Roldán e al. In o ma ion sys ems capabili ies and o ganiza ional agili y
2nd In e na ional Symposium on Pa ial Leas Squa es Pa h Modeling, Se ille (Spain), 2015 1
In o ma ion sys ems capabili ies and o ganiza ional agili y: Unde s anding
he media ing ole o abso p i e capaci y when in luenced by a hie a chy
cul u e
Comple ed Resea ch Pape
José L. Roldán
Uni e sidad de Se illa
[email p o ec ed]
An onio L. Leal-Rod íguez
Uni e sidad Loyola Andalucía
[email p o ec ed]
Ca men Felipe
Uni e sidad de Se illa
[email p o ec ed]
Abs ac
O ganiza ional agili y (OA), as a key dynamic capabili y, is a i m’s abili y o enable sensing
en i onmen al changes and esponding e icien ly and e ec i ely o hem. This s udy
explo es his opic u he by analyzing he pa played by he in o ma ion sys ems
capabili ies (ISC) a iable as an an eceden o OA, and abso p i e capaci y (AC) as a
media o cons uc . Fu he mo e, we es he nega i e mode a ing ole o hie a chy cul u e
(HC) in he AC–OA link. Using pa ial leas squa es (PLS) and he PROCESS mac o, we ind
e idence o hese ela ions p oposed, and he exis ence o a condi ional media ing si ua ion
gene a ed by HC.
Keywo ds: o ganiza ional agili y, in o ma ion sys ems capabili ies, abso p i e capaci y,
hie a chy cul u e, pa ial leas squa es (PLS), media ing analysis, mode a ing e ec ,
condi ional media ion model.
1. In oduc ion
O ganiza ions a e cu en ly acing highly u bulen en i onmen s, which a e mainly
cha ac e ized by s ong doses o dynamism, complexi y and unce ain y. These condi ions
ha e led o hype compe i i e ma ke s whe e he su i al o companies is ce ainly h ea ened.
In such a con ex , knowing he mechanisms ha allow o ganiza ions o de ec , adap and o e
he p ope esponse o he en i onmen ’s changes becomes especially ele an , as his migh
lead i ms o a ain a g ea e success by exploi ing eme ging oppo uni ies and new sou ces o
compe i i e ad an ages. The e o e, he concep o o ganiza ional agili y (OA) appea s as a
key issue conce ning o ganiza ional su i al and success.
Agile o ganiza ions a e hose ha a e able o e ec i ely ope a e wi hin hype compe i i e,
unp edic able and cons an ly changing en i onmen s (Goldman, Nagel & P eiss, 1995). Thus,
OA can be de ined as he i m’s capabili y o sense he changes o he en i onmen and
espond e icien ly and e ec i ely o hem (Ash a i e al., 2005). Assuming he dynamic
capabili ies heo y as a e e ence amewo k (Teece, Pisano, & Shuen, 1997), OA is iden i ied
as one o he key dynamic capabili ies o o ganiza ions in o de o achie e sus ainable
compe i i e ad an ages (Sambamu hy, Bha adwaj, & G o e , 2003) and o su i e in highly
dynamic en i onmen s (Nijssen & Paauwe, 2012).
Roldán e al. In o ma ion sys ems capabili ies and o ganiza ional agili y
2nd In e na ional Symposium on Pa ial Leas Squa es Pa h Modeling, Se ille (Spain), 2015 2
This opic has been a ac ing he a en ion o academic esea ch since he mid-1990s and
has been assessed by mul iple disciplines. This has led o successi e limi a ions o he concep
o agili y, emo ing i om i s gene al aspec o o ganiza ional capaci y (Cha bonnie -Voi in,
2011). In addi ion, o e he las yea s, he ocus o esea ch on he echnological aspec o
business has led o o ge ing o he con ex ual o ganiza ional ac o s equally o e en mo e
ele an , such as cul u e, communica ion and leade ship (C oci o & Yousse , 2003).
Basically, he co e o hese s udies has been he ole o in o ma ion sys ems capabili ies (ISC)
in achie ing a highe le el o OA.
This pape hence aims o co e such esea ch gaps inhe en o OA ha he p io li e a u e
has un il now ailed o do. The e o e, we mean o answe he ollowing ques ions: (1) F om
an inclusi e poin o iew, wha is OA? (2) Wha a e he an eceden s o OA? Can we conside
o he a iables apa om ISC (e.g., abso p i e capaci y)? (3) Wha a e he links be ween
such an eceden s? Do ISC a ec OA di ec ly o h ough an indi ec ela ionship? (4) Could
he p esence o ce ain cul u al alues become a mode a o o he a o emen ioned ela ion?
