Con ac o:
Kamil Kopecký, [email p o ec ed], Palacký Uni e si y Olomouc, Facul y o Educa ion, Žižko o
náměs í 5, Olomouc, 77900, Czech Republic
Dominik Vo áč, dominik. o [email protected], Palacký Uni e si y Olomouc, Facul y o Educa ion, Žižko o
náměs í 5, Olomouc, 77900, Czech Republic
h p:// e is as.um.es/ ei op
Kopecký, K. & Vo áč, D. (2025). Posibilidades de uso de la in eligencia a i icial gene a i a
como apoyo a la educación desde la pe spec i a de los es udian es de p ima ia y secunda ia
en la República Checa. Re is a Elec ónica In e uni e si a ia de Fo mación del P o eso ado,
28(2), 145-159.
DOI: h ps://doi.o g/10.6018/ ei op.661521
Posibilidades de uso de la in eligencia a i icial gene a i a como
apoyo a la educación desde la pe spec i a de los es udian es de
p ima ia y secunda ia en la República Checa
Kamil Kopecký, Dominik Vo áč
Facul y o Educa ion, Palacký Uni e si y in Olomouc
Resumen
El a ículo se cen a en las posibilidades de uso de la in eligencia a i icial gene a i a
(especialmen e LLM) en la educación y la p epa ación en casa de los alumnos de p ima ia y
secunda ia en la República Checa, u ilizando los esul ados de la encues a Los alumnos
checos y la in eligencia a i icial (2024), en la que pa icipa on más de 28.000 encues ados de
oda la República Checa. En nues o a ículo, analizamos las di e encias en el uso de la IA
en e los alumnos de p ima ia y secunda ia, pa a qué ac i idades la u ilizan los alumnos y
cuáles son las o mas más a anzadas de u iliza es as he amien as (den o y ue a de la
escuela) pa a aumen a la e icacia de la educación; nos cen amos en la pe sonalización y la
gami icación (incluida la c eación de juegos educa i os).
Palab as cla e
In eligencia a i icial; g andes modelos lingüís icos; educación de alumnos de p ima ia y
secunda ia; gami icación.
Recepción: 14 de ma zo de 2025
Acep ación: 9 de mayo de 2025
Kamil Kopecký y Dominik Vo áč
146 Re is a Elec ónica In e uni e si a ia de Fo mación del P o eso ado (REIFOP)
Possibili ies o using gene a i e a i icial in elligence o suppo
educa ion om he pe spec i e o p ima y and seconda y
school s uden s in he Czech Republic
Abs ac
The pape ocuses on he possibili ies o using gene a i e a i icial in elligence (especially
LLM) in educa ion and home p epa a ion o p ima y and seconda y school pupils in he Czech
Republic, using he esul s o he su ey Czech Pupils and A i icial In elligence (2024), which
in ol ed o e 28,000 esponden s om all o e he Czech Republic. In ou pape , we look a
he di e ences in he use o AI among p ima y and seconda y school pupils, wha ac i i ies
pupils use AI o , and wha a e he mo e ad anced ways o using hese ools (in and ou o
school) o inc ease he e ec i eness o educa ion.
