Mas e Deg ee P og am in
In o ma ion Managemen
The In luence o Cha GPT on S uden s’ Oppo unis ic Beha io
and Reduced E o
Bea iz Gui a G ilo
Mas e Thesis
p esen ed as pa ial equi emen o ob aining he Mas e Deg ee in In o ma ion Managemen
NOVA In o ma ion Managemen School
Ins i u o Supe io de Es a ís ica e Ges ão de In o mação
Uni e sidade No a de Lisboa
MGI
NOVA In o ma ion Managemen School
Ins i u o Supe io de Es a ís ica e Ges ão de In o mação
Uni e sidade No a de Lisboa
THE INFLUENCE OF CHATGPT ON STUDENTS’ OPPORTUNISTIC BEHAVIOR AND REDUCED
EFFORT
by
Bea iz Gui a G ilo
Mas e Thesis p esen ed as pa ial equi emen o ob aining he Mas e ’s deg ee in
In o ma ion Managemen , wi h a specializa ion in Knowledge Managemen and Business
In elligence
Supe ised by
Mijail Juano ich Na anjo Zolo o , Ph.D., NOVA In o ma ion Managemen School (NOVA IMS)
July, 2024
i
STATEMENT OF INTEGRITY
I he eby decla e ha ing conduc ed his academic wo k wi h in eg i y. I con i m ha I ha e no
used plagia ism o any o m o undue use o in o ma ion o alsi ica ion o esul s along he
p ocess leading o i s elabo a ion. I u he decla e ha I ha e ully acknowledged he Rules
o Conduc and Code o Hono om he NOVA In o ma ion Managemen School.
Bea iz Gui a G ilo
Lisbon, 2024
ii
DEDICATION AND ACKNOWLEDGEMENTS
This disse a ion would no ha e been possible wi hou he endless suppo and
encou agemen o se e al impo an people in my li e.
Fi s ly, I wan o dedica e his disse a ion o my pa en s, Elsa and Rui. Thei uncondi ional
lo e and belie in me ha e been he ock o my academic jou ney. They ha e consis en ly
suppo ed me h ough e e y challenge and obs acle, and p o ided me wi h he oppo uni ies
ha ha e allowed me o each his miles one. I am especially hank ul o hei suppo du ing
hese di icul yea s, as well as hei cons an ai h and belie in my wo k.
I would also like o exp ess my g a i ude o my sis e , Ma ia. She has been a cons an sou ce
o s eng h h oughou his en i e p ocess. Suppo ing me by lis ening and gi ing me
encou agemen du ing momen s whe e I was eeling down and doub ed mysel .
I am also deeply g a e ul o F ancisco, my sou ce o s eng h du ing ough imes and whose
endless suppo has played a c ucial ole in my success. His pa ience and eadiness o help
whene e I aced doub s ha e mean he wo ld o me.
Finally, I wan o hank my supe iso , D . Mijail Na anjo. His guidance and insigh ul eedback
ha e been undamen al o he de elopmen o his disse a ion. His as and hough ul
esponses o my ques ions and his cons uc i e eedback on my d a s ha e been essen ial
h oughou his jou ney. I’m ex emely g a e ul o he suppo and a ailabili y.
iii
ABSTRACT
This s udy challenges he assump ion ha s uden s become less in es ed in asks once hey
achie e an a e age g ade using AI ools like Cha GPT, his leads o g owing conce ns abou
s uden s o e eliance on Cha GPT o comple e hei asks leading o he dec ease o hei
c i ical hinking. This s udy, d aws da a om 249 pa icipan s ga he ed h ough an online
ques ionnai e and analyzes i using Pa ial Leas Squa es S uc u al Equa ion Modeling (PLS-
SEM), p o ing ha while s uden s a e using Cha GPT o quick answe s, hey o en do no
engage deeply o enhance hei lea ning o o achie e hei ull po en ial, hey end o ca e
less and less abou hei asks when hey achie e an a e age g ade by jus using AI. This leads
o he ac ha he e is a g ow h o s uden s oppo unis ic beha io in academic se ings,
whe e s uden s se le o a e age when hey could excel. The s udy poin s ou he need o
educa ional s a egies ha no only p omo e esponsible use o AI ools bu also encou age
deepe engagemen wi h lea ning ma e ials o p e en a decline in academic igo , educe
hei lack o esponsibili y and p omo e academic success.
KEYWORDS
A i icial In elligence; Cha GPT; Educa ion; S uden s Oppo unism; S uden s Pe o mance; AI
Tools; TTF
Sus ainable De elopmen Goals (SDG):

i
TABLE OF CONTENTS
S a emen o In eg i y ........................................................................................................ i
Dedica ion and Acknowledgemen s ................................................................................. ii
Abs ac ............................................................................................................................ iii
Lis o Figu es .....................................................................................................................
Lis o Tables ..................................................................................................................... i
Lis o Abb e ia ions and Ac onyms ................................................................................ ii
1. In oduc ion .................................................................................................................. 1
2. Theo e ical Backg ound ................................................................................................ 3
2.1. E olu ion o Educa ional Technologies ................................................................. 3
2.1.1. T adi ional Lea ning........................................................................................ 3
2.1.2. AI Tool Dis up ion in he Lea ning P ocess .................................................... 4
2.1.3. Oppo unism and Sel -e icacy in S uden s’ Lea ning .................................... 5
2.1.4. E ec i eness o Di e en AI Tools in Lea ning Tasks and Wha In luences S uden s’
Choices…………………………………………………………………………………………………………………………6
2.2. Indi idual Lea ning S yles ...................................................................................... 8
3. Model Building and Hypo hesis De elopmen ............................................................. 9
4. Me hodological App oach .......................................................................................... 15
5. Resul s......................................................................................................................... 19
5.1. Measu emen Model E alua ion ......................................................................... 19
5.1.1. In e nal Consis ency ..................................................................................... 19
5.1.2. Con e gen Validi y ...................................................................................... 19
5.1.3. Disc iminan Validi y .................................................................................... 20
5.2. S uc u al Model E alua ion ................................................................................ 22
6. Discussion ................................................................................................................... 24
6.1. Theo e ical Implica ions ...................................................................................... 24
6.2. P ac ical Implica ions ........................................................................................... 27
7. Conclusions ................................................................................................................. 29
8. Limi a ions and Fu u e Wo k ...................................................................................... 30
Bibliog aphical Re e ences .............................................................................................. 31
Appendix A – NOVA IMS E hics Commi ee App o al .................................................... 37
LIST OF FIGURES
Figu e 1 – Concep ual model ................................................................................................... 10
Figu e 2 – S uc u al model ...................................................................................................... 22
i
LIST OF TABLES
Table 1 - Bene i s and Challenges o AI in s uden lea ning ............................................. 4
Table 2 - Di e en AI ools o lea ning and hei impac on s uden s' lea ning p ocess
e ec i eness ............................................................................................................. 7
Table 3 - Meaning o each a iable p esen ed on he concep ual model ...................... 10
Table 4 - Adap ed ques ions o each cons uc ............................................................. 15
Table 5 - Sample cha ac e is ics ...................................................................................... 18
Table 6 - Cons uc eliabili y and alidi y measu es ...................................................... 20
Table 7 - Fo nell-La cke c i e ion ................................................................................... 21
Table 8 - C oss loadings ................................................................................................... 21
Table 9 - He e o ai -Mono ai Ra io o co ela ions (HTMT) ....................................... 22
ii
LIST OF ABBREVIATIONS AND ACRONYMS
TEL Technology Enhanced Lea ning
AI A i icial In elligence
LLMs La ge Language Models
TTF Task- echnology i
PLS-SEM Pa ial Leas Squa es S uc u al Equa ion Modeling
CA C onbach’s Alpha
CR Composi e Reliabili y
AVE A e age Va iance Ex ac ed
HTMT He e o ai -Mono ai Ra io o co ela ions
VIF Va iance In la ion Fac o
7
ools can be e ec i ely used o suppo s uden s’ pe o mance on speci ic lea ning asks,
depending on he na u e o he ask and he speci ic lea ning objec i es.
