PAPER
ELECTRONIC TEST OR PRINTED EXERCISE BOOK? WHICH INDIVIDUAL MODE OF PREPARATION IS MORE EFFECTIVE?
Elec onic Tes o P in ed Exe cise Book?
Which Indi idual Mode o P epa a ion is
Mo e E ec i e?
h p://dx.doi.o g/10.3991/ijep. 3i1.2306
M. Czenky
Szen Is án Uni e si y, Gödöllő, Hunga y
Abs ac —A Szen Is án Uni e si y we each da abase
managemen in he amewo k o se e al subjec s. Fo
p epa a ion o he da a modeling exam s uden s o one o
hese cou ses p epa ed wi h help he elec onic es s o
Moodle e-lea ning educa ion sys em, while he a endees o
he o he cou se used p in ed da a modeling exe cise book,
we w o e him based on he inspi a ion o s uden s’ demand.
Wi h examina ions based upon s a is ical me hods we
looked o an answe o see whe he one o he p epa a ion
me hods yield signi ican ly be e esul s. We analyzed
whe he he e is any ema kable di e ence be ween he
ad an ages and disad an ages o he wo ways o p epa a-
ion. We made a su ey among s uden s o see which o he
p epa a ion me hods is a o ed by hem. Based on he ou -
come o he analysis we a e going o decide which one o
p epa a ion me hods will be ecommended by us as each-
e s.
Index Te ms—blended-lea ning, e-lea ning, elec onic es ,
exe cise book, assessmen , s a is ical examina ion, su ey
I. INTRODUCTION
A he Facul y o Mechanical Enginee ing o Szen Is -
án Uni e si y we each Da abase managemen (ABK)
subjec o mechanical enginee s uden s specialized o
enginee in o ma ics. The cou se is one semes e long
wi h 2 lec u es pe week. I s main opics a e he da a mod-
eling and he SQL language.
The en i onmen enginee s uden s o he Facul y o
Ag icul u al and En i onmen Enginee ing can a end he
Compu e s udies III. (KM3) cou se, which is also one
semes e long, wi h 2+2 lessons pee week. The main
opics o he subjec a e he da abase managemen and he
CAD. The s uden s s udy da abase managemen in he i s
hal o he semes e , a lec u es hey s udy da a modeling
and a he semina s he SQL language.
The eaching o he da abase managemen ollows he
same syllabus a bo h cou ses. Ou esea ch co e s he
indi idual p epa a ion me hods o he da a modeling opic.
The da a modeling opic can be cascaded down o he
ollowing pa s: undamen al modeling concep s, en i y
ela ionship model, ela ional model, unc ional and mul-
i alued dependences, no maliza ion and dependence
diag ams [2].
In he amewo k o he uni e si y educa ion we gi e
lec u es in he class oom wi h eache aid. A he same
ime he Moodle e-lea ning educa ion sys em is a ailable
also, which allows he use elec onic cu iculum and es s
beside he p in ed cu iculum in he educa ion. So each-
ing o bo h subjec s and da a modeling happens in blend-
ed-lea ning.
The blended-lea ning is an educa ional o m which al-
lows s uden s o see he ad an ages o bo h ace- o- ace
and online lea ning. In his educa ional o m he class-
oom classes acqui s uden s o he independen s udying
he cu iculum, bu a he same ime hey can use he ech-
nology o e-lea ning in he indi idual p epa a ion phase.
This la e plays an impo an ole in he cou se o co e-
spondence cou se, whe e he numbe o he class oom
classes is low. This lea ning me hod allows he eache o
p o ide mo e help o s uden s wi h lowe pe o mance
[7], [10], [12].
The e a e a eas o da abase managemen which can be
augh mo e e ec i ely by he con en ional me hod. The
da a modeling demands high le el o abs ac ion, he
p esen a ion o i s p ocess is mo e e ec i e wi h eache
explana ion in he amewo k o a con en ional lec u e. In
he cou se o eaching SQL language he con ibu ion o
he eache is also impo an in he p esen a ion o he
logic o he language, as s uden s mee wi h decla a i e
p og amming language a i s his ime.
In he cou se o eaching he di e en subjec s he lec-
u es, he consul a ions and he p ac ices o he da abase
design happen in he amewo k o con en ional class-
oom classes. The da abase managemen p ac ices a e held
in compu e labs.
The s uden s acqui e he cu iculum h ough indi idual
lea ning. The eache p o ides hem wi h assis ance a he
consul a ions aken in he da a modeling and SQL opics.
The da a modeling exam pape is a compu a ional es .
Fo his exam pape he indi idual p epa a ion happens
wi h sol ing p ac ice es s (ABK cou ses) o sol ing o he
asks o he exe cise book (KM3 cou se). A he ABK
cou ses nobody lea ned wi h help o he exe cise book,
while 75%-78% o s uden s o KM3 cou se p epa ed wi h
he help o i o he exam pape . 70%-80% o he s uden s
o ABK cou ses sol ed p ac icing es s.
