Using LMS Records to Track Student Performance:

A Case of a Blended Course

Marko Matus

, Josipa Bađari

and Igor Balaban

Faculty of Organization and Informatics, University of Zagreb, Pavlinska 2, 42000 Varaždin, Croatia

Keywords: LMS Logs, Student Performance, Educational Data Analysis, Blended Course.

Abstract: This paper explores the use of Learning Management System (LMS) logs to analyse student performance in

a blended course. The study aims to identify how LMS data can inform teaching strategies and intervention,

focusing on which variables most influence students’ performances. The course was designed using Moodle,

incorporating programmed learning, conditional activities, and assessments like quizzes, flash tests, and self-

assessments.

Data on students' activities, including access logs, quiz scores, and final grades, were collected and analysed.

The results show that students with higher LMS activity, particularly those who completed more self-

assessments and engaged consistently, tended to perform better. However, while self-assessment activities

increased engagement, they had a weaker correlation with final grades compared to midterm exams and flash

tests. A strong positive correlation was found between midterm exam performance and final grades,

highlighting the importance of these assessments for course success. The study suggests that LMS logs can

be a useful tool for teachers to monitor student behaviour and to implement timely interventions to support

struggling students.

1 INTRODUCTION

The extensive adoption of Learning Management

Systems (LMS) in educational institutions has

generated vast amounts of data regarding student

interactions and behaviors during online and blended

courses. These systems record logs of various aspects

of student activities, such as the specific resources

accessed, the timing of these interactions, and in

general, the duration of their engagement with the

variety of resources and activities within LMS. As

teachers increasingly rely on these systems, the

imperative to harness this data for enhancing student

learning outcomes becomes very important in

teaching practice.

The necessity of using LMS data is underscored

by research exploring a deeper understanding of

students' learning contexts and behaviors. Ferguson

(2012) emphasizes the importance of analyzing these

data to optimize learning environments and

processes. Ryabov (2012) demonstrates a positive

https://orcid.org/0009-0009-1254-4957

https://orcid.org/0009-0003-1289-6748

https://orcid.org/0000-0002-4367-9629

correlation between the overall time logged within an

LMS and final academic performance, while Nguyen

(2017) finds significant associations between student

engagement metrics, such as page views and

discussion posts, and learning outcomes.

Furthermore, Wei et al. (2015) explored the impact of

various online activities on academic success,

highlighting the need for teachers to engage with

LMS analytics to foster improved student

performance. The potential of using such data is also

highlighted by studies emphasizing their role in

improving retention rates, predicting performance,

and identifying students at risk of underachievement

(Wong, 2017).

In general, there is a growing need to further

contribute to the field of educational data analytics,

particularly within the higher education sector, where

the effective measurement and improvement of

student performance remains a pressing concern (Jha

et al., 2019). With that respect, this paper aims to

further explore how data recorded in LMS can

Matus, M., BaÄ

Sari, J. and Balaban, I.

Using LMS Records to Track Student Performance: A Case of a Blended Course.

DOI: 10.5220/0013217800003932

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 17th International Conference on Computer Supported Education (CSEDU 2025) - Volume 1, pages 283-290

ISBN: 978-989-758-746-7; ISSN: 2184-5026

283

explain student performance by utilizing student logs

to pinpoint areas of struggle and help teachers to

implement targeted interventions.

2 BACKGROUND

Modern educational institutions use different

Learning Management Systems (LMS) to support

their teaching and learning activities. In recent years,

especially in the COVID and post-COVID era the

researchers and teachers started to realize the

importance of the analysis and use of the LMS

generated logs of teachers and students. Since data in

log files describe how its users interact and

interrelate, the information has been used to create

predictive models for different purposes such as

foreseeing student performance (Conijn et al., 2017),

detecting procrastination (Tuckman, 2005) and

clustering students (Cerezo et al., 2016).

According to Gašević et al. (2016) the prediction

of students at risk of failing a course and the

prediction of students' grades have been based on the

data stored in institutional student information

systems, trace data recorded by LMSs and other

online learning environments; and combinations of

different data sources.