The pape p oceeds as ollows. The nex sec ion p esen s he heo e ical backg ound
oge he wi h he esea ch model and hypo heses. The hi d sec ion comp ises a desc ip ion o
he esea ch me hodology. The ou h sec ion p esen s he esul s o he di e en da a
analyses ca ied ou . Finally, we b ing oge he he discussion and implica ions.
2. Li e a u e Re iew and Resea ch Hypo heses
2.1. App oaching he concep o o ganiza ional agili y
The concep o o ganiza ional agili y has i s oo s in wo p io ela ed concep s:
o ganiza ional adap abili y (a eac i e aspec ) and o ganiza ional lexibili y (a p oac i e
aspec ) (She ehiy, Ka wowski, & Laye , 2007). In his ein, OA in ol es a i m’s abili y o
sense and espond o en i onmen al changes (O e by, Bha adwaj, & Sambamu hy, 2006).
The con ibu ion o Sambamu hy e al. (2003) is qui e ema kable. They s a e ha OA
comp ises h ee in e ela ed dimensions: cus ome agili y (le e aging he oice o cus ome s
o gain ma ke in elligence), pa ne ing agili y (lea ning om business pa ne s o enhance he
i m’s esponse o he ma ke ) and ope a ional agili y ( apid p ocess edesign o exploi
dynamic ma ke place condi ions). The e o e, ollowing Cha bonnie -Voi in (2011), we de ine
OA as he in en ional esponse capabili y de eloped by he o ganiza ion o enable i o ac
e icien ly in a highly u bulen en i onmen , no only by eac ing apidly o change, bu also
h ough i s po en ial o ac ion in an icipa ing and seizing oppo uni ies, in pa icula h ough
inno a ion and lea ning.
2.2. The ela ionship be ween in o ma ion sys ems capabili ies and o ganiza ional agili y
The concep o in o ma ion sys ems capabili ies (ISC) comes om he use o he esou ce-
based heo y in he in o ma ion echnologies (IT) esea ch ield. This heo y enabled he
es ablishmen o a amewo k o assess he s a egic con ibu ion o in o ma ion sys ems (IS)
esou ces o he company (Wade & Hulland, 2004). Unde such a pe spec i e, he i m’s IS
esou ces (asse s and capabili ies) ha a e inimi able and aluable may lead o achie ing
sus ained compe i i e ad an ages (Ra ichand an & Le wongsa ien, 2005).
Bha adwaj (2000) de ines ISC as he i m’s abili ies o mobilize and deploy IT-based
esou ces in combina ion o join ly wi h o he esou ces and capabili ies. These a e skills,
compe ences and abili ies upon which he alue o he physical IT esou ces can be le e aged
(Dohe y & Te y, 2009). Wade and Hulland (2004) desc ibe h ee ypes o ISC: inside-ou
Roldán e al. In o ma ion sys ems capabili ies and o ganiza ional agili y
2nd In e na ional Symposium on Pa ial Leas Squa es Pa h Modeling, Se ille (Spain), 2015 3
(deployed om inside he i m in esponse o ma ke equi emen s and oppo uni ies),
ou side-in (ex e nally o ien ed, placing an emphasis on an icipa ing ma ke equi emen s), and
spanning capabili ies (needed o in eg a e he i m’s inside-ou and ou side-in capabili ies).
The e is a lack o consensus in he scien i ic li e a u e conce ning whe he he impac o
ISC on OA is posi i e o nega i e. On he one hand, some esea che s suppo ing he nega i e
impac a gue ha limi a ions o in lexible IT sys ems may esul in a igidi y which hinde s o
e en impedes he adap a ion o he en i onmen ’s equi emen s (O e by e al., 2006). On he
o he hand, Sambamu hy e al. (2003) conside ha ITs a e gene a o s o he digi al op ions
h ough which OA is posi i ely a ec ed. This is no only because hey allow he c ea ion o
new in o ma ion-based p oduc s and se ices, bu also because hey enable he coo dina ion
o in e nal p ocesses and he building o new in e o ganiza ional ela ionships. Lu and
Ramamu hy (2011) show ha ISC ha e a di ec e ec on agili y, indica ing ha i ms need
o con inually de elop supe io IT capabili ies in o de o success ully manage and exploi
hei esou ces, wi h he aim o building agile o ganiza ions.
Wi h he suppo o his line o he li e a u e, we posi ha p ope ly deployed and
managed ISC can p o ide ools and ins umen s o o ganiza ions o enhance hei capabili ies
o sense and espond o en i onmen al changes. The e o e we pos ula e he ollowing
hypo hesis:
H1: The i m’s in o ma ion sys ems capabili ies (ISC) a e posi i ely linked o i s
o ganiza ional agili y (OA).