Building on p io esea ch ha has demons a ed he po en ial o AI ools o enhance s uden
lea ning (Kooli, 2023), he able 2 aims o enligh en he impac s o commonly used AI ools in
educa ion. To achie e his, and going om Osamo e al. (2023) s udy, whe e i was p esen ed
a wide and de eloping ange o AI-powe ed educa ional ools. Table 2 p esen s a ocused
analysis o he mos equen ly used AI ools and hei an icipa ed e ec s on s uden lea ning
p ocess e ec i eness.
Table 2 – Di e en AI ools o lea ning and hei impac on s uden s’ lea ning p ocess
e ec i eness
AI Tools
Impac on lea ning p ocess e ec i eness
Cha GPT
Cha GPT p o ides s uden s’ wi h
pe sonalized eedback, suppo , in e ac i e
lea ning expe iences and has a big
a ailabili y.
Gemini ( o me ly Google Ba d)
Gemini also p o ides s uden s’ wi h
pe sonalized eedback, suppo , in e ac i e
lea ning expe iences and has a big
a ailabili y.
G amma ly
G amma ly imp o es w i ing skills,
iden i ies g amma ical e o s, and p o ides
sugges ions o imp o emen .
QuillBo
QuillBo is e y e ec i e a summa izing
ex s and ew i ing sen ences.
Duolingo
Duolingo deli e s engaging exe cises and
in e ac i e games o language lea ning.
DeepL
DeepL p o ides accu a e and na u al-
sounding ansla ions, helping s uden s’ o
ead and unde s and o eign ex s mo e
e ec i ely.
SciSpace
SciSpace can in e p e a esea ch pape and
answe ques ions abou i .

8
By assessing he e ec i eness o a ious AI ools o lea ning asks, i is c ucial o ecognize
ha s uden s’ choices o lea ning ools a e in luenced by a a ie y o ac o s. These ac o s go
om indi idual p e e ences, expe iences, sel -e icacy, he na u e o he lea ning ask, he
a ailabili y o ools and hei pe cep ions o he ool’s e ec i eness.
A s udy by Kembe e al. (2004) alida es he Re ised Two-Fac o S udy P ocess Ques ionnai e
(R-SPQ-2F), which can assess s uden s’ p e e ences o deep and su ace app oaches o
lea ning, which may in luence hei ool choices and alida es he Re ised Lea ning P ocess
Ques ionnai e (R-LPQ-2F), which can p o ide insigh s in o s uden s’ mo i es o choosing
speci ic ools. In Dunn L (2002) s udy i is discussed he ole o cogni i e heo ies in lea ning,
emphasizing ha s uden s’ a e mo e likely o choose ools ha align wi h hei cogni i e
p e e ences. And in Hu & Hui (2012) s udy i is examined he e ec s o echnology-media ed
lea ning, inding ha s uden s’ a e mo e likely o engage wi h ools ha a e in e ac i e and
p o ide immedia e eedback. Basically, s uden s’ choices o lea ning ools a e in luenced by a
combina ion o pe sonal ac o s, ask- ela ed ac o s, and en i onmen al ac o s.
2.2. INDIVIDUAL LEARNING STYLES
As p e iously discussed, indi iduals ha e di e en lea ning s yles and s a egies ha wo k bes
o hem. To elabo a e, we will look in o he di e en ypes o lea ning s yles. S a ing o by
he olde lea ning s yle he eading me hod, which is whe e s uden s’ ead a icles, esea ch
pape s, ex books, e c. ha makes hem in e nalize he in o ma ion ha hey wan o lea n,
ano he common ype is he audi o y which is when s uden s’ p e e o lea n by lec u es,
eco ded audio, ideos, e c. (Moussa, 2014).
Subsequen ly, Shabi alyani e al. (2015), men ions ha “1% o wha is lea ned is om he
sense o TASTE, 1.5% o wha is lea ned is om he sense o TOUCH, 3.5% o wha is lea ned is
om he logic o SMELL, 11% o wha is educa ed is om he logic o HEARING and 83% o
wha is lea ned is om he sense o SIGHT” (p. 1).
Thus, mos common a e he isual aids, hese a e s uden s’ ha p e e isual aids such as,
cha s, diag ams, lashca ds, e c. because hese can illus a e he concep s in a clea e way
and make lea ning mo e eal and mo e accu a e. Since isual aids in lea ning end o be he
mos e ec i e me hod, hey a e used in class ooms and o he lea ning en i onmen s. This
same s udy e eals ha isual aids inc ease mo i a ion, cla i ica ion, inc ease he ocabula y,
sa es ime, a oids dullness, and p omo es a di ec expe ience.
9
3. MODEL BUILDING AND HYPOTHESIS DEVELOPMENT
The main goal o his esea ch is o in es iga e he impac o Cha GPT on s uden s’
oppo unis ic s a egies and check i hey only use i exclusi ely o comple e hei educa ional
asks o whe he hey wo k mo e o ge he bes possible pe o mance. Since AI ools, mo e
speci ically Cha GPT con inue o e ol e and a ec a ious aspec s o educa ion,
unde s anding i s po en ial in luence on s uden lea ning ou comes and c i ical hinking skills
is c ucial. Thus, we will be ocusing on he de elopmen o a concep ual model and i s
hypo hesis. In which i will be able o p o ide an unde s anding o he ela ionships be ween
Cha GPT usage, s uden s’ indi idual cha ac e is ics, and po en ial ou comes.
The model inco po a es heo e ical cons uc s inspi ed by p e ious s udies Goodhue &
Thompson (1995), Hoehle & Hu (2012) and Shaw & G ibbins (2005) including sel -e icacy,
e o expec ancy, ask- echnology i , and habi . Along wi h he cons uc s, di e en
hypo heses we e also de eloped based on he concep ual model, and will guide he
implemen a ion o da a collec ion ins umen s and s a is ical analyses. The model was c ea ed
based on he ask- echnology i (TTF) heo y, which based on p e ious s udies aligns bes wi h
he goals and esea ch p oblems ha his s udy in ends o add ess.
