A Learning Analytics Dashboard for Improved Learning Outcomes and

Diversity in Programming Classes

Iris Groher

1 a

, Michael Vierhauser

2 b

and Erik Hartl

Johannes Kepler University Linz, Institute of Business Informatics, Software Engineering, Linz, Austria

University of Innsbruck, Department of Computer Science, Innsbruck, Austria

Keywords:

Learning Objectives, Assurance of Learning, Learning Analytics, Dashboard, Diversity.

Abstract:

The increased emphasis on competency management and learning objectives in higher education has led to

a rise in Learning Analytics (LA) applications. These tools play a vital role in measuring and optimizing

learning outcomes by analyzing and interpreting student-related data. LA tools furthermore provide course

instructors with insights on how to reﬁne teaching methods and material and address diversity in student

performance to tailor instruction to individual needs. This tool demonstration paper introduces our Learning

Analytics Dashboard, designed for an introductory Python programming course. With a focus on gender

diversity, the dashboard analyzes graded Jupyter Notebooks, to provide insights into student performance

across assignments and exams. An initial assessment of the dashboard, applying it to our Python programming

course in the previous year, has provided us with interesting insights and information on how to further improve

our class and teaching materials. We present the dashboard’s design, features, and outcomes while outlining

our plans for its future development and enhancement.

1 INTRODUCTION

In recent years, the systematic management of

competencies and learning objectives has gained

widespread popularity, particularly in higher educa-

tion (Malhotra et al., 2023; Bergsmann et al., 2015).

In this context, Learning Analytics (LA) has become

a major endeavor and a means to track and analyze

learning outcomes and achievement of competencies.

LA is primarily concerned with the measurement, col-

lection, analysis, and interpretation of data related

to students and their learning in order to understand

and ultimately optimize learning outcomes (Scheffel

et al., 2014).

The importance of applications that support teach-

ing and learning has increased signiﬁcantly in re-

cent years, due in no small part to the COVID-19

pandemic. Through the use of technologies such as

online platforms, virtual learning environments, and

learning management systems (LMS), huge amounts

of data are generated, which creates opportunities for

measuring the learning success of students and when

necessary, positively inﬂuencing learning outcomes

https://orcid.org/0000-0003-0905-6791

https://orcid.org/0000-0003-2672-9230

through targeted intervention (Vieira et al., 2018).

Making use of LA can support both students and

educators in many different ways. For students, it

can help to personalize the learning path and enhance

their learning experience. Students can further use LA

to monitor their own progress and the individual feed-

back gained can help them understand their strengths

and weaknesses and make improvements. Educators

can use LA to improve the quality of their courses

and reﬁne their teaching methods, course materials,

and support services. They can identify difﬁculties of

their students and potential drop-outs, measure course

engagement and assessment performance, and mea-

sure learning objectives to ensure that their courses

meet the required standards. LA can further play a

crucial role in supporting diversity in higher educa-

tion by helping educators identify, understand, and

address disparities in student performance, engage-

ment, and outcomes.

As a step towards LA in Programming Education,

and to foster diversity analysis, we have created a

Learning Analytics Dashboard for our introductory

Python programming course. The dashboard takes

graded Jupyter Notebooks (Johnson, 2020) from as-

signments or exams as an input, and provides statis-

tical analyses and visualization of assignments and

618

Groher, I., Vierhauser, M. and Hartl, E.

A Learning Analytics Dashboard for Improved Learning Outcomes and Diversity in Programming Classes.

DOI: 10.5220/0012735000003693

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

In Proceedings of the 16th International Conference on Computer Supported Education (CSEDU 2024) - Volume 2, pages 618-625

ISBN: 978-989-758-697-2; ISSN: 2184-5026

their individual exercises or tasks. Furthermore, we

have put speciﬁc emphasis on the gender diversity

aspect, allowing us to drill down into submissions

and gain valuable insights into how well certain tasks

were performed by different groups of students. Our

main goal was, for us as educators and course in-

structors, to gain insight into the challenges and dif-

ﬁculties our students have with the different topics

covered in our Python course. As we were facing a

gender gap, with respect to course performance and

drop-outs in the previous semesters, which has also

been frequently reported as a major issue (Marquardt

et al., 2023; Rubio et al., 2015; Groher et al., 2022)

in computer science classes, we wanted to ﬁnd out if,

where, and to what extent, female students might face

increased difﬁculties in our course.