Key wo ds
A i icial in elligence; la ge language models; educa ion o p ima y and seconda y school
pupils; gami ica ion
In oduc ion
Gene a i e a i icial in elligence: an in oduc ion o he ield
A i icial in elligence, pa icula ly i s gene a i e o ms, is cu en ly expe iencing a massi e
boom, wi h de elope s eleasing new and inc easingly sophis ica ed e sions o hese
sys ems on an almos weekly basis. The e m gene a i e a i icial in elligence (GenAI) e e s
o a ca ego y o a i icial in elligence capable o au onomously gene a ing new con en , such
as ex , images, audio, and ideo (All o d e al., 2023; Lyu, 2023). In ecen yea s, GenAI has
expe ienced a signi ican su ge, pa icula ly in connec ion wi h la ge language models (LLMs)
(Con e as Kallens e al., 2023; Hough on e al., 2023; Kuma , 2023) and hei inc easing
accessibili y o he public. One o he mos well-known ep esen a i es o hese AI models is
GPT, which has been widely in eg a ed in o a ious eely a ailable ools and con inuously
ained o enhance i s abili y o in e ac wi h humans. The ue expansion o gene a i e AI
occu ed wi h he elease o Cha GPT 3.5 by OpenAI (Ayinde e al., 2023; Holland, 2023;
OpenAI, n.d.) which ma ked a e olu iona y shi and demons a ed o he public he
capabili ies and po en ial o hese ools. While Cha GPT is he mos widely ecognized
example, i is no he only ool le e aging gene a i e AI—compe ing sys ems include Copilo
(Mic oso ), Gemini (Google/Alphabe ), Claude (An h opic), G ok (X), among o he s.
GenAI e y o en uses LLM (Eu opean Commission, 2023; Mic oso , 2023) and is hen able o
answe que ies, edi and pa aph ase ex s, c ea e li e a y esea ch o p oduce o iginal ex s
acco ding o use equi emen s, such as epo s, essays, sho s o ies o scien i ic a icles. In
addi ion o ex gene a ion, i also o e s analy ical unc ions such as iden i ying main ideas,
co ec ing mis akes o sol ing p oblems such as ma hema ical p oblems. Due o i s
e sa ili y, gene a i e AI inds applica ions in many ields, including educa ion a all le els. O
Posibilidades de uso de la in eligencia a i icial gene a i a como apoyo a la educación desde la pe spec i a de los es udian es
de p ima ia y secunda ia en la República Checa
Re is a Elec ónica In e uni e si a ia de Fo mación del P o eso ado (REIFOP) 147
cou se, GenAI can also gene a e pho os o o he images, music, spoken wo d and o he ypes
o digi al con en - including, o example, ideos.
A new ea u e is deep esea ch (OpenAI, 2025), which is an ad anced, mul i-s ep esea ch
p ocess whe e AI ac i ely and au onomously collec s, analyses and syn hesises in o ma ion
om a ious sou ces on he in e ne . I is he e o e no jus a simple answe o a que y, bu
a comp ehensi e app oach ha in ol es speci ying he eques , sys ema ically sea ching
da abases and websi es, hen analysing and e alua ing he collec ed da a, and inally
p oducing a s uc u ed epo summa ising he key indings. This agen -based p ocess allows
o a apid, in-dep h iew o complex opics, which g ea ly s eamlines esea ch asks in bo h
academic and comme cial se ings. Deep esea ch hus ex ends he adi ional capabili ies o
gene a i e AI and akes i o he le el o a ool ha no only gene a es con en bu also
comp ehensi ely suppo s decision-making p ocesses h ough ho ough e i ica ion and
in eg a ion o in o ma ion.
Gene a i e a i icial in elligence in p ima y and seconda y educa ion
Gene a i e AI is apidly e ol ing, and while esea ch on i s applica ion in educa ion is s ill
eme ging, se e al s udies p o ide aluable insigh s in o i s po en ial and challenges. In he
U.S., Dilibe i e al. (2024) ound ha by la e 2023, 18% o K-12 eache s in eg a ed AI in o hei
cu iculum, p ima ily o adap ing con en and ma e ial c ea ion. Simila ly, in U uguay,
Jauhiainen & Gue a (2023) demons a ed how AI-pe sonalized educa ional con en
imp o ed s uden engagemen and lea ning ou comes. In ma hema ics educa ion, Bas ani
e al. (2024) obse ed signi ican imp o emen s in p oblem-sol ing skills when high school
s uden s used AI-guided lea ning ools, hough he e ec i eness depended on he AI’s le el
o in e ac i i y. Resea ch om B azil (Villan & San os, 2023) showed ha Cha GPT-suppo ed
p ojec -based lea ning inc eased s uden mo i a ion and in e disciplina y collabo a ion.