In Goodhue & Thompson (1995) s udy on ask- echnology i es ablished a heo e ical
amewo k ha indica es he in e ac ion be ween he alignmen o in o ma ion sys ems wi h
speci ic asks and he subsequen pe o mance ou comes. Wi h his, Shaw & G ibbins (2005)
sum up in hei s udy ha “(1) ask- echnology i is a ele an concep o p edic in o ma ion
sys ems success (e.g., pe o mance impac s), and ha (2) i is de e mined by an app op ia e
in e play be ween asks, echnology, and indi idual, con ex - ela ed cha ac e is ics” (p. 4).
In his s udy, ask- echnology i is de ined as he ma ch be ween he demands o he asks
assigned o s uden s’, he quali y o AI echnologies a ailable like Cha GPT, and he indi idual
cha ac e is ics in which s uden s’ use hese echnologies. The cha ac e is ics o he asks a e
cha ac e ized by hei na u e. The echnology, ep esen ed by Cha GPT, is de ined by i s
unc ionali ies, which include communica ion, in o ma ion access, and da a p ocessing
capabili ies, and i s adap abili y o s uden s’ con ex s. The indi idual use con ex is
cha ac e ized by hei sel -e icacy, habi s and e o expec ancy. The concep ual model, as
illus a ed in he igu e 1, p oposes an ideal ma ch among hese h ee dimensions wi h hem
being ask cha ac e is ics, echnology cha ac e is ics, and indi idual cha ac e is ics (Shaw &
G ibbins, 2005) which will acili a e a obus ask- echnology i , leading o mo e e ec i e use
o Cha GPT in educa ional se ings.
10
Figu e 1 – Concep ual model
In o de o c ea e a base o his esea ch, i is impo an o in es iga e he aspec s o e e y
a iable in he model and he possible hypo heses ha a e associa ed wi h hem. Fi s ly, by
app oaching each a iable which a e displayed by hei meaning on able 3:
Table 3 – Meaning o each a iable p esen ed on he concep ual model
Va iables
Meaning
Task Cha ac e is ics
The ask cha ac e is ics encompass he
a ibu es o he speci ic ask o goal
indi iduals a e ying o achie e wi h
Cha GPT. I conside s ac o s like ask
complexi y, ambigui y and equi ed
knowledge. Unde s anding hese
cha ac e is ics helps o assess how well he
Cha GPT’s unc ionali ies align wi h he
demands o he ask, ul ima ely in luencing
i s e ec i eness and s uden accep ance.
Technology Cha ac e is ics
This a iable examines he ea u es and
unc ionali ies o Cha GPT i sel . I
app oaches aspec s like ease o use,
accessibili y, eliabili y, lexibili y, and he
a ailable in o ma ion p o ided. Analyzing
hese cha ac e is ics helps o de e mine
how well he ool ma ches he speci ic
needs and capabili ies o he s uden s’,
p omo ing po en ial success o Cha GPT
in eg a ion and ou comes.
11
Sel -e icacy
This a iable acknowledges indi iduals’
con idence in hei abili y o success ully
use Cha GPT and achie e hei expec ed
ou comes. High sel -e icacy can lead o
inc eased e o , de e mina ion, and
posi i e expec a ions, leading o a g ea e
use o Cha GPT.
E o expec ancy
This a iable ocuses on he pe cei ed ease
o di icul y o using Cha GPT o achie e he
expec ed esul s. Lowe e o expec ancy
indica es ha he ool is complex o ime-
consuming, which can discou age i s use
and educe i s po en ial bene i s.
Habi
This a iable examines he equency o
using Cha GPT, o en in luenced by ini ial
expe iences and pe cei ed alue. Recu en
use can lead o inc eased s uden com o
and eliance on he ool, po en ially
p omo ing i s impac on pe o mance.
Task- echnology i
This a iable combines many o he ac o s
app oached abo e, measu ing he deg ee
o which Cha GPT cha ac e is ics align wi h
he demands o he speci ic ask. High ask-
echnology i sugges s a s ong ma ch
be ween ool capabili ies and s uden
needs, which leads o imp o ed
pe o mance and sa is ac ion.
Use o an AI ool
This a iable measu es he equency o
s uden engagemen wi h Cha GPT,
p o iding insigh s in o he in eg a ion and
po en ial impac o his ool.
E o oppo unism
This a iable e alua es he le el o which
s uden s’ do hei academic asks wi h he
aid o jus Cha GPT i sel , u he p o iding
insigh s on s uden s’ o e eliance on he
ool o e hei own unde s anding.
12
Addi ional e o
This a iable e e s o he ex en s uden s’
a e willing o imp o e hei academic asks,
using hei class ma e ials and own
unde s anding, o e he use o Cha GPT.
Pe o mance impac
This a iable e alua es he e ec o using
he AI ool on indi idual pe o mance
measu es, such as ask comple ion ime,
accu acy, o o e all p oduc i i y. Looking a
his ela ionship helps o e alua e he
e ec i eness o Cha GPT.
Mo eo e , i is impo an o unde s and wha ype o ela ionships exis be ween he
a iables. To be e unde s and hem he ollowing hypo hesis ha e been c ea ed:
H1: The cha ac e is ics o a ask a e posi i ely associa ed wi h he i o he echnology.
Acco ding o Shaw & G ibbins (2005) s udy, a ask’s compa ibili y wi h a gi en echnology
needs ca e ul conside a ion o a ious key a ibu es, he complexi y o he ask, and he ime
cons ain in which i mus be execu ed. Basically, he e ec i eness o he echnology in
ques ion, Cha GPT, is signi ican ly enhanced when he e is a co esponden ma ch wi h he
ask cha ac e is ics. The e o e, a bigge ask- echnology i is achie ed when he AI’s
capabili ies a e adap ed o accommoda e he ask’s complica ions, demands, and ime
cons ain s. This alignmen is c i ical, as i di ec ly in luences he e iciency and e icacy wi h
which he ask can be comple ed, le e aging he ull po en ial o Cha GPT.
H2: The cha ac e is ics o he echnology a e posi i ely associa ed wi h he i o he
echnology o he ask.
The hypo hesis H2 app oaches he mul i ace ed concep o ask- echnology i wi hin he
p oposed model and acknowledges he posi i e associa ion be ween echnology
cha ac e is ics and he ask- echnology i . Acco ding o Hoehle & Hu (2012) s udy and he
di e en concep s o i . H2 akes a di ec app oach, hypo hesizing a simple posi i e
ela ionship be ween he ea u es o he echnology and he o e all i wi h he demands o a
speci ic ask. Thus, speci ic ea u es like adap abili y, in e ac i i y, and ease o use con ibu e
di ec ly o a be e " i " by aligning wi h he eques s o he ask. In simple e ms, he mo e
ea u es a echnology has ha align wi h he demands o he ask, he be e sui ed i is o
accomplishing ha ask.