Re e ence [9] sha es expe iences o blended-lea ning
used in dis ance lea ning and esul s o su ey made
among he s uden s. Acco dingly he bigge pa o s u-
den s used he p in ed cu iculum ( ex book, exe cise
book) o he lea ning. Ou o elec onic ools mos o he
s uden s used elec onic es s i s equency is highe han
he equency o he use o elec onic communica ion
o ms (cha , o um, message). The ime o he lea ning
wi h elec onic ools was subs an ially sho e han he
12
h p://www.i-jep.o g
PAPER
ELECTRONIC TEST OR PRINTED EXERCISE BOOK? WHICH INDIVIDUAL MODE OF PREPARATION IS MORE EFFECTIVE?
leng h o ime o lea ning wi h p in ed cu iculum. The
s uden s did no lea n con inuously, bu campaign-like.
The Moodle e-lea ning educa ional sys em is used on
he Depa men o In o ma ics o he Szen Is án Uni e -
si y o 7 yea s and on he uni e si y le el since 2008. To
da e mo e han 800 cou ses can be ound in he uni e -
si y’s Moodle sys em. E e y subjec o he Depa men o
In o ma ics has a cou se in Moodle. The Facul y o Me-
chanical Enginee ing is he acul y on he uni e si y wi h
he highes numbe o online cou ses.
Re e ence [17] desc ibes en yea s o e-lea ning wi hin
he echnical educa ion in he Czech Republic. I compa es
he de ini ion o se e al au ho s o e-lea ning and empha-
sizes wo aspec s o e-lea ning: ”Fi s , i is he educa ional
p ocess suppo ed by in o ma ion and communica ion
echnologies; second, i is a se o echnological ools
suppo ing educa ion.” I p esen s he p ocess o g adual
in oduc ion o e-lea ning on 19 Czech uni e si ies be-
ween 1999 and 2007. Finally i e alua es he esul s o a
su ey on he usage o e-lea ning and inds ha e-lea ning
has become an in eg al pa o he in o ma ion socie y.
Re e ence [1] also s a es ha nowadays online educa-
ion basically eplaced le e cou ses and ideo lec u es,
educa ion o enginee ing wi h online s udies became
widely used and ecognized. The au ho s in addi ion
summa ized he equi emen s o online cou ses:
 hey mus be simila o be e han he class oom ed-
uca ion,
 hey should be a ailable anywhe e and any ime o
any s uden ,
 hey should be a ailable in se e al enginee ing op-
ics.
Re e ence [19] e iews he ad an ages o he usage o
Moodle o eache s and s uden s:
 o he eache s he sys em allows cou se planning
suppo , in o ma ion sha ing wi h o he eache s,
moni o ing s uden s’ ac i i ies, ollow-up o s u-
den s’ lea ning p ocess,
 o he s uden s he sys em helps planning hei indi-
idual wo ks, allows in o ma ion sha ing wi h each
o he , and ensu es eceip o eache eedbacks.
II. ELECTRONIC TESTS IN THE TEACHING OF THE ABK
SUBJECT
We use he Moodle e-lea ning educa ion sys em om
he i s semes e o school yea 2007/2008 a he eaching
o he ABK subjec . We elabo a ed he ques ions o he
elec onic es s based on he wo ks o [2], [15] and [18]. In
he da a modeling es s a small pa o he ques ions a e
heo e ical, he majo i y o hem a e exe cises o p ac ice.
The w i ing o he ques ions las ed o he nex school
yea s, so he numbe o he p ac ice and exam es s
changed yea by yea . We appended eache explana ion
o he majo i y o he ques ions, which we e a ailable o
s uden s a e closing o he es s.
We used a di e en es se ies yea by yea in o de o
mee new aims yea ly. In he i s wo yea s he e we e
only p ac ice and exam es s. In he i s yea he e was
only one p ac ice es wi h 21 andom ques ions and one
exam es . We ga e ela i ely longe ime o sol ing he
p ac ice es , we did no show he co ec answe s a e he
closing he es , and we es ic ed he possible a emp up
o 3 imes. This way we we e going o d i e s uden s o
hinking and inding answe s o he p oblema ic ques-
ions in he cu iculum.
In he second yea he p ac ice es s consis o 4-8 an-
dom ques ions ou o he 15-35 a ailable ques ions in
connec ion wi h he opic. We did no es ic he numbe
o he a emp s. We showed he co ec answe s a e he
closing he es s so ha s uden s can lea n om compa i-
son o hei answe and he co ec one.
TABLE I.
THE ASSESSMENT OF THE MOODLE DATA MODELING TESTS
2007 2008 2009, 2010
numbe o s uden s 54 10 19
numbe o es ques ions 108 102 191
showing he co ec answe s no yes yes
numbe o he p ac ice es s
sol ed by one p ac icing
s uden s 2.8 31.4 33.6
numbe o p ac ice es s 1 5 14
quo ien o he abo e ones
(how many imes a s uden
sol ed a es ) 2.8 6.29 2.40
numbe o non-p ac icing
s uden s 25.93% 30% 10.53%
a e age du a ion o sol ing he
exam es (minu es) 29.47 22.98 36.97
a e age esul o he exam es 71.22% 71.14% 68.01%
use ulness o he es s 3.891 4.251 2
1 on a i e deg ee scale
2 ex assessmen , he 58% o he s uden s answe ed, who conside ed he
es s good
In Table I. we summa ized he cha ac e is ics and e-
sul s o he es s. In 2009 and 2010 s uden s used he same
es se ies; his is he eason o he con ac ion o he e-
sul s.
