MANUFACTURING STRATEGY-TECHNOLOGY RELATIONSHIP
AMONG AUTO SUPPLIERS
1
Césa H. O ega Jiménez (a) & (b)
2
Ped o Ga ido-Vega (b)
José Luis Pé ez Díez de los Ríos (c)
San iago Ga cía González (d)
(a) Uni e sidad Nacional Au ónoma de Hondu as
Ins i u o de In es igaciones Económicas y Sociales (IIES)
Edi icio 5, Plan a Baja, Ciudad Uni e si a ia, Bl d. Suyapa
Tegucigalpa, MDC., Hondu as (Cen al Amé ica).
(b) Uni e sidad de Se illa
Facul ad de Ciencias Económicas y Emp esa iales
Depa amen o de Economía Financie a y Di ección de Ope aciones
G upo de In es igación en Di ección de Ope aciones en la Indus ia y los Se icios
(GIDEAO)
A enida Ramón y Cajal, 1.
41018 – Se illa (Spain).
(c) Uni e sidad de Se illa
Facul ad de Ciencias Económicas y Emp esa iales
Depa amen o de Economía Aplicada I
A enida Ramón y Cajal, 1
41018 – Se illa (Spain)
(d) Uni e sidad de Huel a
Depa amen o de Economía Financie a, Con abilidad y Di ección de Ope aciones
Adminis ación y Modelización de O ganizaciones (G.I.A.M.O.)
Facul ad de Ciencias Emp esa iales.
Plaza de la Me ced, 11.
21002-Huel a (Spain).
1
Acknowledgemen : This s udy is pa o he Spanish Minis y o Educa ion and Science Na ional P og amme
o Indus ial Design (DPI-2006-05531 and DPI 2009 -11148) and he Jun a de Andalucía (Spain) PAIDI (Plan
Andaluz de In es igación, Desa ollo e Inno ación) Excellence P ojec s (P08-SEJ-03841).
2
Co esponding au ho : elephone: +504-22391849, ax numbe : +504-22391849, e-mail: co [email p o ec ed]
2
ABSTRACT
Each manu ac u ing plan has o de elop i s own pa h o success based on con ingencies and
on manu ac u ing p ac ices links. On he basis o he la e , his pape es s he link be ween
wo o he mos impo an manu ac u ing p ac ices a eas, manu ac u ing s a egy (MS) and
echnology, wi hou add essing causali y o hei combined e ec on pe o mance. This is
done by selec ion i , i.e. cong uency adjus men . Howe e , his pape goes beyond g ouping
bo h se s o p ac ices in pai s, by using a mo e gene al selec ion iew e sion, wi h p ac ices
om bo h se s ela ed mul idimensionally and subo dina ed by eg ession analysis o es o
any cong uen pa e n. Reg ession esul s om a wide- anging su ey o au o supplie plan s
show ha , in gene al, MS seems o ha e some kind o impac on echnology, and ha
echnology has some kind o in luence on MS. In addi ion, a s ong cong uency be ween bo h
p ac ices a eas is obse ed when using co ela ion. This sugges s ha when implemen ing o
adjus ing MS o echnology, he o he should also be conside ed; o he wise hey may no
ope a e e ec i ely.
Keywo ds: Cong uency, Selec ion, Fi , Manu ac u ing S a egy, Technology
1. INTRODUCTION
Each manu ac u ing plan mus ind i s own unique pa h o success, based on con ingen
ac o s and he links be ween manu ac u ing p ac ices. P e ious s udies on his opic s ill shed
li le ligh on he easons why he applica ion o he same manu ac u ing p ac ices wo ks well
in some plan s, bu wo se in o he s (P im ose, 1992; Olhage , 1993; Nassimbeni, 1996).
Thus, be o e he selec ion, adap a ion (when equi ed), implemen a ion and in e connec ion o
manu ac u ing p ac ices, he e should also be a s a egic, well-concei ed plan based on he
pa icula si ua ion o he company. Wi hou i , he designed s a egy will no ha e he desi ed
3
e ec : he achie emen o success. All o he abo e should be linked o a planned pa h o
con inuous imp o emen . Hence plan s should be dynamic, cons an ly d awing upon he bes
manu ac u ing p ac ices o hei possible inclusion as pa o he manu ac u ing p ocess.
Such inclusion depends on bo h he con ex o he plan (con ingency) and on he e ec ha
he in oduc ion o new p ac ices will ha e by linking hem o wha he plan is al eady doing
o is planning o do. This esul s in a syne gy o p ocesses designed o achie e a sus ainable
wo ld-class compe i i e ad an age by means o he con inuous imp o emen o he
manu ac u ing capaci y (Sch oede and Flynn, 2001).