H3: Indi idual cha ac e is ics, including sel -e icacy, e o expec ancy and habi a e posi i ely
associa ed wi h he ask- echnology i .

13
Fundamen ally, i s uden s’ ind he echnology’s ea u es, unc ionali ies, and o e all design
o be a o able and aligned wi h hei indi idual p e e ences and alues, hey a e mo e likely
o ake in a good " ask- echnology i " (D’Amb a & Wilson, 2004). This is an icipa ed o
in luence he e ec i eness o he lea ning p ocess. Following Al-Rahmi e al. (2023) s udy,
whe e he alignmen be ween echnology and pe sonal alues leads o echnology adop ion.
I is hypo hesized ha a posi i e pe cep ion o echnology cha ac e is ics will ansla e o a
be e i o he lea ning ask, po en ially leading o enhanced lea ning ou comes. This
posi i e impac could esul om he sel -e icacy, e o expec ancy and habi s aken om
using Cha GPT which is ela ed o he s uden s’ indi idual cha ac e is ics and p e e ences.
H4: A highe ask- echnology i is posi i ely associa ed wi h he usage o he echnology,
including he equency o use o an AI ool, e o oppo unism and addi ional e o belie s.
Hypo hesis H4 p oposes a posi i e associa ion be ween a highe ask- echnology i and he
inc eased use o an AI ool. In Howa d & Rose (2019) s udy says ha indi iduals a e mo e likely
o adop and equen ly use echnologies ha hey pe cei e as aluable and well-sui ed o
hei needs. This hypo hesis sugges s ha when an AI ool aligns wi h he demands and
cha ac e is ics o a speci ic ask, s uden s’ a e mo e likely o ind i use ul and e ec i e.
Consequen ly, hey a e mo e likely o in eg a e i in o hei wo k low and engage wi h i
equen ly. Thus, H4 emphasizes he impo ance o ma ching he capabili ies o an AI ool o
he speci ic needs o he ask a hand.
This i can also in luence s uden belie s abou e o . Since hese ools, Cha GPT, ha e
ea u es ha make asks easie and as e migh lead s uden s’ o belie e hey equi e less
e o o comple e he ask. On he o he hand, hese ools can also p omo e c i ical hinking,
which equi es he s uden s’ no es and class ma e ials o belie e ha hey need o make an
addi ional e o o lea n mo e e ec i ely and ge be e esul s on hei academic asks. Thus,
his i can in luence s uden s’ belie s abou he e o equi ed. Tools ha make asks easie
migh p omo e belie s abou educed e o (e o oppo unism), while some s uden s’ migh
belie e ha addi ional e o will lead hem o ha e a be e pe o mance. Basically, his " i "
ac s as a key d i e o s uden engagemen and in luences he adop ion and use o he
echnology.
H5: A highe ask- echnology i is posi i ely associa ed wi h be e ou comes in e ms o
pe o mance impac .
H5 app oaches he po en ial consequences o a s ong ask- echnology i . Taking up he
unde s anding ha his i comes om he me ge be ween ask cha ac e is ics, echnology
cha ac e is ics, and indi idual cha ac e is ics (Al-Rahmi e al., 2023), i is hypo hesized o ha e
a posi i e associa ion wi h pe o mance impac . Basically, when s uden s’ pe cei e he
echnology as well in eg a ed wi h he lea ning ask and aligned wi h hei needs and
p e e ences, i likely leads o be e lea ning ou comes and inc eased pe o mance on hei
assignmen s.
14
H6: The usage o an AI ool is posi i ely associa ed wi h be e ou comes in e ms o
pe o mance impac .
Hypo hesis H6 app oaches he po en ial bene i s o using an AI ool, p oposing a posi i e
associa ion be ween i s use and imp o ed ou comes in e ms o pe o mance impac . This
aligns wi h he esea ch su ounding echnology-enhanced lea ning and pe o mance,
sugges ing ha e ec i e in eg a ion o AI ools can posi i ely impac s uden s’ expe iences
and achie emen s, as pe o mance is impac ed by he ool’s abili y o display he in o ma ion
wi h he leas amoun o e o s possible (Aljukhada e al., 2014).
Fi s ly, by equen ly using AI ools i can lead o inc eased cogni i e abili ies and op imize
lea ning p ocesses, p o iding capabili ies like pe sonalized eedback, au oma ed da a analysis,
and adap i e con en deli e y. This can p omo e p oblem-sol ing skills, educe he amoun o
in o ma ion ou memo y can p ocess a imes, and ul ima ely lead o imp o ed pe o mance
on asks o assessmen s.
Secondly, AI-powe ed ools can aise o dec ease ou pe o mance depending on ou educed
e o beha io s o jus use he ool o pass on an academic ask and depending on he ex a
e o we a e willing o make o ge he bes g ade possible. Basically, he impac o AI ools
on pe o mance could be posi i e i i p omo es a deepe lea ning, bu nega i e i i
encou ages supe icial use.
15
4. METHODOLOGICAL APPROACH
To assess he s uden s’ willingness o engage in addi ional wo k and hei oppo unis ic
endencies, we c ea ed an elec onic ques ionnai e, he ques ions o hese ques ionnai e
cons uc s we e designed in acco dance wi h he Po uguese g ading sys em. This sys em
e alua es s uden asks on a scale om 0 o 20, whe e 20 ep esen s he highes possible
sco e and 15 is conside ed a "good" a e age g ade.
This knowledge can hen be used o in o m educa ional ins i u ions and guide he
de elopmen o s a egies o maximize he bene i s o Cha GPT while educing i s po en ial
nega i e e ec s. The esea ch design employed in his s udy p o ides a igo ous and e sa ile
app oach o in es iga ing he impac o Cha GPT on s uden s’ po en ial educed e o . The
model also enables igo ous a iable explo a ion, con olled expe imen a ion, and eplica ion
o eal-wo ld scena ios.
The p ima y ool used is a comp ehensi e elec onic ques ionnai e inco po a ing a mix o sel -
e lec i e ques ions. This ques ionnai e is ca e ully designed o e alua e wo dis inc ye
in e connec ed dimensions: s uden oppo unism and po en ial addi ional e o . These
ques ions a e s uc u ed o b ing ou esponses ha e lec he s uden s’ p ac ices, beha io s,
and sel -assessmen in ela ion o hei academic en i onmen .
The cons uc s o he model a e e alua ed in he ques ionnai e h ough a nume ical scale,
each s a emen wi hin he ques ionnai e is measu ed agains a se en-poin Like scale, ha
anges om 1 o 7, whe e 1 ep esen s "S ongly Disag ee" and 7 ep esen s "S ongly Ag ee",
his scaling mechanism helps s uden s’ o p ecisely quan i y hei le el o ag eemen wi h each
s a emen , ensu ing an accu a e e lec ion o hei pe spec i es, as de ailed in able 4.