In this tool demonstration paper, we present our

initial version of the dashboard, its application in our

programming course, and the insights and ﬁndings we

gained when using the LA capabilities of the dash-

board. We also report on the current and future plans

to further expand the capabilities of the dashboard.

The remainder of this paper is structured as fol-

lows. In Section 2 we provide a brief introduction

to the topic of LA in programming education and re-

lated tools and provide a brief introduction to our in-

troductory Python programming course and the main

requirements that stem from this course, guiding the

initial development of our dashboard. In Section 3

we then present the dashboard and its features, with

a concrete application use case in Section 4. Finally,

in Section 5 we discuss enhancements, additional fea-

tures we are planning on adding as part of our ongoing

work, and conclusions.

2 BACKGROUND AND COURSE

SETTING

In this section, we present the background and tools

related to our work, and a brief overview of our in-

troductory programming course, and requirements for

our LA dashboard derived from our experiences.

2.1 Learning Analytics in Programming

Education

Learning analytics has already successfully been ap-

plied in programming education. L

opez-Pernas and

Saqr (L

opez-Pernas and Saqr, 2021) combine data

from different sources, such as learning manage-

ment systems and programming assessment tools

to identify learning patterns among students. The

programming learning platform Artemis integrates

competency-based learning to generate personal-

ized learning paths for individual students (S

olch

et al., 2023). Other work analyzes IDE usage pat-

terns of students to get insights into their skills

and performance (Ardimento et al., 2019). Uta-

machant et al. (Utamachant et al., 2023) assess stu-

dent engagement levels and identify at-risk students

through learning activity gaps. In general, LA has

been a growing issue in recent years with active re-

search and a slew of tools on the commercial market.

Moreover, established LMSs have integrated capabil-

ities into their platforms. For example, Moodle, as

an open-source platform, provides analytics capabil-

ities via a plug-in extension (Moodle, 2023). Moo-

dle Analytics provides several different models (static

and ML-based) that allow generating statistics about,

for example, drop-out risks, activities that are due to

submission, and further predictive models. In this

context, Mwalumbwe et al. (Mwalumbwe and Mtebe,

2017) conducted a study with the intent to develop an

LA tool and analyze data from Moodle LM systems.

Focusing on students as a target user group, Peraic

and Grubisic (Perai

c and Grubi

c, 2022) have pre-

sented a “Learning Analytics Dashboard for students”

(LAD-s), providing visualization for student success

and engagement and further providing predictive an-

alytics capabilities.

Woodclap (Woodclap, 2023) is another platform,

that focuses on virtual classrooms facilitating com-

munication with students on smartphones, messages,

and real-time interaction while monitoring student en-

gagements and providing feedback on teaching tech-

niques. Moreno-Medina et al. (Moreno-Medina et al.,

2023) used this setting with chemical engineering stu-

dents in combination with gamiﬁcation strategies to

assess and improve student participation and moti-

vation. Krusche and Berrezueta-Guzman (Krusche

and Berrezueta-Guzman, 2023) provide an interactive

learning environment for programming classes foster-

ing iterative performance enhancement by real-time

feedback mechanisms. However, the platform does

not integrate support for diversity analysis and is lim-

ited to task-level analysis.

While existing systems offer valuable functionali-

ties for course management and related analysis, they

lack speciﬁc support for assignment-level and task-

level analysis of programming courses. Also, support

for diversity analysis is often limited. This motivated

us to develop a customized dashboard for our setting.

A Learning Analytics Dashboard for Improved Learning Outcomes and Diversity in Programming Classes

619

2.2 Course Setting

We started our introductory Python programming

course in 2021, as part of a new university-wide digi-

talization initiative, where all study programs (includ-

ing non-technical/CS-related ones) should gain some

familiarity with programming and algorithmic think-

ing. As part of this, we took over the programming

education for business students, particularly, business

administration and economics.