Howe e , achie ing a sus ainable compe i i e ad an age, by means o using manu ac u ing
p ac ices, is i sel an e asi e goal: wo ld class plan s may some imes ha e ela i ely poo
implemen a ion le els o p ac ices. In such cases, i may well be ha he success o he plan
will quickly diminish when he condi ions change, as he solid ounda ion o a co ec ly
connec ed ne wo k o p ac ices is no suppo ing he whole. Likewise, he e may be cases
whe e plan s ha e implemen ed a high le el o p ac ices and s ill be unsuccess ul. In he la e
case, he plan s need o conside whe he hey ha e chosen he co ec p ac ices o hei own
ci cums ances and whe he he p ac ices a e app op ia ely linked o he o e all s a egy and
wi h one ano he (Sch oede and Flynn, 2001). On he o he hand, he e ec i e use o
echnological esou ces—amongs o he hings—is essen ial o achie ing a sus ainable
compe i i e ad an age and o inc easing he e ec i eness o he company. The e o e, aking
in o accoun he impo ance o MS and echnology, as well as he p oposi ion ha he lack o
success in some plan s may be pa ially due o a aul y link be ween p ac ices (Sch oede and
Flynn, 2001), he p esen s udy examines he link be ween p ac ices om manu ac u ing
s a egy (MS) and om echnology om an in e na ional au o supplie sec o su ey. The
4
need o in es iga e he in e connec ion be ween s a egy and echnology has also been
s essed by Po e (1983, 1985).
Acco dingly, he p esen pape is p ima ily cen ed on he ollowing esea ch ques ion: A e
he e any links be ween p ac ices om manu ac u ing s a egy and p ac ices om echnology?
This is answe ed by way o explo a o y and con i ma o y esea ch.
A e iew o he li e a u e is made in sec ion 2. Resea ch p oposi ions a e desc ibed in sec ion
3 along wi h hei espec i e hypo heses. The esea ch me hodology o his wo k is explained
in sec ion 4, desc ibing he cons uc s and concep s used. Subsequen ly (sec ion 5), he esul s
a e discussed. Finally, in sec ion 6, some conclusions and inal conside a ions a e ou lined,
highligh ing he implica ions and limi a ions o his s udy.
2. LITERATURE REVIEW
In ela ion o he MS- echnology ela ionships, some au ho s (Ho e and Schendel, 1978;
Po e , 1983; Hayes, 1985; Maidique and Pa ch, 1988; Pa hasa hy and Se hi, 1993; Pa ke ,
2000) p esen a mainly s a ic and unidi ec ional pe spec i e. In his pe spec i e, he causal
ela ionship goes om echnology o s a egy and no ice e sa (since he exis ing echnical
capabili ies should guide he o mula ion o s a egy). Acco ding o his pe spec i e,
compe i i eness in a company’s manu ac u ing echnology is a sp ingboa d o he
de elopmen o s a egy (Pa hasa hy and Se hi, 1993). The e o e, manu ac u ing s a egy
should e lec manu ac u ing capaci ies, including echnological ini ia i es. This a gumen o
complemen a i ies implies ha plan s which y o achie e high e ec i eness om
echnological p ac ices should implemen hese in conjunc ion wi h he app op ia e
manu ac u ing s a egy (e.g. Co be and Van Wassenho e, 1993; Pa hasa hy and Se hi,
1993). Technology is he e o e a ac o ha limi s s a egy in wo ways: 1) he exis ing
5
echnology de e mines he s a egy ha an o ganisa ion can pu sue (I ami and Numagami,
1992), and 2) he company, wan ing o pu sue a di e en s a egy, should expand o change
i s echnological base (Ho e and Schendel, 1978; Maidique and Pa ch, 1988; Pa ke , 2000;
Po e , 1983).
Taking he opposi e iew, o he esea che s (Skinne , 1969; S obaugh and Telesio, 1983;
Dean and Snell, 1996) uphold ha s a egy should de e mine he selec ion o echnology.
Acco ding o his pe spec i e, o an o ganisa ion o be compe i i e, s a egy mus d i e
echnological de elopmen (Po e , 1983). In his way, echnological de elopmen can b ing
bo h a g oup o compe i i e weapons and a deepe echnological base applicable o o he
p oduc s/ma ke s o he plan (I ami and Numagami, 1992; Zah a and Co in, 1993). The
accumula ed esou ces o pas p oduc s/ma ke s may change in o he d i ing o ces behind he
di e si ica ion s a egy o he plan . The ue sou ces o compe i i e ad an age may be
de i ed mo e om consolida ing echnologies wi h manu ac u ing skills in he co e a eas o
compe i ion han om gene a ing p oduc s ha he compe i ion does no an icipa e (see
Chandle , 1962; P ahalad and Hamel, 1990). Thus, he mos impo an plan decisions in
manu ac u ing should be made o imp o e he chosen base o compe i i e ad an age (Hayes
e al., 1988; Ga in, 1993). Manu ac u ing echnology can clea ly be one o hese, since i is a
signi ican elemen in manu ac u ing (Leong e al., 1990; Ma ucheck e al., 1990). Hence, in
o de o use s a egy e ec i ely, echnology should be conside ed h ough i s lens.