Although the LMS logs have emerged as power

tools for capturing and analysing student behaviour

and the data from LMS has contributed to insights

into student learning paths and to predictions about

student performance, the use of LMS data for early

prediction of student performance is still limited (see

Rotelli, Fiorentino & Monreale; 2021; and Baginda,

Santoso & Janus; 2022). Tamada et al. (2022) used

Machine Learning techniques based on logs from the

LMS (Moodle) due to the fact that all interactions in

the LMS generate a log, which stores information in

a database, the amount of data collected is rapidly

increasing in volume and complexity, but also

allowing statistical analysis, data mining, and

building predictive models of school performance

that helps to detect students at risk.

Researching LMS student logs Kadoić & Oreški

(2018) found in their study the correlation between

the number of logs in the Moodle e-course and the

final grades and Felix et al. (2019) found that the

number of interactions with the system, attendance,

and time spent on the platform were essential

variables in predicting student outcomes. Also,

Kaensar & Wongnin (2023) study supports the idea

that student behaviour in online learning platforms

like Moodle affected student performance.

Based on the previous findings this research tries

to provide teachers with further analysis of how to use

LMS logs to identify possibilities to improve learning

design of their blended courses enabling best student

performance.

3 RESEARCH AIMS AND

METHODOLOGY

The main aim of the paper is to explore different ways

of using LMS logs to analyse students’ activities and

describe their behaviour in a blended course.

With that sense, the following research questions

are outlined:

1. In what way can student logs be used to analyse

students’ performance?

2. Which are the variables derived from the LMS

records that most influence student performance?

3. How can teachers use LMS analytics/student logs

as a predictive techniques/tools (e.g. a teacher can

identify areas where students may be struggling

and implement targeted interventions to improve

student outcomes.)

In the initial phase, the online part of the blended

course was developed based on the principles of

programmed learning within the LMS Moodle. A

range of online resources and knowledge assessments

were incorporated, along with conditional activities,

to establish a clear learning pathway for students.

Additional details regarding the programmed

learning principles, conditional activities, and the

course overall can be found in the next section.

Data on students, including their final grades were

gathered from Moodle by exporting student scores

from all quiz activities, including midterms, flash

tests, and self-assessment quizzes. Students'

engagement was obtained from course activity logs.

A Moodle log consists of the time and date it was

accessed, the Internet Protocol (IP) address from

which it was accessed, the name of the student, each

action completed (i.e., view, add, update, or delete),

the activities performed in different modules (e.g., the

forum, resources, or assignment sections), and

additional information about the action.

Following this, descriptive statistics were

employed to analyse and interpret the data.

Additionally, correlation analysis was conducted

among the main components of the dataset to identify

which variables most significantly impact student

performance.

CSEDU 2025 - 17th International Conference on Computer Supported Education

284

4 COURSE DESCRIPTION AND

PARTICIPANTS

The course "Business Informatics" is a first-semester

bachelor course of the specialist study program

"Information Technology in Business Applications"

offered at the University of Zagreb, Faculty of

Organization and Informatics in Varaždin, Croatia.

The course syllabus covers several key topics,

including an introduction to information systems and

their applications in business, a detailed exploration

of computer hardware and software (the fundamental

components of information systems), and

foundational principles of information system

security.

Delivered as a blended course, all teaching

materials and methods have been designed for such

delivery within Moodle. The topics covered in

lectures are supplemented with various online

resources, such as videos and quizzes integrated into

the Moodle. The course content is organized into a

sequence of lessons, prepared as asynchronous

materials for the online component, serving as both a

primary source of information and a mean of

reviewing topics discussed in onsite and in online

lectures.

The structure of the course includes seven

knowledge domains, as shown in Figure 1, while each

knowledge domain consists of several lessons.