With a total of 6 ECTS, the course is split into

a weekly, slide-based lecture with additional live-

coding sessions, and a corresponding weekly exer-

cise where students should apply the concepts from

the previous lecture by solving examples during class

and as part of their homework. Pair programming

is applied during the exercise and by this students

should work together on programming tasks covering

the topic of the lecture. Additionally, homework as-

signments consisting of 5-6 individual tasks are dis-

tributed that have to be completed and submitted by

the students within one week. Tutors manually cor-

rect the assignments, give feedback, and assign points

to the tasks of the assignments. Students are graded

based on the points they receive for the weekly as-

signments and an exam at the end of the semester.

The main challenge, in this case, was that, com-

pared to a computer science study program, where

one can expect a certain level of technical (and mathe-

matical) background, the students participating in our

courses are quite diverse, with different educational

backgrounds and prior knowledge related to program-

ming. For most students, our course was the ﬁrst

time they have written code and/or executed a pro-

gram written by themselves.

For this purpose, we opted for Python as a pro-

gramming language, instead of Java – which is the

standard language for programming education in CS

courses, in conjunction with Jupyter Notebooks. The

weekly assignments are distributed as Jupyter Note-

books and the students submit their solutions as note-

books in Moodle. The ﬁnal exam at the end of the

semester is conducted with the CodeRunner (Lobb,

Richard and Hunt, Tim, 2023) plugin in Moodle.

2.3 Stakeholders & Requirements

In higher education, effective utilization of LA plays

an important role in ensuring effective curriculum

management and enhancing student outcomes. This

is especially important for lecturers, and course in-

structors who engage directly with the course content

and the students. While the initial design of our plat-

form primarily targets the needs of these educators,

there’s also a foresight to expand the platform’s capa-

bilities to incorporate the needs of students and pro-

gram managers in the future. For now, we deﬁned the

following requirements for our dashboard:

R1. Course Management. A fundamental re-

quirement revolves around the management of

courses. This includes functionalities to set

course settings, such as determining the start

and end dates, entering the number of assign-

ments, deﬁning requirements such as the num-

ber of submissions and points for passing the

course, and setting the number of students en-

rolled in the course.

R2. Document Management. The dashboard

should allow for the seamless upload of graded

Jupyter Notebooks, adhering to a deﬁned JSON

format.

R3. Analytical Insight into Assignments. To track

course progress and ensure equitable assess-

ment, there’s a need to provide analytics about

the number of submissions per assignment over

the semester.

R4. Student Data Management. This encompasses

the ability to manage pertinent student data, in-

cluding their names, IDs, gender, and details

about their enrolled study program.

R5. Descriptive Statistics on Performance. For an

in-depth analysis of student performance, edu-

cators require a distribution of points per assign-

ment over the semester for all students. Addi-

tionally, a separate analysis ﬁltered by gender,

visualized using box plots to depict the variabil-

ity and central tendency is needed. For each as-

signment a detailed breakdown into the number

of submissions, average points for the collective

student body, and an analysis separated by gen-

der is necessary. An average effort metric fur-

ther illuminates the student’s engagement levels.

A similar granularity of insights is required for

each task and exam.

R6. Individual Student Analytics. For personal-

ized feedback and support, each student’s pro-

ﬁle should be enriched with their performance

metrics, including points per assignment, aver-

age points, pass status, number of submissions,

and other relevant details.

R7. Export Capability. Recognizing the diverse

uses of such data, there should be a provision for

exporting student-speciﬁc data for further anal-

ysis or reporting purposes.

CSEDU 2024 - 16th International Conference on Computer Supported Education

620

Figure 1: Learning Analytics Dashboard: Assignment-level analysis view.

3 ANALYSIS DASHBOARD

In this section, we present details of our Learning An-

alytics Dashboard by ﬁrst providing an overview of its

main features from this ﬁrst iteration prototype and

then providing a brief overview of the technical de-

tails of the implementation. A short demonstration

video of the main Dashboard features

is available on-

line.

