Digi al Pi acy:
Fac o s ha In luence he In en ion o Pi a e
By
Rúben Emanuel Mou inho Mei eles
Mas e ’s Disse a ion in Economics and Business Adminis a ion
Supe ised by:
P o esso Doc o Ped o José Ramos Mo ei a de Campos
2015
i
Biog aphical No e
Rúben Emanuel Mou inho Mei eles bo n on Ap il 5 h 1992, a ended Escola Básica
2,3 C/ Secundá io de Ca azeda de Ansiães whe e he g adua ed in 2010. On ha same yea
en olled in Facul y o Economics o he Uni e si y o Coimb a, whe e ob ained his
bachelo deg ee in Economics in he summe o 2013.
In Sep embe 2013, seek ou o ob ain his Mas e deg ee in Economics and Business
Adminis a ion a he Uni e si y o Po o, Facul y o Economics, inishing he academic
pa in July 2015 and p esen ing now his disse a ion.
ii
Acknowledgemen s
To my amily, indeed ime uly passes by, i seems like was only yes e day ha his
educa ional jou ney was s a ing, and oday i is in pa he end o i . I am uly g a e ul o
all he oppo uni ies gi en o me, wi hou you help, encou agemen and suppo none o
his could be possible, bu mos impo an ly I would no be you I am.
I would like say hank you o all o my iends. You guys know who you a e and I
am uly glad o he suppo , pa ience, pa icipa ion, and mos o all simply o being
p esen .
To my disse a ion ad iso , P o . D . Ped o Campos I am g a e ul o he
oppo uni y o wo k wi h you and app ecia e he ime and e o dedica ed. Also, hank you
o he sugges ions, ad ices and knowledge ansmi ed.
I am much obliged o all he people ha illed ou he ques ionnai e, all o you made
a huge con ibu ion. Howe e , a special hank you has o go o P o . Ma ia Ma ias.
iii
Abs ac
This disse a ion uses beha io al and economic heo ies o help unde s and some o
he ac o s (a i ude, subjec i e no ms, pe cei ed beha io al con ol, mo al obliga ion, pas
pi acy beha io , punishmen se e i y, punishmen ce ain y, digi al media cos and
pe cei ed alue) ha may in luence an indi idual’s in en ion o pi a e digi al ma e ial. I is
used an expanded amewo k based on he heo y o planed beha io , add essing no only
ac o s capable o in luencing in en ion, bu also using an eceden s o hese ac o s, capable
o in luence in en ion in an indi ec ashion.
This wo k assis s o ul ill he need o s udy digi al pi acy ac oss di e en cul u es,
helping o unde s and how in en ion is di e en ly a ec ed and how policy make s should
adjus policies be ween cul u es. Al hough mos o he ac o s employed a e no new in
pi acy esea ch he e is an excep ion, pe cei ed alue, his ac o was ne e analyzed in his
con ex . Ano he inno a ion o his wo k is he de elopmen o wo models: he i s one
conside s he ull sample and he second conside s only hose who had pi a ed be o e. A
s uden sample has been used and he da a was analyzed using s uc u al equa ion
modeling.
The e we e some di e en esul s be ween he models howe e , he ac o s
pe cei ed beha io al con ol and mo al obliga ion we e signi ican p edic o s o in en ion
in bo h models, bu subjec i e no ms only p esen ed a signi ican e ec in he ull sample
model. Punishmen ce ain y was also a signi ican p edic o o pe cei ed beha io al
con ol in bo h models.
Among he ac o s ha we e no signi ican p edic o s o in en ion was a i ude. I s
an eceden s also showed some mixed esul s, punishmen ce ain y and se e i y did no
p esen a signi ican e ec in bo h models, howe e digi al media cos and pe cei ed alue
we e signi ican p edic o s o a i ude bu only in he ull model. The pi a e model
con i med he exis ence o a signi ican and s ong ela ion be ween pas beha io and
in en ion owa ds digi al pi acy.
The esul s and implica ions a e discussed and o ms o in e en ion a e sugges ed.
i
Resumo
A p esen e disse ação em como obje i o o nece um melho en endimen o sob e
alguns dos a o es que podem a e a a in enção de um indi íduo pi a ea . Usando a heo y
o planned beha io como pon o de pa ida e complemen ando-a com ou as eo ias e
a iá eis ele an es, o am desen ol idos dois modelos capazes de analisa a in enção de
pi a ea .
Es a in es igação em con ibui pa a a necessidade de es uda a pi a a ia digi al
en e cul u as, colabo ando pa a um melho en endimen o de como a in enção de pi a ea é
a e ada e como os deciso es de em ajus a as polí icas en e países. A maio ia dos a o es
analisados não são no os nes a linha de in es igação, con udo a exceção é o a o pe cei ed
alue. Ou a ino ação é o desen ol imen o de dois modelos: um que engloba oda a
amos a (modelo ge al) e ou o apenas os indi íduos que já pi a ea am no passado (modelo
pi a a). Es a sepa ação pe mi e ainda analisa o impac o do ac o pas pi acy beha io na
in enção de pi a ea e compa a esul ados en e modelos. Os dados necessá ios o am
ecolhidos jun o de es udan es e examinados u ilizando a análise de equações es u u ais.
Os esul ados mos a am a exis ência de di e enças signi ica i as en e os modelos,
no en an o alguns a o es ap esen a am um e ei o signi ican e sob e a in enção em ambos,
nomeadamen e os a o es pe cei ed beha io al con ol e mo al obliga ion. Toda ia o a o
subjec i e no ms apenas ap esen ou um e ei o signi ica i o no modelo ge al. O a o
punishmen ce ain y e e em ambos os modelos um e ei o signi ica i o sob e o a o
pe cei ed beha io al con ol.
De en e os que não ap esen a am um e ei o signi ica i o em ambos os modelos
encon a-se o a o a i ude. Po ém, os seus an eceden es demons a am alguns esul ados
dis in os en e os modelos. Os a o es punishmen ce ain y e se e i y não e ela am um
e ei o signi ica i o em ambos os modelos, con udo os a o es digi al media cos e pe cei ed
alue demos a am um e ei o signi ica i o sob e a a i ude, mas apenas no modelo que
conside a a amos a comple a. O modelo pi a a con i mou ainda a exis ência de um e ei o
signi ica i o e o e do compo amen o passado na in enção u u a de pi a ea .
Po im os esul ados são discu idos e são p opos as o mas de in e enção.

Table o Con en s
1. In oduc ion.......................................................................................................................... 1
2. Li e a u e Re iew................................................................................................................. 6
2.1. Pi acy Resea ch ........................................................................................................ 6
2.1.1. Ea ly Pi acy Resea ch ............................................................................................. 6
2.1.2. Digi al Pi acy Resea ch ......................................................................................... 11
2.1.3. Summa y .............................................................................................................. 13
2.2. Model De elopmen : Theo e ical Founda ions and Hypo heses ........................... 14
2.2.1. Theo y o Planned Beha io .................................................................................. 14
2.2.2. Mo al Obliga ion................................................................................................... 17
2.2.3. Pas Pi acy Beha io ............................................................................................. 18
2.2.4. De e ence Theo y ................................................................................................ 19
2.2.5. So wa e and Media Cos ...................................................................................... 21
2.2.6. Pe cei ed Value .................................................................................................... 22
2.2.7. Concep ual Model ................................................................................................. 23
3. Resea ch Me hodology ....................................................................................................... 25
3.1. Ques ionnai e .......................................................................................................... 25
3.1.1. Measu ed Fac o s and Co esponden Sou ces ....................................................... 26
3.2. Da a ......................................................................................................................... 27
3.3. Es ima ion P ocedu e ............................................................................................. 28
4. Resul s ................................................................................................................................ 30
4.1. Fi s Explo a o y Resul s ....................................................................................... 30
4.2. Mul i a ia e Analysis .............................................................................................. 31
4.2.1. Full Model ............................................................................................................ 33
4.2.1.1. Measu emen Model ............................................................................................. 33
4.2.1.2. SR Model.............................................................................................................. 38
4.2.2. Pi a e Model ......................................................................................................... 41
4.2.2.1. Measu emen Model ............................................................................................. 41
4.2.2.2. SR Model.............................................................................................................. 43
5. Discussion and Implica ions ............................................................................................... 45
5.1. TPB Va iables, Mo al Obliga ion and Pas Pi acy Beha io ................................ 45
5.2. Punishmen Ce ain y and Se e i y ....................................................................... 47
i
5.3. Digi al Media Cos and Pe cei ed Value ................................................................ 48
6. Limi a ions and Fu u e Resea ch ...................................................................................... 50
7. Conclusion .......................................................................................................................... 51
8. Re e ences........................................................................................................................... 53
9. Appendix ............................................................................................................................ 60
ii
Lis o Tables
Table 1: Pi acy beha io esea ch ..................................................................................... 13
Table 2: Ques ionnai e ins umen scale ac o s ................................................................ 26
Table 3. Sample demog aphics ......................................................................................... 31
Table 4. Explo a o y ac o analysis ................................................................................. 32
Table 5. Consis ency s a is ics .......................................................................................... 36
Table 6. Squa ed co ela ion be ween ac o s ................................................................... 37
Table 7. Model esul s summa y ...................................................................................... 49
iii
Lis o Figu es
Figu e 1: Concep ual Model.. ........................................................................................... 24
Figu e 2. Final CFA Full Model . ..................................................................................... 35
Figu e 3. Final Full Sample SR Model. ............................................................................ 39
Figu e 4. Final CFA Pi a e Model .................................................................................... 42
Figu e 5. Final Pi a e SR Model. ...................................................................................... 44
5
4. The esul s discussion and implica ions a e in Sec ion 5. Limi a ions and u u e esea ch
di ec ions a e in Sec ion 6 and Sec ion 7 concludes.
A be e unde s anding o consume ’s beha io will help de elop new s a egies and
ideally educe pi acy.

6
2. Li e a u e Re iew
Digi al pi acy is no a new subjec . I has been a ound us o qui e some ime now.
As is indispensable, his sec ion akes on p e ious esea ch and shed a ligh on he li e a u e
ha has been p oduced so a .
I s a s by looking o speci ic a eas o pi acy, like so wa e, music, mo ies and
culmina es on digi al pi acy. While in his i s pa he ocus is on esul s, employed
a iables and heo ies (all in a b ie manne ), he second pa eme ges om he i s one,
which helped selec he esea ch cons uc s ha will be used. This second pa is whe e he
model de elopmen s a s, he heo e ical ounda ions a e lay down and hypo hesis a e
de eloped. I is indeed a mo e speci ic and in-dep h analysis, ha uly dic a es his
in es iga ion pa h.
2.1. Pi acy Resea ch
2.1.1. Ea ly Pi acy Resea ch
The i s majo conce n ega ding copy igh in ingemen was so wa e pi acy.
Resea che s ha e been in es iga ing his phenomenon since he la e 1980s, bu he i s
s udies we e mos ly desc ip i e su eys (Peace e al., 2003; Limayem e al., 2004). One o
he i s empi ical wo ks examining so wa e pi acy, using a model based on a heo e ical
amewo k was Ch is ensen and Eining (1991), applying he Theo y o Reasoned Ac ion
3
(TRA) (Fishbein and Ajzen, 1975). They ound ha a i udes owa d pi acy and subjec i e
no ms we e bo h ela ed wi h he s uden ’s p opensi y o pi a e. Thei in es iga ion
indica ed as well ha his kind o beha io was no seen as inapp op ia e and ha
indi idual’s belie ed ha o he s sha ed he same iew.
Gopal and Sande s (1997) in es iga ed he impac o de e en and p e en i e
measu es on so wa e de elope ’s p o i s, and ound ha p e en i e con ols may ha e a
nega i e impac on p o i s, bu on he o he hand de e en s a egies can po en ially
inc ease hem. They also ound ha de e ence measu es, e hics, sex and age a e ela ed o
3
The TRA exposes human beha io as unc ion o a i ude owa d he beha io and social no ms. Fu he
explana ion is p o ided in he second pa o his sec ion.