2.3. The media ing ole o abso p i e capaci y in he ela ionship be ween ISC and OA.
Wi hin he cu en dynamic en i onmen , o ganiza ional lea ning has become a key success
ac o o i ms. The s udy o abso p i e capaci y (AC) is hence pe ec ly embedded wi hin
such a scena io. Cohen and Le in hal (1990) ini ially de ined AC as he i m’s abili y o
ecognize he alue o new ex e nal knowledge, assimila e i and apply i o comme cial ends.
Zah a and Geo ge (2002) la e de eloped an ex ension o he AC concep , b oadly de ining i
as a se o o ganiza ional ou ines and p ocesses h ough which i ms acqui e, assimila e,
ans o m and exploi knowledge in o de o p oduce a dynamic o ganiza ional capabili y.
These ou ac i i ies a e complemen a y and build upon each o he o p oduce AC.
Liu, Ke, Wei, and Hua (2013) indica e ha ISC a e key ac o s o he de elopmen o
highe o de capabili ies, such as he AC. In ac , some IS unc ions, such as de eloping
knowledge eposi o ies, e ec i e in o ma ion e ie ing mechanisms, o enabling
collabo a ion and communica ion be ween knowledge p oduce s (expe s) and knowledge
seeke s, play a key ole in he i m’s AC enhancemen (Ash a i, Xu, Kuilboe , & Koehle ,
2006). In his ein, Cepeda-Ca ión, Cega a-Na a o and Jiménez-Jiménez (2012) conside
ha ISC suppo o ganiza ions’ AC, since hey enable new knowledge o be combined wi h
pas knowledge in o de o be exploi ed.
The e is s ill is a gap o esea che s conce ning he ie be ween AC and OA. Rega ding
OA, Lu and Ramamu hy (2011) sugges as a u u e esea ch line he s udy o he
mechanisms o de eloping ou ines and s uc u es ha acili a e lea ning and
expe imen a ion, and imp o e capaci y building. A company wi h a s onge AC is mo e
p epa ed o pe cei e changes in he ma ke s and o lea n om expe ience (Malho a, Gosain,
& Sawy, 2005). Ash a i e al. (2005) s a e ha he e a e no empi ical s udies in he li e a u e
o explain how and why in es men in knowledge acquisi ion d i es OA, and p opose wo
concep s - AC and dynamic capabili ies - as enable s o achie ing agili y. Hao, Yu and Dong
(2011) poin ou he media ing ole o AC by i s ansla ing knowledge managemen sys ems
usage in o highe o de o ganiza ional capabili ies, i.e., agili y and inno a i eness
To sum up, ISC lead o an enhanced AC, and a g ea e AC migh imp o e he agili y o
Roldán e al. In o ma ion sys ems capabili ies and o ganiza ional agili y
2nd In e na ional Symposium on Pa ial Leas Squa es Pa h Modeling, Se ille (Spain), 2015 4
he o ganiza ion. Based on his logic and p e ious esea ch, we hus hypo hesize:
H2: The ela ion be ween in o ma ion sys ems capabili ies (ISC) and o ganiza ional
agili y (OA) is media ed by abso p i e capaci y (AC).
2.4. The mode a ing e ec o a hie a chy cul u e in he ela ionship be ween AC and OA.
Di e en o ganiza ional alues gene a e dispa a e knowledge managemen (KM)
beha io s and hese will lead o a ying ou comes (Ala i, Kaywo h & Leidne , 2006). While
some cul u al alues, such as openness and us , can induce posi i e KM beha io s (e.g.,
knowledge con ibu ion and sha ing), which will lead o inno a ion, he e a e o he alues
ha migh lead o dys unc ional KM beha io s (e.g., in o ma ion hoa ding) and, hence,
nega i e esul s such as o ganiza ional igidi y. This means o ganiza ional cul u e may
become a ba ie ha hinde s AC and i s e ec s i i emains excessi ely igid and con ol-
o ien ed.
Using he compe ing alues amewo k (Came on & Quinn, 2011), we ha e ocused on
he hie a chy cul u e (HC), which is o en labeled as bu eauc a ic, op-down ocused, ule-
o ien ed, and by- he-book (Zammu o, Gi o d & Goodman, 2000). This cul u al ypology is
based upon minimal ambigui y le els and an excessi e sense o sa e y, p edic abili y,
e iciency, s abili y, uni o mi y, e c. In sho , i can be sus ained ha HC is a cul u al ypology
essen ially o ien ed owa d e iciency and in e nal con ol. Mo eo e , i s alues a e in e nally
ocused and a e hus mo e aligned wi h keeping a s a ic and igid hie a chical s uc u e han
pu suing business oppo uni ies in he ma ke . We hence hypo hesize:
H3: Hie a chy cul u e (HC) mode a es (dec easing) he link be ween abso p i e capaci y
(AC) and o ganiza ional agili y (OA).
