Student Teacher Interaction While Learning Computer Science: Early

Results from an Experiment on Undergraduates

Manuela-Andreea Petrescu

and Kuderna-Iulian Benta

Department of Computer Science, Faculty of Mathematics and Computer Science, Babes¸-Bolyai University,

Cluj-Napoca, Romania

Keywords:

Learning, Professional Development, Integration, Education, Experiment, Undergraduate.

Abstract:

The scope of this paper was to ﬁnd out how the students in Computer Science perceive different teaching styles

and how the teaching style impacts the learning desire and interest in the course. To ﬁnd out, we designed

and implemented an experiment in which the same groups of students (86 students) were exposed to different

teaching styles (presented by the same teacher at a difference of two weeks between lectures). We tried to

minimize external factors’ impact by carefully selecting the dates (close ones), having the courses in the same

classroom and on the same day of the week, at the same hour, and checking the number and the complexity

of the introduced items to be comparable. We asked for students’ feedback and we deﬁne a set of countable

body signs for their involvement in the course. The results were comparable by both metrics (body language)

and text analysis results, students prefer a more interactive course, with a relaxing atmosphere, and are keener

to learn in these conditions.

1 INTRODUCTION

We propose to validate the importance of interac-

tion, general atmosphere, and teacher’s passion in the

learning process in Computer Science courses. We

created an experiment where the delivered informa-

tion is comparable but the teaching style and interac-

tion are deliberately modiﬁed.

By analyzing the student’s perception - both on

visual level/observed behavior (they were sleepy,

yawns, laying on the benches) and from their col-

lected textual feedback. The feedback was analyzed

manually using thematic analysis.

We wanted to ﬁnd out if there is a difference in

perceived understanding and interest from students

when we use interactive methods combined with el-

ements of nonverbal communication compared to a

classical teaching method (just presenting the infor-

mation). Also, to ﬁnd out if the teaching method-

ology impacted students’ perception related to lec-

ture in terms of affectiveness (they feel enthusiastic,

good, interested or bored, discouraged) and in cogni-

tive terms (they understand better / worse). We cre-

ated a list of nonverbal elements (yawns, laying on

the desks, arms positions, and so on) that we analyzed

https://orcid.org/0000-0002-9537-1466

https://orcid.org/0000-0002-4245-599X

during the lectures and analyzed their frequency. At

the same time, we asked the students for anonymous

feedback.

Education is a key component of society, inﬂuenc-

ing a country’s future development; due to its impor-

tance, countries allocate a part of their expense budget

to the educational system. Except for money, other

factors inﬂuence the attractiveness of a speciﬁc do-

main. For Computer Science and Mathematics do-

main, for example, the interests of secondary schools

students depend on a set of factors: the student’s so-

cioeconomic status, performance, self-efﬁcacy, mo-

tivation, engagement, and task value beliefs (Kahra-

man, 2022; Spieler et al., 2020). Universities noticed

that the number of students that graduate is smaller

compared to the enrolled students, some tried to de-

crease the drop-out rates by offering additional mate-

rials and offering some courses in an online or hybrid

format to minimize the tuition fees and to increase

accessibility. Online courses increase accessibility as

the students can learn whenever they have time, they

can learn at their own pace, and having access to re-

sources before the course allows them to ask more

and more complex questions (Baquerizo et al., 2020);

however neither going online solved all the problems.

Paper (Baquerizo et al., 2020) analyzes methods to

motivate students in an online environment, as (Pe-

Petrescu, M. and Bentasup, K.

Student Teacher Interaction While Learning Computer Science: Early Results from an Experiment on Undergraduates.

DOI: 10.5220/0011844400003470

In Proceedings of the 15th International Conference on Computer Supported Education (CSEDU 2023) - Volume 1, pages 209-216

ISBN: 978-989-758-641-5; ISSN: 2184-5026

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

209

trescu and Sterca, 2022) mentioned students’ difﬁ-

culty in ﬁnding a quiet place for learning.

2 LITERATURE REVIEW

A key to success for companies and countries is hu-

man capital, well prepared, capable people can raise

a company or a country. To have prepared and well-

qualiﬁed people, education, the methods to transmit

information, and information accessibility play a ma-

jor role. Due to this need, different methods were

analyzed and proposed in the literature with the de-

clared scope to increase teaching effectiveness, some

of the methods were strictly related to the computer

science domain (Liu et al., 2022; Salas, 2017; George,

2020). Teaching methods in computer science were

also subjects for other papers/books that analyzed the

impact of teaching styles, some mention teacher col-

laboration methods, technical professional develop-

ment (Wang et al., 2021), or offering guides for com-

puter science instructors in universities (Hazzan et al.,

2020).