Table 4 – Adap ed ques ions o each cons uc
Cons uc
Sou ce
Adap ed Ques ions
Task
cha ac e is ics
(Sa ke &
Valacich,
2010)
1) Ideas o assignmen s o imp o e con en .
2) Re iew key concep s o quiz p epa a ion.
3) Summa ize PDFs o as e unde s anding.
16
Technology
cha ac e is ics
(Yang e al.,
2013)
1) Using Cha GPT aligns well wi h my lea ning s yle.
2) Cha GPT p o ides help ul eedback.
3) Cha GPT is a highly adap able echnology.
Sel -e icacy
(Sun e al.,
2016)
1) I can use Cha GPT o p epa e o a ask independen ly.
2) I can use Cha GPT wi hou elying on guidance om
o he s.
3) I can use Cha GPT e ec i ely e en wi hou p io
expe ience.
4) I can use Cha GPT o a ask wi h jus he online help
a ailable.
E o
expec ancy
(Pa hasa a hy
&
Bha ache jee,
1998; Wixom
& Todd, 2005)
1) I is easy o me o ge he in o ma ion I need om
Cha GPT.
2) I hink mas e ing Cha GPT o asks is simple.
3) I eel ha ope a ing Cha GPT e ec i ely is easy.
Habi
(Bha ache jee
& Lin, 2015;
Venka esh e
al., 2012)
1) Using Cha GPT has become a habi o me.
2) Using Cha GPT comes na u ally o me.
3) I au oma ically hink o using Cha GPT when I need o
comple e a ask.
Task-
echnology i
(Sun e al.,
2016)
1) Cha GPT’s unc ionali ies a e compa ible wi h my
asks.
2) Cha GPT makes ask p epa a ion easie .
3) Cha GPT acili a es he sequencing o asks.
Use o an AI
ool
(Kosi anu i e
al., 2006)
1) How equen ly do I use Cha GPT o academic asks.
E o
oppo unism
(Chen e al.,
2022)
1) I will accep o ha e 15 and no do he manual wo k.
2) I p e e he 15 and dedica e my ime o some hing
else.
3) I is okay o ge 15 i only using Cha GPT.
23
ela ionships a e uly meaning ul (s a is ically signi ican ). Addi ionally, we e alua ed he
p esence o mul icollinea i y, which can dis o esul s. Since he a iance in la ion ac o (VIF)
indica ed no such issue, i con i ms he independence o his s udy’s cons uc s.
Ou indings sugges ha he model a ia ion in s uden pe o mance is 79.3%. This is because
he cons uc "Pe o mance impac " exhibi s a s ong ela ionship and is s a is ically
signi ican (p- alue < 0.05) wi h ask- echnology i (β= 0.483, p- alue= 0.000) and use o an AI
ool (β= 0.455, p- alue= 0.000). Mo eo e , he da a e eals ha he cons uc ask- echnology
i has a posi i e ela ionship and is signi ican ly in luenced by echnology cha ac e is ics (β=
0.261, p- alue= 0.000), and by he s uden s’ indi idual cha ac e is ics being hese sel -e icacy
(β= 0.139, p- alue= 0.025), e o expec ancy (β= 0.242, p- alue= 0.000) and habi (β= 0.274,
p- alue= 0.000), suppo ing he hypo heses H2, H3a, H3b, H3c, H4a, H4b, H5, H6a and H6c.
On he o he hand, i was no possible o demons a e ha pe o mance impac was di ec ly
impac ed by e o oppo unism (β= 0.048, p- alue= 0.111) and ha ask- echnology i was
impac ed wi h addi ional e o (β= 0.124, p- alue= 0.261). In he same way, due o insu icien
s a is ical e idence ask- echnology i was no able o demons a e i s impac on ask
cha ac e is ics (β= 0.115, p- alue= 0.086).

24
6. DISCUSSION
6.1. THEORETICAL IMPLICATIONS
Ha ing es ablished he esul s and indings om he ques ionnai es in he p e ious chap e ,
he discussion chap e analyzes he meaning and implica ions o he esul s da a. Wi h his,
his s udy’s goal is o e alua e he in luence o Cha GPT, a well-known and accessible AI ool
hese days, in s uden s’ wi hin an educa ional se ing, using an adap ed TTF model (Goodhue
& Thompson, 1995). As he indings suppo mos o he p e ious esea ch on ela ed
cons uc s, we can conclude ha 9 ou o he 12 hypo heses a e empi ically suppo ed.
Hypo hesis 1 (H1) is suppo ed by he ela ionship be ween ask cha ac e is ics and ask-
echnology i (β= 0.115, p- alue= 0.086) i doesn’ sugges a s a is ically signi ican impac .
This ela ionship indica es ha he inhe en cha ac e is ics o he asks migh play a ole in
de e mining how well he echnology i s, bu o he ac o s could also be in luencing his i .
These indings imply ha ask cha ac e is ics likely play a ole in ask- echnology i , meaning
he cause o using Cha GPT o a ask depends mo e on o he ac o s like echnology
cha ac e is ics and indi idual cha ac e is ics han he na u e o an educa ional ask.
Looking in o Hypo hesis 2 (H2), he ela ionship be ween he echnology cha ac e is ics and
he ask- echnology i is signi ican and posi i e (β= 0.261, p- alue = 0.000), suppo ing he
ask- echnology i model (D’Amb a & Wilson, 2004; Goodhue & Thompson, 1995; Howa d &
Rose, 2019), meaning ha he e ec i eness o a echnology, in his case Cha GPT in an
educa ional se ing is di ec ly ela ed o how well i mee s he speci ic needs o he ask ha
is in ended o suppo , so he mo e he cha ac e is ics o he echnology align wi h he needs
o he ask, he be e he echnology is in e ms o usabili y and unc ionali y, meaning a
be e i o e all.
As o he indi idual cha ac e is ics hypo heses o his model (H3a, H3b, H3c), which a e sel -
e icacy (β= 0.139, p- alue = 0.025), e o expec ancy (β= 0.242, p- alue = 0.000) and habi
(β= 0.274, p- alue = 0.000), hey a e all s a is ically signi ican and ha e a s ong ela ionship
wi h ask- echnology i meaning ha he mo e sel -e icacy a s uden has, he mo e
con idence in hei abili y o use Cha GPT hey ha e, which signi ican ly imp o es hei
pe cep ion o he Cha GPT i o hei asks. This aligns wi h Bandu a & Adams (1977) s udy
on sel -e icacy, which sugges s ha indi iduals who belie e in hei abili ies a e mo e likely
o emb ace and e ec i ely use new echnologies. I indica es ha s uden s’ wi h high sel -
e icacy will pe cei e a be e i be ween hei needs and Cha GPT, he eby inc easing he
likelihood o adop ing and e ec i ely using i , p o ing H3a. E o expec ancy also has a s ong
posi i e ela ionship o ask- echnology i (H3b), indica ing ha when s uden s’ pe cei e he
use o Cha GPT as easy and equi ing minimal e o , hey pe cei e a be e alignmen wi h
hei academic asks, his aligns wi h Mi chell & Nebeke (1973) s udy. This inding means ha
ein o cing he pe cei ed ease o use plays an impo an ole in echnology accep ance. I
25
implies ha making Cha GPT mo e use - iendly and accessible will enhance i s i o
educa ional asks, he eby inc easing i s adop ion and usage among s uden s’. Las ly, he
signi ican posi i e ela ionship be ween habi and ask- echnology i demons a es ha
s uden s’ who ha e es ablished habi s o using Cha GPT o o he simila echnologies will be
mo e willing o pe cei e a good i be ween Cha GPT and hei asks. Hypo hesis (H3c) aligns
wi h Fio ella (2020) s udy, which emphasizes ha habi ual beha io is a s ong p edic o o
u u e usage. The e o e, s uden s’ accus omed o using AI ools in hei s udies will ind i
easie o in eg a e Cha GPT in o hei wo k low, pe cei ing i as a sui able ool o hei asks.