3. Resea ch Me hod
3.1. Sample and da a collec ion
We ha e chosen inno a i e classi ied sec o s as he popula ion o his s udy. These
indus ies can be conside ed as hype compe i i e, equi ing a lexible and quick esponse
om o ganiza ions. The sec o selec ion has been ca ied ou using he classi ica ion
de eloped by he Spanish Na ional Ins i u e o S a is ics (Co ec, 2009) o high and medium-
high echnology indus ies. This gene a ed a popula ion o 2,360 i ms. An o -line su ey
was used o ga he da a. Because he le el o analysis is he o ganiza ion, he esponden o
he ques ionnai e was a senio managemen membe . A e one mailing e o , he ou come
was 172 usable su eys (a 7.3% esponse a e). The o ganiza ions pa icipa ing belonged
p ima ily o he compu e sys ems design (26.7%), machine y manu ac u ing (18%) and
chemical (17.4%) sec o s. O he indus ies we e included in he manu ac u ing sec o s o
anspo a ion equipmen (8.1%), elec ical equipmen (7.6%), and compu e and elec onic
p oduc s (7%). Acco ding o he Eu opean Union classi ica ion, 23.8% o he i ms
pa icipa ing we e la ge en e p ises, wi h mo e han 250 employees. O he esponden s,
23.8% belonged o he esea ch and de elopmen depa men , ollowed by he ma ke ing
depa men (20.9%), gene al managemen (14%), and he enginee ing depa men (9.3%).
Mos o he esponden s we e male (66%), whe eas women ep esen ed 34%.
3.2. Measu es
The e iew o he li e a u e allowed us o iden i y alida ed measu es o each cons uc .
Roldán e al. In o ma ion sys ems capabili ies and o ganiza ional agili y
2nd In e na ional Symposium on Pa ial Leas Squa es Pa h Modeling, Se ille (Spain), 2015 5
Ou e o s ocused on making he necessa y adjus men s o he con ex o he s udy (i.e., he
Spanish language and se ing). A pilo es o he su ey was conduc ed in o de o assess he
con en alidi y. The ISC a iable, as a supe o dina e mul idimensional cons uc , was
measu ed by eigh i ems adap ed om Wade and Hulland (2004). To assess AC as a
supe o dina e mul idimensional cons uc , we adap ed 21 i ems om Jansen, Van Den Bosch
and Volbe da (2005). OA was also modeled as a supe o dina e mul idimensional cons uc
and measu ed by ele en i ems adap ed om Lu and Ramamu hy (2011), Yang and Liu
(2012), B adley, P a , By d, Ou lay, and Wynn (2012), and Tallon and Pinsonneaul (2011).
The measu emen o he HC a iable used an adap a ion o he scale appea ing in Came on
and Quinn (2011). This a iable has been modeled as a unidimensional cons uc shaped by
six e lec i e i ems. Finally, we con olled he size (numbe o employees) and he age
(numbe o yea s since he ounding) o he i m. All he a iables we e measu ed on he basis
o se en-poin Like scales, excep con ols.
3.3. Da a analysis
We ha e used Pa ial Leas Squa es (PLS) pa h modeling o es he esea ch model. The
choice o PLS is based on he ollowing easons (Roldán & Sánchez-F anco, 2012): (1) The
ocus o he s udy is he p edic ion o he dependen a iables; (2) he sample (n = 172) is no
e y la ge; (3) he esea ch model is complex acco ding o he ype o ela ionships (di ec ,
media ed and mode a ed) desc ibed in he hypo heses and he le els o dimensionali y; (4)
his s udy uses la en a iables sco es in he subsequen analysis o p edic i e ele ance,
pa icula ly in he implemen a ion o he wo-s age app oach o modeling mul idimensional
cons uc s (W igh , Campbell, Tha che , & Robe s, 2012); and (5) he na u e o mos
heo e ical cons uc s is de ined, as we ely on a composi e measu emen model wi h a
e lec i e design app oxima ion, which means ha indica o s and dimensions ep esen
di e en ace s bu a e expec ed o be co ela ed (Hensele , 2014). This way, cons uc s a e
modeled as composi es o hei indica o s wi hou e o e m (Fo nell, 1982). This s udy uses
Sma PLS . 3.2 so wa e (Ringle, Wende, & Becke , 2015) o he PLS analysis, and
PROCESS mac o 2.13 (Hayes, 2013) o he mode a ed media ion analysis.