Howe e , he p esen s udy will go beyond he limi a ions o any single app oach ega ding
he di ec ions o he ela ionships be ween manu ac u ing s a egy and echnology ha can be
explo ed.

6
Thus, he esea ch ques ion o his pape could be nuanced as o how o iden i y he MS
p ac ices ha a ec echnology p ac ices and ice e sa, and o explo e he na u e o hese
ela ionships.
Among he possible models o analyse hese ela ionships, selec ion i
3
has been chosen since
i has p o en o be he bes way o examine how a iables in e ac o explain each o he ’s
designs/implemen a ions (Ge din and G e e, 2004). Addi ionally, selec ion is he mos
common and simples o m o i in he li e a u e (Bu ns and S alke , 1961; Mo se, 1977;
D azin and Van de Ven, 1985; Galunic and Eisenha d , 1994; Meilich, 2006). Fo his,
explo a o y and con i ma o y esea ch based on h ee ela ionships, namely a bidi ec ional
and wo unidi ec ional iews o selec ion (also e med cong uency) will be used. The
adjus men p emise ha is assumed in selec ion is a cong uency be ween bo h p ac ice se s
mu ually in luencing each o he while ope a ing in a plan (see Hannan and F eeman, 1977;
Ald ich, 1979; McKel ey, 1982; Van de Ven and D azin, 1985; D azin and Van de Ven,
1985).
A close look a he way he MS- echnology ela ionships ha e been esea ched e eals ha
only nine s udies om o e 110 pape s compiled in a book edi ed by Sch oede and Flynn
(2001), whose wo main High Pe o mance Manu ac u ing (HPM) esea ch ounda ions we e
con ingency and links be ween manu ac u ing p ac ices, di ec ly deal wi h linkages be ween
p ac ices (Flynn e al., 1992; Flynn, 1994; Mo i a and Sakakiba a, 1994 a, 1994 b; Flynn e
al., 1995; Mo i a and Flynn, 1997; Ahmad, 1998; Mo i a e al., 1999; Cua, 2000).
Fu he mo e, Mo i a and Flynn’s pape (1997) is he only s udy o hese nine ha is di ec ly
conce ned wi h he ela ionship be ween MS and echnology. Howe e , i does no deal wi h
3
Fi could be de ined as he co ela ion be ween wo o mo e ac o s ha leads o a be e esul .
7
his ela ionship in an exclusi e o exhaus i e way, since, on he one hand, i app oaches he
ela ionship o MS (conside ing only s a egic adap a ion) wi h o he p ac ices and, on he
o he hand, i only akes on boa d he concep o echnological adap a ion wi h i s scales. The
au ho s do conclude, howe e , ha he e is an impo an link be ween his echnological
concep and s a egic adap a ion.
Since said book, only h ee wo ks in his same line o HPM esea ch ha e di ec ly examined
his impo an subjec . In hese pape s he e a e indings ha end o con i m he impo ance
o his ela ionship. Ma sui (2002) s udies he con ibu ion o di e en p ac ices (including
MS) in he de elopmen o echnology in h ee p ac ices o p ocess and p oduc echnology
(e ec i e implemen a ion o p ocesses, in e unc ional design e o , simplici y o p oduc
design). Pa s o his esul s cons i u e clea e idence ha he pa icipa ion o manu ac u ing
p ac ices (MS included) in he de elopmen o echnology has a s ong impac on he
compe i i eness o he p oduc ion plan . McKone and Sch oede (2002) seek o de e mine he
ype o companies making use o p ocess and p oduc echnology by aking he ela ionship
wi hin he con ex o he plan ( hey include s a egic aspec s) bu wi hou conside ing
pe o mance. Finally, a pa o Ke oki i and Sch oede ’s (2004) s udy conside s he s a egic
e en uali ies in ol ed in he adop ion and implemen a ion o se e al manu ac u ing p ac ices
o achie e high pe o mance. Howe e , hey include "design o manu ac u abili y" as he
only echnological a iable.
Rega ding he gene al P oduc ion and Ope a ions Managemen (POM) li e a u e, mos o he
p e ious s udies ha e explo ed he ela ionship be ween business s a egy (no MS) and
echnology, ei he in a one-dimensional o a mul idimensional way. Some esea che s ha e
classi ied he essen ial dimensions/p ac ices o echnology ha a e inhe en in a speci ic
s a egy (e.g. Fo d, 1980). On he o he hand, Pa ke (2000) ies o es o cu en and u u e
8
dynamic in e ac ion be ween business s a egy and echnology and i s e ec on he plan ’s
pe o mance, bu wi hou using a ime se ies (a longi udinal s udy).
This li e a u e shows some empi ical in e connec ions be ween speci ic dimensions/p ac ices
o echnology and business s a egy. Some o he disco e ies indica e he need o de e mine
he i /adjus men be ween hese p ac ices (e.g. Pa hasa hy and Se hi, 1993; C o eau and
Be ge on, 2001).