In he model, he hypo hesis ha ela es ask- echnology i and use o an AI ool (H4a) is
signi ican , indica ing a s ong ela ionship (β = 0.686, p- alue = 0.000), meaning ha he mo e
equen ly and e ec i ely an AI ool is used, he be e he i be ween he echnology and
he s uden s’ asks (Goodhue & Thompson, 1995; Howa d & Rose, 2019), his ela ionship
implies ha he ac ual use o he AI ool ein o ces he pe cep ion ha he echnology is well
sui ed o he asks i suppo s, his p o es he hypo hesis and concludes ha p ac ical
expe ience and equen use help s uden s’ unde s and and op imize he ool o hei speci ic
needs. H4b, shows he ela ionship be ween ask- echnology i and e o oppo unism (β =
0.345, p- alue = 0.000) which is s a is ically signi ican , e o oppo unism, e e s o exploi ing
Cha GPT whe e less e o is needed o achie e a desi ed ou come, his has signi ican impac
on he pe cei ed i o he echnology wi h he s uden s’ asks, sugges ing ha he i be ween
he echnology and asks elies on oppo unis ic a emp s o minimize e o . H4c, illus a es
he ela ionship be ween ask- echnology i and addi ional e o (β = 0.124, p- alue = 0.261)
which is no s a is ically signi ican , indica ing ha addi ional e o , o he ex a ime in es ed
beyond he ypical o comple e an educa ional ask, does no signi ican ly in luence how well
he echnology aligns wi h he ask equi emen s. This sugges s ha ask- echnology i is
mo e dependen on inhe en echnology- ask alignmen han on he ex a e o s uden s’ pu
in o making he AI ool wo k o hei needs.
As o he hypo hesis ha lead o his s udy’s ou come, pe o mance impac , which illus a es
he o e all p oduc i i y o s uden s’, he ela ionship be ween his a iable and ask-
echnology i (H5), use o an AI ool (H6a) and addi ional e o (H6c) a e s a is ically
signi ican , (β = 0.483, p- alue = 0.000), (β = 0.455, p- alue = 0.000) and (β = 0.082, p- alue =
0.035), espec i ely. This means ha when he echnology is well ma ched o he asks i
suppo s, he e is a posi i e e ec on pe o mance ou comes, ensu ing ha when he
echnology aligns wi h he equi emen s o he asks leads o a be e i which imp o es
pe o mance and also he mo e an AI ool is used, he g ea e he posi i e impac on
pe o mance (Aljukhada e al., 2014; Goodhue & Thompson, 1995), sugges ing ha s uden s’
who equen ly and e ec i ely use AI ools expe ience imp o emen s in pe o mance, due o
he AI ool’s abili y o enhance p oduc i i y. Despi e addi ional e o ha ing a weake e ec
i is s ill s a is ically signi ican , sugges ing ha s uden s’ end o do a minimal e o in
26
achie ing highe g ades han he ones Cha GPT can gi e hem e en hough he ex a e o
has a sligh posi i e impac on hei asks pe o mance.
On he o he hand, H6b, is he hypo hesis be ween pe o mance impac and e o
oppo unism (β = 0.048, p- alue = 0.111), leads us o conclude despi e no being s a is ically
signi ican , ha minimizing he e o o doing a ask h ough oppo unis ic s a egies does
no di ec ly impac pe o mance in a meaning ul way, indica ing ha inding ways o minimize
e o does no necessa ily ansla e in o signi ican pe o mance imp o emen s.
Finally, he i s esea ch ques ion (RQ1): "How does s uden s’ oppo unis ic use o Cha GPT
a ec hei pe o mance on academic asks?" is di ec ly add essed by he signi ican posi i e
ela ionship be ween ask- echnology i and e o oppo unism which illus a es ha
s uden s’ by aking ad an age o Cha GPT whe e less e o is needed signi ican ly pe cei e a
be e i o he echnology wi h hei asks. This oppo unism, whe e s uden s’ use Cha GPT
o minimize e o , signi ican ly in luences how hey engage wi h hei academic asks,
sugges ing ha while i may enhance ask comple ion e iciency, i does no necessa ily
ansla e o imp o emen s in deepe lea ning ou comes, as indica ed by he non-signi ican
impac on o e all pe o mance (H6b). This implies ha inding sho cu s h ough echnology
does no e lec on be e pe o mance, aising ques ions abou he dep h o lea ning and
unde s anding.
As o he second esea ch ques ion (RQ2): "How does he addi ional e o pe o med by
s uden s’ o imp o e hei g ades in luence hei pe o mance on academic asks when using
Cha GPT?". The non-signi ican ela ionship be ween ask- echnology i and addi ional e o
indica es ha he ex a ime s uden s’ in es beyond he ypical o comple e an educa ional
ask does no signi ican ly in luence how well he echnology aligns wi h ask equi emen s.
Howe e , despi e i s weak e ec , H6c shows ha e en wi h a small inc ease in e o ,
pe o mance is s ill posi i ely impac ed s a is ically. This implies ha mo e wo k imp o es
pe o mance wi hou signi ican ly changing he pe cei ed i o echnology, emphasizing he
impo ance o cau iousness on o e elying oo much on AI ools.
O e all his s udy can be jus i ied wi h he ollowing heo ies, he heo y o planned beha io
and he cogni i e heo y, which oge he p o ide a comp ehensi e amewo k o
unde s anding he dynamics o using educa ional ools like Cha GPT.
The heo y o planned beha io helps explain he beha io al in en ions behind using Cha GPT.
I sugges s ha s uden s’ decisions o use Cha GPT a e in luenced by hei beha io al belie s
(pe cep ions o he ou comes o using Cha GPT), no ma i e belie s (social p essu es and
expec a ions ega ding he use o Cha GPT), and con ol belie s (pe cei ed ease o di icul y
o using Cha GPT). Collec i ely, hese belie s shape s uden s’ in en ions o adop beha io s
hey pe cei e as bene icial o enhancing hei lea ning expe iences wi h Cha GPT (Bosnjak e
al., 2020).