In (Erascu and Mladenovici, 2022) a study took

place while applying Student-centered Learning in

Computer Science. One of the three research ques-

tions was trying to investigate if the student centered

approach is perceived as better compared to a control

group. Pre-tests and post-tests were used. The re-

sults show a positive impact of the student centered

approach on the active learning dimension. However,

our study is oriented more toward the subjective per-

ception of the didactic process. The perspective from

(Makhlouf. and Mine., 2021) is similar to ours in the

sense that the comments are clustered in a small num-

ber of classes with similar meaning to build up an au-

tomated feedback system that facilitates individual-

ized feedback given by the professor to students. By

contrast our work focus on understanding the impact

of nonverbal cues effect on the learning process.

The relationship between emotions and engage-

ment in learning in an e-learning controlled environ-

ment is piloted using three biosensors (El-Abbasy.

et al., 2018): a camera, a 14-channel EEG, and an

eye-tracker. The lab-like setup, the reduced number

of participants (15), and the small number of emo-

tions considered give us inspiration for what we could

use while measuring the learning engagement level in

real-time to improve the quality of teaching. Comple-

mentary to our approach in (Alshaikh. et al., 2021) an

automatic authoring tool based on the Socratic dia-

log is proposed to improve code comprehension. The

authors claim that it improves students’ programming

knowledge by 43% in terms of learning gain.

3 DESIGN OF THE

STUDY/EXPERIMENT

3.1 Experimental Setup

The experiment involved a Computer Science teacher

from a university and students from Computer Sci-

ence. The courses were lectured in English during the

laboratory, it was a ﬁrst-year course called Computer

System Architecture and was structured in two parts:

a theoretical part where different teaching methods

were implemented that lasted for an hour, and the sec-

ond hour was preserved for exercises and examples

and a QA (Question and Answer) session. Because

we considered the students to understand the theoret-

ical part better, all the introduced notions had small

examples during the theoretical part and when possi-

ble, there were mentioned correlations with previous

knowledge/information the students should have ac-

quired. The courses were held in the same location

at the same hour to minimize the environmental im-

pact on the student’s perception. Both courses were

presented using a projector for a presentation, and in

the ﬁrst one, the blackboard was also used to present

other examples. The weather or some speciﬁc events

could impact the student’s perception, but we choose

to have the course as close to one another as possible

(two weeks difference). We were fortunate that there

were no major weather differences during the period.

We selected a time of year that did not have speciﬁc

events (holidays, the ﬁrst days of the semester, con-

certs in the city, and so on). We organized the ﬁrst

course to be the interactive one and the second to be in

a classic ”lecturing” style, because we asked for feed-

back after each course, we wanted them to be able to

analyze the differences (”lecturing” style in our opin-

ion is more commonly used than the interactive style,

where the teacher must do additional efforts to imply

the students).

3.2 Participants

86 students from Computer Science participated in

the experiment, 75 of them were students aleatory

selected from the ﬁrst year, and 11 students were in

the second year of study that did not pass the course

exam in the ﬁrst year and had to retake the course.

They opted to come to a speciﬁc laboratory based

on a timetable and to have the lab with a preferred

teacher. There was no speciﬁc selection depending

on gender or other criteria, the participant set was

randomly selected. The participants were informed

about the experiment, and that their participation was

optional (that’s why the number of students that pro-

vided feedback and participated actively in the survey

CSEDU 2023 - 15th International Conference on Computer Supported Education

210

by providing answers is lower than the overall number

of students that took part in the lectures). The students

were grouped into ﬁve groups, each group having be-

tween 15 and 17 people. For each group was delivered

Lecture 1 and Lecture 2.

The participants were required to provide: 1.

Feedback after each lecture using an anonymous form

that contained questions related to how well they un-

derstood the concepts and the presented information,

2. Give their opinion about the delivery in terms of

interaction, voice volume, body language, the over-

all atmosphere, and if the delivery had an impact on

the interest in learning in that speciﬁc domain. We

decided to have the same participants set for both

lectures/setup so they could appreciate and compare

the teaching styles and the teaching style’s impact on

them because an event or a piece of information could

have a different impact depending on the involved per-

son’s characteristics and mood.