4. Resul s
4.1. Measu emen model
Fi s , he indica o s and dimensions sa is y he equi emen o eliabili y since hei
loadings a e, in gene al, g ea e han 0.7 (Table 1). In o de o accomplish his esul , we
ca ied ou an i em imming p ocess wi h some weak i ems o he AC ins umen . In addi ion,
some i ems o he HC cons uc also had weak loadings. No wi hs anding, we ha e decided o
e ain hem in o de o suppo he con en alidi y o he scale. Due o size limi a ions, we
only show loadings o dimensions.
Second, all mul idimensional cons uc s and dimensions mee he equisi e o cons uc
eliabili y, because hei composi e eliabili ies (CR) a e g ea e han 0.7. Thi d, hese la en
a iables a ain con e gen alidi y since hei a e age a iance ex ac ed (AVE) su passes
he 0.5 le el o a e e y nea o i (Table 1). Las ly, Table 2 shows ha all a iables achie e
disc iminan alidi y ollowing bo h he Fo nell-La cke and he HTMT.90 c i e ion, howe e ,
he AC and OA a iables may ha e a disc iminan alidi y p oblem acco ding o he HTMT.85
c i e ion (Hensele , Ringle & Sa s ed , 2014).

Roldán e al. In o ma ion sys ems capabili ies and o ganiza ional agili y
2nd In e na ional Symposium on Pa ial Leas Squa es Pa h Modeling, Se ille (Spain), 2015 6
Table 1: Measu emen model esul s
Table 2: Measu emen model. Disc iminan alidi y
4.2. S uc u al model
Table 3 includes he main pa ame e s ob ained o he ou models unde s udy in he
s uc u al assessmen . Model 1 desc ibes he signi ican o al e ec (c = 0.642***) o ISC on
OA once he e ec o con ols (age and size) has been conside ed. Model 2 shows how he
di ec e ec o ISC on OA dec eases, al hough i emains signi ican (c’ = 0.236***), when
AC is included. This suppo s H1. Fu he mo e, pa hs a and b1 a e signi ican . The e o e, bo h
he dec emen mani es ed in he di ec e ec (c’) and he signi icance o he eg ession
coe icien s a and b1 would be sugges ing he po en ial exis ence o an indi ec e ec o ISC
on OA ia AC as a media o (H2). None heless, he key condi ion o de e mine such a
media ing e ec is o es he signi icance o a×b1 (Hayes, 2009). Wi h his aim in mind, we
ha e ob ained he alue o his indi ec e ec (a×b1 = 0.401) om Sma PLS, which is
signi ican (Table 4). This ou pu suppo s H2. Consequen ly, we assume a pa ial media ion
o he AC in he ela ion be ween ISC and OA since he di ec (H1 = c’) and he indi ec (H2
= a×b1) e ec s a e bo h signi ican (Ba on & Kenny, 1986). In addi ion, we ha e calcula ed
he a iance accoun ed o (VAF) index (Hai , Hul , Ringle, & Sa s ed , 2014), which
de e mines he size o he indi ec e ec (a×b1) in ela ion o he o al e ec (c). When he
VAF has an ou come be ween 20% and 80%, a pa ial media ion can be expec ed. This occu s
in ou case, gi en ha he VAF o he indi ec e ec is 62.45% (Table 4). Finally, we ha e
sough o ake a u he s ep o wa d by compu ing he s anda dized oo mean squa e esidual
(SRMR), as he oo mean squa e disc epancy be ween he co ela ions obse ed and he
model-implied co ela ions (Hu & Ben le , 1999) o he model wi h he o al e ec and he
Cons uc /Dimension
Loading
CR
AVE
In o ma ion sys ems capabili ies (SMC)
0.934
0.824
Ou side-in capabili ies
0.876
0.908
0.832
Spanning capabili ies
0.937
0.935
0.878
Inside-ou capabili ies
0.910
0.940
0.796
Abso p i e capaci y (SMC)
0.899
0.691
Acquisi ion
0.763
0.839
0.511
Assimila ion
0.802
0.892
0.735
T ans o ma ion
0.919
0.862
0.611
Exploi a ion
0.834
0.793
0.563
O ganiza ional agili y (SMC)
0.921
0.795
Ope a ional agili y
0.860
0.911
0.773
Cus ome agili y
0.946
0.912
0.776
Pa ne ing agili y
0.866
0.885
0.611
Hie a chy cul u e ( e lec i e cons uc )
0.851
0.492
No es: CR: composi e eliabili y; AVE: A e age a iance ex ac ed; SMC: supe o dina e mul idimensional
cons uc
Fo nell-La cke C i e ion
He e o ai -Mono ai Ra io (HTMT)
ISC
AC
OA
HC
Age
Size
ISC
AC
OA
HC
Age
Size
ISC
0.908
ISC
AC
0.665
0.831
AC
0.761
OA
0.638
0.765
0.892
OA
0.716
0.885
HC
0.455
0.467
0.503
0.701
HC
0.508
0.504
0.548
Age
0.064
0.099
0.144
0.157
n.a.