Howe e , challenges pe sis , as highligh ed by Els ad & E iksen (2024) in No way, whe e a
lack o clea AI policies and eache suppo hinde ed adop ion. In he Czech Republic,
Kopecký e al. (2024) ound ha while o e hal o s uden s use AI ools, only a qua e
engage wi h hem egula ly in school se ings. Addi ional esea ch by IPSOS (2025)
emphasized he need o AI-li e a e eache s, wi h 76% o su eyed s uden s acknowledging
he gap in AI in eg a ion. These s udies collec i ely sugges ha AI can enhance pe sonalized
lea ning, engagemen , and e iciency, bu i s e ec i e implemen a ion equi es s uc u ed
policies, eache aining, and e hical conside a ions.
Gene a i e AI is apidly e ol ing, and while esea ch on i s applica ion in educa ion is s ill
eme ging, se e al s udies p o ide aluable insigh s in o i s po en ial and challenges. In he
U.S., Dilibe i e al. (2024) ound ha by la e 2023, 18% o K-12 eache s in eg a ed AI in o hei
cu iculum, p ima ily o adap ing con en and ma e ial c ea ion. Simila ly, in U uguay,
Jauhiainen & Gue a (2023) demons a ed how AI-pe sonalized educa ional con en
imp o ed s uden engagemen and lea ning ou comes. In ma hema ics educa ion, Bas ani
e al. (2024) obse ed signi ican imp o emen s in p oblem-sol ing skills when high school
s uden s used AI-guided lea ning ools, hough he e ec i eness depended on he AI’s le el
o in e ac i i y. Resea ch om B azil (Villan & San os, 2023) showed ha Cha GPT-suppo ed
p ojec -based lea ning inc eased s uden mo i a ion and in e disciplina y collabo a ion.
Howe e , challenges pe sis , as highligh ed by Els ad & E iksen (2024) in No way, whe e a
lack o clea AI policies and eache suppo hinde ed adop ion. In he Czech Republic,
Kopecký e al. (2024) ound ha while o e hal o s uden s use AI ools, only a qua e
engage wi h hem egula ly in school se ings. Addi ional esea ch by IPSOS (2025)
emphasized he need o AI-li e a e eache s, wi h 76% o su eyed s uden s acknowledging
Kamil Kopecký y Dominik Vo áč
148 Re is a Elec ónica In e uni e si a ia de Fo mación del P o eso ado (REIFOP)
he gap in AI in eg a ion. These s udies collec i ely sugges ha AI can enhance pe sonalized
lea ning, engagemen , and e iciency, bu i s e ec i e implemen a ion equi es s uc u ed
policies, eache aining, and e hical conside a ions.
Gene a i e a i icial in elligence o e s eache s, s uden s, bu also, o example, pa en s, a
wide ange o possibili ies - eache s can use i bo h in he p epa a ion phase o lessons, as
well as du ing lessons (e oca ion, awa eness, e lec ion) and, o cou se, as a ool o suppo
s uden s' home p epa a ion. The e a e many possible uses. A he same ime, howe e , i is
always necessa y o obse e he e hical p inciples o AI use - AI should ac i ely suppo he
pupil, no eplace his/he ac i i y. (Ha y, 2023; Okaiye o e al., 2023; Osamo e al., 2023;
UNESCO, 2023).
In p ima y schools, gene a i e AI can be used o c ea e pe sonalised lea ning ma e ials and
in e ac i e asks ha mee he age and indi idual needs o s uden s. Wi h he abili y o
analyse da a on indi idual s uden p og ess, AI makes i easie o eache s o adap he
con en and pace o lessons, p omo ing be e unde s anding o he con en and inc easing
child en's mo i a ion. Such an app oach also helps de elop c ea i e hinking as s uden s a e
encou aged o expe imen wi h new ideas and concep s. Resea ch shows ha he use o
gene a i e ools, o example in homewo k, con ibu es o g ea e pupil engagemen and
mo e e ec i e lea ning. (Kopecký Kamil e al., 2024).