3.3 Lecture Comparison in Terms of

New Information

To have a valid experiment, and to be able to cor-

rectly compare the results of the teaching methods,

we selected two introductory courses: Conversions

and Complementary Code and ASM Arithmetic ex-

pressions that were similar in terms of presented in-

formation (content, difﬁculty, and the number of new

notions). We analyzed the presented information and

structured it into two categories: Build-up and New

Info. In each category, we added the topics that were

discussed and we assigned a difﬁculty level (DF) from

1 to 3, where 1 is easy, 2 is medium and 3 represents

a difﬁcult topic. We summarized the points for each

topic to ﬁnd out what was the overall course difﬁculty.

When we assigned the difﬁculty level we also took

into consideration if a topic was a completely new one

or if it had an increased complexity (build-up) infor-

mation for a subject that the students already should

know. We summarized the difﬁculty levels (DF) and

the results were close (12 to 13) for new information

and (1 to 3) for build-up information. We conclude

that the information presented had a comparable de-

gree of difﬁculty and a comparable number of items.

Below we visualized in Table 1 the topics presented

in each lecture, the build-up type is noted with ”b”.

4 DATA COLLECTION AND

ANALYSIS

The responses were collected anonymously in the

form of open answers, and the students were informed

Table 1: New & build-up topics comparison.

New info - L1 DF Type New info - L2 DF Type

Conversion to base 16

and viceversa

1 b ADC & SBB instruc-

tions

2 b

Conversion to base 8 and

viceversa

1 b IDIV & IMUL 2 b

Addition & Substraction

base 2

1 b CBW & CWD 2 b

Addition & Substraction

base 16

1 b CWDE & CDQ 2 b

Sign bit 1 new Declaring variables 1 new

Complementary code 2 new Declaring constants 1 new

Complement to 2 2 new Push / Pop ﬂags & reg-

istries

2 new

Representation size 1 new Push / Pop ﬂags & reg-

istries

2 new

Asm tools & Program

example

2 new Stack applications 1 new

related to the purpose of the questions and also about

how their responses will be used. The students were

asked to provide optional feedback on a quiz at the

end, but there was no time limit. The response quiz

remained open for two weeks. After this interval, we

considered that we will not get other valid responses.

However, most of the responses (93%) were sent on

the same day - we collected the timestamp by check-

ing the response poll at different intervals.

We opted for open questions as they offer a better

and more profound understanding. We used quantita-

tive methods and more speciﬁc questionnaire surveys

as they were deﬁned in the empirical community stan-

dards (Ralph, Paul (ed.), 2021). These methods were

previously used in Computer Science related studies

(Tichy et al., 1995; Redmond et al., 2013; Petrescu

et al., 2022). For text interpreting, we used thematic

analysis (Braun et al., 2019) to interpret the text and

take into account the recommendations mentioned in

(Kiger and Varpio, 2020) for free text interpretation.

The method was previously used in Software Engi-

neering in other studies: (Cruzes and Dyba, 2011;

Gregory et al., 2015). The reﬂexive approach was ap-

plied as described by (Kiger and Varpio, 2020; Mo-

togna. et al., 2021), and it consisted of the following

steps:

A number of 86 students participated in the exper-

iment, and their feedback was monitored and checked

in two ways:

• By participating voluntarily and providing an-

swers for a quiz - we gathered 73 answers.

• By observing their body language at the end and

during the lectures and looking for quantiﬁable

signs of their interest such as the number of

yawns, the number of questions asked, and for

less quantiﬁable signs such as laying on the desks,

body position, eye movement, and so on.

We asked the following open questions related to

each course to get more information: Q1: Did you

understand the information presented in the lecture?,

Q2: How was the delivery of the course and the teach-

Student Teacher Interaction While Learning Computer Science: Early Results from an Experiment on Undergraduates

211

ing style? and Q3: What was the effect of the teaching

style related to learning interest? What did you like/

dislike about this course? Next, we will break down

and analyze each research question separately.

4.1 Q1:Did You Understand the

Information Presented in the

Lecture?