3.1 Features

The requirements of our initial implementation were

driven by the need to gain insights into students’ abil-

ity to successfully solve assignments, the submission

rates of individual assignments and constituent tasks,

and identifying potential gender gaps (cf. R3, R5),

hence largely by our own requirements. In the fol-

lowing, we provide a brief overview of the function-

ality and show examples in the dashboard (cf. Fig. 1

and Fig. 2 )

• General Analysis and Trends of Assignments. In

our previous programming classes, we have often ex-

perienced drop-outs in the middle of the semester or

even towards the end of the course. Therefore, one of

our main requirements was to get a better overview

of individual assignments (handed out on a weekly

basis), and whether there was a steady number of

handed-in assignments (and successfully completed

https://github.com/TeachingAndLearningSciences/res

ources

tasks within the assignment) or a noticeable decline in

submissions over time. Fig. 1 provides an overview of

the main view of the dashboard. The top part [A] pro-

vides information about the raw submission numbers

for each assignment and allows us to easily identify

if submission numbers are declining for a particular

assignment, or steadily over time. Additionally, the

lower [B] part provides an overview of the results for

each assignment, i.e. the points achieved by students,

and the spectrum (Box Plot) of the results. This helps

us to identify exercises that might be particularly dif-

ﬁcult (where students have received fewer points) or

potential effects of different educational backgrounds

(where we have a broad spectrum of points achieved).

• Detailed Analysis on Assignment and Task Level.

While analysis on the assignment level can provide

some valuable insights into the overall course, it does

not provide sufﬁcient details on how students handle

individual assignments and the topics and ultimately

learning objectives associated with these assignments

(and the constituent tasks). The second analysis level

(cf. Fig. 2), therefore, is concerned with drilling down

into individual assignments and tasks part of the as-

signment (cf. R6). The top part [C] in this case pro-

vides again an overview of submission numbers and

results, whereas the bottom part [D] goes into further

detail for each of the tasks. For each task, we get de-

tailed insights into how well the students performed,

in terms of the points achieved. The ”point-per-point”

visualization (Grouped Bar Charts) provides detailed

insights on the distribution of points for individual

tasks and allows us to identify tasks that might be po-

A Learning Analytics Dashboard for Improved Learning Outcomes and Diversity in Programming Classes

621

Figure 2: Learning Analytics Dashboard: Task-level analysis view.

tentially too difﬁcult or complex.

• Gender Analysis. A cross-cutting concern for all

analysis activities is the aspect of gender. Based on

our previous, multi-year, experience of offering basic

programming classes for various different study pro-

grams, we have observed gender gaps in several of our

courses. Research in this area has shown that precau-

tionary measures and actions can (at least partially)

rectify such issues, for example, by providing ap-

propriate teaching material and assignments (Schmitz

and Nikoleyczik, 2009; Spieler and Slany, 2018).

• Automated Analysis. One of our key requirements

was to facilitate automated analysis of graded note-

books, while retaining manual grading of assignments

performed manually by tutors. Tutors do not only

grade assignments and tasks, but provide individual

feedback about how well a problem was solved, and

give hints and samples when a task could not be com-

pleted. With weekly assignments, the workload for

tutors is already quite high and we did not want to bur-

den them with additional requirements (e.g., entering

results in yet another tool – i.e. our dashboard). In-

stead, we opted for an automated parser, that reads out

assignment/task points (which are entered in a struc-

tured manner) and stores them as JSON information

in the meta-data of the notebook (cf. R2). This in-

formation is then used for subsequent analysis in the

dashboard. As a positive side-effect, this also decou-

ples the dashboard from the speciﬁc structure/format

of the assignments and allows for updated/changed

Jupyter notebooks in the future, as long as the grading

information is provided in the predeﬁned JSON for-

mat. This further contributes to the aspect of general-

izability of the dashboard with potential applications

to other (types) of programming classes (cf. further

discussion in Section 5).

• Other Capabilities. Besides the main analysis

capabilities, additional functionality is related to the

ability to deﬁne courses with respective course set-

tings (e.g., the number of students part of the course,

grading schemes, and number of assignments) (cf.

R1). Additionally, as establishing interfaces to exist-

ing university systems where student data is stored,

is typically challenging, we added the ability to store

basic student information (e.g., names, gender, study

program), to ensure data privacy, stored only locally

on university premises (cf. R4). In conjunction with

this, we also added the ability to export results (cf.

R7) in a standard CSV format, to enable grading in-

formation to be fed back to the existing university

grading system.