7
an indi idual’s p opensi y o pi a e. In a pos e io s udy he au ho s concluded ha he size
o a so wa e indus y is posi i ely ela ed o he go e nmen p opensi y o be an ac i e
o ce in he igh agains pi acy, and ha i is in e sely ela ed o pi acy a es ( ega dless o
a coun y weal h) (Gopal and Sande s, 1998). Consequen ly he exis ence o domes ic
so wa e indus y may be a de e minan ac o agains pi acy.
La e on Gopal and Sande s (2000) es ablished he exis ence o a signi ican e ec
be ween income and global pi acy a es, and hey p oposed global p ice disc imina ion as
he i s line o de ense agains pi acy. Shin e al. (2004), analyzed so wa e pi acy a es o
49 coun ies, conside ing pe capi a GDP and na ional collec i ism as independen
a iables. Finding e idence o a nega i e co ela ion be ween pe capi a GDP and he
so wa e pi acy o a coun y; on he opposi e he ela ionship is posi i e wi h a coun y’s
collec i ism. This suppo s Gopal and Sande s (2000) esul s and implies ha no only
“poo coun ies a e mo e in ol ed in so wa e pi acy, bu also ha high collec i is ic
coun ies a e in ol ed in pi acy” (Shin e al., 2004, p.105).
Tan (2002) ocused his a en ion on he e hical judgmen associa ed wi h so wa e
pi acy, cons uc ing a esea ch amewo k ha inco po a ed se e al beha io al heo ies and
mode a ing a iables capable o in luencing e hical decision-making. His model
4
conside ed he e ec o mo al in ensi y, pe cei ed isk and mo al judgmen , aking also in
accoun he in luence o some mode a ing a iables
5
. Resul s suppo ed he hypo hesis ha
bo h pe cei ed isks and mo al judgmen ha e a nega i e impac on in en ion, in o he
wo ds, he highe he pe cei ed isk/mo al judgmen o consume s, he lowe will be hei
in en ion o pi a e.
Peace e al. (2003) p oposed a amewo k based on he heo y o planed beha io
6
(TPB), complemen ed wi h he expec ed u ili y heo y and de e ence heo y
7
. Using cen al
4
The es ima ion me hod used was he wo-s ep hie a chical eg ession analysis.
5
P ice, gende , age, educa ional a ainmen , income and pas pu chase expe ience.
6
The TPB exposes human beha io as unc ion o a i ude owa d he beha io , subjec i e no ms and
pe cei ed beha io al con ol. Fu he explana ion is p o ided in he second pa o his sec ion.
7
This heo y p oposes ha as punishmen p obabili y and punishmen le el a e inc eased, he le el o illegal
beha io should dec ease (Peace e . al., 2003).
8
ac o s iden i ied by hese heo ies, Peace e al. (2003) p oposed a model o e alua e he
impac on so wa e pi acy done by indi iduals in hei wo kplace. Each ac o iden i ied by
he expec ed u ili y heo y and de e ence heo y (punishmen se e i y, punishmen
ce ain y and so wa e cos ) was included as an an eceden o he a i ude ac o .
Punishmen ce ain y was also conside ed as an an eceden o pe cei ed beha io con ol.
Thei model was es ed using a s uc u al equa ion modeling
8
(SEM) echnique
called pa ial leas squa es (PLS) pa h modeling
9
and accoun ed o 65 pe cen o he
a iance (𝑅2) in so wa e pi acy in en ion. I showed ha a i ude, subjec i e no ms, and
pe cei ed beha io con ol signi ican ly in luence people’s in en ion. A i ude p esen ed
he s onges e ec on pi acy in en ion, and i s p edic ed an eceden s we e ound o ha e a
s ong ela ionship wi h a i ude, also he hypo hesis o punishmen ce ain y as a con ol
belie o pe cei ed beha io con ol was s ongly suppo ed. Simila esul s we e ound by
d’As ous e al. (2005) o online music pi acy, wi h all he ac o s de i ed om he TPB
ha ing a posi i e and s a is ically signi ican impac on he in en ion o engage in pi acy;
addi ionally pas pi acy beha io also had a s ong in luence on in en ion.
Analyzing ac o s ha a ec so wa e pi acy in en ions and i s subsequen esul on
beha io , Limayem e al. (2004) cons uc ed a model
10
based on T iandis’ beha io al
model (T iandis, 1979, c . Limayem e al. , 2004), and ound ha social ac o s, along wi h
pe cei ed consequences had a posi i e ela ionship wi h in en ion o pi a e so wa e, and
ha habi s and acili a ing condi ions a ec he ac ual so wa e pi acy beha io .
Su p isingly in en ions did no led o engagemen . A possible explana ion is ha in en ion
is being o e ide by habi s and acili a ing condi ions. Thei model analyzed using PLS
only explained 17 pe cen o he a iance in pi acy, wi h he au ho s de ending ha u he
esea ch is need be o e s a ing ques ioning p e ious esea ch.
Simila o Limayem e al. (2004), Phau e al. (2014) p oposed no only o iden i y
ac o s capable o in luencing in en ion, bu also he ac ual engagemen in digi al mo ies
8
A de ailed explana ion o SEM is p o ided on sec ion 3.
9
Also e e ed o as PLS-SEM.
10
Thei model elied on ollowing ac o s o explain he beha io al p ocess: habi , a ec , pe cei ed
consequences/belie s, social ac o s, acili a ing condi ions and in en ion.
9
pi acy
11
, using he TPB. The heo y was used in an unusual way, since a i ude owa d
beha io we e measu ed by one’s a i ude owa ds digi al pi acy o mo ies and hei mo al
judgmen . Subjec i e no ms we e measu ed by social habi and pe cei ed beha io con ol
by sel -e icacy.
Thei da a was analyzed using SEM
12
and hey ound ha om he TPB o iginal
de e minan s only a i ude owa ds digi al pi acy o mo ies p esen ed an unexpec ed esul ,
ha ing a nega i e impac on in en ion. Acco ding o he au ho s his may be a i ude being
o e ide by he posi i e in luence coming om mo al judgmen . Mo al judgmen , as
expec ed, had a nega i e impac on in en ion (suppo ing Tan (2002)) and engagemen .
Con a ily social habi posi i ely in luenced indi iduals o pi a e.
A las , bu no leas impo an , Phau’s e al. (2014) esea ch showed a posi i e (bu
weak) ela ion be ween in en ion and he ac ual ac o pi a ing digi al mo ies. This esul
clea ly shows he need o u he s udy he ela ion be ween in en ion and engagemen , a
leas in digi al pi acy. Since p e ious au ho s (e.g. Ajzen (1991)) ound in en ion as an
accu a e p edic o o beha io i sel .
Ano he heo y ha has been used o explain human beha io and so wa e pi acy in
pa icula is he equi y heo y
13
. Douglas e al. (2007) using ecip ocal ai ness, p ocedu al
ai ness
14
and dis ibu i e ai ness
15
as an eceden s o equi y ound ha he i s wo ac o s
we e signi ican de e minan s, and ha equi y had a nega i e and s a is ically signi ican
impac on so wa e pi acy, in o he wo ds, he highe he pe cei ed ai ness/jus ice o he
exchange by he consume , he lowe so wa e pi acy will be.
While many p e ious s udies ha e ocused on so wa e pi acy, o he s ha e
dedica ed hei a en ion o s udy di e en o ma s o digi al pi acy, such as music, mo ies,
11
An impo an limi a ion p esen ed by his esea ch is ha , he ac ual beha io was measu e h ough a p oxy.
12
Howe e hey ail o speci y he SEM echnique used, appea ing o be a co a iance-based SEM.
13
Equi y heo y add esses human pu sui o ai ness and jus ice in a social exchange (Douglas e al., 2007).
14
P ocedu al ai ness “is ep esen ed by he in ol emen and in e ac ion o he p oduce wi h he consume ”
(Douglas e al., 2007, p. 505).
15
Dis ibu i e ai ness “ ela es o pu chase o so wa e by di e en g oups o consume s” (Douglas e al.,
2007, p. 505), o example p ice discimina ion s a egies be ween consume s.
10
ideo games, and o he digi al media. An impo an cha ac e is ic o digi al goods is ha
hey “ha e high ini ial p oduc ion cos s, and e y low - app oaching ze o - ep oduc ion
cos s. They also ha e cha ac e is ics o a public good in ha sha ing wi h o he s does no
educe a consume ’s u ili y o he p oduc ” (Bha acha jee e al., 2003, p. 108). These
p ope ies acili a ed he widesp ead o pi a ed con en wo ldwide.
Bha ach jee e al. (2003) sugges ed ha music pi acy shows a numbe o
simila i ies wi h so wa e pi acy. Acco ding o his au ho , despi e he signi ican p ice
di e ence be ween so wa e and music albums, i is easonable o admi ha demand is
qui e elas ic o bo h, since inc easing he p ice o digi al ma e ial has a s ong posi i e
e ec on pi acy. Fu he mo e, wi h inc easingly highe in e ne connec ions consume ’s
p ice sensi i i y inc eases.
Gopal e al. (2004) sough ou o ha e a be e unde s anding o he beha io
dynamics ha d i e indi iduals o pi a e digi al audio iles, using he concep o pi acy club
size
16
as a p oxy o pi acy le el. They ound ha e hics has a e y s ong ela ionship wi h
club size (e hical indi idual’s will be less likely o sha e pi a ed iles), and ha jus ice is
posi i ely ela ed o e hics, bu ha ing a e y small e ec on club size. In addi ion, he
amoun o money sa ed by using pi a ed con en was a mode a ely s ong p edic o o
pi acy. The au ho concludes ha he high p ice o a p ope licensed audio CD is an
incen i e o pi acy, indica ing ha use s a e ex emely p ice-sensi i e when p esen ed wi h
he possibili y o illegally download an audio ile. Howe e , income did no in luence he
club size.
Thei esul s a e consis en wi h Bha ach jee e al. (2003) ha ound ha income
has a nega i e e ec on pi acy bu only o unknown songs, when he choice is made
ela i ely o a known music, income doesn’ a ec he decision. Fu he mo e hey indica e
ha he gene al e hical model o so wa e pi acy is b oadly applicable o digi al audio
pi acy.
16
Indi iduals wi h simila belie s join oge he o sha e unlicensed ma e ial, bene i ing om sha ing he cos s
incu ed when buying he p ope licensed ma e ial a ma ke p ice, which is hen dis ibu ion o all he club
membe s.

11
2.1.2. Digi al Pi acy Resea ch
Mo e ecen ly Al-Ra ee and C onan (2006), while examining ac o s ha in luence
an indi idual’s a i ude owa d pi a ing digi al ma e ial, ound ha subjec i e no ms
(in luence o signi ican o he s), cogni i e belie s abou he ou come o beha io , pe cei ed
impo ance o he issue, machia ellianism, age, happiness and exci emen we e all
signi ican p edic o s o a i ude. Mo al judgmen , dis ess, and sex we e no signi ican
a iables in luencing a i ude. Thei in es iga ion was suppo ed on he cons uc ha
a i ude is he mos signi ican ac o in luencing beha io al in en ion (e.g. T a imow and
K ys ina, 1996; Peace e al., 2003); he e o e a i ude owa d digi al pi acy was ea ed as a
dependen a iable. Acco ding o hem, unde s anding hese ac o s is impo an because
a i ude can be changed h ough pe suasion and o he means, making i possible o
in luence beha io (in an indi ec ashion). Thus, a be e unde s anding o hese ac o s
could be essen ial in lowe ing pi acy. This s udy also suppo s p e ious esea ch by
showing ha consume s belie e ha digi al ma e ial is o e p iced and ha hey will no be
caugh .
