Learning Analytics provides a methodology for the collection and analysis of learning data. Pedagogical research has always been inspired by ideas from applied psychology to discover and evaluate methods to boost motivation and engagement of students. Past research has shown that people often compare themselves with their peers in various contexts, including education. Social comparison has proven to be an efective motivation factor. Most of the recent research is based on using leaderboards to motivate individual comparison or open social student models to enable comparison with the course average. However, students' preferences towards social comparison can vary. For example, some people tend to compare upwards, while others mostly compare downwards, and some do not rely on social comparison at all.
self-evaluations by comparing oneself with their peers. jective well-being which may enhance their self-esteem. The concept of SC has been observed [3, 4] in children as young as two years. We evaluate our opinions, skills, as young as preschoolers. Verof [ 4] proposed that the information is not available, and then, they do so by comparing themselves with each other. It is found [2] that people may compare downwards to increase their subtools introduce social comparison as a tool through gami- sense of pride of older students is afected more by comabilities, and achievements by comparing ourselves to others to define the self. Due to this, Social Comparison is a strong motivator and has always been leveraged in avenues ranging from commercial advertising to political discourse. Technology Enhanced Learning (TEL) environments provide feedback and supportive interfaces to help the students understand their progress towards the learning goals. Knowingly, or unknowingly, educational ifcation elements such as leaderboards and halls-of-fame.
in 1954 which stands on the premise that humans have an innate desire to evaluate their abilities and opinions.
Proceedings of the Doctoral Consortium of the Seventeenth European 0000-0002-7115-4635 (A. Joshi) CEUR htp:/ceur-ws.org ISN1613-073
CEUR
Workshop Proceedings (CEUR-WS.org) concept of achievement begins in elementary school students, while the social comparison orientation increases as they grow older, the autonomous achievement orientation drops. As they grow up, they emphasize on demonstrating superior performance in comparison to others. In a usual classroom, the reward system provokes students to compare themselves socially. Similar efects were observed by Seidner et. al. [5] who noticed that the paring their performances with their peer rather than mastery.
individual diferences in Social Comparison Orientation.
It was found that two factors were responsible for
explaining 38% and 10% of the variance. These two fectors
reflect the perception of abilities and orientation based
on Social comparison. They explained that such SC
information may help ascertain the SC behavior of individuals
and provide them interventions accordingly. It is also
studied that the demographics like age, sex, race, or
socioeconomic status can be a factor of who students compare
themselves with [
have been found efective in improving engagement and
Though SC is a psychological trait, SC can be further participation. However systems focussed on improving
leveraged or manipulated by a researcher, a teacher, or the SC feedback and studying the efects of SC are
relaa learning aid including learning support systems. SC tively recent topics.
can also serve as a feedback mechanism for self regu- One of the popular works in using comparative
visuallated learning [
Rest of this paper is organized as follows. Section 2 the importance of developing the right visualization with
discusses how TEL systems leverage Social Comparisons, respect to the goal.
their diferent forms and the ideas which have been pro- Progressor [30, 31, 32] introduced Social Visualizations
posed in the past two decades. Section 3 explains the in an interface that helps students find relevant resources.
research plan by highlighting the research problem, re- It was observed that due to social comparison, class
leadsearch questions, explaining the concrete tasks that are ers provide a guidance to the rest of the students, and
planned for the next stage, and details of the TEL sys- eventually lead to more engagement, and thus higher
tem that will be behind this research. Section 4 gives a success rates.
