Exploring and counteracting stage specific course completion risk factors in open higher education distance learning

Digitalisation and lifelong learning education policies keep pushing higher education from classrooms towards flexible distance learning, free from the restrictions of place and time. This change is especially prominent in open university studies, and the development has been further expanded and accelerated under the COVID-19-pandemic lockdown restrictions. Electronic completion methods have replaced their traditional counterparts. The ensued flexibility in study practices has benefitted and supported students’ autonomy in study selection and completion. This development has also facilitated enrolling and participating in higher education, and the availability of online courses has hugely increased.

However, not all changes have facilitated studying or teaching in the distance learning environments. As the face-to-face student support and guidance methods have become inapplicable, the trade-off for flexibility means greater self-guidance and responsibility required from students for adequate learning and study performance. Distance learning requires investment in technical implementation of the flexible teaching and completion methods (Tu & Corry, 2002). The biggest challenge for distance learning seems to be study engagement or student retention—course completion levels are lower in distance learning than in contact teaching (Lee & Choi, 2011; Shah & Cheng, 2019; Woodley & Simpson, 2014)—due to poor pedagogical quality, low social contacts, study motivation or study skills (Lee & Choi, 2011; Lee et al., 2013). Low social contact in distance learning has also been found to reflect impaired learning (Alqurashi, 2019; Kuo et al., 2014). Student dropout is a major problem for institutions’ funding.

Student retention, dropout and attrition are related, widely researched topics in distance learning (for recent reviews, see Delnoij et al., 2020; EADTU, 2024; Elibol & Bozkurt, 2023; Seery et al., 2021; Shaikh & Asif, 2022). It is not always clear what counts as attrition or dropout (Elibol & Bozkurt, 2023). Typically, these concepts refer to a student quitting a degree programme and a problem for student services. When a student does not complete a single course, the concern is in the pedagogical course design (Woodley & Simpson, 2014). Online courses have become a major industry (Bawa, 2016; Seery et al., 2021), yet few dropout studies focus on completion of single courses (Aragon & Johnson; 2008; Hachey et al., 2023; Lee et al., 2013; for a review, see Bawa, 2016), with the exception of massive open online courses (MOOCs) (Chen et al., 2022; Huang et al., 2023). Online course design is the main reason for noncompletion (Rahmani et al., 2024), so it is important to study these reasons, and how the course design can most effectively support completion.

In this article, we focus on course completion risk factors and their variation across completion stages in open higher education distance e-learning. Our particular emphasis is on identifying how and when to support students’ study performance and engagement to prevent student dropout during a course. Addressing these challenges is crucial at the individual, organisational and societal levels. Successful completion of the studies undertaken will, first, ensure that adult learners achieve their learning goals, thus reinforcing their sense of competence. Second, it will enhance the higher education community’s experience of success at work and improve the educational and economic efficiency of the organisation. Engaging individuals in their learning improves institutional success and contributes to the overall improvement of the higher education system. At the societal level, the completion of studies responds to the new skill needs of today’s changing world of work. Hence, several systems for identifying individuals at risk and in need of appropriate student support, for example, using machine learning, have been developed (e.g. Yükselturk et al., 2014). Strengthening study engagement and effectively targeting student support would require an understanding of the risk factors across all study stages.

We studied course completion in open university students participating exclusively in distance learning courses. Typically, the amount of funding granted for a higher education institute at least in Finland depends on its study engagement level. To support course completion, the most significant risk factors should be identified (Valto & Lundell, 2015). In previous studies, the most important preventive factors to open higher education distance learning have been as follows: reconciling work or other life with studies (Maunula et al. 2021; cf. Pekkala, 2004; Repo et al., 2014; Saarenmaa & Virtanen, 2011; Yukselturk & Inan, 2006), students moving to work life before finishing their studies (Maunula et al. 2021; cf. Aho et al., 2012), technical challenges related to the learning environment (Aydin et al., 2019; Gaytan, 2015), lack of academic study skills (Bağriacik Yilmaz & Karataş, 2022; Yukselturk et al., 2014) and passivity in seeking guidance in learning the skills (Maunula et al., 2021), low motivation levels compared with contact teaching (Maunula et al. 2021; cf. Pekkala, 2004; Repo et al., 2014; Törmä et al., 2012), and lack of social support and interaction (Lee & Choi, 2011).

