JOURNAL OF EASTERN EUROPEAN AND CENTRAL ASIAN RESEARCH Vol.10 No.2 (2023)
www.ieeca.o g/jou nal 311
TRANSFORMATION INTO 4PL: THE CASE OF LOCAL LOGISTICS
SERVICE PROVIDERS IN VIETNAM
T ang T uong Diem Le
Tomas Ba a Uni e si y in Zlín, Czech Republic
Felici a Ch omjako á
Tomas Ba a Uni e si y in Zlín, Czech Republic
Vang Dang Quang
Ho Chi Minh Ci y Uni e si y o Technology and Educa ion, Vie nam
ABSTRACT
Fo a ew decades, a new model o logis ics se ice p o ide s (LSP) has appea ed and played he ole o
in eg a ing all ope a ions o he supply chain. This model is known as a logis ics in eg a o , o ou h
pa y logis ics (4PL). 4PL has eme ged as an ideal con igu a ion o en e p ises a ound he wo ld o
e ec i ely u ilize hei esou ces and ob ain cos educ ion ac oss he supply chain. Wi h inc easing
compe i ion among en e p ises, cus ome s’ equi emen s o complica ed se ices, and global supply
chain managemen , he limi a ions o inbound se ices om LSPs ha e become obs acles o hei
de elopmen . As a esul , ans o ma ion in o 4PL is ine i able o LSPs in he global logis ics ma ke .
This pape aims o analyze he ole, cha ac e is ics, and bene i s o 4PL. A model o ans o ma ion in o
4PL o local LSPs in Vie nam is cons uc ed o iden i y impac ing ac o s. The esul s poin ou six
ac o s in luencing he ans o ma ion p ocess, as well as h ee impo an capabili ies o 4PL.
Keywo ds:
logis ics; ou h pa y logis ics; ans o ma ion; logis ics se ice p o ide s; supply chain
DOI: h p://dx.doi.o g/10.15549/jeeca . 10i2.1018
INTRODUCTION
In he la e 1900s and ea ly 2000s, he e ha e
been d ama ic changes in he logis ics ield
which ha e been conside ed o be i al o he
imp o emen o i m p oduc i i y. These
changes consis o he g ow h o Thi d Pa y
Logis ics (3PL), he eme gence o Fou h Pa y
Logis ics (4PL), mo e complica ed pa ne ships,
he inc ease o mul imodal anspo a ion o
goods, he dec ease o logis ics cos s, and alue
chain c ea ion (Li
e al
., 2003). Ma ke s ha e,
nowadays, become highly compe i i e and
u bulen , and a e cons an ly changing. The
logis ics indus y has unde gone a deep
ans o ma ion o o e h ee decades, due o he
p essu e o e e -inc easing expec a ions and
demands om cus ome s.
Fi s ly, logis ics se ice p o ide s (LSPs)
usually in end o hold up hei se ices and
sus ainably ope a e as logis ics solu ion
p o ide s. In he long un, howe e , he endency
o he ma ke c ea es and mo i a es mo e
chances o logis ics en e p ises o un as 4PLS
wi h la ge p ojec s h ough in eg a ed and
coo dina ed ope a ions (Lieb, 2005). LSPs ha e
he e o e been changing hei ope a ions and
T ans o ma ion in o 4PL: The case o local logis ics se ice p o ide s… T ang T uong Diem Le e al.
www.ieeca.o g/jou nal 312
s a egy o become logis ics in eg a o s.
Secondly, Biens ock (2002) poin ed ou ha LSPs’
s a egic hough s conce ning ex e nal
in o ma ion lows a e use ul o help hem e ain
hei places and os e ela ionships wi h hei
cus ome s. Acco ding o Che ne a
e al.
(2015),
he business wo ld is signi ican ly impac ed by
LSPs’ de elopmen , making i essen ial o
en e p ises, managemen p o essionals, and
esea che s o ecognize 4PL’s oppo uni ies and
challenges. Thi dly, conside ing LSPs as na u al
candida es o me ging hei ope a ions o a 4PL,
Visse
e al
. (2004) and Hoek (2006) a i med ha
he ans o ma ion mus be s a ed wi h
comp ehensi e s a egies. Fu he mo e,
g asping he impo ance o logis ics
pe o mance, which di ec ly a ec s cus ome s’
e alua ion o logis ics se ice p o ide s, is
c ucial. Finally, supply chain managemen is
conside ed one o he co e elemen s o success ul
4PL. As i s men ioned by Oli e and Webbe
(1982), supply chain managemen aims o c ea e
alue o cus ome s h ough mul i-en e p ise
in eg a ion and e icien , cos -e ec i e
managemen o lows.