brief outline of the experiments which are planned in the Several ideas related to Open Student Models [33, 34]
next few years. The Section 5 concludes this paper with have been explored. Reading Circle [35] combines the
discussion about the expected outcomes of this research, idea of Open Student Models and Social Comparison to
and the role of those outcomes in designing TEL systems encourage students to read. A textbook reading support
that use SC efectively. interface called Reading Mirror [36] shows SC
information uses a grid-like interface that shows a student’s own
progress and the class average with respect to
section2. Social Comparison in TEL s/chapters of a textbook. It was found that most of the
students felt that SC information altered their behavior
SC has been an active area for research in the past decades. positively. More recent works have used these interfaces
The idea of SC was first studied in depth by Festinger [37] for encouraging motivation and engagement. An
in[
This relates with the idea of Self Regulated Learning A study on the efect of a dashboard widget [ 40] for (SRL) [22], which is described as a cyclical process with Massively Open Online Courses (MOOCs) that provides three stages, namely, Forethought, Performance, and Self- students more crisp infromation about their progress as reflection. Social comparison is active and afects stu- well as SC cues improves the course completion rates. A dents’ decisions and actions at all the three stages [23], dartboard like interface for multidimensional comparison From the past works, it is evident that accessing SC information helps students achieve more motivation and This can be divided into following research questions: engagement, and leads to a higher success rates. Mean- RQ1. What are the current state of the art interactions while it has also been observed [41] that peer comparison to show social comparison and what are their efects on doesn not necessarily improve, but in some cases, hamper students’ learning experience? the motivation of students. In some cases that though RQ2. How is the efect of social comparison on motivastudents prefer personalized recommendations, they may tion and engagement related with personality traits? not find peer comparisons as useful or motivating. It was RQ3. Are there distinct efects of using diferent types also found [42] that students’ SC own preferences do not and direction of SC interfaces with diferent students? necessarily align with their best interests. What are these efects?
This leads to a challenge of analyzing the design as RQ4. How to match a student with a social comparison well as the efect of social comparison at a finer granu- method fine-tuned to promote their learning? larity. Social Comparison for better learning experiences needs further exploration. I plan to explore methods and create adaptive SC interfaces that can be efective tool to promote meaningful learning.
Support System that will be used to supplement students’ learning experience. Some of the initial experiments have been thoroughly discussed and planned to occur in the academic year 2022-23. The students will be asked to use Studylens as a Learning Support System (LSS) that will allow them to attempt ungraded assessment tests related to the concepts in the course, and I will analyze their performance and engagement with respect to the interface provided.
Typically, in most TEL and e-Learning softwares, SC information is usually provided to all the users the same way. However it has been found [43] that demographic and cultural backgrounds have a significant influence on self-construals based on social comparison which may afect the motivating factors. Apart from demographics, the SC orientation and direction also determines whether a person is motivated, challenged, or demotivated by SC information [6, 44]. That is, someone might get inspiration from someone who’s performing better than them, while someone else may feel dissatisfied, or envy. [ 45] mentioned that though we all engage in social comparisons all the time, some people are more concerned and influenced by social comparison than others.
The diferences in perception and efects of SC don’t end with demographics and personality - but even at individual level, they expand over temporal and contextual dimensions. We engage in comparisons with others over time[46, 47] or our own past selves[48].
The popular Social Comparison approaches don’t capture all these dimensions of social comparison and implement a one-size-fits-all solution regardless of individual and contextual diferences. The issue with standard onesize-fits-all approaches is that though they work in some cases, they might afect some users rather negatively.[ 49]
This leads to the idea of a system that adapts the SC interface to the user based on their demographics, social comparison orientation, motivation profile and psychological profile.
is driven by the following research questions: • What are the common ways of conveying SC information in TEL tools? • What are the efects of context and direction of
Social Comparison? • What are common systems that allow students to actively engage with Social Comparison? • What are the unexplored directions of utilizing
SC in Education?
based on SPIDER [50] and PRISMA [51]. SPIDER helps summarize the study on the basis of (S) Sample size, (PI) Phenomenon of Interest, (D) Study Design, (E) Evaluation and (R) Research type. PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analysis) provides a structure for conducting the literature search and summarizing the analysis in a detailed manner. Meanwhile for each research that is included in the study, we also explore what was the SC method used and how was its efect studied. 3.2.2. Development of Studylens
Utrecht University. It is a relatively lean implementation of the system which has been used with several university courses over the past few years [42]. The new implementation has been designed to have only the most necessary features that enable us to closely study the impact of social comparison.