Each lesson ends with a short test, which students

must solve successfully to progress to the next lesson

(a part of conditional activities). Achieving the

required result in all lessons within a knowledge

domain is a prerequisite to access the final self-

assessment quiz at the end of each knowledge

domain. Such a learning path was implemented as a

completion tracking and conditional activities feature

in Moodle. It enables teachers to specify when a

certain activity shall be hidden or enabled for students

according to the planned course design.

Since the students took the self-assessments

outside the class as an optional activity, those results

are not included in their final grade. They are

designed as a student self-monitoring activity

opposite to the formal online midterm exams which

are obligatory and conducted within the Moodle

course. A total of two midterms are performed: the

first in the middle of the semester (Week 8) including

knowledge domains 1 and 2 and the second in the

final week (Week 16) including all knowledge

domains.

Besides the formal tests and self-assessments,

students are also given short flash-tests during

lectures, as warm-up activities covering the content

from the previous domain knowledge. In total, 4-5

flash-tests are provided during the semester.

Students’ final grades are created based on their

results from flash tests, 1st midterm and the 2nd

midterm exam.

During the 2023/2024 academic year, a total of

117 students were enrolled in the course. At the end

of the 16th week a total of 99 students (84.6%)

finished the course out of which 31 were female

students (31.3%) and 68 male students (68.7%).

To analyse students' performance and identify

areas where they may be struggling during the course,

as well as to implement targeted interventions with

the goal of improving student outcomes, the student

activity logs and their scores were exported and

processed

5 RESULTS

The dataset used in this paper was collected during

the first semester of academic year 2023/2024. Since

the course was delivered in blended mode, students

were required to complete part of the activities (e.g.,

view, add, update…) off campus - through LMS (e.g.,

forum posts, self-assessments, lessons completed…).

Since the LMS automatically stored a lot of activity

logs during the course about every student enrolled,

they were exported as a datasheet after the course had

ended. More than 330000 activity logs were exported

Figure 1: Organization of the course across knowledge domains and weeks.

Using LMS Records to Track Student Performance: A Case of a Blended Course

285

for all students enrolled. After the data wash, a total

of 277899 activity logs were prepared for further

analysis within a pivot table including the columns

“Student ID”, “Date and time”, “LMS Module

(Lesson, Forum, Test…)”, “Final Grade Course”,

“Week of the Course” and “Class attendance” that

were analysed.

Furthermore, to analyse the relationship between

monitored activities/objectives, data on the results for

each individual student (based on “Student ID”) were

exported to the new datasheet and later processed in

SPSS Statistics (version 29.0.0.0). The monitored

activities/objectives included: “1st Midterm Exam”,

“2nd Midterm Exam”, “Flash tests”, “Self-

assessment quiz”, “Class attendance”, “Number of

logs” and “Final grade”.

The analysis was started by reviewing the

distribution of logs per week which include access to

lessons, self-assessment attempts, flash tests and

forum views (see Figure 2).

Figure 2: Distribution of logs per week.

It can be noted that the peaks in the graph occur in

four stages of the course: the 5th, 8th, 13th, and 15th

week, when course activity is particularly intensive.

The increases in activities during the 5th and 13th

weeks are linked to the assessments of practical

assignments. Notably, student activities rise steadily

until the 5th week, when their knowledge from the

practical assignments is assessed, before dropping

sharply in the 6th week.

Additionally, there is a noticeable increase in

student preparation between the 7th and 8th week,

coinciding with the schedule of the midterm exam. In

the 13th week, students face a second assessment for

the practical assignments, but this time, there is no

significant drop in activity. In fact, student

engagement is higher during these second practical

assignments compared to the first. A significant

increase in student activity is also noted in the 13th

week, as students prepare for the second midterm

exam and complete any remaining course tasks.

Figure 3: Frequency of logs by grade.

Regarding online activity and final grades, Figure

3 which indicates overall number of logs per grade,

reveals that students with the highest grades were also

the most active in the LMS, recording over 12,000

logs - significantly more than students with lower

grades In Croatia, the grading system ranges from 1

(lowest/fail) to 5 (highest/excellent), but there were

no students in the analysed semester achieving the

highest grade within LMS (chart displays grades 1

and 4, with the size of the populations 29 (1)-42 (2)–

24 (3)–5 (4). It is important to note that in this blended

course, the activity levels of students with lower

grades (1-3) do not differ significantly. This suggests

that, based on their access to resources, it is not

possible to predict their final results, except for the

most active ones. Students who did not meet the

requirements for a grade continuously throughout the

semester were not taken into account.