In his empi ical esea ch, he ac o s
in luencing he ans o ma ion in o 4PL o local
LSPs in Vie nam a e in es iga ed. Fi s , a
li e a u e e iew on a ious de ini ions o 4PL
and cons uc s in he model was conduc ed.
Then, he PLS-SEM model is applied o iden i y
he cons uc s a ec ing local LSPs’ s a egic
ans o ma ion in o 4PL. The indings d awn
om he s udy ha e conside able implica ions
o bo h academic and p ac ical ields alike.
LITERATURE REVIEW
Fou h Pa y Logis ics (4PL)
4PL has been de inied a iously by esea che s
in di e en s udies. The e m 4PL i sel was
in oduced and owned by Accen u e Consul ing
Company (Dolle and Diaz, 2011). Thei
de ini ion o 4PL is s a ed as “an in eg a o ha
assembles he esou ces, capabili ies, and
echnology o i s o ganiza ion and o he
o ganiza ions o design, build and un
comp ehensi e supply chain solu ions.”
Manu ac u e s and e aile s gain ema kable
bene i s hanks o 4PL’s e ec i e coo dina ion
be ween LSPs and hei clien s by managing he
o e all logis ics ac i i ies. In he la e 1990s,
Ga o na men ioned he 4PL concep , which was
ecognized as a combina ion o a ious
esou ces, speci ic capabili ies, and echnological
u iliza ion o assemble and manage
comp ehensi e supply chain solu ions. Yao
(2010) concluded ha 4PL commendably
ope a es an e ec i e, lexible, and easonable
in eg a ion o supply chain ac i i ies.
Papadopoulou
e al
. (2013) discussed 4PL as an
LSP ha c ea es inc easing e olu ion wi hin he
supply chain and ocuses on inno a ion
a ibu es. Pa lic Skende
e al
. (2013)
men ioned 4PL as a join en u e be ween
cus ome s and LSPs.
The e ha e been many o he esea che s who
ha e ocused on 4PL and i s ans o ma ion o
adap o he new, challenging logis ics ma ke
demand, including Li
e al
. (2003), Visse
e al
.
(2004), Ga o na
e al
. (2004), he Supply Chain
Execu i e Boa d (2005), Hoek (2006), Vi aldini
e
al
. (2008), Win (2008), Ji (2008), and Bajec
(2009). 4PL ope a es wi h he aim o e icien ly
u ilizing all esou ces, oge he wi h he
applica ion o in o ma ion echnology (IT) o
simul aneously dec ease he i m’s
backwa dness and inc ease bene i s o all
connec ed pa ies. As such, 4PL c ea es a
compe i i e ad an age in he global logis ics
ma ke , o e ing dominan e ec i eness in i s
se ice p o ision o clien s. Mo eo e , la e
s udies asse ed he impo ance and p eeminen
cha ac e is ics o 4PL in he ole o an e ec i e
and lexible in eg a o h oughou he ne wo k
(Jianming, 2011; Papadopoulou
e al
., 2013).
T ans o ma ion in o 4PL
In ecen yea s, logis ics has been conside ed a
key elemen by manu ac u ing and e ailing
i ms o de elop sys ems wi hin hei supply
chains (Ra ele, 2004). LSPs aim o con inually
p o ide logis ics se ices and de elop in o a
solu ion p o ide ; howe e , logis ics
ansac ions in he ma ke ha e led o a newe
endency o deli e ed p ojec s o be la ge and
mo e complica ed. As a esul , LSPs ha e aced a
challenge in hei ope a ions (Lieb, 2005). The e
a e e en highe expec a ions on he pa o
cus ome s o he expe ized se ices LSPs
p o ide as hey con inue o be mo e p o essional
in all ope a ions. To keep up wi h su i al and
de elopmen , s a egic logis ics solu ions should
be se iously conside ed om he poin o
p inciples and ad an ages o all esou ces
(Biens ock, 2002). Acco ding o Go indan
e al.