Age
0.069
0.102
0.151
0.175
Size
0.199
0.131
0.132
0.125
0.327
n.a.
Size
0.212
0.141
0.137
0.131
0.327
No es: ISC: in o ma ion sys ems capabili ies; AC: abso p i e capaci y; OA: o ganiza ional agili y; HC:
hie a chy cul u e. Fo nell-La cke C i e ion: Diagonal elemen s (bold) a e he squa e oo o he a iance sha ed
be ween he cons uc s and hei measu es (AVE). O -diagonal elemen s a e he co ela ions among cons uc s.
Fo disc iminan alidi y, diagonal elemen s should be la ge han o -diagonal elemen s. n.a.: non-applicable.
Roldán e al. In o ma ion sys ems capabili ies and o ganiza ional agili y
2nd In e na ional Symposium on Pa ial Leas Squa es Pa h Modeling, Se ille (Spain), 2015 7
Table 3: S uc u al model esul s.
Rela ionships
Model 1
Model 2
Model 3
Model 4
Suppo
SRMR c m = 0.054
SRMR c m = 0.047
R2
AC = 0.442 / Q2
AC = 0.297
R2
AC = 0.442
R2
AC = 0.442
R2
OA = 0.426 / Q2
OA = 0.321
R2
OA = 0.621 / Q2
OA = 0.484
R2
OA = 0.636
R2
OA = 0.647
2
H1: ISC ! OA
(c) 0.642*** (12.799) [0,563; 0,727]
(c’) 0.236*** (3.150) [0.106; 0.353]
(c’) 0.199** (2.645) [0.061; 0.305]
(c’) 0.203** (2.769) [0.071; 0.314]
Yes
ISC ! AC = a
0.665*** (15.118) [0.606; 0.748]
0.665*** (15.392) [0.605; 0.746]
0.665*** (15.237) [0.603; 0.745]
AC ! OA = b1
0.603*** (8.387) [0.492; 0.725]
0.563*** (7.806) [0.437; 0.674]
0.528*** (7.280) [0.383; 0.625]
HC ! OA = b2
0.143++ (2.645) [0.080; 0.258]
0.141++ (2.711) [0.069; 0.239]
H3: HC × AC ! OA = b3
-0.117* (-2.073) [-0.269; -0.093]
0.035
Yes
Con ol a iables:
Age
0.115+ (2,150) [0.038; 0.215]
0.075ns (1.915) [0.019; 0.146]
0.059ns (1.518) [-0.002; 0.127]
0.050ns (1.335) [-0.009; 0.114]
Size
-0.031ns [-0.106; 0.057]
-0.019ns (0.340) [-0.095; 0.085]
-0.019ns (0.338) [-0.089; 0.087]
-0.017ns [-0.084; 0.081]
No es: ISC: in o ma ion sys ems capabili ies; AC: abso p i e capaci y; OA: o ganiza ional agili y; HC: hie a chy cul u e; c m: composi e ac o model
alues in pa en heses. Boo s apping 95% con idence in e als bias co ec ed in squa e b acke s (based on n = 5000 subsamples)
***p < .001; **p < .01; *p < .05 (based on (4999), one- ailed es ). (0.05, 4999) = 1.645; (0.01, 4999) = 2.327; (0.001, 4999) = 3.092.
++p < .01; +p < .05; ns: no signi ican (based on (4999), wo- ailed es ). (0.05, 4999) = 1.960; (0.01, 4999) = 2.577; (0.001, 4999) = 3.292.
Table 4: Summa y o media ing e ec es s.