Devising efficient protective countermeasures for course non-completion based on these studies is not a straightforward task and calls for more detailed data about the course completion risk factors. Our approach to this issue is above all practical: to support as efficiently as possible course completion in distance learning. Targeting appropriate countermeasures requires detailed data about the risk factors’ effects at each study completion stage. Instead of formulating specific hypotheses, we explored course completion risk factors in further temporal detail than done in previous studies. We investigated study engagement risk factors in four different course completion stages. We expected that course completion risk factors may differ at different stages of study. Finally, we will discuss practical but science-based suggestions for supporting course completion at different stages. We will specifically discuss how to apply the selfdetermination theory of study motivation in improving completion rates by pedagogical course design (Chen & Jang, 2010; Chiu, 2021; Howard et al., 2021; Hsu et al., 2019; Ryan & Deci, 2018, 2020). According to the theory, student wellbeing and genuine interest in studies depends on experienced competence, autonomy, and social interaction and support. We argue based on our results that genuine learning engagement can be the most effectively supported by designing appropriate completion methods.

To study completion risk factors at different stages of the courses, we sent a Webropol questionnaire by email to undergraduate psychology students at a Finnish open university in the academic year 2020–21 (n = 1,024). We measured the students’ reported course completion performance using a 4-point categorical scale 1 = not initiated, 2 = abandoned (initiated but not submitted), 3 = (submitted but) failed, and 4 = completed (cf. Kember, 1995).

To measure the students’ self-reported levels of study engagement risk factors during their study progress we compiled a questionnaire using a 7-point Likert-scale items (1 = very well to 7 = not at all, including an extra item for not applicable). The questions concerned the following 11 risk factors for course completion, in two categories: instructions and formative feedback, skill/ difficulty levels, completion method, learning material and its availability, learning environment and study schedule (pedagogy-related), study motivation, available time and life situation (student-based). The items were phrased: ‘How well did [risk factor] support completing the course?’ where applicable. Study motivation was probed by an item formulated ‘How motivated were you towards studying on the course?’. Individual personality features and learning styles were also charted, but for clarity and brevity, these are not reported here.

A total of 295 students (M_age = 35.42, standard deviation [SD] = 10.27, 266 females, 26 males, 3 individuals identified with other gender) completed the survey (response rate 28%). Of these, 70 students reported not completing at least one course despite registration. We avoided collecting any personal identity information in the questionnaire. Because of complete data anonymity, our study did not require statement from the local ethical committee according to the committee’s guidelines. To study the adequacy of our sample, a sample-size estimate using an online calculator (https://sample-size.net/logistic-regression-sample-size/) with odds ratios = 2, sample proportion = 0.1/0.9, SD = 1, 2-tailed α-level = 0.05 and β = 0.80, yielded a sample size of ~200 and a more conservative estimate taking the data collinearity into account, ρ² = 0.4, yielded a sample size of ~300. Based on this estimate, the sample is sufficient for detecting effects starting from small to medium sized ones.

We also asked open-ended questions about what supported their course completion, and what prevented them from it with the following four open-ended questions: (1) If a course was not completed, what factors made it difficult to complete the course? (78 responses); (2) What factors specifically supported learning? (183 responses); (3) What kind of completion methods would you have wished? (160 responses), and (4) What kind of further guidance would you have required? (111 responses). Responses to these open questions were analysed using content analysis by identifying and categorising key concepts. In responses to the first two questions, main distinction made was between individual student-based and pedagogy-related factors. The purpose of recognising the key concepts was to complement the results from the quantitative analyses, and to similarly help in designing engaging distance e-learning courses.

We studied course completion risk factors at different completion stages in open higher education distance learning courses. We found that risk factors vary across the course completion stages. Low motivation, inappropriate completion methods and lack of time were associated with not initiating the course completion; conversely, high motivation, appropriate completion method and sufficient available time best supported course completion. Inappropriate completion methods were also associated with task abandonment. The other significant risk factors at this stage, a changed personal life situation, is relatively immune to pedagogical countermeasures. Difficulty level (lack of academic study skills) also identified as a significant risk factor in previous studies (Maunula et al., 2021), did not, surprisingly, risk course completion until at the final stage. The unavailability of study materials, also observed as a significant study engagement risk factor at the latest stage, is probably an institute specific risk factor. Students’ reports about the preventive and supportive factors for studies largely complemented the quantitative results. All in all, study motivation and appropriate distance learning completion method emerged as the most critical factors teachers can affect in designing their distance learning courses. As the most effective measure supporting study engagement may be fostering motivation by designing appropriately motivating distance learning completion methods, we will discuss these countermeasures most extensively below.