(2016), 4PLs ha e high capabili ies in op imizing
T ans o ma ion in o 4PL: The case o local logis ics se ice p o ide s… T ang T uong Diem Le e al.
www.ieeca.o g/jou nal 313
he ne wo k and u ilizing mul i- esou ces.
Mo eo e , 4PLs e ec i ely implemen hei ole
in in eg a ing i ms pa icipa ing in he supply
chain, esul ing in companies coo dina ing wi h
4PL as a s a egic pa ne . Visse
e al
. (2004)
analyzed he ans o ma ion p ocess in o 4PL
and sugges ed ha LSPs a e well p epa ed o
such ansi ions. Hoek (2006) highligh ed se e al
ad an ages in he ansi ion om LSP o 4PL, as
ollows:
• Enhancing added alue se ices and
ejec ing low p o i able ope a ions.
• En iching ela ionships wi h cus ome s and
in ol ing e o s in cus ome s’ supply chains.
• Se ing clien s’ expec a ions and demands
based on high u iliza ion o in o ma ion
sys ems, bu low dependence on owned
physical asse s.
METHODOLOGY
Resea ch da a
In his sec ion, we use s a is ical da a om a
da a collec ion su ey o 414 LSPs in he logis ics
indus y in Vie nam, as s a ed in Table I. To
ensu e he eliabili y and alidi y o he
measu emen index, he au ho s use he
eliabili y analyses o C onbach’s Alpha and
A e age Va iance Ex ac ed (AVE) using
Sma PLS 4 o elimina e a iables
unin e p e able o he esea ch concep .
Table I: Ca ego y o su ey esponden s
Field o logis ics ope a ions
Quan i y
( i ms)
Pe cen age
(%)
T anspo a ion, o wa ding, and wa ehousing
90
21.74
T anspo a ion and wa ehousing
81
19.56
T anspo a ion and o wa ding
62
14.98
T anspo a ion
60
14.49
Thi d-Pa y Logis ics p o ide s (3PL)
55
13.29
Fo wa ding and wa ehousing
43
10.39
Fo wa ding
09
2.17
T anspo a ion, o wa ding, and shipping agency
09
2.17
Fo wa ding and shipping agency
05
1.21
Sou ce: Au ho s’ calcula ions
Sample size
The S uc u al Equa ion Model (SEM) is used o
analyze he ela ionships be ween
ans o ma ion in o 4PL and anspo a ion
capabili y, wa ehouse ope a ions, in o ma ion
echnology (IT) applica ion, human esou ces,
logis ics se ices, anspo a ion in as uc u e,
logis ics ou sou cing ends, compe i ion in he
logis ics ma ke , and policies in he logis ics
indus y. This me hod equi es a la ge numbe o
samples due o i s dependence on sample
dis ibu ion heo y (Rayko and Widaman,
1995). Howe e , Hai
e al
. (1998) a i med ha
he e a e h ee ypes o he sample size used in
SEM, including small size ≤ 100, medium size 100
– 200, and la ge size ≥ 200. The sample size o his
s udy is 414, he e o e, i mee s he sample size
equi emen o he esea ch.
Resea ch gap
The numbe o 4PL se ice p o ide s is
cons an ly inc easing a ound he wo ld. The e
ha e been many esea ch s udies conduc ed
sepa a ely on 4PL’s ole and model, analysis and
compa ison be ween 3PL and 4PL, sugges ions
o he ansi ion om 3PL o 4PL, and model o
con lic esolu ions on 4PL de elopmen .
Fo he logis ics indus y in Vie nam, he e has
only been esea ch conduc ed on s a egic
de elopmen in o 3PL o local p i a e logis ics
companies in Vie nam. O he s udies ha e
mainly ocused on he po en ial and p ospec s o
he logis ics indus y in Vie nam, including
s udies on he supply chain and logis ics o
Vie nam in he con ex o in e na ional economic
in eg a ion, assessing he Na ional logis ics
sys em o Vie nam, sus ainable de elopmen o
logis ics in Vie nam in he 2020-2025 pe iod,
human esou ce managemen o logis ics in
T ans o ma ion in o 4PL: The case o local logis ics se ice p o ide s… T ang T uong Diem Le e al.
www.ieeca.o g/jou nal 314
Vie nam, and using he op imiza ion algo i hm
o e alua e and p edic he business pe o mance
o logis ics companies. Ne e heless, he e has
no been any esea ch on he ans o ma ion
om LSP o 4PL, especially in he con ex o local
LSPs in Vie nam.