27
The cogni i e load heo y explains how educa ional ools like Cha GPT can op imize cogni i e
p ocessing. This heo y is conce ned wi h he managemen o wo king memo y capaci y
du ing lea ning asks. I sugges s ha e ec i e educa ional ools can help educe unnecessa y
cogni i e load, he eby eeing up cogni i e esou ces o highe -o de hinking. Howe e , he
heo y also challenges he assump ion ha a high load is in a iably bene icial. I poin s ou
ha lea ning can occu e en wi hou addi ional demands on wo king memo y, indica ing ha
he in eg a ion and au oma ion o knowledge can happen e icien ly wi h well-designed
educa ional in e en ions (Schno z & Kü schne , 2007).
These heo ies collec i ely poin ou he complexi y o in eg a ing echnology like Cha GPT in
educa ional se ings, emphasizing he need o ools ha a e bo h easy o use and e ec i e in
educing cogni i e o e load, while s ill p omo ing obus lea ning and e en ion.
6.2. PRACTICAL IMPLICATIONS
The indings sugges ha while addi ional e o can be bene icial, s uden s’ gene ally do no
end o in es in i . Ins ead, hey o en ake a mo e oppo unis ic app oach, elying on AI ools
like Cha GPT o comple e hei asks o achie e a sa is ac o y g ade a he han a emp o
he bes possible ou come, he lack o signi ican impac o addi ional e o indica es ha
s uden s’ ely on wha Cha GPT p o ides, and hey’ e ypically con en wi h hose esul s, no
showing a s ong endency o do ex a wo k o maximize hei g ades, encou aging s uden s’
o engage deeply wi h hei cou sewo k beyond using AI ools o quick answe s can be a good
solu ion o his, since his can be done by designing assignmen s ha equi e c i ical hinking
and pe sonal inpu , pushing s uden s’ o apply hei unde s anding mo e ully.
The esea ch indings also o e se e al p ac ical implica ions o educa ional se ings o
in eg a e AI ools. Task cha ac e is ics play a signi ican ole in echnology i , bu i ’s equally
impo an o conside he echnology’s ea u es and indi idual s uden cha ac e is ics,
educa ional se ings should p io i ize AI ools ha align wi h educa ional asks, emphasizing
usabili y, unc ionali y, and ea u es ha di ec ly add ess s uden needs, his ocus will ensu e
a good i and maximize he AI ools po en ial.
De eloping p og ams o enhance s uden sel -e icacy and con idence in using AI ools
signi ican ly imp o es pe cei ed ask- echnology i , ins i u ions mus p o ide use - iendly
in e aces and clea ins uc ions o p omo e a posi i e s uden expe ience and encou age i s
adop ion, aining p og ams ha build s uden s’ con idence h ough wo kshops, hands on
ac i i ies, and men o ships will likely imp o e AI ools adop ion, e ec i eness and
p oduc i i y.
Regula use o AI ools helps s uden s’ de elop habi s and build amilia i y, leading o a be e
pe cei ed i . S uden s’ wi h expe ience using simila echnologies will ind in eg a ing ools
28
like Cha GPT easie , le e aging exis ing s uden expe ise will acili a e a smoo he ansi ion,
allowing s uden s’ o pe sonalize hei use o he ool and op imize i o speci ic asks,
ul ima ely leading o mo e e ec i e lea ning.
Aligning echnology wi h asks and encou aging equen , e ec i e AI ool use posi i ely
impac s s uden pe o mance, he emphasis should be on empowe ing s uden s’ o use AI
ools s a egically o enhance hei pe o mance and lea ning ou comes.

29
7. CONCLUSIONS
This disse a ion aimed o s udy he impac o mode n AI ools, pa icula ly Cha GPT, in
educa ion. As hese echnologies a e apidly e ol ing, i ’s impo an o unde s and hei
in luence. The indings o his s udy poin o he impo an ole o ask- echnology i in
s uden s educa ional pe o mance and echnology adop ion.
The s udy e eals ha while s uden s’ ask cha ac e is ics modes ly a ec ask- echnology i ,
and a iables such as echnology cha ac e is ics, sel -e icacy, e o expec ancy and habi s,
ha e a g ea e in luence. These a iables con ibu e o a pe cep ion o a s ong alignmen
be ween Cha GPT and educa ional asks, acili a ing ease o use and habi ual in eg a ion.
Signi ican ly, he equen use o AI ools like Cha GPT op imizes hei e icacy o s uden s
speci ic needs, al hough addi ional e o does no signi ican ly enhance ask- echnology i ,
emphasizing a endency owa ds oppo unis ic beha io a he han s i ing o highe
academic achie emen s.
This sugges s a endency o s uden s o p io i ize minimal e o s a egies using Cha GPT o
achie e sa is ac o y g ades, e en hough deepe engagemen could lead o be e lea ning
ou comes. Educa ional app oaches ha encou age c i ical hinking and pe sonal engagemen
wi h cou se ma e ials, beyond simply elying on AI o quick answe s, can be a aluable
solu ion o p omo e deepe lea ning and mo i a e s uden s o s i e o excellence.
Thus, his esea ch o e s impo an insigh s in o how AI ools like Cha GPT impac educa ion,
and impo an insigh s o he need o p omo e e ec i e echnology usage me hods ha
enhance s uden s’ pe o mance, maximize hei lea ning ou comes and incen i e he wan o
an in-dep h knowledge abso p ion.
30
8. LIMITATIONS AND FUTURE WORK
This s udy acknowledges se e al limi a ions. Fi s ly, one o he limi a ions his s udy had was
he sho numbe o samples and ime we had o collec hem. This da a was collec ed om
a ious uni e si ies, deg ees and le els o s udy o s uden s’ abo e 18, i would be in e es ing
i his s udy’s indings we e applied o example in s uden s’ wi h di e en na ionali ies,
cul u es and adi ions.
Secondly, his ques ionnai e was c ea ed o be answe ed anonymously and in an hones way
bu his me hod can also in oduce bias, as s uden s’ migh no accu a ely e lec hei use
and pe cep ion o AI ools. So u he wo k o imp o e his would be by employing mixed-
me hod app oaches, like in e iews o di ec obse a ion, his way mo e deepe insigh s could
be p o ided in o how s uden s’ ac ually use AI ools like Cha GPT beyond hei ques ionnai es
answe s.
Fu he mo e, ocusing only on Cha GPT migh limi he unde s anding o how he es o he
AI ools ha aid s uden s’ in hei educa ional asks, impac s uden s’ pe o mance wi hin
hei oppo unis ic beha io s and will o do ex a wo k o ge be e g ades. Thus, compa a i e
s udies on mul iple AI ools could gi e in e es ing indings no only on s uden s’ oppo unism
bu also e alua ing i hese ools a e in ac mo e e ec i e han p oblema ic in educa ional
se ings.
Las ly, his s udy may no deeply analyze unde lying psychological o beha io al easons o
s uden s’ o e eliance on AI ools. So o u u e wo k inco po a ing psychological models
could be bene icial o unde s anding wha d i es s uden s’ o use AI ools.