3.2 Implementation

To facilitate easy access and availability to a broad

range of users, we decided to implement our Learn-

ing Analytics Dashboard as a web application using

JavaScript. The components of the dashboard are

structured in a 3-tier architecture, presentation, logic,

and data layer with central data storage. For this pur-

pose, we use a PostgreSQL database where informa-

tion in courses, and results extracted from the Jupyter

notebooks are stored. As the dashboard uses sensi-

ble data concerning students and learning outcomes,

we refrained from using cloud services, but deployed

our application as containers using Docker

. The core

implementation, presentation, and logic use Node.js

Next.js

, and React

https://www.docker.com

https://nodejs.org/en

https://nextjs.org

https://react.dev

CSEDU 2024 - 16th International Conference on Computer Supported Education

622

4 APPLICATION EXAMPLE &

DISCUSSION

As an initial assessment of the usefulness of our

Learning Analytics Dashboard application, we used

it to analyze one iteration of our Python course in the

summer semester of 2022. In this section, we present

the insights and ﬁndings and further discuss its limi-

tations and potential threats that need to be taken into

account.

4.1 Analyzed Python Course

As part of this initial validation, we used the

dashboard to analyze students’ performance in 10

homework programming assignments throughout the

course. The assignments covered a range of topics

from variables and data types, to advanced modules

like NumPy and Pandas, as well as object orientation.

Fig. 3 (top) shows the number of submissions for

each assignment and the distribution of points re-

ceived for all students. The bottom part shows the

detailed results for two tasks part of Assignment 9.

The analysis of the assignments with the help of our

dashboard revealed several key insights.

• Submission Trends. While there was a slight de-

crease in the number of submissions for Assignments

9 and 10, we did not observe a signiﬁcant drop-out.

The lower numbers for Assignments 9 and 10 can be

attributed to the fact that only 8 out of 10 submissions

were mandatory.

• Assignment Metrics. For the ﬁrst three assign-

ments, covering basic concepts, data types, and sim-

ple programs, we observed a high average score and

little spread, However, further into the semester and

with increasing complexity, from Assignment 4 on-

ward, we observed a lower average score and a higher

spread in points.

• Assignment Drill-Down. The ability to further an-

alyze constituent tasks of an assignment also revealed

some interesting insights, particularly for assignments

where we already observed a signiﬁcant spread in

points. Particularly, for Assignment 9, which covered

the topic of modules including Math, NumPy, Mat-

plotlib, and Pandas, we observed the largest variabil-

ity in scores. Notably, one-third of the students scored

less than 1 out of 3 points in the pandas task, whereas

half of the students reached the maximum score of 3

(cf. Fig. 3 – bottom part). For Assignment 10, which

focused on object orientations, the median score in-

creased, indicating better understanding compared to

the previous assignment.

• Gender Gap. Throughout the semester and across

all assignments, in contrast to our initial assumptions,

Matplotlib

Pandas

Figure 3: Results from our analysis – Trends in Assignment

Submission/Points (top) and detailed results for tasks (Pan-

das and Matplotlib) in the modules assignment (bottom).

no signiﬁcant gender-based performance gap was de-

tected. We explicitly designed this course in an inclu-

sive way based on our previous ﬁndings in an intro-

ductory Java course. Further analysis of exam results

and time spent on homework assignments, however,

are required to further conﬁrm the gender equality in

our Python course.

4.2 Implications and Limitations

Using the visualization capabilities of the dashboard

we were able to identify issues during the semester

pertaining to assignments and the topics covered in

the class. Parts of these ﬁndings were later – at the

end of the semester – also conﬁrmed through a ques-

tionnaire we sent out to students, where we specif-

ically asked for issues/challenges they experienced,

and improvements for the class. Assignment 9, cov-

ering different modules like Pandas and Matplotlib,

seemed to be particularly difﬁcult for the students in

our course. We, therefore, developed additional ma-

terial and planned an extra lesson in the following

semester. Also, the rule that students only need to

submit 8 out of 10 exercises leads to the fact that

many students drop the last two exercises (covering

modules and object orientation). As a result, we plan

A Learning Analytics Dashboard for Improved Learning Outcomes and Diversity in Programming Classes

623

to change this rule in future semesters especially the

topic of modules is needed in courses of subsequent

semesters.

• Limitations. This preliminary validation does not

capture other potential factors affecting student per-

formance, such as attendance, participation in tuto-

rials, or speciﬁc educational backgrounds hence, we

can only draw limited conclusions about the learning

outcomes of the course. However, the primary pur-

pose was to assess the usefulness of our dashboard

and the initial set of visualizations and statistical anal-

yses that are provided. Furthermore, we so far only

covered one semester, but after initial positive results,

our future plans to extend and apply the dashboard to

subsequent iterations and other programming classes

(cf. Section 5) will provide us with additional data and

relevant stakeholders for our tool.