Resea ch hypo heses
H1: High anspo a ion capabili y os e s
local LSPs’ s a egic ans o ma ion in o
4PL in Vie nam.
One o he key elemen s o logis ics se ice is
he anspo a ion capabili y o LSPs. Pa k (2011)
poin ed ou ha he main elemen s o
compe i i eness be ween companies include
human esou ces, anspo a ion capabili y,
inance, da abase, and asse s. T anspo a ion
planning p o ides he oppo uni y o i ms o
maximize cos -e ec i eness when i e icien ly
in eg a es collabo a i e pa ne s (Mason
e al
.,
2007). Logis ics ope a ions could p o ide high-
quali y se ices when LSPs ha e ad anced
anspo a ion sys ems. Due o he ema kable
signi icance o anspo a ion, he e alua ion o
he e iciency o anspo a ion modes ha e been
conduc ed by many schola s, including Smi h
and Nash (2014), Mandic
e al
., (2014),
Chakh ou a and Pojani (2016), Rodse h (2017),
and Cui and Li (2017a, 2017b, 2017c).
H2: E ec i e wa ehouse ope a ions ha e a
posi i e in luence on local LSPs’ s a egic
ans o ma ion in o 4PL in Vie nam.
Wa ehouse ope a ions a e i al in logis ics
se ices since wa ehouse p ocesses pe o m
ac i i ies in he supply chain, including ma e ial
s o age, ma e ial di ision, packaging, ga he ing,
and alloca ing. The s udy o Kłodawski
e al
.
(2017) showed a li e a u e e iew on a ious
s ochas ic models o analyzing wa ehouse
ope a ions and ele an wa ehouse s a egies
h ough esea ch conduc ed by Le-Duc and
Kos e (2005). Kłodawski
e al
. (2017) poin ed
ou ha con inuous and p ope wa ehouse
ope a ions a e e y impo an o signi ican ly
impac ing he whole supply chain.
H3: The absolu e le el o ad anced IT
applica ion has a posi i e impac on local
LSPs s a egic ans o ma ion in o 4PL in
Vie nam.
Schola s (Sabhe wal and Jeya aj, 2015;
Chaysin
e al
., 2016) ha e also highligh ed he
ole and use ulness o IT applica ions in
dis ibu ion, especially whe e hey aim a cos -
e ec i eness, se ice quali y, and small s ock.
S udies execu ed by many schola s, including
Pinna
e al
. (2010), E angelis a
e al
. (2012),
Ghobakhloo and Hong (2014), and Wong
e al
.
(2016), poin ed ou he enhancemen o logis ics
pe o mance based on IT u iliza ion. Acco ding o
Sau age (2003), IT in es men would lead la ge
LSPs o achie e supe io ad an ages and become
leade s in he global logis ics ma ke .
H4: High-quali ied human esou ces c ea e a
posi i e in luence on local LSPs’ s a egic
ans o ma ion in o 4PL in Vie nam.
A skilled wo k o ce is essen ial o unning a
s able and complica ed logis ics sys em. Bene i s
gained om human esou ce pe o mance
including ec ui ing, aining, and assessmen , all
o which help i ms imp o e he e ec i eness o
he whole supply chain and enhance compe i i e
ad an ages (Hall
e al
., 2013). Okeudo (2012),
h ough da a collec ed om LSPs, concluded ha
LSPs will imp o e hei pe o mance i he e is an
inc ease in in es men in human esou ces. Kam
e al
. (2010) s udied he ela ionship be ween
human pe o mance and logis ics capabili ies
and poin ed ou ha bene i s and pe o mance
managemen mechanisms inc ease he
commi men and capabili ies o LSPs.
H5: Ad anced logis ics se ices ha e a
posi i e impac on local LSPs’ s a egic
ans o ma ion in o 4PL in Vie nam.