Thus, some o hese s udies ha e indeed p oposed in eg a ed models ha desc ibe i s
be ween se e al dimensions/p ac ices o echnology and business s a egy (Maidique and
Pa ch, 1988; Zah a and Co in, 1993). Howe e , hey ha e no empi ically shown i he e is a
ela ionship o mu ual adap a ion in he design and implemen a ion o MS and echnology
p ac ices, which ensu es ha only wo ld class o ganisa ions will su i e hanks o he
exis ence o a supposed isomo phic p ocess be ween he wo p ac ice a eas (selec ion i ).
In conclusion, al hough he abo e s udies ha e inc eased he gene al unde s anding o
s a egy- echnology ela ionships, hey ha e no examined he possible cong uency/selec ion
aspec s o his appo . Mo eo e , al hough hey ha e had an in luence on he gene a ion o
ideas conce ning he ela ionships be ween s a egy and echnology, o da e he co esponding
empi ical alida ions ha e been minimal and he e ha e been e en ewe ega ding MS, since
mos o hese pas pape s analyse ela ionships om a business s a egy pe spec i e. Wi h his
in mind, i is possible o conclude ha : 1) p e ious esea ch has undamen ally been o ien ed
owa ds heo y and 2) he possible impac o a selec i e ela ionship be ween MS and
echnology has no been well documen ed.
Due o he abo e, i is no clea whe he he ela ionship be ween MS and echnology is
inhe en ly selec i e in i s na u e. The e o e, he p esen wo k ies o shed mo e ligh on his
9
subjec by e i ying a possible cong uency be ween MS and echnology (T) p ac ices, aking
da a om an au o supplie su ey conduc ed in en coun ies.
3. ANALYTICAL FRAMEWORK AND HYPOTHESES
One impo an ocus o POM esea ch in ecen yea s has been linkages be ween
manu ac u ing p ac ices. D awing on his, his pape ies o ind whe he he a iables (in ou
case, a se o 3 MS p ac ices and ano he o 3 T p ac ices) show a ce ain deg ee o
cong uency. Thus, his is di e en om add essing he ela ionship o how hese same
a iables in luence pe o mance (i.e. uni e sal pe spec i e) o om inding whe he bo h
p ac ice se s in e ac ing wi h each o he a ec pe o mance (i.e. in e ac ion pe spec i e, whe e
one o he se s in e ac s wi h he o he ) (Ha mann and Moe s, 1999; Lu and Shields, 2003).
The undamen al di e ence compa ed o he cong uency/selec ion iew is ha in he wo
la e (uni e sal and in e ac ion) he esea che is no p ima ily in e es ed in examining how
a iables in e ac o explain each o he ’s designs, bu in showing ha some combina ions a e
mo e ela ed o highe pe o mance han o he s.
In he con ingency li e a u e, he selec ion o m o i en isions p ima ily a linea
co espondence be ween he s uc u al and con ingency a iables. Thus, as a s a ing-poin o
he adjus men be ween bo h p ac ice se s, his concep o i could be desc ibed as he
co ela ion be ween wo o mo e ac o s ha leads o a be e esul (see Venka aman and
P esco , 1990; Milg om and Robe s, 1995; Cua e al., 2001). In keeping wi h his,
p oposi ions o he ela ionship s udied he e a e i s desc ibed and hen hei espec i e
hypo heses a e p esen ed. On he basis o he i and mis i concep s, his pape will he e o e
add ess he concep s in ol ing he di ec ela ionship be ween he wo se s o p ac ices in
ques ion using a bi a ia e selec ion model.
16
4.2. Measu emen
All o he measu emen s used in his s udy we e pe o med using pe cep ual scales, each
consis ing o se e al ques ions (i ems). Each ques ion was answe ed using a se en-poin
Like scale. Re e se-wo ded i ems we e e e se sco ed.
Con en alidi y was ensu ed h ough bo h a ep esen a i e collec ion o i ems, as well as a
me hod o es cons uc ion (Nunnally, 1967). A comp ehensi e e iew o he ex an li e a u e
was used o he ep esen a i e lis o i ems. The es cons uc ion me hod ollowed
ques ionnai e p epa a ion, pilo es ing, s uc u ed in e iews, ansla ion, and back ansla ion
when he ques ionnai es we e adminis e ed in coun ies whose mo he ongue was no
English.
Fo cons uc alidi y, he i ems o each ac o we e checked o see i hey loaded on o one
ac o . Fo his, wi hin-scale ac o analysis was pe o med o es whe he each scale om
bo h manu ac u ing p ac ice se s o med co esponding unidimensional measu es, as ollows:
h ee scales we e used o measu e MS p ac ices acco ding o he de ini ion o MP p ac ices
desc ibed ea lie . An i em was dele ed i i loaded on o a second ac o . All ac o loadings o
he scales we e abo e 0.60, much highe han he cu -o alue o ± 0.40 (Hai e al., 1998). A
simila p ocedu e was used o cons uc he echnology p ac ices se wi h i s h ee scales (all
o he ac o loadings we e abo e 0.70 excep o one (0.476, bu s ill highe han cu -o )).