Studylens provides students a dashboard for exploring their expertise of the concepts covered in a course. It allows students to attempt self-assessment quizzes, which are associated with concepts and misconceptions that are part of a course inventory. When a student chooses to view the results of the test, they are shown their performance over each concept, and accordingly, remidiation 1–8 material is recommended. Figure 2 shows the organization of a course into constituent Learning Goals each of which have one or more Concepts related to it.
When the next Learning Goal is activated, the student can take the self assessment test. The Knowledge Map is updated that helps student get feedback about their expertise of the topic. Figure 3 shows the current version which is expected to be further updated. In the social comparison setting, the student is shown the average performance of their peers as well. For research purposes, visibility of the social comparison widgets is configurable to provide a diferent view to each student based on their experiment group. The student can explore their knowledge and take the right remedial action through a list of learning resources.
At the time of writing, Studylens is planned to be used as a part of courses related to Evolutionary Biology at Utrecht University in The Netherlands. The courses are conducted over three-month terms and expected to be taken by 120-480 students. Studylens is recommended to the students as a self evaluation tool that can help them find their strong and weak points, and recommend remediation material to improve their understanding of the topics. 3.2.3. Technical Details
at the backend. The database is MySQL, and the front end is based on a popular Javascript framework that provides a highly extensible component based design. The system is designed to provide user interfaces based on the experiment groups a user is allotted to.
A minimalistic Learning Record Store (LRS) is implemented in the database that stores users’ activities in terms of actor (the student), verb (loading an activity, answering a question etc), and object (question or learning material). At later stage, this may be replaced by a full lfedged LRS based on research requirements.
we have planned a project to explore diferent methods for personalized support with focus on exploring SC as a vehicle to motivate students with non-mandatory educational content. Studylens will be used as a learning support system for three courses at the department of Biology. These are all related to Evolutionary Biology at Year 1 and Year 2 of their undergraduate degree program.
Students will be able to take formative tests and a dashboard (Knowledge Map) will display their progress of mastery with respect to the Concepts covered in these courses. The interactive dashboard will help them explore the topics, learning goals, concepts better and de
We have thoroughly analyzed the privacy and ethical concerns related to any experiment of this kind. To mitigate the privacy risks, we have devised that the user’s details in the system will be synthetic, and the teachers of the courses would map the user ids with actual students in the class. Meanwhile the teachers will not have access to the database or any internals of our system. This creates a safe barrier, thus allow anonymity during the data col
The first set of experiments will be concluded by the middle of 2023 which would give adequate insights on the factors that determine the efect of SC on students’ motivation and learning outcome. This will help in the development and refinement of efective interfaces for conveying SC information. This would be followed by providing adaptive SC interfaces to the students and comparing their efects with respect to student controlled and static social comparison.
This research will contribute to the empirical knowledge in Technology Enhanced Learning and Pedagogy domains. The outcomes of this research will allow us to gain a thorough and coherent understanding about how Social Comparison afect diferent behavior profiles, and create a system that adapts to a learner’s behavior and provides them the Social Comparison cues that will motivate them.
The learning support system being built as a part of this research, Studylens, will be used to help students towards Self Regulated Learning. Though the experiments that are planned in the next year are related to Biology students, the tool and the ideas are domain independent and can be easily applied to other subjects and areas like computer science and soft skill training. We believe this research would lead to adaptation of SC methods that help the students achieve their learning goals. lection and analysis. Meanwhile use of the software, data collection, and participation in the motivation profiling surveys will be voluntary.
The aim of any TEL system is to devise the ideas that lead to discovery of more efective learning methodologies. This research can potentially impact how the Social Comparison information is visualized and used in learning softwares.