase their workload. Similar concerns have been found in other studies. For example, Howell et al. [24] found that academics were uncertain about the responsibility they should bear after releasing the findings of LA to students. The follow-up efforts to help the students deal with the negative reports based on LA analyses may have had a significant impact on their workload. To solve this problem, clear responsibility, instructions, and procedures should be provided to the teachers to facilitate their follow-up with the students. It is preferable that an intervention unit to offer necessary and timely assistance and consultancy to students be set up to reduce the workload of the teaching staff. One of the participants mentioned this mechanism in her university which has been functioning effectively and has provided considerable help to students in need.

The participants mentioned that some of the students do not like the idea that the university can track their digital footprints and they are concerned about the issue of data protection and privacy. However, most of the participants indicated that their students were informed that their personal data and data generated from their study in the university would be collected and used for analysis and report purposes once they were admitted to a programme. The practices of the interviewees’ institutions showed that there may not be an option for students to opt out of data collection for LA. The need for informed consent and the option of opting out should be provided in the universities in the planning of LA [25].

In some of the institutions, LA is a new field to be explored. Some participants found it difficult and time-consuming to consolidate data from different sources and make an integral use of them. In particular, one mentioned that sometimes there is no clue to identify the useful and important data in the huge dataset. These comments revealed that the data collection for LA needs cooperation from different departments and units within the university, where support and coordination from senior management is considered of great importance. Hiring data analytics specialists to form a specific working group for learning analytics would be a solution to managing the collected data in an efficient way.

  1. 4.

    The senior managers interviewed from most of the institutions provided sufficient support for the development of learning analytics.

Most of the participants expressed that the senior management at their universities have a positive view of LA. Some of them provided administrative support to coordinate the data collection for the LA projects, while others offered research funds for a pilot study on LA. However, in one university in Japan, the participant said that the senior management was sceptical about the effectiveness of LA and did not provide any support for it.

  1. 5.

    Most of the participants possessed a positive attitude towards learning analytics and consider it as an effective tool for higher education.

The participants expressed that research and the application of LA should become one of the foci in tertiary education. More useful predictive models should be discovered and the findings should be used not only for monitoring students, but more importantly for intervention when students at risk are identified. It is hoped that it can help to personalise the students’ learning process and improve their learning experiences and performance.

5 Conclusions

The present study reveals the trends of LA in Asia, which are not adequately addressed in the previous research. It contributes to showing that tertiary institutions in Asia, though starting late, have gradually become aware of the usefulness and importance of LA. In the interviews, two of the eight participants reported that their universities have been using LA for policy-making and student retention. Four of the participants’ institutions have been conducting preliminary research on LA and plan to make use of it in the near future. The other two participants reported that their universities were not in favour of LA or even doubted its effectiveness, and therefore, no support was received from the senior management for its implementation.

In the interviews, no participant mentioned the views of students. It has been pointed out that students’ views have been missing for a long time in the decision-making on LA and attention should be paid to engaging them in such a process [26]. The students should be empowered and enabled to become one of the designers of their own learning experiences so that they can gain control over their own learning. For developing LA in Asian universities, it is therefore recommended that students’ perspectives should be taken into account in decision-making.

The follow-up intervention based on the results of LA may not be welcomed by students. The students at risk may not want to be identified, as the negative feedback from LA may damage their psychological well-being by labelling them and lowering their self-esteem [24]. Therefore, how to deal properly with the students’ data and provide feedback to them needs a more considerate and informed approach.

Despite the relatively small sample size, the findings of the present study suggest that the development of LA has been slow but is gradually progressing in Asia, and the tertiary institutions in Asia are generally positive towards it. The challenges for the institutions, as raised by the participants in this study, have also been reported and addressed in the relevant literature. It can be expected that LA will gain a more important position on the agenda of higher education development in the continent.