Bo h he MS and echnology p ac ice se s a e concep ualised and de ined as unidimensional
cons uc s. Meanwhile, a eliabili y analysis was conduc ed a he plan le el o each scale o
e alua e in e nal consis ency. The eliabili y o he scales was measu ed by C onbach’s alpha
acco ding o Nunnally (1978) and all we e g ea e han 0.7 ( he co esponding analysis wi h
an accep able deg ee o eliabili y and alidi y will be p o ided upon eques ).

17
4.3. Me hods
The unc ional o m o selec ion i is linea co espondence be ween MS and echnology. Some o he
ad an ages o his model a e i s simple p ocedu e and he ac ha i does no equi e he measu e o a
hi d a iable as an ou come. In addi ion, ope a ionalising he selec ion me hod is e y s aigh o wa d
using co ela ion, eg ession, analysis o a iance (ANOVA), and so on. This s udy uses bo h
co ela ion and eg ession.
4.3.1. Co ela ion
The ypical es ing scheme associa ed wi h he selec ion app oach is he assessmen o simple
co ela ion be ween each pai o MP a iables (e.g. Aiken and Hage, 1968; Cohn and Tu yn, 1980;
Damanpou , 1991). Thus, he i s me hod is he mos common in selec ion and in his pape consis s
o g ouping bo h se s o a iables in pai s, whe e a se ies o canonical co ela ion analyses could
demons a e whe he he se o echnology p ac ices used he e is cong uen wi h he MS se .
Hypo hesis H1 equi es he s eng h o he ela ionship be ween wo se s o a iables o be es ed.
Canonical co ela ion analysis is used o es his ela ionship. I cons uc s a weigh ed linea
combina ion o he a iables in each o he wo se s being co ela ed, wi h weigh s selec ed o
maximise he co ela ion be ween he wo weigh ed ec o s, o canonical a ia es. One o he
ad an ages o canonical co ela ion analysis is ha i equi es only mul i a ia e no mali y o he
a iables in he da a se s. In addi ion, canonical co ela ion pe mi s he use o mul iple dependen
a iables.
Th ee c i e ia we e conside ed o assess he s eng h o he o e all ela ionship desc ibed by canonical
co ela ion analysis (Hai e al., 1998): 1) le el o s a is ical signi icance; 2) magni ude o he
canonical co ela ion coe icien ; and 3) edundancy measu e o he pe cen age o a iance explained
by he wo se s o a iables. The i s canonical pai comp ises he wo canonical a ia es ha ha e he
s onges ela ionship wi h each o he , and is su icien e idence o ejec he null hypo hesis.
Fo he signi ican canonical pai s, canonical c oss-loadings a e calcula ed as he co ela ion be ween
18
each o he o iginal a iables in one se and he weigh ed canonical a ia e om he o he se o
a iables. This se o c oss-loadings is used o in e p e he s eng h o each o he a iables in
explaining he ela ionship wi h he o he se as a whole.
Na u ally, canonical co ela ion analysis is easible i you do no wan o conside one se o a iables
as he ou come and he o he se as p edic o a iables. This pape he e o e p esen s he ollowing
me hod.
4.3.2. Reg ession
The e migh be some limi a ions o he use o canonical co ela ion analysis o es ing he p oposed
hypo heses. The main basis o his is ha he conside a ion o a iables om he wo domains in
isola ed pai s and he ex apola ion o he indings o in e ences associa ing he oo domains a e
p oblema ic. This p oblem, howe e , is o e come o an ex en by using eg ession as a mo e gene al
e sion o he selec ion app oach, consis en wi h he de ini ion o cong uency. Fi has been widely
measu ed h ough eg ession coe icien s in he selec ion pe spec i e (see Simons, 1987; Kaplan and
Mackey, 1992; Hai e al., 1998). This analysis no only shows he gene al di ec ion o he associa ion,
bu also de e mines he deg ee o which he independen a iables a ec he dependen a iables.
He e, as opposed o ea ing he a iables as independen pai s, se s o a iables om he wo domains
a e ela ed, essen ially depic ing a cong uen pa e n in a mul i-a ibu e con igu a ion, whe e he
p ac ices o he wo MP se s a e ela ed mul idimensionally and subo dina ed by mul i a ia e mul iple
eg ession analysis (MMRA), in o de o obse e whe he hey ollow a cong uen pa e n. This ype
o eg ession is used when you ha e wo o mo e a iables ha a e o be p edic ed om wo o mo e
p edic o a iables. F om he esea ch a iables, his me hod will p edic i s ly echnology (T) om
MS (H2), and secondly MS om T (H3). In bo h cases, his pape uses hei speci ic p ac ices.