In 2008, he same au ho s (C onan and Al-Ra ee, 2008), using a s uden sample
om a business college, sough o analyze ac o s ha in luence an indi idual’s in en ion o
pi a e so wa e and media, a emp ing o o e a be e unde s anding o digi al pi acy
beha io . An eceden s o digi al pi acy beha io we e in es iga ed using an ex ended TPB
model, which included mo al obliga ion and pas pi acy beha io in addi ion o he o iginal
TPB de e minan s. I was (sepa a ely) hypo hesized ha indi iduals wi h highe a i ude,
subjec i e no ms, pe cei ed beha io al con ol and pas pi acy occu ences will co espond
wi h a g ea e in en ion o pi a e; on he con a y, highe mo al obliga ions co espond wi h
a lowe in en ion. The esul o he SEM analysis indica ed ha hei model explained 70.8
pe cen o he a iance in digi al pi acy in en ion, wi h only subjec i e no ms no being a
signi ican p edic o o in en ion.
Al-Ra ee and Dash i (2012) also a gue ha indi idual’s in en ion ega ding digi al
pi acy could change be ween cul u es. Using wo samples om di e en cul u es (Uni ed
S a es and Middle Eas ) hey de eloped a model expanding he TPB amewo k wi h mo al
12
obliga ion. The model (analyzed using PLS) p esen ed subs an ial explana o y powe in
bo h cul u es, wi h only he a iable subjec i e no ms in he U.S. no being a signi ican
p edic o o in en ion. As expec ed he a iables had a di e en impac on people’s
in en ion: in he U.S. sample, in en ion was s ongly a ec ed by hei abili y o pi a e and
mo al obliga ion, whe e in he Middle Eas sample one’s a i ude was he o emos
impo an ac o , ollowed by he abili y o pi a e.
Thei wo k shows ha cul u e may ha e a signi ican impac in in en ion, and
subsequen ly in indi iduals’ beha io when i comes o pi a e digi al media. I also
highligh s he need o s udy digi al pi acy ac oss di e en cul u es, since policies should be
adjus ed ( ine uning) o each coun y by go e nmen s and copy igh o ganiza ions.
This esea ch will examine digi al pi acy using he TPB as amewo k, since i
shown i sel as a eliable model o in es iga e beha io al in en ions associa ed wi h digi al
pi acy. Howe e TPB will be ex end using he expec ed u ili y heo y, he de e ence
heo y, as well as o he p o en beha io al cons uc s like e hics and pas beha io .
People s ill may ask why in es iga e digi al pi acy as a whole. The answe is ha
i ’s easonable o assume ha any indi idual capable o download a music ile is capable o
download any o he ype o ile. Al hough some migh say ha downloading so wa e and
ideo games is only hal o he job, because he nex s ep is o ins all hem, i is also ue
ha mos uploade s include u o ials ha each how o ins all he illegal ma e ial. Thus his
addi ional ba ie is easily o e aken. I is also e y common o an indi idual o ind in he
same websi e links/ o en s o download music, mo ies, so wa e and o he digi al ma e ial.
Finally, s o age capabili ies and in e ne connec ion speed, ba ie s poin ed in he pas as
de e en s o pi acy, a e no longe a p oblem (a leas in de eloped coun ies). E en mo e
he s o age ba ie is now o ally obli e a ed since he s eaming o unlicensed ma e ial like
music, mo ies and TV shows is becoming commonly used.
13
2.1.3. Summa y
The ollowing able syn hesizes he p esen ed pi acy esea ch.
Table 1: Pi acy beha io esea ch (sou ce: au ho )
E hical
Issue
Resea che
Fac o s In luencing
In en ion o Pi acy
Theo e ical
Unde pinning
Me hodology
Main Resul s
So wa e
Pi acy
Ch is ensen
and Eining
(1991)
A i ude
Subjec i e no ms
Teo y o easoned
ac ion
Chi-squa e
s a is ics
Mul iple
eg ession
analysis
A i ude owa d pi acy
and subjec i e no ms
we e di ec ly ela ed o
so wa e pi cy.
So wa e
Pi acy
Tan (2002)
Mo al in ensi y
Pe cei ed isk
Mo al judgmen
Res ’s ou -
componen model
Jones’ issue-
con ingen model
Two-s ep
hie a chical
eg ession
analysis
Pe cei ed isks and
mo al judgmen had a
nega i e impac on
in en ion.
So wa e
Pi acy
Peace e al.
(2003)
A i ude
Subjec i e no ms
Pe cei ed beha io al
con ol
Punishmen se e i y
Punishmen ce ain y
So wa e cos
Teo y o easoned
ac ion
Theo y o planned
beha io
Expec ed u ili y
heo y
De e ence heo y
PLS-SEM
𝑅2= 0.65; TPB
componen s p esen ed a
posi i e impac on
in en ion.
All he an icipa ed
hypo heses we e
suppo ed.
So wa e
Pi acy
Limayem e
al. (2004)
Habi
A ec
Pe cei ed
consequences/belie s
Social ac o s
Facili a ing
condi ions In en ion
T iandis’
beha io al model
PLS-SEM
Social ac o s and
pe cei ed
consequences had a
impac on in en ion.
Habi and acili a ing
condi ions had a
impac on he ac ual
beha io .
So wa e
Pi acy
Douglas e al.
(2007)
Recip ocal ai ness
P ocedu al ai ness
Dis ibu i e ai ness
Equi y
Equi y heo y
Co a iance
SEM
Equi y had a nega i e
and s a is ically
signi ican impac on
pi acy.
Music
Pi acy
Gopal e al.
(2004)
Age
Gende
E hical Index
Jus ice
Money Sa ed
Expec ed u ili y
heo y
De e ence heo y
Co a iance
SEM
Club size is posi i ely
in luenced by gende
and money sa ed,
while nega i ely
in luenced by he
emaining ac o s.
Music
Pi acy
d’As ous e
al. (2005)
A i ude
Subjec i e no ms
Pe cei ed beha io al
con ol
Pas beha io
Pe sonal
consequences
E hical
p edisposi ions
Theo y o planned
beha io
Mul iple
eg ession
analysis
Tes o
media ion
p ocedu e
TPB componen s
p esen ed a posi i e
impac on in en ion.
Pe sonal consequences
and e hical
p edisposi ions
p esen ed a nega i e
ela ionship wi h
a i ude.
Pas beha io showed a
posi i e ela ionship
wi h a i ude and
in en ion.
14
E hical
Issue
Resea che
Fac o s In luencing
In en ion o Pi acy
Theo e ical
Unde pinning
Me hodology
Main Resul s
Mo ies
Pi acy
Phau e al.
(2014)
A ec
A i ude
Mo al judgmen
Social habi
Sel -e icacy
In en ion
Theo y o planned
beha io
Neu alisa ion
heo y
SEM
A i ude and mo al
judgmen had a
nega i e impac on
in en ion. Mo al
judgmen also had a
nega i e impac on
engagemen .
The ac ual ac o
pi a ing was posi i ely
in luenced by social
habi and in en ion.
Digi al
Pi acy
Al-Ra ee and
C onan
(2006)
Subjec i e no ms
Cogni i e belie s
Pe cei ed impo ance
Machia ellianism
Age
Happiness and
exci emen
Mo al judgmen
Dis ess
Gende
Theo y o planned
beha io
S epwise
eg ession
analysis
𝑅2= 0.436
Only mo al judgmen ,
dis ess, and sex we e
no signi ican a iables
in luencing a i ude.
Digi al
Pi acy
C onan and
Al-Ra ee
(2008)
A i ude
Subjec i e no ms
Pe cei ed beha io al
con ol
Mo al obliga ion
Pas pi acy beha io
Teo y o easoned
ac ion
Theo y o planned
beha io
SEM
R2= 0.708
Only subjec i e no ms
we e no a signi ican
p edic o o in en ion.
Digi al
Pi acy
Al-Ra ee and
Dash i (2012)
A i ude
Subjec i e no ms
Pe cei ed beha io al
con ol
Mo al obliga ion
Teo y o easoned
ac ion
Theo y o planned
beha io
PLS-SEM
Only subjec i e no ms
in he U.S. sample we e
no a signi ican
p edic o o in en ion.
2.2. Model De elopmen : Theo e ical Founda ions and Hypo heses
2.2.1. Theo y o Planned Beha io
The heo y o planned beha io (Ajzen, 1985, 1991, 2002a) is a well known,
ecognized and empi ically suppo ed heo y o p edic ing in en ions and beha io
(A mi age and Conne , 2001). The heo y eme ged om he heo y o easoned ac ion
(Fishbein and Ajzen, 1975), which was designed o p edic beha io s ha a e unde
oli ional con ol, his is, beha io s ha a pe son can decide a will o pe o m. Howe e , i
is clea ha mos o he beha io s a e no unde oli ional con ol, in e nal ac o s (e.g.
in o ma ion, skills, abili ies, powe o will) and ex e nal ac o s (e.g. lack o ime and
oppo uni y) can comp omise in en ion and ul ima ely he beha io . In esponse o his
21
2.2.5. So wa e and Media Cos
I appea s ha economic incen i es play a majo ole in consume ’s beha io
decision, wi h so wa e and media p ice being a de e minan ac o . Al hough o he
economic ac o s like income, money sa ed and pe cei ed cos -bene i a e no he a ge o
his in es iga ion, hey we e used by p e ious esea che s and ep esen an impo an
insigh o a consume ’s decision p ocess.
So wa e pi acy a e was ound o ha e a signi ican nega i e co ela ion wi h pe
capi a GDP (and pe capi a GNP) mainly in poo coun ies, wi h in es iga o s inding an
in lec ion poin a USD 6000, whe e income le el below he in lec ion poin e eal a
s onge nega i e ela ion (Gopal and Sande s, 2000; Shin e al., 2004). Acco ding o Gopal
and Sande s (2000) his e eals an impo an p oblem: people wi h low income canno
a o d high so wa e p ices, hus pi acy is in luenced by he signi ican p ice di e en ial
be ween legal and pi a ed con en
17
. They p opose add ess his p oblem h ough global
p ice disc imina ion, which acco ding o hem is capable o maximize de elope ’s p o i s
18
and c ea e incen i es o go e nmen ac ion (e.g. en o cemen o copy igh laws). Peace e
al. (2003) also ound e idence suppo ing his ype o s a egies, wi h so wa e cos ha ing
a s ong posi i e ela ionship wi h one’s a i ude owa d pi acy.
I is hen expec ed ha so wa e p ice will ha e an impo an ole in he decision-
making p ocess, since so wa e packages usually a e he mos expensi e digi al goods, bu
su p isingly in music, p ice is also an impo an ac o . The highe he p ice, he s onge is
he posi i e e ec on pi acy, poin ing o a qui e elas ic demand, as in so wa e
(Bha acha jee el al., 2003; Gopal e al., 2004). In he mo ion pic u e indus y, consume ’s
pe cei ed cos -bene i has a posi i e impac on in en ion o buy pi a ed con en , indica ing
as well ha educing he p ices o mo ie DVDs would mos likely ha e a nega i e impac
on pi acy (Wang, 2005).
17
A he ime was usual o buy physical pi a ed con en (e.g. CDs) as opposed o downloading. These di ec
cos s a e he one’s e e ed by he au ho s, howe e oday pi a es usually download all hei unlicensed
con en om he in e ne , incu ing only in indi ec cos s, as ha ing a PC wi h in e ne connec ion (we belie e
hese a e indi ec cos s because he o dina y pe son will no p ima ily use hei PC o pi a e digi al ma e ial).
18
I de elope s make hei so wa e mo e a o dable, i ’s expec ed ha mo e people will buy i .

22
In a gene al way, consume s seem o belie e ha digi al media is o e p iced, using
pi acy as a mean o sa e money (AI-Ra ee and C onan, 2006). So i appea s ha e en when
he p ice o a digi al good is low, and p obably does no ep esen an economic bu den, i
s ill has an impac on he decision-making p ocess.