4.3 Discussion

While our analytics dashboard offers many possibil-

ities for enhancing Python programming education,

it also raises several concerns that require attention.

One of the most important issues regarding the im-

plementation of our Learning Analytics Dashboard is

the concern for student privacy. The dashboard col-

lects and analyzes various types of data, including as-

signment points, task points, and gender information.

While this data is important for educational insights,

it also raises questions about the conﬁdentiality and

anonymity of student information. Ensuring that the

data is securely stored and accessed only by autho-

rized personnel is vital. Additionally, the dashboard

must comply with relevant data protection regulations

to ensure student privacy.

Ethical considerations extend beyond data pri-

vacy. The gender analysis feature, for instance, could

inadvertently preserve stereotypes or biases if not

carefully designed and interpreted. There is also the

ethical question of how the data should be used. For

example, should low performance of students trigger

an automatic alert to educational staff, or should the

data only serve as an analytical tool for course im-

provement? Balancing data utility and ethical consid-

erations is crucial in this case.

The risk of data misinterpretation is inherent in

any analytics tool. In the educational context, incor-

rect interpretation of the dashboard’s data could lead

to misplaced educational interventions. For exam-

ple, a gender-based performance gap in assignment

points might be wrongly attributed to pedagogical is-

sues when external factors could be inﬂuencing the

data. Therefore, it is essential to provide adequate

training for lecturers and program managers who will

be interpreting the dashboard’s data. Contextualizing

the data with qualitative insights is also recommended

to avoid simplistic or misleading conclusions.

Future work should focus on addressing these is-

sues through a combination of technical safeguards,

policies, and user education to ensure that the dash-

board serves as an effective, ethical, and secure edu-

cational tool.

5 CONCLUSION AND FUTURE

WORK

The rapid development of LA in higher education em-

phasizes the need for systematic management of com-

petencies and learning objectives. In this tool demon-

stration paper, we introduced an innovative Learning

Analytics Dashboard speciﬁcally designed for an in-

troductory Python programming course. This dash-

board not only aims to assist educators in pedagogi-

cal decisions but also focuses on the critical area of

gender diversity within the course setting.

Our initial application which we used for our own

analysis, provided a series of valuable insights into

student performance and engagement, pointing out

speciﬁc challenges regarding the topics of modules

in Python. We could not detect a signiﬁcant gender

gap and drop-out rates in our course. These insights,

even with our initial prototype, already demonstrated

the power of LA as not just a reactive tool for un-

derstanding student performance, but as a proactive

mechanism that allows for targeted interventions to

enhance educational equality.

Future work will expand on these initial successes.

We plan to enhance the dashboard’s capabilities to

include more diversiﬁed analytics features, poten-

tially adding support for analyzing different educa-

tional backgrounds. We further plan to add support

for competency management and the establishment

of links between competencies and assignment and

exam tasks and to analyze competency coverage of

the tasks and competency achievements of students

in the course. We are currently also working on sup-

port to increase the degree of automation. This in-

cludes a dedicated grading-support plug-in in Visual-

Studio for tutors, that generates the JSON data and

automatically uploads notebooks to the dashboard

when graded. Furthermore, we plan on going beyond

Jupyter notebook-based Python courses, and support

for analysis capabilities over multiple semesters. The

current version of our dashboard focuses on educa-

tors as our primary stakeholders. In the future, we

will also provide views for students to monitor their

progress in the course.

CSEDU 2024 - 16th International Conference on Computer Supported Education

624

REFERENCES

Ardimento, P., Bernardi, M. L., Cimitile, M., and Ruvo,

G. D. (2019). Learning analytics to improve coding

abilities: a fuzzy-based process mining approach. In

Proc. of the 2019 IEEE International Conference on

Fuzzy Systems, pages 1–7. IEEE.

Bergsmann, E., Schultes, M.-T., Winter, P., Schober, B.,

and Spiel, C. (2015). Evaluation of competence-based

teaching in higher education: From theory to practice.

Evaluation and Program Planning, 52:1–9.

