Research

NLP methods in EdRes
Labour market research
Digitalisation of education
Student support
Phenomengraphy


Computer-assisted phenomenography: testing the use of NLP methods for analysis of students' conceptual understanding

The volume and complexity of qualitative data generated by students, such as reflections on threshold concepts, present significant challenges for educators seeking to provide meaningful feedback. Advanced methods for analyzing large qualitative datasets, including natural language processing (NLP) (Kovanović et al., 2018; Prieto et al., 2019), offer promising solutions. However, existing studies often lack robust theoretical foundations, making it difficult to systematically understand students’ conceptual (mis)understandings. This project proposes a groundbreaking approach by integrating phenomenography and variation theory with advanced NLP techniques, specifically topic modeling, to analyze students’ reflections on threshold concepts. While phenomenography and variation theory have been applied to studying conceptual understanding (Han & Ellis, 2019; Prosser, 1994; Reed, 2006), their combination with advanced analytical methods remains unexplored. This project aims to fill this gap by using theory-supported topic modeling to identify variations in students’ conceptual understanding and uncover mechanisms behind these variations. The project objectives are:

1. To identify and code critical disciplinary concepts needing curriculum design attention based on threshold concepts.
2. To analyze variations in student understanding using NLP topic modeling and variation theory.
3. To create a phenomenographic model representing a nested hierarchy of students’ conceptual understanding.
4. To explore how different reflective approaches correlate with levels of conceptual understanding.
5. To develop and test a novel computer-assisted approach for analyzing students’ conceptual understanding, integrating advanced NLP techniques with phenomenography.

By combining data-driven analytics with phenomenography and variation theory, this study could lead to better-informed instructional strategies and more personalised educational interventions. Additionally, it will challenge traditional qualitative research norms, proposing innovative workflows for analyzing qualitative educational data.

The project has been accepted to present at the EARLI 2023 conference. The project description can be found l here.  

Labour market research

My interest in labour market research includes understanding how decision-making processes unfold, examining the impact of incentive contracts on job satisfaction, and exploring the implications of precarious working conditions.

Traditional economic models assume rational behaviour on the labour market, but research from behavioural economics and neuroeconomics shows that individuals deviate from rationality due to bounded cognitive abilities, risk aversion, and emotions. I am interested in investigating how these elements influence decisions in labour markets, specifically in regard to job offers and salary negotiations. Additionally, I am interested in understanding gender differences in decision-making in the labour market context, to contribute to explaining experimental data on risky choices and provide evidence on whether complex job roles may be designed to exclude certain groups, thereby reinforcing labour market inequalities.

I investigate how incentive contracts, like pay-for-performance schemes, impact academics’ job satisfaction. I aim to understand whether intrinsic factors affect job satisfaction in the presence of the wage, and whether the enrollment into the pay-for-performance enhances or diminishes employee satisfaction.

I also study the rise of precarious employment in academia, such as part-time and fixed-term contracts. I’m particularly interested in how these contracts are distributed across different academic roles and levels of seniority, and how they affect various groups, including gender.

My research aims to highlight disparities and inform HR practices and policy reforms to address labour market inequalities.

Research outcomes

MOOCs adoption

Scholars argue that MOOCs are “the most significant technological advance of the millennium in the pedagogic part of higher education” (Teece, 2018, p. 98), suggesting they could revolutionize higher education (Larionova et al., 2018). The Russian higher education system has been undergoing significant digitalization over the past decade (Bekova et al., 2020), with the COVID-19 pandemic further accelerating the adoption of online learning technologies.

Russian institutions are increasingly replacing classroom courses with online ones via platforms like Universarium, OpenEdX, and Uniweb (Semenova & Rudakova, 2016). However, little is known about how Russian academics adopt MOOCs and other online courses in their teaching, except for a recent study by Sukhostavtseva and Rudakov (2021). Previous research has addressed the pros and cons of MOOCs (Zakharova & Tanasenko, 2019) and their use for professional development (Roshina et al., 2017) but has not examined the profiles of academics who create or are interested in using MOOCs and online courses.

This project analyzed data from the “Monitoring of Education Markets and Organisations” 2020 and reported on three aspects of MOOCs and online course adoption by Russian academics: MOOC use in teaching, online course authorship, and the intent to design personal online courses.

Research outcomes

Student Support

It is widely recognised that student support is critical for overcoming barriers to learning and ensuring learner engagement, motivation, and success in online higher education. Although many support strategies are available for review, there have been no attempts to systematically analyze them in relation to the different stages of student learning. As a result, there is a lack of understanding of where and when student support can be embedded into the online learning curriculum. This project aimed to address this issue and answered the following research question: What support strategies can be offered to online students at different phases of the learning cycle? In doing so, this project developed a framework for embedding support strategies into different stages of the online learning cycle that can be used by online learning designers and educators.

By bringing together research on reported support strategies and interventions, this project aimed to develop a framework for embedding support interventions into various stages of the online learning cycle that can be used by online learning designers and educators.

Research outcomes

Phenomenographic research

    Research outcomes