I pi acy beha io is modeled h ough he expec ed u ili y om choosing be ween
illegal download, pu chase, o do wi hou he digi al good, a a ional agen will choose he
u ili y unc ion
19
ha maximizes his expec ed u ili y. The e o e u ili y is used as a way o
desc ibe his p e e ences among he al e na i es, and he co esponden cha ac e is ics o
each al e na i e (Va ian, 2009). Conside ing he expec ed cos s and bene i s, he will selec
he al e na i e ha he belie es is associa ed wi h he mos desi able ou come. I pi acy
yields a posi i e su plus, despi e being nega i ely a ec ed by he isk inhe en o
punishmen ce ain y and se e i y (among o he ac o s), a lowe p ice would dec ease he
payo , ce e is pa ibus.
The cos o digi al ma e ial can be inco po a ed in o he TPB as an an eceden o
a i ude by he same easons appoin ed in he de e ence heo y. I is he e o e expec ed ha
a highe he inancial cos will co espond o a highe a i ude owa ds pi acy, due o he
highe expec ed payo . As such, can be hypo hesized ha :
H9: Digi al media cos will ha e a posi i e in luence on a i ude owa d pi a e
digi al ma e ials.
2.2.6. Pe cei ed Value
I is expec ed ha he highe he p ice he highe will be he a i ude owa ds pi acy,
howe e he pe cei ed p ice may no be enough o e alua e a digi al good, and in his way
ano he ac o was added o cap u e a b oade se o pe cei ed cha ac e is ics.
This ac o is pe cei ed alue, and helps us unde s and i consume s pe cei e digi al
goods as high alue p oduc s, ha a e wo hy o hei inancial cos , o on he o he end, he
19
We may look a his unc ion as an o dinal u ili y unc ion, howe e he u ili y unc ion only exis s i a
consume p e e ences espec he ollowing axioms: comple eness, e lexi i y, ansi i i y and con inui y. The
i s ee axioms ende he beha io o a a ional agen .
23
ime, e o and isk associa ed wi h pi a e hem. So wha is alue? When someone is
e alua ing he alue o a ce ain good, hey a e o ming hei own cons uc , hus pe cei ed
alue is an abs ac concep ha is highly pe sonal and indi idualis ic (Zei haml, 1998;Chu
and Lu, 2007). Zei haml (1998, p.14) de ined i as he “consume ’s o e all assessmen o
he u ili y o a p oduc based on pe cep ions o wha is ecei ed and wha is gi en”.
The e o e, i a consume belie es ha a p oduc has a low (o high alue), i is he ne esul
be ween he assessed gains (e.g. in insic a ibu es, olume, quali y) and sac i ices (e.g.
money, ime, e o ).
P e ious au ho s ha e s udied pe cei ed alue in e y di e se p oduc s o se ices,
and ound e idence o a posi i e ela ion be ween pe cei ed alue and consume
willingness- o-buy (o pu chase in en ions) (Dodds e al.1991, Chu and Lu, 2007).
Howe e , no one e e (a leas as a as we know) applied his concep o digi al pi acy and
so we may expec ha he highe he pe cei ed alue, he lowe will be one’s a i ude o
pi a e. The e o e, i is hypo hesized ha :
H10: Pe cei ed alue will ha e a nega i e in luence on a i ude owa d pi a e digi al
ma e ials.
I is also expec ed ha p ice will ha e an in luence on pe cei ed alue howe e , he
di ec ion (posi i e o nega i e) o ha in luence is unce ain. Dodds e al. (1991) old us
ha p ice has a double unc ion, i may se e has an indica o o sac i ice, leading o a
nega i e impac on he pe cei ed alue, and a he same ime can be an indica o o quali y,
since highe p ices lead o highe pe cei ed quali y and as a esul o a highe pe cei ed
alue. This adeo o ms an indi idual’s pe cep ion o alue, wi h he au ho s inding
some mixed esul s o he ela ion p ice-quali y, bu suppo o a nega i e ela ion
be ween p ice and a buye ’s pe cep ion o alue, as he p ice inc eases he pe cei ed alue
dec eases.
2.2.7. Concep ual Model
An easy and simple way o summa ize all he pos ula ed hypo heses is o obse e
he concep ual model p esen ed in he nex page. This concep ual model uly ep esen s
24
no one, bu wo models: a i s one will conside he ull sample, bu no e alua ing he
e ec o pas pi acy beha io in in en ion (Full Model); and a second one, ha has been
ob ained by adding pas pi acy beha io and, as consequence, will only conside s hose
who had pi a ed (Pi a e Model). The dashed pa h be ween pas pi acy beha io and
in en ion is mean o indica e exac ly his, since his ac o will only be in one o he
models.
Figu e 1: Concep ual Model. Expanded om Peace e al. (2003) and C onan and Al-Ra ee (2007).
25
3. Resea ch Me hodology
A e a e iew o pi acy esea ch and he s uc u al heo ies on which his wo k is
buil , i is now ime o b ing he concep ual model ou o he pape . This sec ion makes his
equi ed nex s ep, o pu i simply, answe s o he ollowing h ee ques ions: how was he
ques ionnai e de eloped; how he la en a iables ( ac o s) will be measu e using p ima y
da a; and how he collec ed da a will be analyzed.
This sec ion is di ided in h ee pa s. In i s one is co e ed he de elopmen o he
ques ionnai e, being highligh ed he measu ed ac o s as well he co esponding sou ces.
The second pa add esses da a collec ion and he las one explains a echnique called
s uc u al equa ion modeling. This echnique will be used o alida e he da a and elabo a e
he models hus, an in dep h look o SEM is indispensable.
3.1. Ques ionnai e
The da a used in his esea ch was collec ed using a ques ionnai e w i en in
Po uguese, which can be ound on Appendix A. I was also c ea ed an online e sion
20
o
acili a e dis ibu ion and each as many people as possible. In o de o a oid
misin e p e a ions he ini ial page explained wha digi al pi acy is, and how he
ques ionnai e should be illed.
Indi iduals we e asked o olun a ily pa icipa e, hei anonymi y and
con iden iali y being assu ed by he au ho . These aspec s had o be assu ed because digi al
pi acy is an illegal ac and his esea ch has a s ong e hical componen , as such hese
measu es may help o acili a e esponses bu also, and mo e impo an ly, u h ul ones.
These conce ns we e also e y impo an in he decision o no measu ing beha io i sel ,
bu using ins ead in en ion as a p oxy o hei p edic ed digi al pi acy beha io , since i
would be impossible o iden i y he esponden s o a ollow-up ques ionnai e.
20
The online e sion was iden ical o he pape e sion, bu wi h some isual modi ica ions o be e
accommoda e i o he online pla o m
26
To help ensu e measu emen eliabili y and alidi y, all he ac o s and
measu emen a iables used we e based on p e ious alida ed esea ch, as we can see on
Table 2, bu some adjus men s we e necessa y o con o m he indica o s o his esea ch.
A p elimina y e sion o he ques ionnai e was de eloped and p e- es ed in one
ocus g oup discussion
21
, as well dis ibu ed o indi iduals ha ga e hei eedback. This
was a necessa y and e y impo an s ep o ensu e ha esponden s unde s and all he
ques ions. O e all, he eedback was posi i e, wi h some punc ua ion and wo ds/sen ences
changed due o hei ambiguous s a emen . The ins uc ions o ill in he ques ionnai e also
eme ged om he p e- es , conside ing ha hose who we e no amilia wi h Like scales
did no unde s and immedia ely wha was being asked.
3.1.1. Measu ed Fac o s and Co esponden Sou ces
Following he hypo heses de eloped and he heo e ical ounda ions on ha hey a e
cons uc ed, i is ime o speci y how he unobse ed a iables p esen ed on he concep ual
model (Figu e 1) will be measu ed. All he ac o s and co esponden indica o s ha will be
used a e lis ed in Table 1, wi h all he i ems being sco ed on a se en-poin Like scale,
anging om “s ongly ag ee” o “s ongly disag ee” in almos all indica o s.
Table 2: Ques ionnai e ins umen scale ac o s
Fac o
Sou ce
No. o indica o s
Indica o loca ion
on ques ionnai e
In en ion (INT)
C onan and Al-Ra ee (2008);
Peace e al. (2003)
3
Page 3; Se 1
A i ude (ATT)
C onan and Al-Ra ee (2008)
4
Page 2; Se 1
Subjec i e No ms (SN)
C onan and Al-Ra ee (2008)
3
Page 3; Se 2
Pe cei ed Beha io al
Con ol (PBC)
C onan and Al-Ra ee (2008)
5
Page 2; Se 2
Mo al Obliga ion (MO)
C onan and Al-Ra ee (2008)
3
Page 5; Se 1
21
The ocus g oup discussion ook place in ea ly Ma ch, whe e ou s uden s colleagues pa icipa ed.

27
Fac o
Sou ce
No. o indica o s
Indica o loca ion
on ques ionnai e
Pas Pi acy Beha io (PPB)
C onan and Al-Ra ee (2008);
Au ho
2
Page 2; Se 3
Punishmen Se e i y (PS)
Peace e al. (2003)
2
Page 3; Se 3
Punishmen Ce ain y (PC)
Peace e al. (2003)
2
Page 4; Se 1
Digi al Media Cos (DMC)
Peace e al. (2003)
3
Page 4; Se 2
Pe cei ed Value (PV)
Dodds e al. (1991)
3
Page 4; Se 3
No e: The ques ionnai e can be ound on Appendix A
3.2. Da a
Da a was collec ed using an online ques ionnai e and a pape one. This decision
may ha e a biasing e ec on he esul s, howe e , i should be mino and negligible.
The URL o he online ques ionnai e was sen by e-mail o 28 715 s uden s o
Uni e si y o Po o, while he pape one was adminis e ed o 79 s uden s du ing egula
class ime in Ca azeda de Ansiães high school. The ques ionnai e was online du ing he
mon h o Ap il and he pape e sion was also dis ibu ed in he middle o he same mon h.
A o al o 590 ques ionnai es we e collec ed. F om hese, wen y-se en had missing da a
which led o a sample o 563 ques ionnai es wi h comple e da a.
The use o a s uden sample was deemed app op ia ed in he con ex o his esea ch
o ou main easons:
a) P e ious esea che s ha e shown ha digi al pi acy is gene alized
among he s uden s (Im and Van Epps, 1991; C onan and Al-Ra ee,
2008);
b) S uden s samples ha e been used in se e al pi acy s udies (Peace e
al., 2003; Gopal e al., 2004; Limayem e al., 2004; D'As ous e al.,
2005; Wang, 2005; AI-Ra ee and C onan, 2006; Lysonski and
Du asula, 2008; C onan and Al-Ra ee, 2008; Al-Ra ee and Dash i,
28
2012; Phau e al., 2014), hus using a s uden sample will acili a e
compa isons be ween s udies;
c) Today’s s uden s will be omo ow’s wo k o ce; and
d) Since i is di icul o use andom sampling me hodologies due o he
scope o he wo k, s uden s cons i u e a good a ge popula ion o
con enience sampling.
3.3. Es ima ion P ocedu e
S uc u al Equa ion Modeling (SEM) was used in his in es iga ion. SEM is a
echnique o “speci y, es ima e, and e alua e models o linea ela ionships among a se o
obse ed a iables in e ms o a gene ally smalle numbe o unobse ed a iables” (Shah
and Golds ein, 2006, p.149). Howe e we should no look o SEM as a echnique, bu
ins ead as se o ela ed p ocedu es design o e alua e how well a p oposed concep ual
model is consis en ( i s) wi h he da a (Kline, 2011). Fu he mo e, SEM allows mul iple
exogenous and endogenous a iables o be es ima ed simul aneously (Ande son and
Ge bing, 1988), his ep esen s a majo ad an age o e mul iple eg ession.
Why use SEM? SEM has been conside ed a be e (and bes sui ed) echnique o
heo y es ing and de elopmen han es ima ion me hods ha analyze a single equa ion a a
ime, because (when all he p e equisi es a e ul illed) he es ima ion me hods employed by
SEM p o ide a mo e e icien and consis en pa ame e es ima es; i also deals wi h he
o e all model i (Ande son and Ge bing, 1988; Kline, 2011). This echnique is also
commonly used in pi acy esea ch, o example, as we saw was used by Peace e al. (2003),
Limayem e al. (2004), Douglas e al. (2007), C onan and Al-Ra ee (2008), and Al-Ra ee
and Dash i (2012).