Groher, I., Vierhauser, M., Sabitzer, B., Kuka, L., Hofer, A.,

and Muster, D. (2022). Exploring diversity in intro-

ductory programming classes: an experience report.

In Proc. of the ACM/IEEE 44th International Confer-

ence on Software Engineering: Software Engineering

Education and Training, pages 102–112. IEEE.

Johnson, J. W. (2020). Beneﬁts and pitfalls of jupyter

notebooks in the classroom. In Proc. of the 21st an-

nual Conference on Information Technology Educa-

tion, pages 32–37. ACM.

Krusche, S. and Berrezueta-Guzman, J. (2023). Introduc-

tion to programming using interactive learning. In

Proc. of the 2023 IEEE 35th International Confer-

ence on Software Engineering Education and Train-

ing, pages 178–182. IEEE.

Lobb, Richard and Hunt, Tim (2023). Moodle CodeRun-

ner. https://moodle.org/plugins/qtype coderunner.

[Online; Accessed 01-10-2023].

opez-Pernas, S. and Saqr, M. (2021). Bringing synchrony

and clarity to complex multi-channel data: A learn-

ing analytics study in programming education. IEEE

Access, 9:166531–166541.

Malhotra, R., Massoudi, M., and Jindal, R. (2023). Shift-

ing from traditional engineering education towards

competency-based approach: The most recommended

approach-review. Education and Information Tech-

nologies, 28(7):9081–9111.

Marquardt, K., Wagner, I., and Happe, L. (2023). Engag-

ing girls in computer science: Do single-gender inter-

disciplinary classes help? In 2023 IEEE/ACM 45th

International Conference on Software Engineering:

Software Engineering Education and Training (ICSE-

SEET), pages 128–140. IEEE.

Moodle (2023). Moodle Analytics. https://docs.moodle.or

g/402/en/Analytics. [Online; Accessed 01-10-2023].

Moreno-Medina, I., Pe

nas-Garz

on, M., Belver, C., and

Bedia, J. (2023). Wooclap for improving student

achievement and motivation in the chemical engi-

neering degree. Education for Chemical Engineers,

45:11–18.

Mwalumbwe, I. and Mtebe, J. S. (2017). Using learning

analytics to predict students’ performance in moodle

learning management system: A case of mbeya uni-

versity of science and technology. The Electronic

Journal of Information Systems in Developing Coun-

tries, 79(1):1–13.

Perai

c, I. and Grubi

c, A. (2022). Development and eval-

uation of a learning analytics dashboard for moodle

learning management system. In Proc. of the 2022

HCI International Conference - Late Breaking Pa-

pers. Interaction in New Media, Learning and Games,

pages 390–408, Cham. Springer Nature Switzerland.

Rubio, M. A., Romero-Zaliz, R., Ma

noso, C., and

de Madrid, A. P. (2015). Closing the gender gap in

an introductory programming course. Computers &

Education, 82:409–420.

Scheffel, M., Drachsler, H., Stoyanov, S., and Specht, M.

(2014). Quality indicators for learning analytics. Jour-

nal of Educational Technology & Society, 17(4):117–

132.

Schmitz, S. and Nikoleyczik, K. (2009). Transdisciplinary

and gender-sensitive teaching: didactical concepts and

technical support. International Journal of Innovation

in Education, 1.

Spieler, B. and Slany, W. (2018). Female teenagers and

coding: Create gender sensitive and creative learning

environments. In Constructionism 2018: Construc-

tionism, Computational Thinking and Educational In-

novation, pages 405–414.

olch, M., Aberle, M., and Krusche, S. (2023). Integrat-

ing competency-based education in interactive learn-

ing systems. In Companion Proc. of the 13th Interna-

tional Learning Analytics and Knowledge Conference,

pages 53–56.

Utamachant, P., Anutariya, C., and Pongnumkul, S. (2023).

i-ntervene: applying an evidence-based learning ana-

lytics intervention to support computer programming

instruction. Smart Learning Environments, 10:37.

Vieira, C., Parsons, P., and Byrd, V. (2018). Visual learning

analytics of educational data: A systematic literature

review and research agenda. Computers & Education,

122:119–135.

Woodclap (2023). Woodclap. https://www.wooclap.com/

de/. [Online; Accessed 01-10-2023].

A Learning Analytics Dashboard for Improved Learning Outcomes and Diversity in Programming Classes

625