Mangan
e al
. (2008) s a ed ha “Logis ics
in ol es ge ing, in he igh way, he igh
p oduc , in he igh quan i y and igh quali y, in
he igh place a he igh ime, o he igh
cus ome a he igh cos .” In line wi h his
concep , logis ics se ices consis o i e main
componen s: IT, s ock, anspo a ion,
wa ehousing, and packaging. These ac i i ies
s a wi h supplie s and end wi h cus ome s.
Logis ics se ices a e designed by LSPs o ensu e
ha clien s a e se ed wi h he lowes cos s and
highes e iciency (Badenho s -Weiss & Waugh,
2014). O e s in logis ics se ices ha a e
deli e ed by LSPs ha e imp o ed o he poin o
ad anced se ice p o ision wi h complex alue-
added anspo and wa ehousing ac i i ies
(Sel ia idis and Sp ing, 2007).
H6: Good anspo a ion in as uc u e
os e s local LSPs’ s a egic
ans o ma ion in o 4PL in Vie nam.
Fechne (2011) s a ed ha anspo a ion
in as uc u e, consis ing o land, ha bo , and
T ans o ma ion in o 4PL: The case o local logis ics se ice p o ide s… T ang T uong Diem Le e al.
www.ieeca.o g/jou nal 315
ai way sys ems, as well as in o ma ion and
communica ion echnology (ICT), ha e high
signi icance in he logis ics se ice indus y. The
linea logis ics in as uc u e o nodes in
anspo a ion e ec i ely suppo s LSPs in
p o iding logis ics ac i i ies, including
packaging, wa ehousing, deli e y, and
ansshipmen . Logis ics pe o mance, cos -
e ec i eness, and quali y assu ance would be
enhanced when he e is a well-in es ed
anspo a ion sys em, c ea ing a ema kable
ad an age o LSPs no only in ope a ion
e iciency, bu also in hei posi ion in he
logis ics ne wo k.
H7: The g ow h o he logis ics ou sou cing
end c ea es a posi i e impac on local
LSPs’ s a egic ans o ma ion in o 4PL in
Vie nam.
The g ow h o logis ics ou sou cing has
a i med a powe ul end in he global ma ke .
Solaki
e al
. (2013) poin ed ou ha inancial
sa ings, esou ce ulness, and comp ehensi e
ope a ions a e co e ac o s o companies in
using logis ics ou sou cing. Ou sou cing is
conside ed one o he i al s a egies o an
en e p ise’s business o ake ad an age o
ou side esou ces o non-co e ac i i ies and o
ocus on s a egic unc ions o long- e m
de elopmen . O he schola s ha e s a ed ha
logis ics ou sou cing c ea es chances o
en e p ises o achie e hei aims, such as cos -
e ec i eness, p oduc ion enhancemen ,
esou ce ulness, and business g ow h (Aimi,
2007; Ba dhan e al., 2006; Lau and Zhang, 2006).
H8: The inc ease o compe i ion in he
logis ics ma ke causes a posi i e
in luence on local LSPs’ s a egic
ans o ma ion in o 4PL in Vie nam.
Recen ly, LSPs, unde he equi emen s o
economic globaliza ion, ha e ocused on
enhancing hei capabili ies in o de o gain hei
clien s’ sa is ac ion. The su i al and g ow h o
LSPs depend on he success ul u iliza ion o hei
capabili ies and co e esou ces (Lu & C.S, 2007).
As a esul , he es ablishmen o a de elopmen
s a egy has been ecognized o be signi ican o
LSPs in an inc easingly compe i i e ma ke .
Acco ding o he 2017 epo on logis ics in
Vie nam, issued by The Minis y o Indus y and
T ade o he Socialis Republic o Vie nam,
compe i ion in he global logis ics ma ke has
become mo e ie ce. La ge logis ics se ice
p o ide s h oughou he wo ld hold a ound 15%
o he global logis ics ma ke sha e.
H9: Comple ed go e nmen policies in he
logis ics indus y c ea e posi i e
s imula ion o local LSPs’ s a egic
ans o ma ion in o 4PL in Vie nam.
The mechanism and policies o logis ics
indus y de elopmen ha e a i al signi icance
o LSPs. Acco ding o esea ch conduc ed by Jin
(2012) and Liu
e al
. (2013), he e a e wo le els
o de elopmen policies o he logis ics indus y,
consis ing o na ion and egion. Na ional logis ics
policies a e issued o encou age he de elopmen
o he mac oeconomy, while egional logis ics
policies mainly ocus on he speci ic
cha ac e is ics o di e en egions.