Among he ou cou ses examined by e e ence [19]
h ee ha e es s. I is salien , ha in wo cou ses he num-
be s o es a emp s by s uden is 5.03 and 4.4 o a he
4.32, while in he hi d cou se 2.44 and 2.04. Fo compa i-
son in he cou ses examined by us hese alues a e 2.8,
6.29 and 2.4, bu he middle alue is no ypical, i is due
o he ex ao dina y ac i i y o one s uden in a small sized
g oup.
In 2008 a he end o he school yea we e alua ed he
esul s, as well as we examined he lea ning p ac ice o
s uden s [3]. Al hough s uden s p ac iced subs an ially
mo e han in he p e ious yea , he esul s o he exam es
we e p ac ically unchanged in he i s wo yea s. In 2008
one s uden caused he high le el o s uden p ac ices who
sol ed he es s much mo e imes han he o he s uden s
o he cou se.
Examining he lea ning p ac ice we expe ienced ha in
he beginning o he semes e s uden s p ac iced mo e han
in he second hal o he cou se, u mos in he week o he
exam es . Du ing he day he mos ac i e pe iods we e he
ea ly e ening hou s and he one o wo hou s a e mid-
nigh .
In 2009 we p epa ed h ee addi ional es s wi h new
ques ions in all opics. We wished o measu e he le el o
unde s anding o he cu iculum o he cou se wi h he
help o he es s “be o e he p ac ice”. S uden s had o
sol e he es s a he end o he classes. We es ablished
“con ol be o e p ac ice” es s o measu e he e iciency o
p ac ice, which had he same con en as he “be o e he
p ac ice” es s, we asked he s uden s o sol e hem a e
iJEP – Volume 3, Issue 1, Janua y 2013
13
PAPER
ELECTRONIC TEST OR PRINTED EXERCISE BOOK? WHICH INDIVIDUAL MODE OF PREPARATION IS MORE EFFECTIVE?
he p ac ice. The measu emen o he acquisi ion o he
knowledge happened wi h “a e he p ac ice” es s, hese
we e he exam es s.
In 2009 we inc eased he numbe o he p ac ice es s in
he ollowing opics: no maliza ion, dependences and
dependence diag ams. We limi ed he ime a ailable o
sol ing he es s, and showed he co ec answe s a e
closing i , while we did no limi he numbe o a emp s.
We used he same es se ies in 2010 as in 2009.
In 2010 we e alua ed again he lea ning p ac ice o he
s uden s [5], [6]. We expe ienced ha he s uden s lea ned
hence o wa d campaign-wise ins ead o con inuous lea n-
ing. Only hal o he s uden s sol ed he es s which meas-
u ed he e iciency o p ac ice. They used mos o he ime
o sol e he dependency and no maliza ion es s, hese a e
he mos di icul opics o he da a modeling. Du ing he
day he ac i e pe iods we e he same as a he p e ious
examina ion.
Acco ding o da a o Table I. he esul o all h ee yea s
can be e alua ed as sa is ac o y, bu in he las pe iod he
e iciency has dec eased. The eason o he no imp o ing
esul s can be he ollowings:
 The s uden s p ac iced less.
 The ime de o ed o sol e he es s dec eased.
 The s uden s do no ha e he same abili ies.
 Fo pa o s uden s he adi ional educa ional o m
would be mo e sui able, ha is lea ning om a book,
and p ac ice wi h he help o an exe cise book.
Acco ding o he da a o Table I. in 2009 and 2010 s u-
den s sol ed mo e es s han in p e ious yea s, bu since in
his pe iod we e he numbe o p ac ice es s he highes ,
he numbe o a emp s pe es sol ed by s uden is
sligh ly lowe in his pe iod. We aise an issue as eache :
did he s uden s p ac ice less o is he numbe o es s oo
high in he hi d pe iod?
Table II. shows he dis ibu ion o he es s by opics.
The e is mo e han one es in he ela ional model opic,
due o he impo ance and he ex en o he opic. Mo e
es s belong also o he no maliza ion opic which is
caused by he adequa e lea ning o he algo i hm o he
no maliza ion. A his opic each es comes along he
no maliza ion o a able. Fi e sample asks can’ be con-
side ed oo much, due o he impo ance o he opic.
The ime de o ed o sol e he es s dec eased in he
consecu i e yea s, see Table III. This may ake us o he
conclusion ha he aim o achie ing be e esul s changed
and a he became ecei ing he co ec answe s as soon
as possible [6].
Al hough we did no make a su ey o compa e he
abili y o he s uden s o he di e en cou ses, bu i has o
be men ioned ha he ma ks o 2009’ cou se we e wo se
han he ma ks o he o he cou ses. In 2009 he e was no
be e ma k hen sa is ac o y. The s uden s o his cou se
lagged wi h one yea om uni e si y sample syllabus.
Thei esul s in luenced he esul s o hi d pe iod oo.
The majo i y o he s uden s o ABK cou ses ound he
es s use ul du ing he p epa a ion and enjoyed sol ing he
es s, bu he e we e some s uden s in each cou se who
we e in e es ed in he exe cise book. Pa ly because o
his, on he o he hand o imp o e he e iciency o he
eaching we began o compile a da a modeling exe cise
book.