Ano he bene i o AI in p ima y schools is he au oma ion o eedback and assessmen o
s uden wo k (Spec o , 2023), allowing eache s o spend mo e ime in e ac ing di ec ly wi h
child en. Indeed, ools based on gene a i e AI can quickly analyse and co ec asks, helping
o iden i y weaknesses in pupils' knowledge and design exe cises ailo ed o hei needs. This
imp o es he quali y o eaching and educes he adminis a i e bu den on educa o s.
Ne e heless, i is c ucial o ensu e ha he esul s gene a ed by AI a e egula ly checked and
supplemen ed by human supe ision (Galindo-Domínguez e al., 2024). A he same ime, i
should be no ed ha he de aul se ing o AI ools is no en i ely neu al, e.g. Cha GPT o en
p o ides eedback on e.g. s uden wo k, bu his is o low quali y. The e o e, he AI needs o
be ins uc ed ia p omp s o be uly objec i e in i s e alua ion.
A i icial in elligence can also in e ac au onomously wi h he pupil, i.e. i can au oma ically
in en asks o pupils, pupils pe o m hem, he AI e alua es hem and mo i a es pupils
u he . In case o ailu e, i can hen guide hem o he co ec solu ion. This capabili y can
be applied, o example, in he de elopmen o simple (bu also mo e complex) educa ional
games o gami ied lea ning si ua ions (B andon, 2023; Tulsiani, 2024). The ange o
applica ions o gene a i e AI is indeed e y b oad. On he o he hand, howe e , he
assignmen needs o be hough h ough so ha i is uly unc ional and suppo s, no
eplaces, lea ne ac i i y. (Kopecký Kamil, 2024)
In seconda y schools, he possibili ies o using gene a i e AI a e expanding owa ds
suppo ing mo e complex asks and p ojec s. S uden s can use AI o w i e epo s, c ea e
mo e complex ex s o e en gene a e isual con en , allowing hem o be e unde s and
and apply heo e ical knowledge in p ac ice. Teache s can use hese ools o p epa e
eaching ma e ials, c ea e es s, and au oma e ou ine asks, eeing hei hands o guiding
discussions and suppo ing s uden s indi idually. This app oach p omo es he de elopmen
o c i ical hinking as s uden s a e encou aged o alida e and modi y AI-gene a ed con en .
On he o he hand, i mus be aken in o accoun ha AI can also be used by s uden s o sol e
gi en asks, wi h he s uden himsel ob aining he solu ion wi hou demons a ing his
knowledge and skills. In his case, he impac o AI on he pupil is al eady nega i e.
We mus no o ge how o use AI o suppo s uden s wi h special educa ional needs. Wi h
he abili y o analyze la ge olumes o da a and adap o indi idual s uden needs, AI enables
Posibilidades de uso de la in eligencia a i icial gene a i a como apoyo a la educación desde la pe spec i a de los es udian es
de p ima ia y secunda ia en la República Checa
Re is a Elec ónica In e uni e si a ia de Fo mación del P o eso ado (REIFOP) 149
he c ea ion o pe sonalized lea ning p og ams ha ake in o accoun he unique needs o
each s uden . Fo example, adap i e lea ning pla o ms can ack s uden p og ess in eal-
ime and au oma ically adjus he di icul y o assignmen s o p o ide speci ic ma e ials o
a eas in which a s uden needs imp o emen .
In addi ion, AI o e s assis i e echnologies such as ex - o-speech ools o speech
ecogni ion o help s uden s wi h lea ning disabili ies such as dyslexia be e unde s and
lea ning ma e ials. AI-powe ed i ual assis an s and cha bo s can p o ide ins an eedback
and suppo , inc easing he accessibili y o educa ional esou ces and allowing s uden s o
wo k a hei own pace. These echnologies no only imp o e he accessibili y o educa ion
bu also p omo e inclusi i y by ensu ing ha e e y s uden has equal oppo uni ies o lea n
and de elop.