However, a different conclusion can be made if

we take a closer look at the distribution of activities

related to self-assessment quizzes for each course

week presented in Figure 4. Over the 16-week period,

the activity levels of all students are generally low for

the most weeks, with increased activity around weeks

8 and 15 when midterms are taking place. It is evident

that students with lowest grades exhibited minimal

activity in the early weeks, with a slight increase in

week 8 and in week 15, reaching an average of 3.5

self-assessment tests completed per student (out of 7).

Students with grade 2 demonstrated somewhat better

activity, particularly around week 8, and significantly

increased their self-assessment completion by week

15, similar to students with a grade of 3 (averaging 6

out of 7).

Interestingly, students with grade 3 had similar

activity levels to those with grade 4 around week 8,

indicating that they have completed both self-

assessment quizzes covering the material for the first

midterm exam. However, the most notable difference

among the grades appears in week 15, when grade 4

CSEDU 2025 - 17th International Conference on Computer Supported Education

286

Figure 4: Self-assessment test activity based on student grades.

students completed all 7 self-assessments, while

grade 3 students averaged around 6, suggesting that

most grade 3 students did not complete all of the self-

assessments.

During the course, a total of 7 self-assessment

quizzes were available as shown in Figure 5. As

mentioned earlier, the first midterm exam included

domains 1 and 2, "Information systems” and

“Information Systems Security” respectively.

Students have started completing the first self-

assessment quiz in the 3rd week, with the highest

number of attempts in the 8th week, during the

preparation for the midterm exam.

It can also be observed that students continued

working on the first two self-assessments until the

end of the semester, particularly as preparation for the

second midterm exam, basically since the second

midterm exam included some questions from the first

two knowledge domains. Although the first of the

remaining self-assessments was available from the

7th week, it can be noted that students learned

“periodically”, having activity peaks only around

midterm periods. This is also supported by the fact

that students started accessing self-assessments in the

13th week as a way of preparation for the second

midterm.

Finally, the last part of analysis refers to

reviewing the correlations between the monitored

activities/objects. The correlation matrix in Table 1

shows several significant relationships between

various factors that contribute to monitoring student

performance.

Since the Final grade is calculated as the sum of

the points earned on the 1st and 2nd Midterm exams,

Flash tests, and the Self-assessment quiz, a strong

Table 1: Correlation matrix between factors that contribute

to monitor student success.

Correlations

1ME 2ME FT SAQ CA NL FG

1ME 1 0,396

0,274

0,275

-0,063

0,366

0,709

2ME 0,396

1 0,283

0,040 -0,040 0,128 0,773

FT 0,274

0,283

1 0,160 -0,068 0,211

0,311

SAQ 0,275

0,040 0,160 1 -0,068

0,595

0,241

CA -0,063 -0,040 -0,068 -0,068 1 0,042 -0,021

NL 0,366

0,128 0,211

0,595

0,042 1 0,352

FG 0,709

0,773

0,311

0,241

0,021

0,352

Correlation is significant at the 0.01 level,

Correlation is

significant at the 0.05 level

Explanation: 1st Midterm exam (1ME), 2nd Midterm exam

(2ME), Flash tests (FT), Self-assessment quiz (SAQ),

Class attendance (CA), Number of logs (NL), Final grade

(FG)

positive correlation between the 2nd midterm exam

and the final grade (r = 0,773, p=0,01), predicting that

the performance on the 2nd midterm exam is the key

predictor in achieving success in the course. Also, the

midterm exams show strong positive correlation with

final grade, which indicates the importance of the 2

major assessments as a key component in achieving

overall success in the course. Moderate positive

correlation was found between flash tests and the

Using LMS Records to Track Student Performance: A Case of a Blended Course

287

Figure 5: Self-assessment quiz attempts per week of the course.