III. EXERCISE BOOK IN THE TEACHING OF KM3 SUBJECT
We published he da a modeling exe cise book in p in
in Sep embe 2010 con aining 300 exe cises. Pa o hese
exe cises could be ound in he elec onic es s also, while
he o he pa consis ed o new exe cises. In he exe cise
book 50 ques ions a e heo e ical, 250 a e p ac icing,
sho e modeling asks [4].
In he au umn semes e o 2010 we also ecommended
he use o he exe cise book o he s uden s o he ABK
cou se, bu hey chose he elec onic es s. On he o he
hand he s uden s o KM3 cou se p epa ed o he exam
pape wi h he help o he exe cise book. The exam pape
was an elec onic es in his case oo, be o e exam he
KM3 s uden s could only sol e a sample es consis ing o
22 ques ions in he Moodle sys em o y his exam
me hod.
We summa ized he ad an ages and disad an ages o
he wo indi idual p epa a ion me hods based on ou ex-
pe ience and [14] in Table IV.
TABLE II.
DISTRIBUTION OF THE PRACTICE TESTS BY TOPIC
Topic Numbe o
heo e ical
es s
Numbe o
p ac ice
es s
undamen al modeling concep s 1
E-R model 1
ela ional model 1 2
dependences 1 1
no maliza ion 1 5
dependence diag ams 1
TABLE III.
THE CHANGE OF THE TIME DEVOTED FOR TEST SOLVING
Yea Used ime/ a ailable
ime P opo ion o s u-
den s
1/6 23.5%
1/3 53%
2007
2/3 23.5%
1/5 81%
2008 2/5 19%
1/4 74%
2009, 2010 1/3 26%
TABLE IV.
THE ADVANTAGES AND DISADVANTAGES OF USE OF TEST AND EXERCISE
BOOK
Ad an age o disad an age Elec onic
es Exe cise
book
i is adap ed o indi idual lea ning yes yes
i needs compu a ional in as uc u e yes no
he s uden maps ou his ime by onesel yes yes
i is a ailable wi hou e e ence o ime
and geog aphy posi ion yes yes
s uden s can ocus on hei own lea ning
needs yes yes
he s uden may consul wi h he eache
om ime o ime (cha ) yes no
i can be used mo e imes yes yes
i is gene ally ee o cha ge o he
s uden s yes no
eamwo k no yes
he s uden can make a no e beside he
asks no yes
he s uden can sea ch be ween ques ions no yes
14
h p://www.i-jep.o g
PAPER
ELECTRONIC TEST OR PRINTED EXERCISE BOOK? WHICH INDIVIDUAL MODE OF PREPARATION IS MORE EFFECTIVE?
A he elec onic es s he necessi y o he compu a-
ional in as uc u e can’ be conside ed as a disad an age,
because e e y s uden has a compu e . The disad an age is
ha he exe cise book is no ee o cha ge, bu he exe -
cise book is cheap compa ed o Hunga ian in o ma ion
echnological books and he s uden s ha e 20% discoun
a he pu chase. The e is consul a ion oppo uni y o
s uden s lea ning om exe cise book oo, bu i did no
happen in an elec onic way, bu in class oom a a p ede-
e mined ime.
Two ac s suppo ed he use o he exe cise book: we
expe ienced ha s uden s we e lea ning in eams, which
was no ypical o he solu ion o he elec onic es s.
Du ing he consul a ion he s uden s may w i e so no es
in o he exe cise book, he ask and he no es emained
a ailable la e oo. We held an assessmen a e he solu-
ion o he elec onic es s oo, bu in his case he ques-
ions and he no es a e no a ailable oge he a he same
ime, because s uden s can’ sea ch in he ques ion bank,
and he e is no enough ime o pu down he ques ions.
IV. THE METHOD OF THE EXAMINATION
Fo he assessmen o he esul s o wo subjec s on one
hand we compa ed he esul s o he exam es s o he
ABK 2007-2008 and he KM3 cou ses, on he o he hand
we examined he esul s o 20-20 ques ions o he exam
es s o he ABK 2009-2010 and he KM3 cou ses. A he
ABK cou ses in 2007 and 2008 he exam es s we e he
same, and he s uden s o he ABK 2009 and ABK 2010
had o sol e he same exam es s, o his we con ac ed
he esul s.
The ci cums ances o w i ing he exam es s o he
ABK 2007-2008 and he KM3 cou ses we e simila as we
gene a ed he ABK 2007-2008 exam es om p ac iced
ques ions, and he KM3 exam es om he ques ions o
he exe cise book. This did no mean in ei he o wo cases
ha he s uden s knew all he ques ions, because hey
sol ed only a pa o he ques ions o he ques ion bank
du ing he p ac icing es s, while he s uden s p epa ed
om exe cise book also deal wi h only pa o he ques-
ions o i (see Table XI.).
In he cou se o examining he esul s o he exam es s
o he ABK 2007-2008 and KM3 cou ses he main ques-
ions a e he ollowings:
 May he esul s o wo g oups be cha ac e ized simi-
la ly?