This eg ession is "mul i a ia e" because he e a e h ee ou come a iables (scales o p ac ices) om
one o he MP se s. I is a "mul iple" eg ession because he e a e h ee p edic o a iables (scales o
p ac ices) o he co esponding MP se . This pape does no ecommend his eg ession me hod o
simul aneous equa ions, because i may cause he eg ession coe icien s o be biased. The e o e each
o he es s (i.e. MS o T and T o MS) is mu ually exclusi e.
19
MMRA is a logical ex ension o he mul iple eg ession concep o allow o mul iple esponse
(dependen ) a iables. Mul i a ia e eg ession es ima es he same coe icien s and s anda d e o s as
would be ob ained using sepa a e OLS eg essions o each ou come a iable. Howe e , he OLS
eg essions will no p oduce mul i a ia e esul s, no will hey allow o es ing o coe icien s ac oss
equa ions. On he o he hand, mul i a ia e eg ession, being a join es ima o , also es ima es he
be ween-equa ion co a iances. This means ha i is possible o es coe icien s ac oss di e en
ou come a iables. Hence, MMRA allows o mul i a ia e es s o a collec ion o wo o mo e
esponses, each in wo o mo e p ac ices. In o he wo ds, i allows es ing o wo o mo e esponses o
Ys p edic ed by wo o mo e p ac ices o Xs.
Finally, he e a e a leas wo issues o conside when applying MMRA in his pape :
1. The esiduals om mul i a ia e eg ession models a e assumed o be mul i a ia e no mal.
This is analogous o he assump ion o no mally dis ibu ed e o s in uni a ia e linea
eg ession (i.e. OLS eg ession).
2. The ou come a iables should be a leas mode a ely co ela ed o he mul i a ia e
eg ession analysis o make sense.
5. RESULTS
A wo-s ep p ocedu e was used when pe o ming he da a analysis. Fi s , canonical
co ela ion analysis was pe o med o es he mul i a ia e ela ionship ac oss he a iables
ep esen ing T and MS p ac ices (H1). The signi icance o his es p o ided he basis o wo
se ies o indi idual and mu ually exclusi e mul i a ia e mul iple eg ession analyses—one o
each o he nex wo hypo heses (H2 and H3). This is a eg ession p ocedu e ha enabled
assessmen o he e ec o all h ee p ac ices o an MP domain on all h ee p ac ices om he
o he MP domain.
20
The canonical co ela ion analysis indica ed a signi ican mul i a ia e ela ionship ac oss MS
and T a iable se s, hus lending suppo o he ela ionship hypo hesis H1. The s a is ical
analysis ega ding he selec i e i be ween MS and echnology p ac ice se s h ough he
associa ion o canonical co ela ion be ween hese a iables allows o he deduc ion o he
combina ions ha desc ibed he ollowing esul s in Table 1.
Take in able 1
Table 1 shows he esul s o a canonical co ela ion analysis be ween h ee echnology
p ac ices and ano he h ee manu ac u ing s a egy- ela ed p ac ices ep esen ing he main
ope a ions managemen a eas. Only he i s canonical pai was s a is ically signi ican . The
canonical co ela ion (0.77) is high. Al hough he e a e no guidelines abou he minimum
accep able alue o he edundancy index, gene ally he highe he alue o he index he
be e . Thus, he e is e idence o he impac be ween he MS and T p ac ice se s, since he
edundancy index shows ha close o hal o he a iance in he T p ac ices se is explained by
he i s canonical a iables o MS- ela ed p ac ices and ha a ound one hi d o he a iance
in he MS p ac ices se is explained by he i s canonical a iables o T- ela ed p ac ices.
These esul s indica e ha he e is a e y s ong ela ionship be ween MS p ac ices and T
p ac ices.
T adi ionally, canonical pai s ha e been in e p e ed by examining he sign and he magni ude
o he canonical weigh s. Howe e , hese weigh s a e subjec o conside able ins abili y due o
sligh changes in sample size, pa icula ly whe e he a iables a e highly co ela ed.
Canonical c oss-loadings ha e been sugges ed as a p e e able al e na i e o he canonical
weigh s (Hai e al., 1998). The canonical c oss-loadings show he co ela ions o each o he
dependen a iables wi h he independen canonical a ia e, and ice e sa. Table 1 shows he
21
canonical c oss-loadings o he i s canonical pai . A loading o a leas 0.31 is conside ed
signi ican ly di e en om ze o a he 5% signi icance le el (G aybill, 1961). Acco ding o
his c i e ion, each o he MS a iables is signi ican ly ela ed o he T canonical a ia e
(canonical a ia e ep esen ing p ac ices). On he o he hand, all T a iables (p ac ices) a e
signi ican ly ela ed o he MS canonical a ia e (canonical a ia e ep esen ing p ac ices). I
is impo an o s ess ha he manu ac u ing-business s a egy linkage is he mos impo an
ac o in accoun ing o he i s canonical a iable o T- ela ed p ac ices, bu he o he wo
MS p ac ices a e no a behind. On he o he hand, e ec i e p ocess implemen a ion shows
he highes co ela ion wi h he i s canonical a iable o MS- ela ed p ac ices, a in ad ance
o he o he wo T p ac ices.