To al e ec on OA (Model 1)
Di ec e ec s on OA (Model 2)
Indi ec e ec on OA (Model 2)
BCCI
BCCI
BCCI
Pa h
Lowe
Uppe
Pa h
Lowe
Uppe
Poin es ima e
Lowe
Uppe
Sig
VAF
ISC (c)
0.642***
12.271
0.554
0.723
H1: ISC (c')
0.236***
3.150
0.106
0.353
H2: ab1 ( ia AC)
0.401***
7.527
0.324
0.499
Yes
62.45%
Con ol a iables
Age
0.115+
2.150
0.038
0.215
Age
0.075ns
1.915
0.019
0.146
Size
-0.031ns
0.631
-0.106
0.057
Size
-0.019ns
0.340
-0.095
0.085
No es: ISC: in o ma ion sys ems capabili ies; AC: abso p i e capaci y; OA: o ganiza ional agili y.
BCCI: Bias co ec ed con idence in e al. Boo s apping based on n = 5000 subsamples.
VAF: Va iance accoun ed o .
***p < .001 (based on (4999), one- ailed es ). (0.05, 4999) = 1.645; (0.01, 4999) = 2.327; (0.001, 4999) = 3.092.
+p < .05; ns: no signi ican (based on (4999), wo- ailed es ). (0.05, 4999) = 1.960; (0.01, 4999) = 2.577; (0.001, 4999) = 3.292.
Roldán e al. In o ma ion sys ems capabili ies and o ganiza ional agili y
2nd In e na ional Symposium on Pa ial Leas Squa es Pa h Modeling, Se ille (Spain), 2015 8
model wi h he indi ec e ec . Following Hensele e al. (2014), we ha e de e mined he
SRMR o a composi e ac o model. This p o ides he exac i o he composi e ac o
model, hus cons i u ing a con i ma o y composi e analysis. Model 1 ( o al e ec ) achie es a
SRMR composi e ac o model o 0.054, which means an app op ia e i assuming he usual
cu -o o 0.08 (Hu & Ben le , 1999). Howe e , he SRMR composi e ac o model o Model
2 is s ill be e : 0.047. This would imply an addi ional suppo o he media ing ole o AC.
The mode a ion hypo hesis (H3: b3) o he hie a chy cul u e (HC) in he pa h be ween AC
and OA is es ed using he p oduc -indica o echnique (Chin, Ma colin, & News ed, 2003).
Model 3 includes HC and Model 4 adds he in e ac ion e m (HC×AC = b3) (Table 3). The
esul seems o suppo H3 (b3 = −0.117*) (Table 3, Model 4) (Figu e 1). Mo eo e , he
o e all e ec size o b3 achie es an 2 alue o 0.035, which exceeds he minimum h eshold
o 0.02 (Chin, Ma colin, & News ed, 2003).
Figu e 1: Model wi h a condi ional indi ec e ec (Model 4)
The suppo o H3 oge he wi h he signi ican indi ec e ec (a×b1) gene a es he
eme gence o a mode a ed media ion (Hayes, 2013). This in ol es he dependence o he
indi ec e ec (a×b1) on he alue o HC (b3), which would ac as a mode a o a iable. As
AC's e ec on OA is con ingen on he HC a iable, so is ISC's indi ec in luence on OA.
Following Hayes (2013), such an indi ec impac is a × (b1 + b3HC).
In o de o es ima e his condi ional indi ec e ec , we ha e applied he PROCESS mac o
de eloped by Hayes (2013). Using la en a iable sco es om Sma PLS 3 as inpu ,
PROCESS p oduces es ima es and bias-co ec ed 95% boo s ap CI o he indi ec e ec a
di e en alues o HC as a mode a ing cons uc . Table 5A shows ha he indi ec e ec o
ISC on OA h ough AC is consis en ly posi i e and dec eases as he HC alues inc ease. A
95% CI bias-co ec ed boo s ap o he condi ional indi ec e ec is abo e ze o o he
di e en alues o HC. This indi ec impac is signi ican in all he scena ios analyzed. Hence,
AC pa ially media es ISC’s in luence on OA, al hough his indi ec e ec dec eases as HC
inc eases i s alue. Finally, Table 5B con ains an index o mode a ed media ion (−0.0512)
(Hayes, 2015), which is also signi ican .
Las o all, he signi ican di ec e ec (non-hypo hesized) o HC on OA (Models 3 and 4,
Table 3) dese es o be commen ed on. In spi e o i s nega i e mode a ing in luence on he
pa h be ween AC and OA (Model 4), we ind some e idence o a posi i e in luence o he
alues associa ed wi h HC on he OA achie ed by i ms.