Interestingly, in our study, difficulty level or lack of academic skills became critically risky not earlier than at the last stages of course completion, at which the students may already be too passive in seeking support (Maunula et al., 2021). Alternatively, they may not be aware of their skill deficiencies before assessment. For example, at our institute, students are required to use complex citation systems even in time pressured and time limited electronic exams, using of which elicits exam anxiety in students (Dikmen, 2022), and they may not early enough realise the academic competency required for completion. Student profiling charting the academic skill level and targeted support measures, in the form of educational and technical support (Gaytan, 2015; Muljana & Luo, 2019), might prevent course dropouts at this stage. Low student preparedness can be counteracted by offering the students preparatory modules in the learning environment, or orientation courses (EADTU, 2024, ch. 3).

The challenge of Finnish open university education has traditionally been poor availability of teaching materials due to large numbers of students, although digitalisation and e-learning materials have alleviated the challenge. The curricula of the Open University are based on those of degree students. The curricula designers cannot always reconcile the need extent with sufficient availability of learning materials, especially in open distance learning, which can significantly affect the amount of funding for the universities. Opting for open access online textbooks as course material may be one effective way of alleviating inaccessibility of course materials.

There are a few limitations to our study. First, it was carried out only with psychology students; the stagespecific dynamics of risk factors revealed by the present study may differ in other subjects from the ones found in the present sample. Second, the results probably reflect problems inherent in the national higher education system, and those elsewhere may differ (see, e.g. Aydin et al., 2019). Larger studies are needed to draw generalised conclusions. Third, it is not infrequent in open higher education that students enroll into a course out of curiosity and without intention of completion (Kember, 1995), for example, in MOOCs, suffering from great numbers of dropout students. Nevertheless, we did not consider this dimension of study engagement here. Furthermore, it is probable that the data were biased towards the more achieved than average students: 28% of the students in our sample reported non-completion of one or more courses, while the actual non-completion rate is about 40%. Although personalisation is important for course design, we did not include individual temperamental or other psychophysiological variables in our study, and furthermore, targeted studies are required to uncover the potential effects of individual features on course completion (e.g. Ertem & Gokalp, 2022). Finally, we neglected a potential preventive factor of student dropout that is self-efficacy or self-regulation capacity (Lee et al., 2013; Daumiller et al., 2023), which is typically interrelated with emotions (Asikainen et al., 2017) and motivation (Smit et al., 2017), and an important determinant of student satisfaction (Lysitsa & Mavroeidis, 2024). Nevertheless, motivation predicts most of the study engagement and significantly study performance (e.g. Froiland & Worrell, 2016); motivational regulation does not directly affect performance, so its effect is probably complex (see, e.g. Schwinger et al., 2009). As our questionnaire did not probe this important factor, so its potential direct and indirect impacts must be left for further studies.

In conclusion, study engagement remains a critical issue for open higher education distance e-learning. Responding to the evolving societal needs, study degrees and in-service training are under pressure to serve work life more directly than before. Higher education professionals strive to design shorter study and degree formats, for example, in the form of microcredentials (Lauder & Mayhew, 2020), offered in the form of scalable distance learning courses. Maunula et al. (2021) emphasised the need for continuous learning and the decline in study engagement in digitised distance learning being a major challenge for these studies. All this highlights the significance of engaging course and instructional design that supports students’ wellbeing and autonomous learning (Hachey et al., 2023; Lysitsa & Mavroeidis, 2024; Rahmani et al., 2024), especially in open higher education distance e-learning for adult students that value autonomy (Chu & Tsai, 2009). Proficiency in online course design also increases instructor satisfaction and protects instructors from burnout (Yang & Du, 2024). Based on our results, study engagement risk factors differ across course completion stages, and this result is critical for counteracting these factors by engaging course design. Countermeasures should target inspiring and fostering motivation and motivational regulation by designing appropriate learning and completion methods. This could be implemented by creating a communal learning environment with increased testing and instructor and peer feedback, as well as flexibility in completion and study methods.