Obse ed a iables a e usually used as an indi ec measu e o unobse ed a iables,
and a e ypically e e ed o as an indica o (o measu emen a iable), while unobse ed
a iables a e no mally called la en a iables ( ac o s o esea ch cons uc s), and gene ally
co espond o hypo he ical cons uc s o ac o s (Ge en e al., 2000; Kline, 2011).
29
A SEM model combines a measu emen model and a s uc u al model. The
measu emen model (a con i ma o y ac o analysis model) is an a p io i model (de eloped
om heo e ical expec a ions) ha iden i ies he la en a iables and hei co esponden
indica o s (Ge en e al., 2000; Kline, 2011). The s uc u al model (a pa h model) ep esen s
he hypo hesized e ec p io i ies, howe e dissimila om pa h models hese e ec s can,
and usually in ol e la en a iables (Ge en e al., 2000; Kline, 2011).
The mos common SEM model, ha will be used in his wo k, is a s uc u al
eg ession model (SR model), also known as LISREL model. This is conside ed a
co a iance-based SEM (CB-SEM) (Ande son and Ge bing, 1988; Ge en e al., 2000),
whe e “model i ing o compa e he co a iance s uc u e
22
i o he esea che ’s model o a
bes possible i co a iance s uc u e” is used (Ge en e al., 2000, p. 26). The i be ween
he da a and he concep ual model is assessed h ough a se ies o model i es s, as he a io
o chi-squa e o deg ees o eedom, he goodness o i index (GFI), he adjus ed goodness
o i index (AGFI), and he oo mean esidual (RMR). On he o he hand (a he indi idual
pa h le el) cons uc alidi y and eliabili y a e assessed using con i ma o y ac o
analysis
23
(CFA) (see Ge en e al., 2000; Kline, 2011; Ma ôco, 2014).
The de aul me hod o es ima ion in SR models is he maximum likelihood (ML)
es ima ion, whe e “es ima es a e he ones ha maximize he likelihood ( he con inuous
gene a ion) ha he da a ( he obse ed co a iances) we e d awn om his popula ion”
(Kline, 2011, p.154). I is assumed ha a iables a e con inuous and no mally dis ibu ed.
Howe e , as his assump ion is equen ly elaxed, he a iables will be measu ed using a
Like scale. The e o e, i will be assumed (as Ma ôco, 2014) ha as long as he numbe o
ca ego ies o scale poin s used is high (a leas i e) and ha he dis ibu ion is close o a
no mal dis ibu ion hey can be ea ed as con inuous a iables.
22
Co a iance s uc u e is he pa o a SEM ha ep esen s hypo heses abou a iances and co a iances
(Kline, 2011).
23
The CFA should show con e gen alidi y and disc iminan alidi y, o he wise he measu emen model
mus be especi ied.
30
4. Resul s
As he i le abo e sugges s, i is now ime o ocus on esul s. This sec ion is di ided
in wo pa s, s a ing wi h a desc ip i e analysis. This ype o analysis despi e i s simplici y
is e y impo an , allowing o a sample o e iew. Fi s we ake a look a sample
demog aphics and hen pas pi acy beha io . Pas beha io was also segmen ed acco ding
o demog aphic cha ac e is ics.
The second pa is whe e he mul i a ia e analysis begins. An explo a o y ac o
analysis is ca ied ou ( o ha e a deepe iew a he da a) ollowed by SEM. To alida e he
measu emen models a con i ma o y ac o analysis is implemen ed. Gi en an accep able
measu emen model, he second s ep is o iden i y and speci y he s uc u al model. The
inal SR models (s uc u al model + measu emen model) a e hen p esen ed and e alua ed.
All he esul s we e ob ained using SPSS S a is ics 21 (essen ially o desc ip i e da a
analysis) and subsequen ly AMOS 21 ( o SEM).
4.1. Fi s Explo a o y Resul s
A i s desc ip i e analysis shows ha mo e han hal we e emale s uden s and
37.8% (213 s uden s) we e male, he a e age age was 23 yea s. The majo i y o he
s uden s (83.3%) we e ei he bachelo o mas e s uden s, and wi h 79.9% o he s uden s
e ealing ha , hey do no do any hing else besides s udying. Abou 75% o he s uden s
epo ed ha ing pi a ed p e iously, om hese 40.4% disclosed ha hey do pi a e a lo , and
25.4% does i in a daily base o almos daily. Table 3 ga he s all he p esen ed in o ma ion
and o e s a mo e de ailed iew.
The da a also shows ha almos 81% o he men admi ed o pi a e, while his
numbe was lowe o he women, bu s ill e y high (71.1%). Ano he in e es ing way o
look a pi acy pas beha io is o b eak i down by educa ion le el. Only 9.6% o he high
school s uden s admi ed ha hey ne e had pi a ed, which ep esen s he lowes alue o
all, as o he eaming (Doc o al, Mas e ’s and Bachelo ’s s uden s) hey all p esen ed
simila alues, be ween 25% and 27.6%.
37
Cons uc alidi y is used o assess i he used a iable uly measu e/ ep esen s he
cons uc ha we wan o e alua e (O’Lea y-Kell and Voku ka, 1998; Ma ôco, 2014). Since
ac o alidi y was al eady examined emains o es ablish con e gen and disc iminan
alidi y. The i s one occu s when indica o s load signi ican ly on hei co esponding
ac o s, his means ha he beha io o an indica o is essen ially explained by i s
co esponden ac o , he las one is a measu e o how unique each se o indica o s is, hus
disc iminan alidi y assess he co ela ions be ween he ac o s (Ma ôco, 2014).
Con e gen and disc iminan alidi y we e analyzed using he a e age a iance
ex ac ed (AVE) o each cons uc , as desc ibed in Fo nell and La cke (1981). Acco ding
o Hai e al. (1998) an AVE ≥ 0.5 is an adequa e indica o o con e gen alidi y, and as
we can see on Table 5 all cons uc s p esen ed and sui able AVE. On he o he end, we
ul ill he equi ed condi ion o disc iminan alidi y when he squa ed co ela ion be ween
wo ac o s is equal o lowe han he indi idual AVE o hem (Fo nell and La cke , 1981).
Compa ing he a e age a iance ex ac ed pe ac o wi h he co esponden squa ed
co ela ion alues on Table 6 we can see ha he p e ious condi ion is accomplished.
Table 6. Squa ed co ela ion be ween ac o s
Fac o s
Squa ed
Co ela ion
In en ion (INT)
and
A i ude (ATT)
0,309
In en ion (INT)
and
Pe cei ed Beha io al Con ol(PBC)
0,452
In en ion (INT)
and
Mo al Obliga ion (MO)
0,484
In en ion (INT)
and
Subjec i e No ms (SN)
0,490
In en ion (INT)
and
Pe cei ed Value (PV)
0,005
In en ion (INT)
and
Digi al Media Cos (DMC)
0,036
In en ion (INT)
and
Punishmen Se e i y (PS)
0,062
In en ion (INT)
and
Punishmen Ce ain y (PC)
0,158
A i ude (ATT)
and
Pe cei ed Beha io al Con ol(PBC)
0,172
A i ude (ATT)
and
Mo al Obliga ion (MO)
0,448
A i ude (ATT)
and
Subjec i e No ms (SN)
0,259
A i ude (ATT)
and
Pe cei ed Value (PV)
0,000
A i ude (ATT)
and
Digi al Media Cos (DMC)
0,036
A i ude (ATT)
and
Punishmen Se e i y (PS)
0,020
A i ude (ATT)
and
Punishmen Ce ain y (PC)
0,065
Pe cei ed Beha io al Con ol(PBC)
and
Mo al Obliga ion (MO)
0,266

38
Fac o s
Squa ed
Co ela ion
Pe cei ed Beha io al Con ol(PBC)
and
Subjec i e No ms (SN)
0,398
Pe cei ed Beha io al Con ol(PBC)
and
Pe cei ed Value (PV)
0,023
Pe cei ed Beha io al Con ol(PBC)
and
Digi al Media Cos (DMC)
0,026
Pe cei ed Beha io al Con ol(PBC)
and
Punishmen Se e i y (PS)
0,040
Pe cei ed Beha io al Con ol(PBC)
and
Punishmen Ce ain y (PC)
0,184
Mo al Obliga ion (MO)
and
Subjec i e No ms (SN)
0,423
Mo al Obliga ion (MO)
and
Pe cei ed Value (PV)
0,012
Mo al Obliga ion (MO)
and
Digi al Media Cos (DMC)
0,014
Mo al Obliga ion (MO)
and
Punishmen Se e i y (PS)
0,058
Mo al Obliga ion (MO)
and
Punishmen Ce ain y (PC)
0,141
Subjec i e No ms (SN)
and
Pe cei ed Value (PV)
0,022
Subjec i e No ms (SN)
and
Digi al Media Cos (DMC)
0,041
Subjec i e No ms (SN)
and
Punishmen Se e i y (PS)
0,115
Subjec i e No ms (SN)
and
Punishmen Ce ain y (PC)
0,270
Pe cei ed Value (PV)
and
Digi al Media Cos (DMC)
0,196
Pe cei ed Value (PV)
and
Punishmen Se e i y (PS)
0,002
Pe cei ed Value (PV)
and
Punishmen Ce ain y (PC)
0,006
Digi al Media Cos (DMC)
and
Punishmen Se e i y (PS)
0,000
Digi al Media Cos (DMC)
and
Punishmen Ce ain y (PC)
0,017
Punishmen Se e i y (PS)
and
Punishmen Ce ain y (PC)
0,231
4.2.1.2. SR Model
Gi en an accep able measu emen model, he second s ep is o iden i y and speci y
he s uc u al model, his ype o s a egy ( wo-s ep) helps ensu e ha he measu emen
model is co ec ly alida ed (Ma ôco, 2014). In his sec ion is in oduced he inal adjus ed
(using he modi ica ion indices) SR model (s uc u al model + measu emen model) and
e alua ed i s adjus men quali y using a se o i s a is ics.
39
Figu e 3. Final Full Sample SR Model. Pa h coe icien es ima es a e epo ed as s anda dized (∗∗p<0.01; ∗0.01≤p≤0.05; ns (no signi ican )p>0.05).
40
The o e all model i is sa is ac o y (𝑋2𝑑𝑓
⁄= 2.14; 𝐶𝐹𝐼 = 0.970; 𝐺𝐹𝐼 =
0.934;𝑅𝑀𝑆𝐸𝐴 = 0.046;𝑃[𝑟𝑚𝑠𝑒𝑎 < 0.05] = 0.890) howe e , i ails he model chi-
squa e es (𝜒2=447.852,𝑑𝑓 = 209, 𝑝 = 0.000). Acco ding o Ma ôco (2014) he 𝜒2
es is hea ily in luenced by he sample size (among o he ac o s, e.g. co ela ion be ween
obse ed a iables), so a model can be ejec ed despi e uly p esen ing a good adjus men
o he da a simply because o he sample size. When he sample p esen s a conside able
dimension (𝑛 > 400) he 𝜒2 es e y o en leads o he w ong conclusion, in o he wo ds,
i is e y likely o be signi ican (𝑝 < 0.05).This may be happening on he p esen ed model.
Because o his p oblem and o he s, esea che s ha e de eloped o he absolu e i indices
such as he goodness o i index (GFI) and oo mean e o o app oxima ion (RMSEA).
GFI p esen ed a good alue a 0.934, while RMSEA was accep able a 0.046. The ela i e
i index CFI (compa a i e i index) was 0.970, hus showing e idence o a good model i .