Me hodology
The p ima y objec i e o his s udy is o explo e how p ima y and seconda y school s uden s
in he Czech Republic use gene a i e a i icial in elligence (GenAI), especially la ge language
models (LLMs), in bo h school- ela ed and ex acu icula con ex s. The esea ch aims o:
• Iden i y which GenAI ools a e mos commonly used by s uden s.
• Examine how s uden s’ age, gende , and ype o school in luence AI usage pa e ns.
• Analyze he con ex s in which eache s in e ac wi h GenAI.
Resea ch ins umen and da a collec ion p ocedu e
Da a we e collec ed using a cus om-designed ques ionnai e c ea ed in Google Fo ms, which
was dis ibu ed ia email o eache s a all p ima y and seconda y schools ac oss he Czech
Republic. Pa icipa ion in he esea ch was olun a y, wi h eache s deciding whe he o
in ol e hei s uden s in he s udy. Da a collec ion ook place be ween 2 Sep embe 2024
and 5 Decembe 2024. No pilo s udy was conduc ed p io o he ques ionnai e’s dis ibu ion.
The su ey was de eloped based on expe consul a ion and p e ious esea ch expe ience.
Ini ially, 28,745 esponses we e eco ded; howe e , ollowing da a cleaning - whe e
esponses om uni e si y s uden s and indi iduals unde he age o 10 we e excluded - a o al
o 27,336 alid esponses we e analyzed. The da a we e p ocessed using Julius AI alongside
o he analy ical ools, including Mic oso Excel.
Ques ionnai e s uc u e
The esea ch ins umen ollowed a mixed-me hods design, inco po a ing bo h quan i a i e
me hods (e.g., closed and scaled ques ions) and quali a i e me hods (e.g., open-ended
ques ions). This app oach acili a ed a nuanced and mul idimensional analysis o s uden s'
engagemen wi h gene a i e AI.
The ques ionnai e was o ganized in o se e al hema ic sec ions, each add essing speci ic
aspec s o AI use:
1. Demog aphics – The ini ial sec ion collec ed undamen al esponden in o ma ion,
including biological sex, age, egion o esidence, and school ype.
2. Gene a i e AI in Leisu e Time – This sec ion explo ed s uden s’ engagemen wi h
gene a i e AI ou side he school en i onmen . Pa icipan s we e asked o speci y
which AI ools hey used (e.g., Cha GPT, Midjou ney, DALL-E, S able Di usion, Adobe
Kamil Kopecký y Dominik Vo áč
150 Re is a Elec ónica In e uni e si a ia de Fo mación del P o eso ado (REIFOP)
Fi e ly, Mic oso Designe ) and o wha pu poses, such as comple ing homewo k,
c ea ing p esen a ions, w i ing academic pape s, o engaging in c ea i e ac i i ies.
3. AI in Educa ion – This sec ion examined he p esence and applica ion o AI in
educa ional se ings. Responden s indica ed whe he AI was discussed o u ilized in
school and iden i ied he p ima y use s ( eache s, s uden s, o bo h). Fu he mo e,
hey de ailed speci ic AI applica ions, such as gene a ing g aphics, p oducing ex , o
suppo ing school p ojec s.
4. AI and he Fu u e – Pa icipan s assessed he signi icance o AI p o iciency and
exp essed hei pe spec i es on whe he gene a i e AI could e en ually eplace
eache s. They we e also asked o jus i y hei opinions ega ding his po en ial shi .