 Ha e he wo da a ows same dis ibu ion?
 Is he esul o ei he g oup signi ican ly be e ?
Fi s we calcula ed he a e age ime de o ed o sol e
he exam es and a e age sco es as well as he co ela ion
coe icien s o compa e he esul s o he cou ses. A e
his we ep esen ed he sco es in he unc ion o imes
de o ed o es sol ing on diag am and de e mined ha 3-
3 clus e s in o which he alues can be assigned. Finally
we execu ed he signi icance es o he co ela ion coe i-
cien s oo.
We analyzed oo i he da a ows a e no mally dis ib-
u ed, because he esul s o he exam pape s a e gene ally
no mally dis ibu ed. To decide whe he he wo da a ows
ha e he same dis ibu ion we made a homogenei y es .
The a e age sco es al eady show which g oup’s esul
is be e , bu o e i y his we ep esen ed he dis ibu ion
o he esul s on diag am oo. Finally we made an inde-
pendence es o decide whe he he esul s a e signi i-
can ly dissimila .
The exam es s o he ABK 2009-2010 and KM3
cou ses we e p incipally simila . We examined 20-20
ques ions o hese exam es s since hese ques ions can be
s ood in o pai s based on hei ype. Wi h examina ion o
he esul s o hese ques ions we a e going o answe wo
ques ions:
 In his case is he esul o KM3 g oup signi ican ly
be e oo?
 A e he e any opics which can be conside ed easy o
di icul equally o bo h g oups?
In his case we calcula ed he a e age sco es oo hen
we made an independence es o decide whe he he e-
sul s o wo g oups a e signi ican ly di e ing. By ou
opinion he e is no a disad an age ha he exam es o
ABK 2009-2010 cou se did no gene a e om p ac icing
ques ion, because in p ac icing es s and in “con ol be o e
p ac ice” es he s uden s o his cou se sol ed same ype
ques ions han ques ions o he exam es s.
V. THE RESULTS OF STATISTICAL AND DATA MINING
EXAMINATIONS
We execu ed he examina ions by [11], [13], [16]. In
he cou se o he examina ion o he esul s o he ABK
2007-2008 and KM3 cou ses i s we calcula ed he a e -
age ime de o ed o sol e he exam es and a e age
sco es, hen we de e mined he co ela ion coe icien s,
see Table V.
To decide whe he he e is a ela ionship be ween he
ime used and sco es o he ABK 2007-2008 g oup we
analyzed he signi icance o he co ela ion coe icien .
We made he examina ion wi h da a o KM3 g oup oo.
Ou null hypo hesis is he ollowing in bo h cases:
H0: he co ela ion coe icien is ze o
Table VI. shows he esul s o he examina ions.
TABLE V.
AVERAGE TIMES AND SCORES IN EXAM TESTS
Da a ABK 2007-
2008 KM3
a ailable ime (minu es) 56 35
a e age ime used (minu es) 39,83 24.31
a e age ime used (%) 71.13% 69.48%
achie able sco e 35 25
a e agely sco e 24.92 19.206
a e age esul (%) 71.2% 76.82%
co ela ion coe icien 0.0670837 -0.2591699
TABLE VI.
RESULTS OF THE SIGNIFICANCE ANALYSIS OF THE CORRELATION COEF-
FICIENTS
ABK 2007-2008 KM3
co ela ion coe i-
cien 0.0670837 -0.2591699
n 64 72
0.529409809 -2.24508148
signi icance le el 0.05 0.05
deg ee o eedom 62 70
alue in able 1.998971498 1.994437086
iJEP – Volume 3, Issue 1, Janua y 2013
15
PAPER
ELECTRONIC TEST OR PRINTED EXERCISE BOOK? WHICH INDIVIDUAL MODE OF PREPARATION IS MORE EFFECTIVE?
We ep esen ed he achie ed sco es in he unc ion o
imes de o ed o es sol ing on diag am in case o bo h
g oups, see Figu e 1. We pu he eg ession lines, hei
equa ions and de e mina ion coe icien s on he diag am
oo.
Beside he a e ages a bo h g oups we de ined hose
h ee- h ee clus e s, in o which he alues can be assigned,
see Table VII., and we depic ed hei cen e on he igu e
1. These clus e s show endencies be e han he a e age
alues.
The esul o he exam pape o a g oup is gene ally
no mally dis ibu ed. We examined wi h Chi-squa ed es
whe he he esul s o exam es s o wo g oups a e no -
mally dis ibu ed? We ound wi h 95% p obabili y ha
nei he o hem is no mally dis ibu ed.
We made a homogenei y es oo o de e mine whe he
wo da a ows ha e he same dis ibu ion. Ou null hy-
po hesis is he ollowing:
H0: wo da a ows ha e he same dis ibu ion
Table VIII. shows he esul s o he examina ions.
We p epa ed a diag am showing he dis ibu ion o he
esul s ins ead o he usual equency diag am, see Figu e
2.
We made an independence es in o de o decide
whe he KM3 g oup's esul is signi ican ly be e . Ou
null hypo hesis is he ollowing:
H0: he e is no ela ionship be ween he esul and he
way o he p epa a ion, meaning he da a ows a e inde-
penden .
Table IX. shows he esul s o he independence es .