Scales o measu emen
In he p esen s udy, scales o measu emen
a e es ablished o 10 cons uc s, wi h a o al o
31 indica o s. All indica o s a e measu ed on a
i e-poin Like scale (1 = Ve y high, 2 = High, 3
= Medium, 4 = Low, 5 = Ve y low). The speci ic
cons uc s and indica o s a e as ollows:
Table 2: Resea ch cons uc s and indica o s
Cons uc s
Indica o s
T anspo a ion capabili y (TRA)
TRA1: Owned means o anspo a ion
TRA2: Speed o anspo a ion na ionally and
in e na ionally
TRA3: Connec ion o anspo a ion chain and logis ics
se ices
Wa ehouse ope a ions (WOP)
WOP1: Scale o owned wa ehouse
WOP2: Technology applica ion
WOP3: Ra e o e o s
WOP4: C oss-docking u iliza ion
Ad anced IT applica ion (ITA)
ITA1: Highly quali ied IT human esou ces

T ans o ma ion in o 4PL: The case o local logis ics se ice p o ide s… T ang T uong Diem Le e al.
www.ieeca.o g/jou nal 316
Cons uc s
Indica o s
ITA2: Ad anced IT in as uc u e
ITA3: S ong pa ne ing ela ionship be ween IT and
logis ics se ice managemen
Human esou ces (HMR)
HMR1: Specialized compe ence
HMR2: Planning and con olling capabili y
HMR3: Lea ning and in eg a ing compe ence
Logis ics se ices (LOS)
LOS1: P o ision o di e si ica ion and s a egy
cus omiza ion o logis ics se ices
LOS2: P o ision o alue-added se ices o cus ome s
LOS3: Logis ics se ice cos s
T anspo a ion in as uc u e (INF)
INF1: Ai po in as uc u e
INF2: Ha bou in as uc u e
INF3: Land in as uc u e
The g ow h o logis ics ou sou cing
end (OUT)
OUT1: T end o logis ics ou sou cing
OUT2: Size o an o ganiza ion adop ing logis ics
ou sou cing
OUT3: Le els o logis ics ou sou cing
Compe i ion in he logis ics indus y
(COM)
COM1: Numbe o i als who a e LSPs in he logis ics
ma ke
COM2: Ma ke sha e o LSPs in he logis ics ma ke
COM3: Types o i als’ logis ics se ice p o ision: 2PL,
3PL, 4PL
Policies in he logis ics indus y (POL)
POL1: Policies in he logis ics indus y
POL2: Suppo ed policies o LSPs
POL3: Di ec ions and s a egies o go e nmen o he
de elopmen o he logis ics indus y
T ans o ma ion in o 4PL (4PL)
4PL1: Value chain c ea ion
4PL2: In eg a ion o mul iple 3PL p o ide s’ ac i i ies
4PL3: Managemen compe ence in global supply chain
Sou ce: Au ho s’ s udy
RESULTS AND DISCUSSION
Fi s , he eliabili y o he esea ch model was
assessed. Known as he coe icien
de e mina ion, he R-Squa ed o mula de ines
he deg ee o which he a iance in he
dependen a iable can be explained by
independen a iables. F om he esul s o
boo s apping, he R-Squa e o he model is
0.644, and he R-Squa e Adjus ed is 0.636 (as
shown in Figu e 1). This means ha 63,6% o he
a ia ion in he dependen a iable (4PL) is
explained by independen a iables (TRA, ITA,
HRM, LOS, COM, OUT, WOP, INF, and POL). Wi h
his esul , he eliabili y o he esea ch model is
demons a ed.
Table 2: Con inued
T ans o ma ion in o 4PL: The case o local logis ics se ice p o ide s… T ang T uong Diem Le e al.
www.ieeca.o g/jou nal 317
Figu e 1: Model o ans o ma ion in o 4PL
Sou ce: Au ho ’s es ima ions om Sma PLS
Based on he ecommenda ions o PLS-SEM
heo y and he li e a u e o Hai e al. (2017), he
cons uc s’ eliabili y le els a e e alua ed using
Dijks a-Hensele ’s ho, along wi h C onbach’s
alpha, coe icien s. As shown in Table III, all
alues exceed he h eshold o 0.5 and indica e
s ong coe icien s o he cons uc ’s eliabili y,
as sugges ed by Bagozzi & Yi (1998) and Hai e
al. (2019).