These esul s o he in e na ional au o supplie plan s suppo hypo hesis 1 since he e is a
cong uency displayed h ough a ela ionship be ween Manu ac u ing S a egy and
Technology. Thus, he success o manu ac u ing indus ies may o en be a ibu ed o he
links be ween hei own pa icula p ac ices: echnology- ela ed p ac ices mus be
accompanied by MS- ela ed p ac ices, which is one o he mos impo an easons why some
manu ac u ing companies achie e a desi able e ec i eness le el in he global ma ke place.
The e o e, canonical co ela ion analysis p o ides a good es o he o e all ela ionship
speci ied by he hypo hesis, as well as a basis o u he eg ession analysis o he e ec s o
he indi idual a iables.
Nex , wo sepa a e mul i a ia e mul iple eg ession analyses (one pe each MP se as a
p edic o ) we e pe o med o es hypo heses H2 and H3. Thus, wo s ages o bo h
independen eg essions will be shown, he i s s age ocusing on he mul i a ia e es s and
he second on he es s o be ween-subjec s e ec s. The second s age o MMRA may be
ea ed in a simila way o mul iple linea eg ession. Thus, in line wi h Umana h and Kim’s

22
(1992) and Umana h’s (2003) conclusions on cong uency and om he i s pa o he
MMRA, equa ions [1] and[2] we e used, whe e MS ep esen s Manu ac u ing S a egy and T,
Technology. The MS and T indexes 1, 2, and 3 ep esen he h ee co esponding p ac ices o
each se o manu ac u ing p ac ices (3 MS p ac ices and 3 T p ac ices) explained in sec ion
4.1 (page 15), he βs a e he i coe icien s associa ed wi h hei espec i e a iables, i=1-3
ep esen s he same h ee p ac ices abo e o each se o MS and T manu ac u ing p ac ices,
and ε is he e o .
MSi = βmsi + βmsiT1T1 + βmsiT2T2+ βmsi 3T3 + εmsi [1]
Ti = β i + β ims1MS1 + β ims2MS2 + β ims3MS3 + ε i [2]
The selec ion pe spec i e is suppo ed by he s a is ical signi icance o β associa ed wi h he
in e es independen a iable (MS1, MS2 and MS3 o equa ion 1 and T1, T2 and T3 o
equa ion 2).
Thus, o he i s MMRA wi h MS as a p edic o , mul i a ia e es s gi e he ollowing o he
MS p ac ices: Pillai’s T ace, Wilks' Lambda, Ho elling's T ace and Roy's La ges Roo a e all
signi ican . All p ac ices om MS collec i ely may p edic he p ac ices o T as an ou pu .
Hence, o he i s s age o he i s eg ession, all MS p ac ices a e signi ican o po en ially
p edic ing he T se , o in o he wo ds, all 3 MS p ac ices may p edic all 3 T p ac ices (Table
2a). Following up, Table 2b shows all esul s o he second s age o he i s MMRA (equa ion
1) using a ows o indica e signi ican ela ionship di ec ions om es s o be ween-subjec s
e ec s. In iew o he o egoing esul s ( ega dless o he ac ha he e does no seem o be
comple e cong uency), his pape could conclude wi h ese a ions (MS does no in luence
23
T3) ha hypo hesis 2 has been pa ially p o en: manu ac u ing s a egy in luences
echnology.
Take in able 2
Table 2b shows he esul s o he i s model (equa ion 1) in mo e de ail. The columns
ep esen MS p ac ices, which we e es ed o see whe he each p ac ice p edic ed he ows as
echnology p ac ices. The consequen es ima ed pa ame e s om his es show echnology
p ac ices ha a e in luenced by he manu ac u ing s a egy p ac ices. Thus, only o mal
s a egic planning (MS2) does no signi ican ly p edic in e unc ional design e o (T1),
p obably due o some ype o es ic ion caused by planning. MMRA showed ha bo h MS1
and MS3 ha e posi i e impac s on T1 (βs a e 0.300 and 0.328). In he nex ow, he e ec i e
p ocess implemen a ion (T2) ow shows ha his is signi ican ly p edic ed by all he
manu ac u ing s a egy p ac ices (a he 1%, 10% and 5% signi icance le els, espec i ely).
MMRA calcula ions showed ha all MSs ha e posi i e impac s (βs a e 0.309, 0.169 and
0.281) on T2. Finally, echnology supplie in ol emen (T3) does no seem o be signi ican ly
p edic ed by any o he manu ac u ing s a egy p ac ices, possibly due o he ac ha i is
some hing ha he company canno comple ely con ol (con ex ual ac o s ela ed o
supplie s). This can all be summa ised as ollows. 5 ou o 9 con igu a ions a e signi ican :
• MS (all bu MS2) p edic s T1
• MS (all i s 3 MS’s) p edic s T2MS does no p edic T3
The ollowing possible unidi ec ional cong uency ela ionships a e he e o e ob ained:
1. Manu ac u ing s a egy (excep MS2) in e unc ional design e o .
Manu ac u ing s a egy  e ec i e p ocess implemen a ion.