We calcula ed h ee con ingency coe icien s. The alue
o he Pea son con ingency coe icien is 0.284028 (1), he
alue o he Tschup o con ingency coe icien is
0.148114 (0.707107), while he alue o he C ame con-
ingency coe icien is 0.296227 (1), in he pa en hesis a e
he possible highes alues.
A he 20-20 ques ions examina ion o he ABK 2009-
2010 and KM3 g oups i s we calcula ed he a e age
sco es, hen we examined whe he he da a ows a e no -
mally dis ibu ed. The a e age sco e o he ABK 2009-
2010 g oup is 16.18, while he a e age sco e o he KM3
g oup is 17.62. So he esul o he KM3 g oup is be e
wi h 6.25%. A he no mally es we ound so ha esul o
ABK 2009-2010 g oup is no mally dis ibu ed wi h 0.05
signi icance le el, bu he esul o he KM3 g oup is no .
We made he independence es oo Table X. shows he
esul s o i .
We calcula ed he a e age sco es by opic oo, hen we
di ided i wi h he maximum achie able sco e in he opic
and depic ed he alues on diag am, see Figu e 3.
TABLE VII. THE RESULT OF THE CLUSTERING
Clus e s o KM3 cou se
Time Sco e Head %
clus 0 32.477 14.705 15 20.8%
clus 1 25.672 21.823 31 43.1%
clus 2 17.999 18.683 26 36.1%
Clus e s o ABK 2007-2008 cou se
Time Sco e Head %
clus 0 44.729 25.226 44 68.8%
clus 1 29.738 16.259 6 9.4%
clus 2 28.778 27.678 14 21.9%
Figu e 1. Time-sco e cha
Figu e 2. The dis ibu ion o he esul s
Figu e 3. Resul s by opics
TABLE VIII.
THE RESULT OF THE HOMOGENEITY TEST
2 11.93409003
deg ee o eedom 4
signi icance le el 0.05
c i ical alue 9.487729037
TABLE IX.
THE RESULT OF THE INDEPENDENCE TEST – ABK 2007-2008 AND KM3
calcula ed p obabili y 0.004247
2 15.230086
deg ee o eedom 4
signi icance le el 0.05
c i ical alue 9.487729
TABLE X.
THE RESULT OF THE INDEPENDENCE TEST – ABK 2009-2010 AND KM3
calcula ed p obabili y 0.03554033 0.03554033
2 10.3084525 10.3084525
deg ee o eedom 4 4
signi icance le el 0.05 0.03
c i ical alue 9.48772904 10.7118983
16
h p://www.i-jep.o g

PAPER
ELECTRONIC TEST OR PRINTED EXERCISE BOOK? WHICH INDIVIDUAL MODE OF PREPARATION IS MORE EFFECTIVE?
VI. THE ASSESSMENT OF THE EXAMINATIONS
Da a o he Table V. show ha he s uden s o he ABK
2007-2008 g oup de o ed by 1.65% mo e ime o he
sol ing o he exam es han he s uden s o he KM3
g oup, bu hei a e age esul was wo se by 5.2%. The
co ela ion coe icien o he KM3 g oup is nega i e i s
absolu e alue shows ha he e is a e y weak ela ionship
be ween he ime used and he achie ed sco e. I s nega i e
alue shows so ha hose s uden s who de o ed mo e ime
o he sol ing exam es achie ed wo se esul han hose
s uden s who sol ed exam es in less ime. In he ABK
g oup he alue o he co ela ion coe icien s is almos
ze o, i e e s on e y weak ela ionship oo.
Based on he esul o he signi icance analysis o he
co ela ion coe icien s (Table VI.) we can claim wi h
95% p obabili y ha a ABK 2007-2008 g oup he co e-
la ion coe icien can be conside ed ze o, hus he e is no
signi ican ela ionship be ween ime de o ed o sol e he
es and he achie ed sco e. Bu a KM3 g oup he co ela-
ion coe icien is no ze o.
The clus e ing shows so ha abou hal o he s uden s
o he KM3 g oup achie ed be e esul wi h medium
ime spen on i . In con as o his 70% o he s uden s o
he ABK 2007-2008 g oup achie ed medium esul wi h
he longes ime spen (Figu e 1 and Table VII.).
The analysis indica es ha he esul s o wo g oups do
no wo k simila ly. Nei he o wo da a ow is no mally
dis ibu ed. The homogenei y es indica es a 0.05 signi i-
cance le el ha wo da a ows do no ha e he same dis-
ibu ion.
The a e age sco e is be e in KM3 g oup and he e-
quency diag am o Figu e 2 shows he be e esul o
KM3 g oup oo. A KM3 g oup 35% o he s uden s
achie ed he bes esul while in ABK 2007-2008 only
11% o hem achie ed i . Howe e in ABK 2007-2008
g oup mo e esul s we e be ween 61%-85% han in KM3
g oup. All in all we can say ha he esul o KM3 g oup
is be e .