Model wi h a condi ional indi ec e ec (Model 4)
In o ma ion
sys ems
capabili ies (ISC)
b1 = 0.528***
Abso p i e
capaci y (AC)
R2 = 0.442
O ganiza ional
agili y (OA)
R2 = 0.647
H1(+) = c’ = 0.203***
Size
Age
H2(+) = ISC ! AC ! OA = a × (b1 + b3HC) 0.050ns
-0.017ns
*** p < 0.001, * p < 0.05 (based on (4999), one- ailed es )
++ p < 0.01,+ p < 0.05, ns: no signi ican (based on (4999), wo- ailed es )
H3(-) = b3 = -0.117*
Hie a chy
cul u e (HC)
AC × HC
b2 = 0.141++
a = 0.665***
Roldán e al. In o ma ion sys ems capabili ies and o ganiza ional agili y
2nd In e na ional Symposium on Pa ial Leas Squa es Pa h Modeling, Se ille (Spain), 2015 9
Table 5: Condi ional indi ec e ec analyses
5. Discussion
O ganiza ions mus de elop capabili ies which connec hem pe manen ly o hei
en i onmen and hei s akeholde s, and which enable hem o ans o m cap u ed in o ma ion
in o apid and p ecise esponses. This pape has posi ed OA as a clea example o his ype o
mechanism. This s udy also con ibu es o enhancing he ecen esea ch on he i m’s
s a egic e o s and endea o s o ind mechanisms ha lead o imp o ing OA, p oposing a
model wi h ISC and AC as i s an eceden s. Fi s ly, we ind suppo o he di ec ela ionship
be ween ISC and OA, a link ha was no comple ely clea in he p e ious li e a u e.
Secondly, om a media ion poin o iew, we also p o ide e idence o he exis ence o an
indi ec e ec o ISC on OA h ough AC. Resul s e eal ha he in luence o ISC on OA is
mo e an indi ec han a di ec e ec . This means ha he i m’s ISC will impac on OA
enhancemen conce ning he ex en o which i is capable o gene a ing AC. Thi dly, we ind
suppo o he hypo hesis sugges ing he nega i e mode a ing ole o HC on he AC-OA link.
This wo k p esen s some signi ican academic implica ions. Fi s , while p e ious s udies
ha e sugges ed ha ISC can enable o ganiza ional agili y (Sambamu hy e al., 2003; Lu &
Ramamu hy, 2011), empi ical e idence has been sca ce, ocused on pa ial aspec s and
lacking a amewo k o explain how ISC in luence a i m’s OA (T inh-Phuong e al., 2012).
This wo k ies o ill his gap, by de eloping and es ing a comp ehensi e model, all o
whose a iables a e de ined in hei mos inclusi e o m o o ganiza ional capabili ies.
Second, ou esul s shed ligh on he exis ing gap conce ning he po en ial impac o ISC
on OA enhancemen , and h ough which mechanisms hey ac . Fu he mo e, acco ding o ou
esul s, we conclude ha AC pa ially media es he ISC-OA ela ionship
Thi dly esul s also indica e he coun e e ec o HC on he AC-OA link, ac ing as a
mode a o a iable ha dec eases he di ec link men ioned. Al hough he p esence o cul u al
alues associa ed wi h HC can hinde he posi i e e ec o AC on OA, he indi ec
ela ionship emains signi ican .
This nega i e mode a ing e ec su p isingly con as s wi h a non-hypo hesized di ec
A) Condi ional indi ec e ec o ISC on OA a alues o HC as mode a o
BCCI
Media o
HC
E ec
Boo SE
Lowe
Uppe
AC
-1.0029
0.4079
0.0604
0.3025
0.5412
AC
0
0.3565
0.0596
0.2531
0.4877
AC
1.0029
0.3051
0.0668
0.1881
0.4502
No e: Values o HC (mode a o ) a e he mean and plus/minus one s anda d de ia ion (SD) om mean
BCCI
Media o
HC
E ec
Boo SE
Lowe
Uppe
AC
-1.3581
0.4261
0.0623
0.3146
0.5628
AC
-0.6497
0.3898
0.0589
0.2863
0.5203
AC
0.0638
0.3533
0.0596
0.2507
0.4856
AC
0.6185
0.3248
0.063
0.2162
0.4634
AC
1.2783
0.291
0.0697
0.1702
0.4419
No e: Values o HC (mode a o ) a e 10 h, 25 h, 50 h, 75 h, and 90 h pe cen iles
B) Index o mode a ed media ion
BCCI
Media o
Index
SE (Boo )
Lowe
Uppe
AC
-0.0512
0.0223
-0.0994
-0.01
No es: ISC: in o ma ion sys ems capabili ies; AC: abso p i e capaci y; OA: o ganiza ional agili y; HC:
hie a chy cul u e. Con ol a iables: Age and Size on OA. BCCI: Bias co ec ed con idence in e al.
Boo s apping based on n = 5000 subsamples.