We tried to structure the information in a logical and

easy-to-understand way, where one presented topic is

tightly related to the previous one, thus creating an

easy-to-follow and clear presentation. The presented

information was doubled by exercises (9 answers in

total appreciated integrating exercises in the theoreti-

cal part): ”we got involved in the exercises too”. In

the answers provided for the course structure and con-

tent, we classiﬁed the keywords into two classes, one

related to content and structure and one related to the

content’s applicability Table 2.

Table 2: Keywords class.

Item Class Selected keywords

Structure Well structured, Easy to understand, Clear, On the point,

Good, Exercises

Applicability Useful, Valuable, Changed the perspective,

We analyzed the prevalence of the keywords in the

received answers, for each course, each student pro-

vided three answers related to structure and content,

delivery, and effect. Sometimes content related key-

words appeared in two out of three, or even in all

three answers, sometimes more keywords appeared

in one answer. We measured them to establish the

general impact of a feature and then we compared

the results for each lecture. We analyzed the an-

swers, each text could contain none, one or more key-

words (clearly, well structured, and so on) that were

related to a positive appreciation. When we sum-

marized all the keywords, they were more than the

number of the students, so the percentage compared

to the number of the students was 114.58%. Same

method was performed when analyzing the responses

for the second lecture and we obtained 74.07%. The

positive appreciation related to content and structure

was higher in the ﬁrst course compared to the sec-

ond course (114.58% vs 74.07%). We took into con-

sideration two factors: in the ﬁrst course the students

were more motivated to provide longer answers, thus

increasing the prevalence of speciﬁc keywords (num-

ber of words/answer/course), and having 74.07% of

answers appreciating positively the structure and the

content in the second course, we considered that both

courses scored high in terms of content and structure,

i.e. ”everything was explained clearly and if I had any

questions they were answered”,vs. ”The information

was structured well, but the delivery is not as helpful

compared to more active and interactive one”.

However, the keyword reﬂecting the overall appli-

cability was much higher in the ﬁrst course, we be-

lieve the difference lies in the information taught: in

the ﬁrst lecture the students found out why we have

speciﬁc data types, information generally useful ”The

content was interesting and it changed the way I view

computers and programming”. In the second lecture,

they found only instructions related to Assembly Pro-

gramming Language, and considered that the topics

presented in lecture 2 do not have any applicability.

Figure 1: Applicability and Structure Keywords Prevalence.

Based on their answers, we concluded that:

A. The information was structured clearly and easily.

B. The students appreciated the applicability of the

presented information in the ﬁrst course, thus rein-

forcing the fact that students are more interested in

practical aspects (when they realize the theoretical

part’s applicability).

4.2 Q2: How Was the Delivery of the

Course and the Teaching Style?

For the second question, we identiﬁed two major

classes: one that refers to the delivery itself contain-

ing keywords such as pace, interactivity, and the vol-

ume of the voice, and one that refers to the teacher’s

involvement: passionate, calm, open-minded. We

counted the prevalence of the ﬁrst class in the answers

for the ﬁrst lecture (35 mentions), the relatively sim-

ilar prevalence was obtained for the second lecture,

for the same class, but the appreciation had a nega-

tive connotation (31 mentions): no interaction, blunt,

boring, bad delivery. In both cases there were key

items with neutral connotation: delivered straight, or

”Okay”. The Table 3 reﬂects the resulting keywords

for the second question.

CSEDU 2023 - 15th International Conference on Computer Supported Education

212

Table 3: Delivery Keywords Prevalence.

Item Class Selected keywords (number of appearance)

Content Delivery

Lecture 1

Interactive / dynamic (9), Well delivered (9), Right

pace (6), Presentation ok /good (4), Open discus-

sion (2), Check understanding (2), Answers (2),

PPT and white board (1).

Content Delivery

Lecture 2

No interaction/interactivity (19), Pace too fast (5),

Blunt/Boring (4), Ok (4), Bad delivery (1), Mono-

tone (1), Delivered straight (1)

Teacher’s attitude

Lecture 1

Relaxed Atmosphere (9), Friendly (5), Active (4),

Calm (3), Cherry attitude (1), Nice (1), Helpful (2),

Funny (2), , Involved (2), Open-minded (1)

Teacher’s attitude

Lecture 2

Cold and distant way (2), No emphasis (1)

The answers from the ﬁrst lecture provided in-

formation not only about the delivery but also their

perception related to a supportive environment: ”the

information was delivered relatively good, you made

sure we were not behind”. We also got short answers

related more to the lecture difﬁculty: ”Pretty basic

stuff, she made it feel interesting though”. In the

second lecture, there were still examples presented,

but the delivery changed: sitting down, fewer body

movements and gestures, less eye contact, and so on.