Table 3: Cons uc Reliabili y and Validi y
Cons uc
C onbach's
alpha(α)
Dijks a-Hensele 's
ho_A
Composi e
Reliabili y
A e age a iance
ex ac ed (AVE)
COM
0,886
0,891
0,930
0,816
HRM
0,878
0,887
0,925
0,806
ITA
0,896
0,909
0,936
0,829
OUT
0,713
0,717
0,839
0,634
LOS
0,894
0,912
0,934
0,825
POL
0,872
0,891
0,921
0,795
4PL
0,767
0,768
0,866
0,683
TRA
0,715
0,805
0,847
0,662
INF
0,864
0,875
0,916
0,785
WOP
0,907
0,920
0,935
0,783
Sou ce: Au ho ’s es ima ions om Sma PLS
T ans o ma ion in o 4PL: The case o local logis ics se ice p o ide s… T ang T uong Diem Le e al.
www.ieeca.o g/jou nal 318
The S anda dized Roo Mean Squa e Residual
(SRMR) alue was also conside ed in o de o
measu e he app op ia e le el o he model o
he esea ch con ex . Acco ding o Hu and
Ben le (1999), he SRMR alue mus be lowe
han 0.08 o 0.1. Mo eo e , Hensele
e al
. (2014)
also a i med ha he SRMR alue de e mines
“goodness o i ” in PLS-SEM. This alue is
measu ed o a oid model misspeci ica ion. Table
IV shows he SRMR alue o he esea ch model
as 0.049, again demons a ing he
app op ia eness o he model.
Table 4: S anda dized Roo Mean Squa e Residual (SRMR)
O iginal sample (O)
Sample mean (M)
95%
99%
Sa u a ed Model
0.049
0.034
0.037
0.038
Es ima ed Model
0.049
0.034
0.037
0.038
Sou ce: Au ho ’s es ima ions om Sma PLS
Rega ding indica o loadings o la en
cons uc s, he eliabili y o indica o s mus gain
ou e loadings highe o equal o he h eshold o
0.5 o mee he s anda d o eliabili y, while
composi e eliabili y mus be highe o equal o
0.7 (Hulland, 1999). All i ems in he model a e
loaded meaning ully and sa is ac o ily o hei
co esponding cons uc s. Values a e p esen ed
in Table 5.
Con e gen Validi y is used o e alua e he
s abili y o scales. Fo nell and La cke (1981)
poin ed ou ha A e age Va iance Ex ac ed
(AVE) mus be highe o equal o 0.5 o indica e
sa is ac o y con e gen alidi y. The cons uc s
in he s udy ha e minimum o maximum alues
om 0.634 o 0.829. The e o e, hese alues a e
sa is ac o y. The de ails a e s a ed in Table 5.
Table 5: Indica o s’ Ou e Loadings, Cons uc s’ Composi e Reliabili y, and A e age Va iance
Ex ac ed (AVE)
Cons uc
Indica o
Ou e Loadings
Composi e Reliabili y
A e age Va iance
Ex ac ed (AVE)
4PL
4PL1
0.801
0.866
0.683
4PL2
0.833
4PL3
0.843
COM
COM1
0.953
0.930
0.816
COM2
0.871
COM3
0.884
HRM
HMR1
0.950
0.925 0.806
HMR2
0.861
HMR3
0.879
INF
INF1
0.865
0.916
0.785
INF2
0.909
INF3
0.884
ITA
ITA1
0.958
0.936
0.829
ITA2
0.883
ITA3
0.889
LOS
LOS1
0.960
0.934
0.825
LOS2
0.870
LOS3
0.893
OUT
OUT1
0.799
0.839
0.634
OUT2
0.826
OUT3
0.763
POL
POL1
0.857
0.921
0.795
POL2
0.926
POL3
0.891
TRA
TRA1
0.929
0.847
0.662
TRA2
0.511
TRA3
0.928
WOP
WOP1
0.940
0.935 0.783
WOP2
0.857
WOP3
0.850
WOP4
0.890
Sou ce: Au ho ’s es ima ions om Sma PLS
T ans o ma ion in o 4PL: The case o local logis ics se ice p o ide s… T ang T uong Diem Le e al.
www.ieeca.o g/jou nal 319
A Disc iminan Validi y assessmen aims o
ensu e ha a e lec i e cons uc has he
s onges ela ionships wi h i s indica o s in he
PLS pa h model (Hai e a ., 2017), whe ein he
squa e oo o AVE mus be highe han he
La en Va iable Co ela ions. The squa e oo s o
he AVE o he esea ch cons uc s a e shown in
bold diagonals in Table 6.