31
BIBLIOGRAPHICAL REFERENCES
Akkoyunlu, B., & Soylu, M. Y. (2008). In e na ional Fo um o Educa ional Technology & Socie y
A S udy o S uden ’s Pe cep ions in a Blended Lea ning En i onmen Based on Di e en
Lea ning S yles. Sou ce: Jou nal o Educa ional Technology & Socie y, 11(1), 183–193.
h ps://doi.o g/10.2307/jeduc echsoci.11.1.183
Aljukhada , M., Senecal, S., & Nan el, J. (2014). Is mo e always be e ? In es iga ing he ask-
echnology i heo y in an online use con ex . In o ma ion and Managemen , 51(4),
391–397. h ps://doi.o g/10.1016/j.im.2013.10.003
Alonso, F., López, G., Man ique, D., & Viñes, J. M. (2005). An ins uc ional model o web-
based e-lea ning educa ion wi h a blended lea ning p ocess app oach. B i ish Jou nal o
Educa ional Technology, 36(2), 217–235. h ps://doi.o g/10.1111/j.1467-
8535.2005.00454.x
Al-Rahmi, W. M., Al-Adwan, A. S., Al-Maa ouk, Q., O hman, M. S., Alsaud, A. R., Almog en, A.
S., & Al-Rahmi, A. M. (2023). In eg a ing Communica ion and Task–Technology Fi
Theo ies: The Adop ion o Digi al Media in Lea ning. Sus ainabili y (Swi ze land), 15(10).
h ps://doi.o g/10.3390/su15108144
Baa man, L. K. J., & De B uijn, E. (2011). In eg a ing knowledge, skills and a i udes:
Concep ualising lea ning p ocesses owa ds oca ional compe ence. In Educa ional
Resea ch Re iew (Vol. 6, Issue 2, pp. 125–134).
h ps://doi.o g/10.1016/j.edu e .2011.03.001
Baidoo-Anu, D., & Owusu Ansah, L. (2023). Educa ion in he E a o Gene a i e A i icial
In elligence (AI): Unde s anding he Po en ial Bene i s o Cha GPT in P omo ing Teaching
and Lea ning. In Jou nal o AI (Vol. 52, Issue 7). h ps://doi.o g/10.61969/jai.1337500
Bandu a, A., & Adams, N. E. (1977). Analysis o Sel -E icacy Theo y o Beha io al Change’. In
Cogni i e The apy and Resea ch (Vol. 1, Issue 4). h ps://doi.o g/10.1007/BF01663995
Basbe h, F., Manajemen, P., As ul, M., & Ib ahim, H. (2018). Fou Hou s Basic PLS-SEM A S ep
by S ep Guide Wi h Video Clips Fo S uden and Schola .
h ps://www. esea chga e.ne /publica ion/344954763
Bha ache jee, A., & Lin, C. P. (2015). A uni ied model o IT con inuance: Th ee complemen a y
pe spec i es and c osso e e ec s. In Eu opean Jou nal o In o ma ion Sys ems (Vol. 24,
Issue 4, pp. 364–373). Palg a e Macmillan L d. h ps://doi.o g/10.1057/ejis.2013.36
Bilad, M. R., Yaqin, L. N., & Zubaidah, S. (2023). Recen P og ess in he Use o A i icial
In elligence Tools in Educa ion. Ju nal Peneli ian Dan Pengkajian Ilmu Pendidikan: E-
Sain ika, 7(3), 279–315. h ps://doi.o g/10.36312/esain ika. 7i3.1377
32
Bosnjak, M., Ajzen, I., & Schmid , P. (2020). The heo y o planned beha io : Selec ed ecen
ad ances and applica ions. In Eu ope’s Jou nal o Psychology (Vol. 16, Issue 3, pp. 352–
356). PsychOpen. h ps://doi.o g/10.5964/ejop. 16i3.3107
B own, T. B., Mann, B., Ryde , N., Subbiah, M., Kaplan, J., Dha iwal, P., Neelakan an, A., Shyam,
P., Sas y, G., Askell, A., Aga wal, S., He be -Voss, A., K uege , G., Henighan, T., Child, R.,
Ramesh, A., Ziegle , D. M., Wu, J., Win e , C., … Amodei, D. (2020). Language Models a e
Few-Sho Lea ne s. h ps://commonc awl.o g/ he-da a/
Cas ono a, J. A. (2002). Disco e y Lea ning o he 21 s Cen u y: Wha is i and how does i
compa e o adi ional lea ning in e ec i eness in he 21 s Cen u y?
h ps://www.academia.edu/10871733/Disco e y_Lea ning_ o _ he_21s _Cen u y_Wh
a _is_i _and_how_does_i _compa e_ o_ adi ional_lea ning_in_e ec i eness_in_ he_
21s _Cen u y
Chen, J., Liu, Z., & Hu, R. (2022). Resea ch on he In luence o Tole ance o Oppo unis ic
Beha io s o Channel Bounda ies on he Choice o Response S a egies. F on ie s in
Psychology, 13. h ps://doi.o g/10.3389/ psyg.2022.823416
D’Amb a, J., & Wilson, C. S. (2004). Use o he wo ld wide web o in e na ional a el:
In eg a ing he cons uc o unce ain y in in o ma ion seeking and he Task-Technology
Fi (TTF) model. Jou nal o he Ame ican Socie y o In o ma ion Science and Technology,
55(8), 731–742. h ps://doi.o g/10.1002/asi.20017
D no šek, M., Wincen , J., & Ca don, M. S. (2010). En ep eneu ial sel -e icacy and business
s a -up: De eloping a mul i-dimensional de ini ion. In e na ional Jou nal o
En ep eneu ial Beha iou and Resea ch, 16(4), 329–348.
h ps://doi.o g/10.1108/13552551011054516
Dunn L. (2002). LEARNING AND TEACHING BRIEFING PAPERS SERIES Ox o d Cen e o S a
and Lea ning De elopmen .
www.b ookes.ac.uk/se ices/ocsld/ esou ces/b ie ing_pape s
Du a, S., & Chak abo y, T. (2023). Thus Spake Cha GPT. Communica ions o he ACM, 66(12),
16–19. h ps://doi.o g/10.1145/3616863
Dwi edi, Y. K., Kshe i, N., Hughes, L., Slade, E. L., Jeya aj, A., Ka , A. K., Baabdullah, A. M.,
Koohang, A., Ragha an, V., Ahuja, M., Albanna, H., Albash awi, M. A., Al-Busaidi, A. S.,
Balak ishnan, J., Ba le e, Y., Basu, S., Bose, I., B ooks, L., Buhalis, D., … W igh , R. (2023).
“So wha i Cha GPT w o e i ?” Mul idisciplina y pe spec i es on oppo uni ies,
challenges and implica ions o gene a i e con e sa ional AI o esea ch, p ac ice and
policy. In e na ional Jou nal o In o ma ion Managemen , 71.
h ps://doi.o g/10.1016/j.ijin omg .2023.102642