Resea ch sample
The inal da ase (a e da a cleaning) comp ised esponden s aged 10 o 18 yea s ( emale =
51.79%, male = 48.21%). The age dis ibu ion was analyzed o assess i s con o mi y wi h a
no mal dis ibu ion. The mean age o esponden s was 14.18 yea s (SD = 1.99). Skewness
(0.05) and ku osis (-0.65) alues indica ed a ela i ely la dis ibu ion. Al hough bo h he
Shapi o-Wilk and Ande son-Da ling es s ejec ed he null hypo hesis o no mali y (p < 0.001),
isual inspec ion o his og ams sugges ed ha he dis ibu ion was su icien ly close o
no mal o p ac ical pu poses.
Figu e 1.
Age dis ibu ion wi h no mal cu e.
n = 27,336
Regionally, he highes p opo ion o esponden s came om he Cen al Bohemian Region
(14.61%), ollowed by he Mo a ian-Silesian Region (11.35%) and he Sou h Mo a ian Region
(11.31%). In con as , he Ka lo y Va y Region had he lowes ep esen a ion (0.82%).
Consequen ly, he da a dis ibu ion ac oss egions was no ably une en.
In e ms o educa ional le els, he majo i y o esponden s (60.0%) we e en olled in p ima y
and lowe seconda y educa ion (g ades 1–9). The emaining 40% we e dis ibu ed ac oss
Posibilidades de uso de la in eligencia a i icial gene a i a como apoyo a la educación desde la pe spec i a de los es udian es
de p ima ia y secunda ia en la República Checa
Re is a Elec ónica In e uni e si a ia de Fo mación del P o eso ado (REIFOP) 151
a ious ypes o seconda y educa ion. Seconda y Voca ional Schools wi h a g adua ion exam
(“Ma u i a”) accoun ed o 18.17% o esponden s, ollowed by Eigh -Yea G amma Schools
(8.70%) and Fou -Yea G amma Schools (7.38%). Seconda y Voca ional Schools awa ding
App en iceship Ce i ica es ep esen ed 3.62% o he sample, while Six-Yea G amma
Schools cons i u ed 1.81%. A small ac ion o esponden s (0.33%) epo ed a ending o he
ypes o educa ional ins i u ions.
Resul s
Iden i ica ion o he mos equen ly used ools
The ques ionnai e included a ques ion speci ically ocused on gene a i e a i icial in elligence
ools, allowing esponden s o choose om a p ede ined lis o six een widely used ools
(e.g., Cha GPT, Mic oso Copilo ) o o speci y addi ional ools o hei own. This
combina ion o a closed selec ion and an open-ended esponse enabled a mo e
comp ehensi e analysis o he AI ools mos equen ly u ilized by Czech s uden s.
Fo a mo e de ailed examina ion, we selec ed he six mos used ools, each o which was
epo ed by a leas 2% o esponden s: Cha GPT, Pho oma h, Mic oso Copilo , Google
Gemini, Mic oso Image C ea o , and Suno. These ools e lec he di e si y o AI
applica ions - beyond he h ee la ge language models, he selec ion includes a g aphic
design ool (Mic oso Image C ea o ), a specialized ool o sol ing ma hema ical p oblems
(Pho oma h), and an AI-powe ed music gene a ion ool (Suno).
Age- ela ed pa e ns in AI ool usage
To be e unde s and how s uden s engage wi h AI ools, we analyzed usage pa e ns ac oss
di e en age g oups, as illus a ed in Figu e 2.
Figu e 2.
GenAI ools usage by age
n = 27,336
Kamil Kopecký y Dominik Vo áč
152 Re is a Elec ónica In e uni e si a ia de Fo mación del P o eso ado (REIFOP)
Chi-squa e es s con i med s a is ically signi ican associa ions be ween age and he usage
pa e ns o all six AI ools (p < 0.05), indica ing ha adop ion a es a y ac oss di e en age
g oups.