final grade (r=0,311, p=0,01) which indicates that

active student engagement (flash tests were solved

during lessons) have a moderate impact in achieving

positive final grade. Also, the number of logs which

indicates an overall number of logs per student, and

final grade have a moderate positive correlation

(r=0,352, p=0,01). Although the self-assessment quiz

demonstrated a strong correlation with the number of

logs (r = 0,595, p = 0,01), correlation to the final grade

was weaker (r = 0,241, p = 0,05). This suggests that

although the self-assessment activities may

encourage higher levels of engagement, they do not

necessarily lead to improved final grades directly.

6 DISCUSSION

Re RQ1 discussing in what way can student logs be

used to analyse students’ performance it is evident

that in a blended environment, where students

alternate between face-to-face and online teaching

and learning, LMS logs can help teachers to identify

how well students are balancing both delivery modes.

These logs capture activities like lesson access, quiz

attempts, participation in forums, and interactions

with learning materials, providing a comprehensive

view of students’ online learning behaviour.

However, some additional tracking elements are

needed to capture their face-to-face activity, as in this

case we used class attendance and flash-tests written

during face-to-face lectures. Besides these

information, the teacher gets clear insight about the

student time management (when do students usually

approach specific resources) and after the 1st midterm

they are able to identify potentially at-risk students.

They are also able to note which parts of the course

content are visited more frequently than others, and

which ones might not be visited at all, leading them

to revise those materials or the course requirements.

The analysis of LMS logs allows teachers to track

student progress and intervene early to support

students in need. In the case of a blended course, it is

important to note that the number of logs may not

reflect real student engagement and knowledge. The

data in this example showed a very slight difference

in the number of logs between students with final

grades 1 to 3. However, the self-assessment activity

is notably different for students with different final

scores/grades, which is fully in line with study from

Schön (2022) who has also shown that completion

rates of online quizzes can predict final exam

performance.

Re RQ2 aimed at identifying the variables that

most influence the students’ performance using

course-agnostic LMS log data, the correlation matrix

in the results reveals that midterm exam performance

and the number of logs are significant predictors of

final grades, highlighting research that demonstrates

the predictive power of LMS data in early

identification of students at risk of underperformance

(Gašević et al., 2016; Tamada et al., 2022).

Based on the analysis of correlations, several

variables have been identified as influential on

student performance:

1. Number of Logs: The overall number of

interactions in the LMS has been shown to correlate

with higher grades. This is also supported by Kadoić

& Oreški (2018) who found a positive relationship

between the number of Moodle log entries and final

grades, indicating that students who frequently

engage with course content tend to perform better.

Such finding is also supported by other studies (e.g.

Conijn et al., 2017; Baginda et al. 2022) where LMS

log frequency was shown to be a common indicator

of performance.

100

120

1345678910111213141516

Number of self-assessment quiz attempts

Week of the course

Knowledge Domain: Information

Systems

Knowledge Domain: Memory

Unit

Knowledge Domain: Basic

Computers Principles

Knowledge Domain: Computer

Software

Knowledge Domain: Information

Systems Security

Knowledge Domain: Central Unit

of a Computer

Knowledge Domain:

Input/Output Software

Midterm

Exam

Midterm

Exam

CSEDU 2025 - 17th International Conference on Computer Supported Education

288

2. Engagement in Self-Assessment Quizzes:

In the given study, students with higher grades

consistently completed more self-assessment quizzes,

especially around midterm periods. Although self-

assessment quizzes were highly correlated with

overall activity (r = 0,595), they had a lower direct

correlation with final grades, suggesting that while

they might boost engagement, they may only

indirectly influence performance.

3. Participation in Flash Tests: Flash tests,

as mandatory, but brief in-course assessments, have a

positive relationship with final grades (r = 0,311),

highlighting the impact of frequent, low-stakes

testing on learning outcomes. Rotelli, Fiorentino &

Monreale (2021) suggested that these micro-

assessment engagements are valuable for reinforcing

material and maintaining consistent engagement,

which contributes to better academic outcomes.