Some students managed to ﬁgure out some of the

teacher’s delivery methods that inﬂuenced the presen-

tation: ”The information was hard to follow. There

was no emphasis put on the more important informa-

tion”, others stated that ”I managed to understand the

information in its entirety. The content in itself was

logical and the required explanation or clariﬁcation

of the written theory was provided by the professor”.

The second class keywords that were identiﬁed

can be traced more to the lecturer’s personality than

to the teaching style: active, friendly, calm, helpful,

and even funny: ”The atmosphere was relaxed and

the course was taught in a fun way”

It was interesting that all these key items appeared

in the answers from the ﬁrst course, and none ap-

peared in the answers related to the second course.

There were no key items reﬂecting personality char-

acteristics. However, key items were referring to the

overall presentation ”It was bland, boring, but I man-

aged to understand”.

For this section, we can conclude that:

A. The presentation mode inﬂuenced the perceived

difﬁculty level of the presented information. The stu-

dents found it more difﬁcult to concentrate and not

lose focus during the course when the course was not

presented in an interactive way (even if there were

exercises and examples and the teacher answered the

questions).

B. Students reacted positively to the lecture’s person-

ality, appreciating a more ”open discussion” lecture

than a classical one. In the classical ”just words”

style (as one of the students mentioned”, the students

were bored ” it felt like a YouTube tutorial, tutorials

are useful but not enjoyable”.

4.3 Q3: What Was the Effect of the

Teaching Style Related to Learning

Interest? What Did You Like/

Dislike About this Course?

We expected that teaching style to have an impact on

the overall interest. We analyzed the teaching style

effect on students by two methods:

• We analyzed their body language during and at

the end of the course to look for signs.

• We asked for their feedback in the form of open

questions.

In the interactive lecture, when they were actively

asked to answer questions and constantly provoked

related to their knowledge or to make logical deduc-

tions, the students were active, we counted a small

number of yawns only at the beginning of the courses

that took place from 8 am. In the other course, fa-

tigue, sleepiness, and uninterested signs of body lan-

guage appeared approximately half an hour after the

course started, and much faster (after 10-15 minutes)

in the courses that took place at 8 am. Next, in Ta-

ble 4 can be seen a comparison between the observed

body language signs, two lectures took place in the 8

am - 10 am interval, two lectures took place in the 10

am-12 am interval, and one lecture took place at 12

am - 2 pm interval for each topic. So in total, we an-

alyzed 10h lectures, by observing and counting how

many times a speciﬁc behavior appeared.

Table 4: Body language signs observed during the lectures.

Body Signs

That Ap-

peared (per

lecture)

L1 (8.00-

10.00)

L1 (10.00-

12.00)

L2 (8.00-

10.00)

L2 (10.00-

12.00)

Laying on

the back,

related posi-

tions

Only in the

ﬁrst 10-15

min, 22%

from stu-

dents

No signs After ﬁrst

10-15 min,

at 68% from

students

After ﬁrst

30 min, at

64% from

students

Yawns Only in the

ﬁrst 15 min

(9% of stu-

dents)

No yawns Appeared at

30% of stu-

dents, more

than once

Appeared at

25% of stu-

dents, more

than once

Unfocused

eye look

Only in

the ﬁrst

10-15min,

at 13% of

students

No signs no-

ticed

Appeared to

90% from

students

After 30

min, at 85%

of students

We also performed a text analysis by manually

processing the students’ responses to the open ques-

tions, and the results were congruent with the body

Student Teacher Interaction While Learning Computer Science: Early Results from an Experiment on Undergraduates

213

language that was noticed during the lectures. Most

of the students appreciated the interactive lecture, and

some of them mentioned elements of interactivity: ”I

enjoyed the delivery and the fact that the class was in-

teractive. The presentation gave everything a bit more

structure”. Some answers reﬂected more the effect,

even mentioning causes that inﬂuenced the learning

environment: ”I feel really good at this lab because

the atmosphere is really friendly and open-minded. I

really enjoy it.”.