O e all, AI ool usage gene ally inc eased wi h age, wi h he mos no able ise obse ed o
Cha GPT, which was used by 17.29% o 10-yea -olds and 71.27% o 18-yea -olds. Pho oma h
ollowed a simila end, peaking a 43.85% among 17-yea -olds. Mic oso Copilo , Mic oso
Image C ea o , and Suno also demons a ed highe adop ion among olde s uden s,
al hough wi h less p onounced di e ences. In con as , Google Gemini exhibi ed an in e se
end, showing he highes usage a e among 10-yea -olds (22.25%) and he lowes among 18-
yea -olds (11.20%).
Gende - ela ed pa e ns in AI ool usage
We also examined whe he he usage o selec ed AI ools di e ed based on gende , as
illus a ed in Figu e 3.
Figu e 3.
AI ools usage by gende .
n = 27,336
Boys and gi ls engage wi h AI ools di e en ly, as indica ed by ou analysis o usage pa e ns
and s a is ical compa isons. Fo each ool, usage pe cen ages we e calcula ed sepa a ely o
boys and gi ls and wo-sample - es s (assuming unequal a iances) we e employed o
de e mine whe he he obse ed di e ences we e s a is ically signi ican .
The analysis e ealed signi ican gende -dependen dispa i ies in he adop ion o ce ain
ools. Fo ins ance, Cha GPT displayed an app oxima ely 18% highe usage a e among boys
compa ed o gi ls, and he co esponding - es con i med ha his di e ence was
s a is ically signi ican (p < 0.05). Simila ly, Mic oso Copilo also demons a ed signi ican
di e ences a o ing boys, while Pho oma h exhibi ed a s a is ically signi ican highe usage
among gi ls.
Posibilidades de uso de la in eligencia a i icial gene a i a como apoyo a la educación desde la pe spec i a de los es udian es
de p ima ia y secunda ia en la República Checa
Re is a Elec ónica In e uni e si a ia de Fo mación del P o eso ado (REIFOP) 153
In con as , al hough ools such as Google Gemini, Mic oso Image C ea o , and Suno
p esen ed di e ences in u iliza ion be ween he wo gende g oups, hese di e ences did
no consis en ly each s a is ical signi icance. Impo an ly, an age-adjus ed analysis was also
conduc ed o de e mine whe he he age o use s migh con ound he obse ed usage
di e ences. Welch’s - es compa ing he mean ages o boys (M = 14.16, SD = 1.97) and gi ls
(M = 14.20, SD = 2.01) e ealed no s a is ically signi ican di e ence (p = 0.072). This inding
indica es ha age does no play a con ounding ole in he obse ed ends.
Type o school
Ano he a iable analyzed was he compa ison o di e en school ypes. To minimize he
impac o age on AI usage ac oss di e en ypes o schools, we selec ed only esponden s
aged 13 and olde o his analysis. We ocused on he ou main ypes o schools in he Czech
Republic.
1. P ima y Schools and Lowe Seconda y Schools (n = 10,986) – he s anda d compulso y
educa ion ins i u ions, whe e he pupils a e ypically aged 6 o 15.
2. G amma Schools (n = 4,591) – selec i e schools ha culmina e in a inal examina ion
(“Ma u i a”), which quali ies s uden s o uni e si y admission. Admission o hese
schools is compe i i e, wi h s uden s applying a di e en s ages: a e he 5 h g ade
(Eigh -Yea G amma School), a e he 7 h g ade (Six-Yea G amma School), o a e
he 9 h g ade (Fou -Yea G amma School).
3. Seconda y Voca ional Schools (n = 4,959) – hese ins i u ions p o ide specialized
seconda y educa ion and conclude wi h he “Ma u i a” exam, which is a p e equisi e o
uni e si y admission. These schools p ima ily en oll s uden s om he age o 15.
4. Seconda y Voca ional Schools wi h App en iceship Ce i ica es (n = 986) – hese schools
o e aining in speci ic ades (e.g., cook, au o mechanic) bu do no include he
“Ma u i a” exam, meaning g adua es a e no eligible o uni e si y s udies.
Figu e 4.
Using AI a di e en school le els (13+ yea s)
n = 21522