Analyzing each speci ic pa h in Figu e 3, we can see ha h ee o he pa hs we e no
signi ican (5% was conside ed as he c i ical le el o signi icance), hese we e:
"Punishmen Ce ain y → A i ude" (𝛽𝐴𝑡𝑡.𝑃𝐶 = 0.009;p = 0.846);
"Punishmen Se e i y→ A i ude"(𝛽𝐴𝑡𝑡.𝑃𝑆 = 0.021;p = 0.624); and
"A i ude→ In en ion"(𝛽𝐼𝑛𝑡.𝐴𝑡𝑡 = 0.087;p = 0.089).
The emaining TPB componen s had a signi ican bu mode a ed e ec on
in en ion(𝛽𝐼𝑛𝑡.𝑆𝑁 = 0.257;p < 0.01; 𝛽𝐼𝑛𝑡.𝑃𝐵𝐶 = 0.358;p < 0.01), wi h mo al obliga ion
also ha ing a signi ican bu mode a ed e ec on in en ion(𝛽𝐼𝑛𝑡.𝑀𝑂 = −0.307;p < 0.01).
A i ude had a signi ican bu mode a ed e ec on pe cei ed beha io al con ol(𝛽𝑃𝐵𝐶.𝐴𝑡𝑡 =
0.349;p < 0.01), and a i ude emaining an eceden s exhibi ed a signi ican bu small
e ec (𝛽𝐴𝑡𝑡.𝑃𝑉 = 0.118; p = 0.016; 𝛽𝐴𝑡𝑡.𝐷𝑀𝐶 = 0.155;p < 0.01). Punishmen ce ain y
had a mode a ed e ec on pe cei ed beha io al con ol(𝛽𝑃𝐵𝐶.𝑃𝐶 = −0.348;p < 0.01).
Finally, he pa h "Mo al Obliga ion → A i ude" was he one ha p esen ed he highe pa h
coe icien (𝛽𝐴𝑡𝑡.𝑀𝑂 = −0.699;p < 0.01).
The inal ull model explains 63% o he a iance in digi al pi acy in en ion.
41
4.2.2. Pi a e Model
This second model conside s only hose s uden s who had pi a ed and he e o e, he
sample was smalle adding up o 421 en ies.
4.2.2.1. Measu emen Model
The inal CFA pi a e model is p esen ed in Figu e 4, while he ini ial one is
p esen ed in Appendix C. To ge o his inal model we ye again ollow Ma ôco (2014).
The i s s ep was o analyze ac o alidi y, un o una ely he same h ee indica o s
(SN2, PV1, DMC3) ailed again o ul ill he equi ed condi ions and we e emo ed om
he model. The Skew and Ku osis coe icien s showed adequa e alues ha made possible
o admi a no mal dis ibu ion o almos all obse ed a iables, he excep ion was PBC4
and consequen ly was emo ed. The exis ence o ou lie s assessed by Mahalanobis squa e
dis ance displayed i e cases wi h alues sugges ing ha hese we e ou lie s, so he CFA
was conduc ed wi hou hem.
The model was hen adjus ed using he modi ica ion indices (g ea e han 11), a se
o sugges ed modi ica ions we e ela ed o he co a iance be ween he e o e ms o he
ollowing indica o s: ATT1 (e4) and ATT3 (e6); ATT3 (e6) and ATT4 (e7); MO1 (e13)
and MO3 (e15). The sugges ed ajec o ies we e added o he model since ha all ela ions
a e be ween i ems ha load on he same ac o . The co ela ion be ween he e o s may be
occu ing because o he simila i y o wo ding and con en .
The i be ween he da a and he inal CFA pi a e model was o e all conside ed as
good (𝑋2𝑑𝑓 = 1.761;
⁄𝐶𝐹𝐼 = 0.969; 𝐺𝐹𝐼 = 0.926;𝑅𝑀𝑆𝐸𝐴 = 0.043;𝑃[𝑟𝑚𝑠𝑒𝑎 <
0.05] = 0.966;𝑀𝐸𝐶𝑉𝐼 = 1.580 ).
Es ablished a good model i , cons uc eliabili y, con e gen and disc iminan
alidi y we e analyzed. All ac o s p esen ed an adequa e eliabili y and demons a ed
con e gen as well disc iminan alidi y (see Appendix D).
42
Figu e 4. Final CFA Pi a e Model (𝐗𝟐𝐝𝐟 = 𝟏.𝟕𝟔𝟏;
⁄𝐂𝐅𝐈 = 𝟎.𝟗𝟔𝟗; 𝐆𝐅𝐈 = 𝟎. 𝟗𝟐𝟔;𝐑𝐌𝐒𝐄𝐀 = 𝟎.𝟎𝟒𝟑;𝐏[𝐫𝐦𝐬𝐞𝐚 <
𝟎.𝟎𝟓] = 𝟎. 𝟗𝟔𝟔;𝐌𝐄𝐂𝐕𝐈 = 𝟏. 𝟓𝟖𝟎 ).

43
4.2.2.2. SR Model
The inal SR pi a e model (Figu e 5) e ealed a sa is ac o y model i (𝜒2=
447.852, 𝑑𝑓 = 209,𝑝 = 0.000; 𝑋2𝑑𝑓
⁄= 1.960; 𝐶𝐹𝐼 = 0.957; 𝐺𝐹𝐼 = 0.910; 𝑅𝑀𝑆𝐸𝐴 =
0.048;𝑃[𝑟𝑚𝑠𝑒𝑎 < 0.05] = 0.694). Howe e , when we examine each speci ic pa h we can
see ha six pa hs we e no signi ican , hese we e:
i. "Pe cei ed Value → A i ude" (𝛽𝐴𝑡𝑡.𝑃𝑉 = 0.096;p = 0.078);
ii. "Digi al Media Cos → A i ude" (𝛽𝐴𝑡𝑡.𝐷𝑀𝐶 = 0.083;p = 0.083);
iii. "Punishmen Ce ain y → A i ude" (𝛽𝐴𝑡𝑡.𝑃𝐶 = 0.056;p = 0.279);
i . "Punishmen Se e i y→ A i ude"(𝛽𝐴𝑡𝑡.𝑃𝑆 = −0.006;p = 0.897);
. "A i ude→ In en ion"(𝛽𝐼𝑛𝑡.𝐴𝑡𝑡 = 0.042;p = 0.465); and
i. "Subjec i e No ms → In en ion" (𝛽𝐼𝑛𝑡.𝑆𝑁 = 0.084;p = 0.118).
The decision o use he s anda d le el o signi icance ins ead o 10% culmina ed in
he ejec ion o some hypo hesis ha o he wise would no be ejec ed, his is ue o bo h
models and i is impo an o ake in accoun when compa ing he conclusions he e
p esen ed wi h o he au ho s.
The emaining TPB componen had a signi ican bu weak e ec on
in en ion(𝛽𝐼𝑛𝑡.𝑃𝐵𝐶 = 0.124;p = 0.003). Mo al obliga ion had a signi ican bu mode a ed
e ec on in en ion(𝛽𝐼𝑛𝑡.𝑀𝑂 = −0.304;p < 0.01), while pas pi acy beha io p esen ed a
subs an ial e ec on in en ion (𝛽𝐼𝑛𝑡.𝑃𝑃𝐵 = 0.490; p < 0.01). Punishmen ce ain y and pas
pi acy beha io had a signi ican e ec on pe cei ed beha io al con ol, he i s ha ing a
mode a ed e ec and he second a signi ican
one (𝛽𝑃𝐵𝐶.𝑃𝐶 = −0.237;p < 0.01; 𝛽𝑃𝐵𝐶.𝑃𝑃𝐵 = 0.490;p < 0.01). Finally, he pa h
"Mo al Obliga ion → A i ude" was he one ha p esen ed he highe pa h
coe icien (𝛽𝐴𝑡𝑡.𝑀𝑂 = −0.712;p < 0.01).
The inal pi a e model explains 70% o he a iance in digi al pi acy in en ion.
44
Figu e 5. Final Pi a e SR Model. Pa h coe icien es ima es a e epo ed as s anda dized (∗∗p<0.01; ∗0.01≤p≤0.05; ns (no signi ican )p>0.05)
45
5. Discussion and Implica ions
5.1. TPB Va iables, Mo al Obliga ion and Pas Pi acy Beha io
A i ude owa d he beha io is a pe sonal ac o ha e alua es an indi idual’s
p edisposi ion owa d pe o ming digi al pi acy. I was hypo hesized ha indi iduals wi h a
mo e posi i e a i ude owa ds pi acy will co espond o a g ea e in en ion o pi a e digi al
ma e ials. Howe e , con a y o expec a ions a i ude was no a signi ican p edic o o
in en ion in bo h models, as so hypo hesis H1 is ejec ed. This may be due o he in luence
o mo al obliga ion, which had a s ong nega i e e ec on a i ude in bo h models and
migh diminished a i ude’s posi i e e ec on in en ion and co esponden signi icance.
This e ec was no expec ed, bu i makes sense, sugges ing ha i someone iews digi al
pi acy as mo ally w ong, hen his a i ude would be nega i ely in luenced.
This esul does no suppo p e ious esea che s ha ha e ound a i ude a
signi ican p ecu so o in en ion (see o example Peace e al., 2003; D'As ous e al., 2005;
Al-Ra ee and Dash i, 2012), p esen ing a mode a ed/high e ec on i . Usually, is conside ed
ha by al e ing a i ude i should be possible o educe pi acy, hus making a i ude a e y
impo an a iable in he igh agains pi acy and making i mo e di icul since a i ude was
no a signi icance ac o . I may also indica e ha indi iduals pi a e despi e p esen ing an
un a o able a i ude owa d digi al pi acy.
The emaining TPB componen s in he ull sample model p esen ed he expec ed
ou come. I was hypo hesized ha indi iduals wi h a highe le el o subjec i e no ms
suppo i e/(pe cei ed con ol o e pe o mance) o pi acy will co espond o a g ea e
in en ion o pi a e digi al ma e ials. The esul s showed ha subjec i e no ms and pe cei ed
beha io al con ol had a signi ican bu mode a ed e ec on in en ion. As such, hypo heses
H2 and H3 a e no ejec ed, and we conclude ha : i) he app o al o digi al pi acy by
iends, amily (o any signi ican o he s) posi i ely a ec he indi idual’s in en ion; ii) ha
subjec s ha ind easy o pi a e and ha e he oppo uni y o do so, will mos likely ha e a
g ea e in en ion o pi a e digi al ma e ials. The pi a e model yield a simila esul ega ding
pe cei ed beha io al con ol, bu he o he a iable, subjec i e no ms, was no a signi ican
46
p edic o o in en ion. Thus, i is possible ha hose who ha e pi a ed be o e may no be
in luenced by pe cei ed social p essu es.
Hypo hesis H4 s a es ha he highe he eeling o mo al obliga ion, he lowe is an
indi idual in en ion o pi a e digi al ma e ials. Examining he esul s, his hypo hesis is no
ejec ed o bo h models, wi h mo al obliga ion ha ing a signi ican and nega i e e ec on
in en ion. This nega i e ela ion enables o conclude ha indi iduals wi h a highe sense o
mo ali y will end o ha e a lowe in en ion owa ds pi a ing. As we can see, i appea s ha
mo al obliga ion and pe cei ed beha io al con ol play a key ole in digi al pi acy, being
signi ican p edic o s o in en ion in bo h models, making he connec ion be ween hem and
p esen ing hemsel es as he ideal ac o s o “a ack”. A possible app oach is o use an
indi idual’s mo al obliga ion o eelings o guil o show ha pi acy is no only a ec ing
company’s ea nings bu ul ima ely is a majo issue o he whole socie y wi h all o us
losing, no allowing mo e jobs (o e en des oying cu en one’s) and axes ha could be
used o di ec ly imp o e people’s li es.