However, no charm and no teaching method can

overcome the call of nature, in the course that took

place during the lunch period, 20% of them men-

tioned they were hungry: ”I want to learn more

and I’m also hungry..”, we even had a response that

mentioned: ”Really dizzy, hungover from last night

party”. Curiously, there were no such mentions for

the second course (which was held during the same

hours). We assumed based on the shortness of an-

swers that they just decided to provide less informa-

tion.

As for the second part of the experiment, the stu-

dents mentioned that they were bored and missed the

interest, we had short, neutral answers: ”Presented in

a linear way”, ”It was okay”, or little longer answers:

”Monotone, kind of boring, not interactive”. Because

they participated in both courses (same teacher, dif-

ferent teaching styles), they performed a comparison

of the teaching styles, thus offering reliable informa-

tion related to the effect of the teaching style. They

realized the difference and emit conclusions based on

their experience: ”I lost my concentration level in the

middle of the course and I missed the interaction with

the teacher” and even recognized doing completely

something else: ”I disliked the teaching style, at some

points I even realized I completely zoned out and just

worked on other stuff on my laptop, completely ignor-

ing the teacher.” Item classes are detailed in the next

table.

Table 5: Classes of items.

Needs extra

work

”need more time, have to reread, need to learn

more, information was difﬁcult, less clear”

Easy ”easy to understand, easy, understand all, not

overwhelming”

Environment ”less stressed, relaxed atmosphere, monotonous,

boring, attention not drawn ”

Student’s

perception

”entertained, liked, enjoy, surprised, no fear, no

boring” or ”did not like, felt discouraging, not pay

attention”

Student’s

reaction

”waiting for next courses, interested”, ”impact

learning desire”, ”worried about course difﬁculty”

To see how much the teaching style impacted how

the students perceived the difﬁculty (even if the top-

ics had a similar level of difﬁculty), we checked how

many students reported the information to be easy for

the ﬁrst course compared to the second, 60.42% of the

answers considered the information delivered in the

ﬁrst course to be ”easy” compared to 40.74% of the

answers for the second course. The numbers corre-

late with the number of answers that mentioned: ”ex-

tra work” is needed: 20.83% for the ﬁrst course com-

pared to 44.44% for the second course: Fig 2.

Figure 2: Perceived Course Difﬁculty.

Regarding the environment and the course atmo-

sphere, positive characterizations were received for

the ﬁrst lecture ”relaxed atmosphere, less stressed”,

which appeared in 18.75% of the answers for the ﬁrst

course. Negative characterizations were received for

the second course ”monotonous, boring” (14.81% of

the total number of answers).41.67% of the answers

for the ﬁrst course mentioned the students ”liked, en-

joyed, entertained” as 25.93% of the answers from

the second course mentioned ”did not like, felt dis-

couraging”, Fig 3. Moreover, 18.75% of the an-

swers in the ﬁrst course mentioned that they are ”in-

trigued to learn more” even if they are worried about

the course complexity (6.25% of the answers). In

the second course, 37.04% of answers mentioned that

the teaching style impacted their learning desire: ”in

time I think it would impact my learning desire.”,”The

switch in teaching style felt discouraging compared to

previous labs”, ”It impacted the learning desire quite

hard, making me not paying attention, and lose fo-

cus”.

Based on their answers, we concluded that:

A. The teaching style impacted their attention and as

a consequence, their desire to learn.

B. The teacher’s ability to relate and to create a re-

laxing calm environment is mentioned by 18.75% of

them, so this factor is important for a large majority.

C. There might be a problem with the overall student-

teacher relationship in the high-school cycle as sev-

eral students appreciated the fact that ”the questions

were answered”, ”we could ask without FEAR”, and

”open talk discussion”, all these suggest a more pro-

found problem.

CSEDU 2023 - 15th International Conference on Computer Supported Education

214

Figure 3: Student’s Perception.

D. No teaching style and no abilities can come against

nature’s call (hungry, sleepiness, and so on).

5 THREATS TO VALIDITY

According to ACM (Association for Computing Ma-

chinery) standards for software engineering research

(Ralph, Paul (ed.), 2021), we were aware that we

need to address possible threats to validity that could

impact the obtained results and we decided to ana-

lyze the following: research ethics, target population

and participant selection, drop-out measures, envi-

ronmental threats, methods to decrease the subjective

form of data processing, hypotheses are missing.