4. Midterm scores: The scores from

structured assessments showed the strongest

correlation with final grades, especially the 2nd

midterm (r = 0,773). This aligns with Gašević et al.

(2016), who found that scores on significant

assessments derived from LMS logs are critical

predictors of final performance. This variable acts as

a summative reflection of students’ knowledge and

learning throughout the course, meaning that based

on the 1st midterm the teachers could detect failing

students.

5. Course Material Access and Forum

Participation: This research has found a moderate

correlation between access to materials and forums

and the final grade. That might be related to the fact

that some students had printed materials and were not

assessing LMS. A positive relationship between

access to materials and final grade was also supported

by Baginda et al. (2022) who identified that accessing

core LMS features, such as course materials,

assignments, and forums, was strongly associated

with higher grades. Regular interaction with these

resources suggests proactive learning and

engagement with course content. This is consistent

with the findings of Li et al. (2018), who emphasized

that students who frequently interact with learning

resources and engage in forums demonstrate higher

levels of comprehension and academic performance.

The last interesting finding in respect to RQ2 in

this research revealed that course participation does

not affect student performance, which is probably

related to the fact that students were required to attend

at least 65% of face to face lectures.

Re RQ3 aimed at identifying areas where teachers

can use LMS logs to enable better student

performance this research highlights that students

with higher activity levels, such as frequent self-

assessment attempts and access to resources,

generally achieved better grades. By monitoring these

logs, teachers can detect early signs of

underperformance, such as a lack of engagement

before assessments, and intervene accordingly. This

is supported by research from Gašević et al. (2016)

and Tamada et al. (2022), which emphasize the use of

LMS data for predicting at-risk students. Predictive

models based on these logs can enable teachers to

offer timely support, such as additional resources or

feedback, improving students’ chances of success. As

seen, more interactive materials (self-assessments,

flash-tests…) could provide students with more

opportunities to self-test and perform better and

teachers with more data for analytics and prompt

reaction and course redesign.

Within the context of this course, where around

30% of students fail the course during the continuous

monitoring, the conclusions provide teachers with the

clear guidelines on how to redesign and when to react

and provide students with more stimuli to

successfully conclude the course. Further analysis of

the student feedback on course delivery, content and

available (self-)assessment options will enable deeper

analysis and improvements of the course.

7 CONCLUSION

This study, conducted in an institution with limited

resources, sought to identify patterns in student

engagement and performance using LMS log data.

The results demonstrate the viability of using

accessible and affordable methods for monitoring

student progress in a blended learning environment.

Key findings show that the majority of student

activity is concentrated around major assessments, as

well as support the fact that the students who engage

more consistently within LMS are also generally

performing better. However, while self-assessment

activities correlated with higher levels of

engagement, they did not strongly predict final

grades.

Importantly, this study confirms the potential for

institutions to leverage existing data to provide timely

feedback for students at risk of underperforming,

allowing for interventions such as adjusted teaching

methods or additional assignments tailored to both

advanced and struggling students. These findings can

help teachers make predictions for upcoming

semesters, offering live recognition of both at-risk

and high-achieving students. Since this was one of the

first courses where students interacted with LMS

Using LMS Records to Track Student Performance: A Case of a Blended Course

289

since entering higher education, the implications of

this research can be beneficial to other teachers,

potentially yielding long-term positive effects for

students across the study programme.

To further support educators in applying these

findings, we recommend the integration of automated

alerts within the LMS platforms to identify and notify

students at risk based on engagement metrics.

Future research could explore student

perspectives by incorporating surveys,

complementing the log data with qualitative insights

into student experiences and engagement. Analysing

students’ perceptions of course components,

perceived workload, and their reasons for

engagement patterns could provide insights to refine

predictive models and develop more effective

teaching interventions.

ACKNOWLEDGEMENTS

This work has been fully supported by the Croatian

Science Foundation under the project IP-2020-02-

5071.