A las , i was hypo hesized ha he e is a posi i e ela ionship be ween pas pi acy
beha io and in en ion. This was indeed ue, wi h pas pi acy beha io p esen ing a
subs an ial e ec on in en ion, hypo hesis H5 was no ejec ed. As so, i is expec ed ha
indi iduals ha pi a ed digi al ma e ial in he pas a e mo e likely o incu in he same
in en ions. Pas pi acy beha io also e ealed a signi ican and s ong posi i e ela ion wi h
pe cei ed beha io al con ol, his ela ion shows ha wi h expe ience we ge com o able
doing a ce ain ask, ou sense o con ol ge s highe . Indeed, 40.4% o he s uden s
disclosed ha hey pi a e a lo , and 25.4% does i in a daily base o almos daily, all his
shows ha pas beha io has a s ong and de e minan in luence on con ol and in en ion.
Nowadays we can access he in e ne i ually anywhe e and download wha e e
we wan o e en gi e o de s o ou compu e a home o s a a download, making pi a ing
so easy ha can become ecu en and ul ima ely a habi . This makes pas pi acy beha io a
e y di icul ac o o add ess. A sugges ion is o es ic he numbe o places whe e
people can access websi es ha acili a e he download o s eaming o pi a e con en . Fo
53
8. Re e ences
AI-Ra ee, S. and C onan, T. P. (2006), "Digi al Pi acy: Fac o s ha In luence A i ude
Towa d Beha io ", Jou nal o Business E hics, Vol. 63 No. 3, pp. 237-259.
Ajzen, I. (1985), "F om in en ions o ac ions: A heo y o planned beha io ", in Kuhl, J.
and Beckman, J. (Eds.), Ac ion-con ol: F om cogni ion o beha io . Sp inge ,
Heidelbe g, Ge many, pp. 11- 39.
Ajzen, I. (1991), "The heo y o planned beha io ", O ganiza ional Beha io and Human
Decision P ocesses, Vol. 50, pp. 179-211.
Ajzen, I. (2002a), "Pe cei ed beha io al con ol, sel -e icacy, locus o con ol, and he
heo y o planned beha io ", Jou nal o Applied Social Psychology, Vol. 32 No. 4,
pp. 665-683.
Ajzen, I. (2002b), "Residual e ec s o pas on la e beha io : Habi ua ion and easoned
ac ion pe spec i es", Pe sonali y and Social Psychology Re iew, Vol. 6 No. 2, pp.
107- 122.
Ajzen, I. (2014), "The heo y o planned beha iou is ali e and well, and no eady o e i e:
a commen a y on Snieho a, P esseau, and A aújo-Soa es", Heal h Psychology
Re iew.
Ajzen, I. and Madden, T. (1986), "P edic ion o goal-di ec ed beha io : A i udes,
in en ions, and pe cei ed beha io al con ol", Jou nal o Expe imen al Social
Psychology, Vol. 22, pp. 453-474.
Al-Ra ee, S. and Dash i, A. (2012), "A c oss cul u al compa ison o he ex ended TPB: The
case o digi al pi acy", Jou nal o Global In o ma ion Technology Managemen, Vol.
15 No. 1, pp. 5-24.
Ande son, J. C. and Ge bing, D. W. (1988), "S uc u al equa ion modeling in p ac ice: A
e iew and ecommended wo-s ep app oach", Psychological bulle in, Vol. 103 No.
3, pp. 411.
A mi age, C. J. and Conne , M. (2001), "E icacy o he Theo y o Planned Beha iou : A
me a-analy ic e iew", B i ish Jou nal o Social Psychology, Vol. 40 No. 4, pp.
471-499.

54
Becke , G. S. (1968), "C ime and Punishmen : An Economic App oach", Jou nal o
Poli ical Economy, Vol. 76 No. 2, pp. 169-217.
Becke , J. U. and Clemen , M. (2006), "Dynamics o Illegal Pa icipa ion in Pee - o-Pee
Ne wo ks—Why Do People Illegally Sha e Media Files?", Jou nal o Media
Economics, Vol. 19 No. 1, pp. 7-32.
Ben le , P. and Specka , G. (1979), "Models o a i ude-beha io ela ions", Psychological
Re iew, Vol. 86 No. 5, pp. 452-464.
Bha acha jee, S., Gopal, R. and Sande s, G. (2003), "Digi al music and online sha ing:
So wa e pi acy 2.0?", Communica ion o he ACM, Vol. 46 No. 7, pp. 107-111.
BSA (2012), "Shadow Ma ke : 2011 BSA Global So wa e Pi acy S udy", in. Business
So wa e Alliance.
BSA (2014), "The Compliance Gap: BSA Global So wa e Su ey", in. Business So wa e
Alliance.
Came on, S. (1988), "The Economics o C ime De e ence: A Su ey o Theo y and
E idence", Kyklos, Vol. 41 No. 2, pp. 301-323.
Cen o de Es udos Aplicados da Uni e sidade Ca ólica Po uguesa (2012), "Impac o
Económico da Pi a a ia In o má ica em Po ugal".
Ch is ensen, A. and Eining, M. (1991), "Fac o s in luencing so wa e pi acy: Implica ions
o accoun an s", Jou nal o In o ma ion Sys ems, Vol. 5 No. 1, pp. 67-80.
Chu, C., Lu, H. (2007), "Fac o s in luencing online music pu chase in en ion in Taiwan: An
empi ical s udy based on he alue‐in en ion amewo k", In e ne Resea ch, Vol.
17 No. 2, pp. 139 – 155.
Conne , M. and A mi age, C. (1998), "Ex ending he Theo y o Planned Beha io : A
Re iew and A enues o Fu he Resea ch", Jou nal o Applied Social Psychology,
Vol. 28 No. 15, pp. 1429–1464.
C onan, T. P. and Al-Ra ee, S. (2008), "Fac o s ha In luence he In en ion o Pi a e
So wa e and Media", Jou nal o Business E hics, Vol. 78 No. 4, pp. 527-545.
D'As ous, A., Colbe , F. and Mon pe i , D. (2005), "Music pi acy on he web – how
e ec i e a e an i-pi acy a gumen s? E idence om he heo y o planned
beha iou ", Jou nal o Consume Policy, Vol. 28, pp. 289-310.
55
Danahe , B., Smi h, M. and Telang, R. (2014), "Pi acy and Copy igh En o cemen
Mechanisms", Inno a ion Policy and he Economy, Vol. 14 No. 1, pp. 25-61.
Dansby, A. (2002), "Naps e Files o Bank up cy", a ailable a :
h p://www. ollings one.com/music/news/naps e - iles- o -bank up cy-20020604
(accessed Decembe 2 2014).
De Vany, A. and Walls, W. (2007), "Es ima ing he E ec s o Mo ie Pi acy on Box-o ice
Re enue", Re iew o Indus ial O ganiza ion, Vol. 30 No. 4, pp. 291-301.
Dodds, W. B., Mon oe, K. B., and G ewal, D. (1991), "E ec s o P ice, B and, and S o e
In o ma ion on Buye s' P oduc E alua ions", Jou nal o Ma ke ing Resea ch, pp.
307-319.
Douglas, D., C onan, T. and Behelb, J. (2007), "Equi y pe cep ions as a de e en o
so wa e pi acy beha io ", In o ma ion & Managemen , Vol. 44 No. 5, pp. 503-512.
Eh lich, I. (1973), "Pa icipa ion in illegi ima e ac i i ies: A heo e ical and empi ical
in es iga ion", The Jou nal o Poli ical Economy, pp. 521-565.
Eh lich, I. (1996), "C ime, Punishmen , and he Ma ke o O enses", The Jou nal o
Economic Pe spec i es, Vol. 10 No. 1, pp. 43-67.
E nes o. (2013), ""The Hobbi " Mos Pi a ed Film o 2013", a ailable a :
h p:// o en eak.com/ he-hobbi -mos -pi a ed- ilm-o -2013-131231/, accessed
Decembe 2 2014.
Fishbein, M. and Ajzen, I. (1975), Belie , A i ude, In en ion, and Beha io : An
In oduc ion o Theo y and Resea ch, MA: Addison-Wesley, Reading.
Fo nell, C., La cke , D. (1981), "E alua ing S uc u al Equa ion Models wi h Unobse able
Va iables and Measu emen E o ", Jou nal o Ma ke ing Resea ch, Vol. 18 No. 1,
pp. 39-50.
Ge en, D., S aub, D. and Boud eau, M.-C. (2000), "S uc u al equa ion modeling and
eg ession: Guidelines o esea ch p ac ice", Communica ions o he associa ion o
in o ma ion sys ems, Vol. 4 No. 1, pp. 7.
Ge bing, W., and Ande son, J. (1988), "An upda ed pa adigm o scale de elopmen
inco po a ing unidimensionali y and i s assessmen ", Jou nal o Ma ke ing
Resea ch, Vol.25, pp. 186-192.
56
Gopal, R., Sande s, G., Bha acha jee, S., Ag awal, M. and Wagne , S. (2004), "A
Beha io al Model o Digi al Music Pi acy", Jou nal o O ganiza ional Compu ing
and Elec onic Comme ce, Vol. 14 No. 2, pp. 89-105.
Gopal, R. and Sande s, L. (1997), "P e en i e and de e en con ols o so wa e pi acy",
Jou nal o Managemen In o ma ion Sys ems, Vol. 13 No. 4, pp. 29-47.
Gopal, R. and Sande s, L. (1998), "In e na ional so wa e pi acy: Analysis o key issues
and impac s", In o ma ion Sys ems Resea ch, Vol. 9 No. 4, pp. 380–397.
Gopal, R. and Sande s, L. (2000), "Global so wa e pi acy: You can’ ge blood ou o a
u nip", Communica ions o he ACM, Vol. 43 No. 9, pp. 83-89.
Go such, R. and O be g, J. (1983), "Mo al obliga ion and a i udes: Thei ela ion o
beha io al in en ions", Jou nal o Pe sonali y and Social Psychology, Vol. 44 No.
5, pp. 1025-1028.
Hai , J. F., Ande son, R. E., Ta ham, R. L. and Black, W. C. (1998), Mul i a ia e Da a
Analysis (5 h ed.). Englewood Cli s: P en ice-Hall.
Hai , J. F., Ringle, C. M. and Sa s ed , M. (2012), "Pa ial Leas Squa es: The Be e
App oach o S uc u al Equa ion Modeling?", Long Range Planning, Vol. 45 No. 5–
6, pp. 312-319.
Im, J. H. and Van Epps, P. D. (1991), "So wa e pi acy and so wa e secu i y in business
schools: an e hical pe spec i e", ACM SIGMIS Da abase, Vol. 22 No. 3, pp. 15-22.
Jackson, D. L. (2003), "Re isi ing Sample Size and Numbe o Pa ame e Es ima es: Some
Suppo o he N:q Hypo hesis", S uc u al Equa ion Modeling: A
Mul idisciplina y Jou nal, Vol. 10 No. 1, pp. 128-141.
Jacobs, R., Heu elman, A., Tan, M. and Pe e s, O. (2012), "Digi al mo ie pi acy: A
pe spec i e on downloading beha io h ough social cogni i e heo y", Compu e s
in Human Beha io , Vol. 28 No. 3, pp. 958-967.
Jö eskog, K. G. (1990), "New de elopmen s in LISREL: Analysis o o dinal a iables
using polycho ic co ela ions and weigh ed leas squa es", Quali y & Quan i y, Vol.
24 No. 4, pp. 387-404.
Kline, R. B. (2011), P inciples and P ac ice o S uc u al Equa ion Modeling, Guil o d
P ess.
57
Liebowi z, S. J. (2007), "How Reliable is he Obe holze -Gee and S ump Pape on File-
Sha ing?", in. Uni e si y o Texas a Dallas.
Liebowi z, S. J. (2010), "The Obe holze -Gee/S ump File-Sha ing Ins umen Fails he
Laugh Tes ", in. Uni e si y o Texas a Dallas.
Limayem, M., Khali a, M. and Chin, W., Wynne. (2004), "Fac o s Mo i a ing So wa e
Pi acy: A Longi udinal S udy", IEEE T ansac ions on Enginee ing Managemen ,
Vol. 51, pp. 414-425.
Lysonski, S. and Du asula, S. (2008), "Digi al pi acy o MP3s: consume and e hical
p edisposi ions", Jou nal o Consume Ma ke ing, Vol. 25 No. 3, pp. 167-178.
