The rapid shift towards online learning environments has dramatically increased the self-directed nature of higher education, making effective goal setting an essential skill. In her research, Gabrielle Martins van Jaarsveld investigates how learning analytics can be best applied to support students in effective goal setting.
Having obtained her master’s degree in psychology two years ago, and now teaching some of the bachelor courses she attended herself, Gabrielle has first-hand experience of the speed with which digitalisation is changing higher education. “The way students get their course materials, how assessments are carried out, it has already changed quite a lot again,” she says. “There is nobody looking over your shoulder. More than ever before, students have to direct their own learning activities.” Whereas some students thrive in these environments that heavily focus on self-regulated learning (SRL), others could use some support. “It all starts with effective goal setting. I want to use the opportunities offered by digitally enhanced learning to provide those students the support they need in setting effective goals.”
More than ever before, students have to direct their own learning activities, set their own goals
Effective goals
‘I want to study for seven hours’ is not an effective goal as it should be much more detailed and include a plan towards achieving it. Therefore: ‘This week I will study for one hour every morning, watching my lectures and making notes. This will help me be better prepared for my classes next week’. Gabrielle aims to build a tool that will help students through the process of effective goal setting. “Providing such support is a fine balancing act,” she says. SRL theory states that, to feel motivated to work towards a goal, students need to feel some form of autonomy in what they do and how they do it.” A second major challenge, common to all SRL support tools, is that the students who are most inclined to use them tend to already be high achievers – very organised, good planners. “The idea is to make the support tools as personalised as possible so it will reach a broad audience, hopefully including those students who most need it.”
(More than) a delivery method
Gabrielle spent the first year of her PhD reviewing goal setting intervention literature, from home. “I started in the middle of COVID-19, a very strange time to start a new job,” she says. “It was a year before I met my boss in person. But LDE-CEL has a very welcoming PhD community – a good place to share our expertise and our struggles. And, truth be told, I quite liked being away from all the office buzz while doing all this intense reading and writing.” The review showed that technology for the purpose of goal setting intervention tends to be used as a delivery method. “Typically, it comes down to replacing pen and paper by an online form using the exact same questions. My goal is to use technology to make these interventions adaptive and personalised.”
A chatbot can reach out to students, to reach those most in need of support
Chatbot
By now, two years into her PhD, she has settled on her tool being a chatbot (also called a conversational agent). “Next to providing the option of making the tool adaptive and personalised, it also increases engagement,” she says. “A chatbot can reach out to students, for example by sending a small notification, rather than requiring students to initiate the interaction.” She uses a design-based research approach to continually improve her chatbot in an iterative process. An early implementation will be used to test if the quality of goal setting improves with the chatbot offering guidance about how to set goals and/or providing feedback as to the quality of their goals. The bot combines scripting with natural language processing, and she built it herself using online libraries. “I’m in the fortunate position where my topic falls on the border between educational technology and educational sciences.”
Personalisation
Next, she will use learning analytics to personalise her chatbot. At its core, learning analytics makes use of trace data collected by a digital learning platform – a student has watched a certain video or has scrolled through the course manual. But it can also include self-reported data and even personality traits. Gabrielle: “We are particularly interested in perfectionism and self-efficacy, and how these relate to the kinds of goals students are setting. Highly perfectionistic students tend to set far too difficult goals, and we want to see if this pattern emerges in our current research. We can then adapt the chatbot to offer them feedback that may help them in setting more realistic goals.”
Closing the gap
Right now, her test population consists of bachelor students, mostly from the Erasmus School of Social and Behavioural Sciences, but she intends to broaden this scope. “Technology focussed degrees, such as aerospace engineering, may have a somewhat different student population when compared to the social sciences. We want our chatbot to be generalisable.” She feels strongly about making her research available to a broad audience – students, teachers and the general public. But most of all, at the end of her PhD, she wants to have built a tool, based on research and grounded in educational theory, that will be of actual use to students. Gabrielle: “In my experience, there is a disconnect in educational technology between research and practice. I want to close this gap.”
I want to close this gap