Research ethics: At the beginning of the experi-

ment (at each course) the students were informed that

they will take part in an experiment that consists in

teaching two courses in two different styles, and that

their participation in the survey is optional and anony-

mous. We also informed them about the scope of the

study and how the collected data will be used.

Target population and participant selection: The

groups are formed in alphabetical order based on the

student’s surname, there were 5 aleatory groups se-

lected for the experiment (from a total of 14 groups

of students enrolled in Computer Science - English

line), and all of them were required to participate in

the study. Thus, we assured that the selection was

aleatory.

Drop-out measures: As being optional, a number

of students decided not to participate, and they did

not provide answers; we did not want to use any con-

straint methods, and we could not inﬂuence the drop-

out rates.

Environmental threats or inappropriate design for

the conditions under which the experiment took place:

human subjects can be inﬂuenced in many ways,

that’s why we wanted to control as many external fac-

tors as possible: we used the same classrooms, the

same times of day, and close dates. We used the same

materials (video projector), and we took into consid-

eration course difﬁculty (to same a similar grade of

difﬁculty in terms of new information and complex-

ity). All these factors were taken into consideration so

their perception/opinion was inﬂuenced only by dif-

ferent teaching styles.

Data processing: the process used for data pro-

cessing was in concordance with deﬁned processes

and was also used in other computer science papers.

We deﬁned two distinct methods to validate the re-

sults, by observing behavior in a countable manner

(counting the number of yawns) and by having a text

analysis of the received responses.

Hypotheses are missing: We assumed that teach-

ing style and teacher-students relation impacts the

learning desire even in a Computer Science course.

6 CONCLUSIONS AND FUTURE

WORK

We wanted to ﬁnd out how the students perceive dif-

ferent teaching styles and how the teaching style im-

pacts the learning desire and interest in the course. To

be able to have valid data, we used the same envi-

ronment (classes, course hours, close dates), and the

same students participated and were exposed to both

teaching styles, thus being able to evaluate and com-

pare them. We paid attention to the introduced con-

cepts, they were comparable in terms of number and

complexity. We tried to take into consideration all the

aspects that could inﬂuence the outcome and could

become a threat to the validity of the results. We ana-

lyzed the effects and the results of the teaching styles

in two methods - the ﬁrst one: asking for students’

feedback after each course and the second one: deﬁn-

ing a set of countable behavioral signs (yawns, lay-

ing on the bench). Both methods returned the same

result: students were more attracted to the interac-

tive teaching style and did not show behavioral signs

of boredom. Teacher behavior manager to create a

”calm and relaxing atmosphere” in the ﬁrst course. In

the second course, when the teaching style was de-

liberately changed, to a non-interactive one, the signs

of boredom could be measured and the students re-

ported a lack of interest, daydreaming, and even doing

a completely different thing. A more rigorous com-

parison by using different natural language process-

ing techniques could be used to show the correlation

between the two performed analyses. We plan to de-

ﬁne a metric to measure more accurately the student’s

interest related to the teaching style and also to repli-

cate the experiment for a whole semester so we can

evaluate the results using exam outcomes. A multi-

Student Teacher Interaction While Learning Computer Science: Early Results from an Experiment on Undergraduates

215

modal affective state monitoring could be designed on

the class level to pervasively measure students’ emo-

tions and mood while learning using wearable devices

(Benta et al., 2015).

Funding. The publication of this article was sup-

ported by the 2022 Development Fund of the Babes¸-

Bolyai University.

REFERENCES

Alshaikh., Z., Tamang., L., and Rus., V. (2021). Ex-

periments with auto-generated socratic dialogue for

source code understanding. In Proceedings of the 13th

International Conference on Computer Supported Ed-

ucation - Volume 2: CSEDU,, pages 35–44. INSTICC,

SciTePress.

Baquerizo, G., M

arquez, F., and Tobar, F. (2020). Motiva-

tion in online teaching. 16:316–321.

Benta, K.-I., Cremene, M., and Vaida, M.-F. (2015). A

multimodal affective monitoring tool for mobile learn-

ing. In 2015 14th RoEduNet International Conference

- Networking in Education and Research (RoEduNet

NER), pages 34–38.

Braun, V., Clarke, V., Hayﬁeld, N., and Terry, G. (2019).

Thematic Analysis, pages 843–860. Springer Singa-

pore.