Table 6: Disc iminan Validi y
Cons uc
COM
HMR
ITA
OUT
LOS
POL
4PL
TRA
INF
WOP
COM
0.903
INF
0.110
0.886
POL
0.124
0.219
0.892
OUT
0.177
0.130
0.111
0.796
LOS
0.074
0.282
0.201
0.114
0.908
ITA
0.152
0.283
0.271
0.206
0.311
0.911
4PL
0.421
0.332
0.275
0.426
0.444
0.600
0.826
TRA
0.184
0.362
0.268
0.206
0.410
0.324
0.460
0.814
HMR
0.233
0.072
0.233
0.242
0.052
0.106
0.340
0.115
0.898
WOP
0.194
0.371
0.249
0.169
0.385
0.262
0.381
0.538
0.204
0.885
Sou ce: Au ho ’s es ima ions om Sma PLS
To e alua e whe he he e a e ela ionships
be ween cons uc s, he S uc u al Equa ion
Model is used. When he - alue is highe han
1.96, i means ha he signi ican le el is lowe
han 5% (p- alue < 0.05). Ou e Weigh s a e
c i e ia showing he ela i e con ibu ion o each
indica o . In he S uc u al Equa ion Model
(SEM), Ou e Weigh s a e o en lowe han Ou e
Loadings (Hai e al., 2014). To e alua e whe he
indica o s con ibu e o he es ablishmen o
la en a iables, boo s apping should be used. In
his s udy, he so wa e Sma PLS 4 was used o
build a S uc u al Equa ion Model wi h 5,000
boo s ap samples.
A e comple ing he es ablishmen o he
esea ch model, implemen ing he assessmen o
he model is essen ial. The e a e a ious ways o
e alua e he eliabili y o he esea ch model.
Fi s , he sample may be di ided in o wo sub-
samples, wi h one used o building he esea ch
model and he emaining used o ee alua e he
eliabili y o ha esea ch model. The eliabili y
o he model may also be assessed by collec ing
mo e samples.
Ande son and Ge bing (1988), howe e ,
s a ed ha SEM equi es a la ge numbe o
samples and consumes exo bi an ime and cos s
o esea che s. Schumake and Lomax (2016)
hen assumed ha boo s apping is app op ia e
o apply due o i s epea ed sample me hod,
while he ini ial samples emain o be majo
pa s. The boo s apping me hod uses he
ob ained sample da a om he s udy o esample
i a ious imes o c ea e many simula ed
samples. The sampling dis ibu ions a e
conside ed he ounda ion o con idence
in e als and hypo hesis es ing, wi h he - alue
calcula ed based on he dis ibu ions o c ea ed
samples.
Table 7: Pa h Coe icien and Cons uc Rela ionships
E ec
Boo s apping esul s
Empi ical
ema ks
O iginal
Coe icien
Mean alue
S anda d
de ia ion
- alue p- alue
COM -> 4PL
0,244
0,245
0,033
7,345
0,000
Suppo ed
HRM -> 4PL
0,165
0,165
0,032
5,082
0,000
Suppo ed
ITA -> 4PL
0,384
0,383
0,032
11,912
0,000
Suppo ed
OUT -> 4PL
0,208
0,209
0,035
6,022
0,000
Suppo ed
LOS -> 4PL
0,208
0,208
0,039
5,352
0,000
Suppo ed
TRA -> 4PL
0,127
0,127
0,040
3,137
0,002
Suppo ed
POL -> 4PL
-0,009
-0,008
0,031
0,282
0,778
No suppo ed
INF -> 4PL
0,056
0,056
0,034
1,640
0,101
No suppo ed
WOP -> 4PL
-0,002
-0,002
0,038
0,054
0,957
No suppo ed