REFERENCES

Cenka, B. A. N., Santoso, H. B., & Junus, K. (2022).

Analysing student behaviour in a learning management

system using a process mining approach. Knowledge

Management & E-Learning, 14(1), 62-80.

Cerezo, R., Sánchez-Santillán, M., Paule-Ruiz, M. P., &

Núñez, J. C. (2016). Students' LMS interaction patterns

and their relationship with achievement: A case study

in higher education. Computers & Education, 96, 42-

54.

Conijn, R., Snijders, C., Kleingeld, A., & Matzat, U.

(2016). Predicting student performance from LMS data:

A comparison of 17 blended courses using Moodle

LMS. IEEE Transactions on Learning Technologies,

10(1), 17-29.

Felix, I., Ambrosio, A., Duilio, J., & Simões, E. (2019,

February). Predicting student outcome in moodle. In

Proceedings of the Conference: Academic Success in

Higher Education, Porto, Portugal (pp. 14-15).

Ferguson, R. (2012). Learning analytics: drivers,

developments and challenges. International Journal of

Technology Enhanced Learning, 4, 304–317.

Gašević, D., Dawson, S., Rogers, T., & Gasevic, D. (2016).

Learning analytics should not promote one size fits all:

The effects of instructional conditions in predicting

academic success. The Internet and Higher Education,

28, 68-84.

Jha, S., Jha, M., & O’Brien, L. (2019, December).

Analysing Computer Science Course Using Learning

Analytics Techniques. In 2019 IEEE Asia-Pacific

Conference on Computer Science and Data

Engineering (CSDE) (pp. 1-6). IEEE.

Kadoić, N., & Oreški, D. (2018, May). Analysis of student

behavior and success based on logs in Moodle. In 2018

41st International Convention on Information and

Communication Technology, Electronics and

Microelectronics (MIPRO) (pp. 0654-0659). IEEE.

Kaensar, C., & Wongnin, W. (2023). Analysis and

Prediction of Student Performance Based on Moodle

Log Data using Machine Learning Techniques. Int. J.

Emerg. Technol. Learn., 18(10), 184-203.

Nguyen, V. A. (2017). The impact of online learning

activities on student learning outcome in blended

learning course. Journal of Information & Knowledge

Management, 16, 1750040.

Panigrahi, R., Srivastava, P. R., & Sharma, D. (2018).

Online learning: Adoption, continuance, and learning

outcome—A review of literature. International Journal

of Information Management, 43, 1-14.

Rotelli, Daniela & Fiorentino, Giuseppe & Monreale,

Anna. (2021). Making Sense of Moodle Log Data.

10.48550/arXiv.2106.11071.

Ryabov, I. (2012). The efect of time online on grades in

online sociology courses. MERLOT Journal of Online

Learning and Teaching, 8, 13–23

Schön, S., Leitner, P., Ebner, M., Edelsbrunner, S., &

Hohla, K. (2021, September). Quiz feedback in massive

open online courses from the perspective of learning

analytics: role of first quiz attempts. In International

Conference on Interactive Collaborative Learning (pp.

972-983). Cham: Springer International Publishing.

Tamada, M. M., Giusti, R., & de Magalhães Netto, J. F.

(2021, October). Predicting student performance based

on logs in moodle LMS. In 2021 IEEE Frontiers in

Education Conference (FIE) (pp. 1-8). IEEE.

Tuckman, B. W. (2005). Relations of academic

procrastination, rationalizations, and performance in a

web course with deadlines. Psychological reports,

96(3_suppl), 1015-1021.

Wei, H. C., Peng, H., & Chou, C. (2015). Can more

interactivity improve learning achievement in an online

course? Effects of college students’ perception and

actual use of a course-management system on their

learning achievement. Computers & Education, 83, 10–

21.

Wong, B. T. M. (2017). Learning analytics in higher

education: an analysis of case studies. Asian

Association of Open Universities Journal, 12(1), 21-40.

CSEDU 2025 - 17th International Conference on Computer Supported Education

290