Cruzes, D. S. and Dyba, T. (2011). Recommended steps for

thematic synthesis in software engineering. In 2011

International Symposium on Empirical Software En-

gineering and Measurement, pages 275–284.

El-Abbasy., K., Angelopoulou., A., and Towell., T. (2018).

Measuring the engagement of the learner in a con-

trolled environment using three different biosensors.

In Proceedings of the 10th International Confer-

ence on Computer Supported Education - Volume 1:

CSEDU,, pages 278–284. INSTICC, SciTePress.

Erascu, M. and Mladenovici, V. (2022). Transferring learn-

ing into the workplace: Evaluating a student-centered

learning approach through computer science students’

lens. In Cukurova, M., Rummel, N., Gillet, D.,

McLaren, B. M., and Uhomoibhi, J., editors, Proceed-

ings of the 14th International Conference on Com-

puter Supported Education, CSEDU 2022, Online

Streaming, April 22-24, 2022, Volume 2, pages 442–

449. SCITEPRESS.

George, M. L. (2020). Effective teaching and examination

strategies for undergraduate learning during covid-19

school restrictions. Journal of Educational Technol-

ogy Systems, 49(1):23–48.

Gregory, P., Barroca, L., Taylor, K., Salah, D., and Sharp, H.

(2015). Agile challenges in practice: A thematic anal-

ysis. In Lassenius, C., Dingsøyr, T., and Paasivaara,

M., editors, Agile Processes in Software Engineering

and Extreme Programming, pages 64–80. Springer.

Hazzan, O., Lapidot, T., and Ragonis, N. (2020). Guide to

teaching computer science. Springer.

Kahraman, N. (2022). Middle school boys’ and girls’ career

aspirations in science and mathematics. BO

GAZ

IC¸

University Journal of Education, 39:1–30.

Kiger, M. E. and Varpio, L. (2020). Thematic analysis

of qualitative data: Amee guide no. 131. Medical

Teacher, 42(8):846–854.

Liu, M., Gorgievski, M. J., Qi, J., and Paas, F. (2022).

Increasing teaching effectiveness in entrepreneurship

education: Course characteristics and student needs

differences. Learning and Individual Differences,

96:102147.

Makhlouf., J. and Mine., T. (2021). Mining students’ com-

ments to build an automated feedback system. In

Proceedings of the 13th International Conference on

Computer Supported Education - Volume 1: CSEDU,,

pages 15–25. INSTICC, SciTePress.

Motogna., S., Suciu., D., and Molnar., A. (2021). Inves-

tigating student insight in software engineering team

projects. In Proceedings of the 16th International

Conference on Evaluation of Novel Approaches to

Software Engineering - ENASE,, pages 362–371. IN-

STICC, SciTePress.

Petrescu, M. and Sterca, A. (2022). Agile methodology in

online learning and how it can improve communica-

tion. a case study. pages 542–549.

Petrescu, M. A., Borza, D. L., and Suciu, D. M. (2022).

Findings from teaching entrepreneurship to under-

graduate multidisciplinary students: case study. In

Proceedings of the 4th International Workshop on Ed-

ucation through Advanced Software Engineering and

Artiﬁcial Intelligence, pages 25–32.

Ralph, Paul (ed.) (2021). ACM Sigsoft Empirical Standards

for Software Engineering Research, version 0.2.0.

Redmond, K., Evans, S., and Sahami, M. (2013). A large-

scale quantitative study of women in computer science

at stanford university. In Proceeding of the 44th ACM

technical symposium on Computer science education,

pages 439–444.

Salas, R. P. (2017). Teaching entrepreneurship in computer

science: Lessons learned. In 2017 IEEE Frontiers in

Education Conference (FIE), pages 1-7. IEEE.

Spieler, B., Oates-Indruchova, L., and Slany, W. (2020). Fe-

male students in computer science education: Under-

standing stereotypes, negative impacts, and positive

motivation. Journal of Women and Minorities in Sci-

ence and Engineering, 26:473–510.

Tichy, W. F., Lukowicz, P., Prechelt, L., and Heinz, E. A.

(1995). Experimental evaluation in computer science:

A quantitative study. Journal of Systems and Software,

28(1):9–18.

Wang, S., Bajwa, N., Tong, R., and Kelly, H. (2021). Tran-

sitioning to Online Teaching, pages 177–188.

CSEDU 2023 - 15th International Conference on Computer Supported Education

216