Ma ôco, J. (2014), Análise de Equações Es u u ais: Fundamen os eó icos, So wa e &
Aplicações, Repo Numbe , Lda.
Mason, R. (1986), "Fou E hical Issues o he In o ma ion Age", Managemen In o ma ion
Sys ems Qua e ly, Vol. 10 No. 1, pp. 5-12.
Mic oso . (n.d.), "Reg esso às Aulas", a ailable a :
h p://www.mic oso s o e.com/s o e/mseea/p _PT/ca /S uden /ca ego yID.6623640
0, accessed 28 Sep embe 2015.
Nagin, D. S. and Poga sky, G. (2001), "In eg a in Cele i y, Impilsi i y, and Ex alegal
Sanc ion Th ea s In o a Model o Gene al De e ence: Theo y and E idence",
C iminology, Vol. 39 No. 4, pp. 865-892.
Neglia, G., Reina, G., Honggang, Z., Towsley, D., Venka a amani, A. and Danahe , J.
(2007), "A ailabili y in Bi To en Sys ems", in INFOCOM 2007. 26 h IEEE
In e na ional Con e ence on Compu e Communica ions. IEEE, pp. 2216-2224.
Obe holze ‐Gee, F. and S ump , K. (2007), "The E ec o File Sha ing on Reco d Sales:
An Empi ical Analysis", Jou nal o Poli ical Economy, Vol. 115 No. 1, pp. 1-42.
O'Lea y-Kelly, S., Voku ka, R. (1998), "The empi ical assessmen o cons uc alidi y",
Jou nal o Ope a ions Managemen , Vol. 16, pp. 387-405.
Peace, A., Galle e, D. and Thong, J. (2003), "So wa e Pi acy in he Wo kplace: A Model
and Empi ical Tes ", Jou nal o Managemen In o ma ion Sys ems, Vol. 20, pp.
153-177.
58
Phau, I., Lim, A., Liang, J. and Lwin, M. (2014), "Engaging in digi al pi acy o mo ies: a
heo y o planned beha iou app oach", In e ne Resea ch, Vol. 24 No. 2, pp. 246-
266.
Ramos, J. (2015), "Ne lix em Po ugal: como, quando e po quê", a ailable a :
h p://exp esso.sapo.p /economia/2015-06-12-Ne lix-em-Po ugal-como-quando-e-
po que, accessed Sep embe 2 2015.
Shah, R. and Golds ein, S. M. (2006), "Use o s uc u al equa ion modeling in ope a ions
managemen esea ch: Looking back and o wa d", Jou nal o Ope a ions
Managemen , Vol. 24 No. 2, pp. 148-169.
Shin, S., Gopal, R., Sande s, G. and Whins on, A. (2004), "Global so wa e pi acy
e isi ed", Communica ions o he ACM, Vol. 47 No. 1, pp. 103-107.
Siwek, S. (2007), "The T ue Cos o Copy igh Indus y Pi acy o he U.S. Economy", in.
Ins i u e o Policy Inno a ion.
Smi h, M. D. and Telang, R. (2010), "Pi acy o p omo ion? The impac o b oadband
In e ne pene a ion on DVD sales", In o ma ion Economics and Policy, Vol. 22 No.
4, pp. 289-298.
Snieho a, F. F., P esseau, J. and A aújo-Soa es. (2014), "Time o e i e he heo y o
planned beha io ", Heal h Psychology Re iew, Vol. 8 No. 1, pp. 1-7.
Spo i y. (n.d.), "In o ma ion", a ailable a : h ps://p ess.spo i y.com/p /in o ma ion/,
accessed 2 Sep embe 2015.
Sudle , H. (2013), "E ec i eness o an i-pi acy echnology: Finding app op ia e solu ions
o e ol ing online pi acy", Business Ho izons, Vol. 56 No. 2, pp. 149-157.
Tan, B. (2002), "Unde s anding consume e hical decision making wi h espec o pu chase
o pi a ed so wa e", The Jou nal o Consume Ma ke ing, Vol. 19 No. 2, pp. 96-
111.
T a imow, D. and K ys ina, F. (1996), "The impo ance o subjec i e no ms o a mino i y
o people: Be ween subjec s and wi hin-subjec s analyses", Pe sonali y and Social
Psychology Bulle in, Vol. 22 No. 8, pp. 820-828.

59
T iandis, H. C. (1979), "Values, a i udes, and in e pe sonal beha io ",in P oc. Neb aska
Symp. Mo i a ion, 1979: Belie s, A i udes and Values, Lincon, NE, 1980, pp. 159-
295.
Va ian, H. R. (2009), In e media e Mic oeconomics: A Mode n App oach, W.W. No on &
Company.
Wang, C. (2005), "Fac o s ha in luence he pi acy o DVD/VCD mo ion pic u es",
Jou nal o Ame ican Academy o Business, Vol. 6 No. 1, pp. 231-237.
Williams, K. R. and Hawkins, R. (1986), "Pe cep ual esea ch on gene al de e ence: A
c i ical e iew", Law and Socie y Re iew, pp. 545-572.
Zei haml, V. (1998), "Consume Pe cep ions o P ice, Quali y, and Value: A Means-End
Model and Syn hesis o E idence", Jou nal o Ma ke ing, Vol. 52, pp. 2-22.
60
9. Appendix
Appendix A.: Ques ionnai e
Ques ioná io Sob e Pi a a ia Digi al
Es e ques ioná io em como p opósi o auxilia o desen ol imen o de uma disse ação de mes ado,
que p ocu a in es iga a pi a a ia digi al no nosso país, endo sido desen ol ido no âmbi o do
Mes ado em Economia e Adminis ação de Emp esas da Faculdade de Economia da Uni e sidade
do Po o. Mesmo que nunca enha pi a eado qualque ipo de ma e ial digi al pode á p eenche o
ques ioná io sem qualque di iculdade. Ag adeço a sua pa icipação, pois é indispensá el pa a o
desen ol imen o do meu abalho e solici o que esponda a odas as ques ões com a máxima
since idade.
A pa icipação nes e es udo é olun á ia, e se á assegu ada a con idencialidade de odas as
espos as.
Idade:____
Sexo (M/F):____
Es udan e de:
Ensino Secundá io
Ensino P o issional
Licencia u a
Mes ado
Dou o amen o
Pós-G aduação
Ocupação:
Es udan e a empo in ei o
T abalhado -Es udan e
Ou a:____________________
O que é a Pi a a ia Digi al?
En ende-se po Pi a a ia Digi al o download/cópia de o ma ilegal de so wa e e ichei os de
media p o egidos po di ei os de au o . Tais ichei os podem se ilmes, musica, ídeo jogos,
en e ou os.
Como p eenche o ques ioná io?
 O seguin e ques ioná io em como obje i o a alia o g au de conco dância ou
disco dância com cada uma das ques ões/a i mações ap esen adas.
 Todas as epos as se ão medidas numa escala de 7 pon os. Ao seleciona a
quad ícula p óxima dos ex emos, signi ica que conco da com o adje i o (ou
a i mação) p óximo da sua escolha, à medida que se a as a dos ex emos a sua
conco dância ai diminuindo, ep esen ando a quad ícula cen al, indi e ença
(ou neu alidade).
 Responda a cada ques ão selecionando a opção p e endida com uma c uz.
61
Conside o que de um modo ge al que a pi a a ia digi al é:
Fa o á el
☐
☐
☐
☐
☐
☐
☐
Des a o á el
Bené ica
☐
☐
☐
☐
☐
☐
☐
P ejudicial
Sensa a
☐
☐
☐
☐
☐
☐
☐
Insensa a
Boa
☐
☐
☐
☐
☐
☐
☐
Má
O seguin e conjun o de pe gun as p ocu a a e i a sua capacidade pa a pi a ea
Pa a mim pi a ea ma e ial digi al se ia/é:
Mui o Fácil
☐
☐
☐
☐
☐
☐
☐
Mui o Di ícil
Que endo acilmen e pode ia pi a ea ma e ial digi al
Conco do
Plenamen e
☐
☐
☐
☐
☐
☐
☐
Disco do
Comple amen e
Conside o-me capaz de pi a ea ma e ial digi al
Conco do
Plenamen e
☐
☐
☐
☐
☐
☐
☐
Disco do
Comple amen e
Tenho os ecu sos necessá ios (ex. compu ado , ligação à in e ne , e c.) pa a pi a ea
ma e ial digi al
Conco do
Plenamen e
☐
☐
☐
☐
☐
☐
☐
Disco do
Comple amen e
Se quise sou capaz de encon a ma e ial digi al pa a pi a ea
Conco do
Plenamen e
☐
☐
☐
☐
☐
☐
☐
Disco do
Comple amen e
O seguin e conjun o de pe gun as encon a-se elacionado com a sua a i ude de
pi a a ia no passado e a sua in enção de pi a ea no u u o
Pi a eei ma e ial digi al no passado (Se esponde não, sal e as duas p óximas ques ões)
Sim
☐
☐
Não
Quan o ma e ial digi al pi a eou?
Mui o
☐
☐
☐
☐
☐
☐
☐
Pouco
Qual a equência com que pi a eou ma e ial digi al
Dia iamen e
☐
☐
☐
☐
☐
☐
☐
Espo adicamen e
62
Tenciono pi a ea ma e ial digi al num u u o p óximo
Ce amen e
que Sim
☐
☐
☐
☐
☐
☐
☐
Ce amen e
que Não
Se i esse a opo unidade, pi a ea ia ma e ial digi al
Ce amen e
que Sim
☐
☐
☐
☐
☐
☐
☐
Ce amen e
que Não
Fa ei odos os es o ços pa a pi a ea ma e ial digi al num u u o p óximo
Ce amen e
que Sim
☐
☐
☐
☐
☐
☐
☐
Ce amen e
que Não
As ques ões abaixo p ocu am o nece um melho en endimen o da opinião daqueles
que conside a como impo an es na sua ida (exemplo: amilia es e amigos)
As pessoas que conside o impo an es na minha ida pensam que não de o pi a ea
ma e ial digi al
Conco do
Plenamen e
☐
☐
☐
☐
☐
☐
☐
Disco do
Comple amen e
Rela i amen e à pi a a ia digi al, penso que de o aze o que as pessoas que me são
impo an es conside am como co e o
Conco do
Plenamen e
☐
☐
☒
☐
☐
☐
☐
Disco do
Comple amen e
Se pi a ea ma e ial digi al, a maio ia das pessoas que me são impo an es i iam:
Não se Impo a
☐
☐
☐
☐
☐
☐
☐
Desap o a
O seguin e conjun o de pe gun as es á elacionado com a se e idade da punição, no
caso de se apanhado a pi a ea
Se osse apanhado a pi a ea penso que a punição se ia:
Mui o Ele ada
☐
☐
☐
☐
☐
☐
☐
Mui o Baixa
Se osse apanhado a pi a ea , se ia se e amen e punido
Conco do
Plenamen e
☐
☐
☐
☐
☐
☐
☐
Disco do
Comple amen e
69
Appendix E.: Model i es s and co esponden s e e ence alues.
Model Fi Tes
Re e ence Values
𝑿𝟐 and p- alue
The lowe he be e ; 𝑝 > 0.05
𝑿𝟐𝒅𝒇
⁄
< 5 − Bad i
]2;5] − Accep able i
]1;2] − Good i
~ 1 − Ve y good i
CFI
GFI
< 0.8 − Bad i
]0.8;0.9] − Accep able i
]0.9;0.95] − Good i
≥ 0.95 − Ve y good i
RMSEA
and
p- alue (𝑯𝟎: 𝒓𝒎𝒔𝒆𝒂 ≤
𝟎.𝟎𝟓)
> 0.10 − Unaccep able i
]0.05;0.10] − Accep able i
≤ 0.05 − Ve y good i
𝑝 ≥ 0.05
MECVI
To compa e models. The lowe he be e
Sou ce: Ma ôco (2014, p.55).