Acco ding o he esul o he independence es (Table
IX.), wi h 95% p obabili y he da a ows a e no inde-
penden , so he p epa a ion me hod does in luence he
esul . Al hough he Chi-squa e alue is no a p ope indi-
ca o o he s eng h o he ela ionship o wo da a ows,
bu i is well known, ha i s small alue shows he inde-
pendence o he da a ows, while i s highe alue e e s, as
in ou case, on ha he da a ows a e no independen . All
h ee con ingency coe icien s calcula ed by us di e om
ze o, which shows a medium weak ela ionship, so he
da a ows a e no independen . Thus he esul o he KM3
g oup is signi ican ly be e .
A he examina ion o he 20-20 ques ions exam es o
he ABK 2009-2010 and KM3 g oup we ound ha he
esul o he KM3 g oup is be e . Acco ding o analysis o
he dis ibu ion he wo da a ows does no ha e he same
dis ibu ion. The independence es shows ha he wo
da a ows a 0.05 signi icance le el a e no independen ,
on he o he hand a 0.03 and smalle signi icance le el
hey a e independen . I he le el o accep ance is be ween
1% and 3%, hen he s a is ic p ac ice ejec s he null
hypo hesis. So he esul o he second independence
analysis con i ms he esul o he i s independence es .
Thus he mode o he lea ning in luences he esul .
Acco ding o Figu e 3 showing he esul by opics i
can be seen ha i is ue o nei he o he g oups ha all
esul a e be e han he o he . Nei he o he opics can be
conside ed easy o di icul equally o bo h g oups.
Thus we can say based on ou analysis ha he lea ning
om exe cise book p o ed mo e e ec i e.
VII. SURVEY
A he end o he semes e s we made a su ey, in which
beside o he opics we asked he s uden s how use ul hey
ound he p epa a ion wi h he help o p ac ice es s and
wi h he use o he exe cise books. In cou se o he su ey
we ollowed [8]'s guide.
We ou lined he summa y o answe s o he ABK
cou ses al eady in Table I. We asked all s uden s o KM3
g oup o answe ou ques ions. 66% o he KM3 g oup’
s uden s ook pa in he su ey, i is andom who did no
ill up he ques ionnai e hus we can conside he su ey
ep esen a i e. F om among he ques ions wo a e in e es -
ing o us:
 How many imes did hey sol e he exe cises o he
exe cise book o a he how many pe cen o he ex-
e cises did hey sol e mo e imes? (Table XI.)
 Which p epa a ion me hod did hey conside mo e
use ul, and which one would hey ecommend o he
nex college yea s uden s? (Table XII.)
Acco ding o he da a o Table XI. a qua e o he s u-
den s did no use he exe cise book. 1/5 o he s uden s
sol ed all exe cises o he exe cise book, a qua e o hem
sol ed mo e han 75% o he exe cises, while he o he
s uden s a ended o less han hal o he exe cises. The
numbe o he solu ions o exe cises o he second ime o
se e al imes is dec easing. Only ew pe cen o he s u-
den s sol ed all exe cises mo e imes (4.3%).
Acco ding o da a o he Table XII. mo e han hal o
he s uden s conside bo h p epa a ion manne s use ul.
Among hose s uden s who ma ked only one o he p epa-
a ion manne a li le bi mo e s uden s conside ed he
exe cise book use ul. Two hi d o he s uden s ecom-
mended bo h o he p epa a ion manne s o he s uden s
o he nex colleges yea . The p opo ion o hose s uden s
is nea alike, who sugges only one o o he p epa a ion
manne o he u u e s uden s.
TABLE XI.
THE RESULTS OF THE KM3 SURVEY I.
How many ques ions o
he exe cise book did you
sol e (in pe cen age)?
How many pe cen ages o
he ques ions o he exe cise
book did you sol e mo e
imes?
did no use he
exe cise book 23.4% 19.1%
100% 21.3% 4.3%
75%-100% 27.7% 8.5%
50%-75% 6.4% 12.8%
25%-50% 14.9% 17.0%
1%-25% 6.4% 38.3%
TABLE XII.
THE RESULTS OF THE KM3 SURVEY II.
The use o exe cise book o
sol ing o he es s did
suppo be e you
p epa a ion?
Which p epa a ion
me hod would you ec-
ommend o he s uden s
o he nex college yea ?
nei he o hem 2.1%
exe cise book 23.4% 15.2%
p ac ice es 19.1% 17.4%
bo h 53.2% 67.4%
iJEP – Volume 3, Issue 1, Janua y 2013
17
PAPER
ELECTRONIC TEST OR PRINTED EXERCISE BOOK? WHICH INDIVIDUAL MODE OF PREPARATION IS MORE EFFECTIVE?
VIII. SUMMARY
The analysis made based on basic s a is ic ea u es indi-
ca ed ha he esul s o he s uden s lea ning wi h he help
o exe cise book a e be e . The alues o he con ingency
coe icien s indica e ha he me hod o he p epa a ion
de ines he esul s on a medium le el. All o he wo inde-
pendence analysis shows a signi ican ly be e esul a he
s uden s lea ning om exe cise book. The e is no an
essen ial di e ence be ween he ad an ages and disad an-
ages o he p epa a ion me hods, jus like no di e ence in
he le el o suppo among s uden s p e e ing p epa a ion
me hods. Based on hese we would sugges he use o he
exe cise book as a eache , bu we lea e he decision o
he s uden s due o he di e en p e e ences o s uden s.
REFERENCES
[1] J: Bou ne, D. Ha is, and F. Mayadas, "Online Enginee ing Educa-
ion: Lea ning Anywhe e, Any ime", Jou nal o Enginee ing Edu-
ca ion, Janua y 2005, pp. 131-146
[2] M. Czenky, Ada modellezés, SQL és Access alkalmazás, SQL
Se e és ADO [Da a modeling, applica ion o he SQL and he
Access, SQL Se e and ADO]. Budapes : Compu e Books Kiadó,
2005.
[3] M. Czenky, “Hallga ók anulási szokásainak és e edményességé-
nek elemzése” [The analysis o s uden s' lea ning habi s and hei
e iciency]. P oceedings o Ag ia Média 2008 Kon e encia, Ege :
EKF Líceum Kiadó. pp. 324-341, 2008.
[4] M. Czenky, Ada modellezési példa á [Exe cise book o da a
modeling]. Budapes : Compu e Books Kiadó, 2010.
[5] M. Czenky, “Hallga ói e ékenységek elemzése a Moodle
napló ájljai alapján” [The analysis o he s uden s ac i i ies based
on he Moodle’s log iles]. Pos e p esen ed a he MoodleMoo
2010 Kon e encia, Gödöllő, Hunga y, 2010.
[6] M. Czenky, “Moodle esz ek használa ának e edményesség izs-
gála a” [The examina ion o he e iciency o he usage o Moodle
es s]. P oceedings o II. Ok a ás-In o ma ikai Kon e encia, Buda-
pes : ELTE Eö ös Kiadó. pp. 59-64, 2010.
[7] C. D. Dziuban, P. D. Moskal, and J. Ha man, “Highe educa ion,
blended lea ning, and he gene a ions: Knowledge is powe : No
mo e” In J. Bou ne & J. C. Moo e (Eds.), Elemen s o Quali y
Online Educa ion: Engaging Communi ies. Needham, MA: Sloan
Cen e o Online Educa ion, 2005.
[8] I. Falus, J. Ollé, Az empi ikus ku a ások gyako la a [The p ac ice
o he empi ical esea ches]. Budapes : Nemze i Tanköny kiadó,
2008.
[9] S. Fo gó, Z. Hause , L. Kis-Tó h, Cs. Komló and B. Szabó, “A
blended lea ning ( egyes ípusú) anulás ámoga ási o mák
be eze ése, és ha ékonyságának izsgála a az eg i Esz e házy
Ká oly Főiskolán” [The in oduc ion and he examina ion o e i-
ciency o he blended-lea ning suppo ed o ms a he Esz e házy
Ká oly Collage in Ege ] P oceedings o Ag ia Media Kon e encia,
Ege , pp. 193-217, 2005.
[10] D. R. Ga ison and H. Kanuka, “Blended lea ning: Unco e ing i s
ans o ma i e po en ial in highe educa ion” In e ne and Highe
Educa ion 7, pp. 95–105, 2004. h p://dx.doi.o g/10.1016/j.iheduc.
2004.02.001
[11] J. Han and M. Kambe , Da a mining: Concep s and Techniques.
New Yo k: Mo gan-Kau man, 2006.
[12] A. Ha ding, D. Kaczynski and L. Wood, “E alua ion o blended
lea ning: analysis o quali a i e da a”. P oceedings o he Blended
Lea ning in Science Teaching and Lea ning, Uni e si y o Syd-
ney. pp. 56-62, 2005.
[13] L. Hunyadi, Gy. Mund oczó and K. Vi a, S a isz ika [S a is ics],
Budapes : Aula, 1996.
[14] O. Hu e , G. Magya , J. Mlina ics, E-lea ning 2005, Budapes :
Műszaki Köny kiadó, 2005.
[15] L. Ko ács, Ada bázisok e ezésének és kezelésének módsze ana
[The me hod o he design and managemen o he da abases], Bu-
dapes : Compu e Books, 2004.
[16] W. R. Pes man, Ma hema ical s a is ics: an in oduc ion. Be lin,
New Yo k: de G uy e , 1998.
[17] I. Simono a and M. Bilek, “Ten Yea s o eLea ning wi hin he
Enginee ing Educa ion in he Czech Republic”, In e na ional
Jou nal o Enginee ing Pedagogy, Volume 2, Issue 3, 2012, pp.
29-39
[18] J. D. Ullmann and J. Widom, Ada bázis endsze ek – Alap e és
[Da abase sys ems - basics].Budapes : Panem KFT, 2008.
[19] C. Viegas, A. Ma ques, G. R. Al es and C. C. Lobo, “Engaging
S uden s by Moodleing a Cou se? Case S udies a he Poly echnic
o Po o – School o Enginee ing”, In e na ional Jou nal o Engi-
nee ing Pedagogy, Volume 2, Issue 3, 2012, pp. 40-46
AUTHOR
M. Czenky is an assis an p o esso a he Depa men
o In o ma ics, Szen Is án Uni e si y, Gödöllő, Hun-
ga y, 2100 (e-mail: ma a.czenky@ -online.hu).
Recei ed 11 Oc obe 2012. Published as esubmi ed by
he au ho 18 Decembe 2012.
18
h p://www.i-jep.o g