24
Thus, hese esul s show ha all o he MS a iables (excep MS2, which is pa ial) in he
model ha e s a is ically signi ican ela ionships wi h he join dis ibu ion o in e unc ional
design e o and e ec i e p ocess implemen a ion. The e o e, i can be said ha
manu ac u ing s a egy in luences echnology o a ce ain deg ee, as e lec ed in mos
p ac ices, and ha as a esul , hypo hesis H2 has been pa ially ul illed.
Fu he mo e, whils s essing ha bidi ec ional ela ionships a e no wi hin he scope o his
pa o he s udy, hypo hesis 3 was independen ly p o en o a ce ain deg ee in he MMRA
second s age: echnology in luences manu ac u ing s a egy (Table 3b). This will be explained
in de ail below.
Thus, in he i s s age o he second independen eg ession, only T1 is no signi ican o
po en ially p edic ing he MS se (possibly due o coo dina ion p oblems), o in o he wo ds
bo h T2 and T3 may p edic all 3 MS p ac ices (Table 3a).
Take in able 3
As in Table 2b, Table 3b se s ou he esul s o he second model (Equa ion 2). In his case,
ows a e he echnology p ac ices, whe e T2 and T3 we e each
5
es ed o check whe he hey
migh in luence he manu ac u ing s a egy p ac ices (columns). Thus, only e ec i e p ocess
implemen a ion (T2) p edic s an icipa ion o new echnologies (MS1) a he 1% signi icance
le el (MMRA showed a β o 0.657), p obably due o echnology p ocesses ha a ec his
an icipa ion s a egy. All echnology p ac ices bu T1 signi ican ly p edic o mal s a egic
planning, MS2 (a he 1% signi icance le el), since in e unc ional ac ions may equi e mo e
oom o wo k. MMRA showed T2 and T3 bo h had posi i e impac s (βs a e 0.794 and 0.157).
Finally, i can be seen ha all he echnology p ac ices excep o in e unc ional design e o
5
The i s s age o MMRA, mul i a ia e es , showed T1 was no signi ican .
25
(T1) (see Table 3b) signi ican ly p edic he manu ac u ing s a egy and business s a egy link,
MS3 (a he 1% and 5% signi icance le els), possibly due o s a egies being somewha igid.
MMRA showed ha bo h T2 and T3 ha e posi i e impac s (βs o 0.660 and 0.103).
As wi h he o he eg ession, on his basis i can be s a ed ha 5 ou o 9 con igu a ions a e
signi ican :
• T (all bu T1) p edic s MS2.
• T2 p edic s MS1
• T (all bu T1) p edic s MS3
The e o e, he unidi ec ional ela ionships can be summa ised as ollows:
1. Technology (excep T1)  o mal s a egic planning
2. E ec i e p ocess implemen a ion  an icipa ion o new echnologies.
3. Technology (excep T1)  manu ac u ing s a egy-business s a egy link
Thus, hese esul s show ha he T a iables (excep T1) in he model ha e a s a is ically
signi ican ela ionship wi h join dis ibu ion o MS p ac ices (e ec i e p ocess
implemen a ion is he only echnology p ac ice wi h a s a is ically signi ican ela ionship
wi h he MS p ac ice, an icipa ion o new echnologies). The e o e, i can be s a ed ha
echnology in luences manu ac u ing s a egy o a ce ain deg ee, hus pa ially con i ming
hypo hesis H3.
6. CONCLUSIONS AND FINAL CONSIDERATIONS
32
Ha mann, F., Moe s F., 1999. Tes ing con ingency hypo heses in budge a y esea ch: an e alua ion o he use o
mode a ed eg ession analysis. Accoun ing, O ganiza ions and Socie y, 24, 291-315
Hayes R.H., 1985. S a egic planning- o wa d in e e se?. Ha a d Business Re iew, 63, 67-77.
Hayes R.H., Wheelw igh S.C., Cla k K.B., 1988. Dynamic manu ac u ing: C ea ing he lea ning o ganisa ion.
New Yo k: F ee P ess.
Ho e , C. W., Schendel, D., 1978. S a egy Fo mula ion: Analy ical Concep s. S . Paul, MN: Wes Publishing.
I ami, H., Numagami, T., 1992. Dynamic In e ac ion Be ween S a egy and Technology. S a egic Managemen
Jou nal, l, 13, 119–36.
Kaplan, S.E., Mackey, J.T., 1992. An examina ion o he associa ion be ween o ganisa ional design ac o s and
he use o accoun ing in o ma ion o manage ial pe o mance e alua ion. Jou nal o Managemen Accoun ing
Resea ch, 4, 116–30.
Ke oki i, M., Sch oede , R.G., 2004. Manu ac u ing p ac ices, s a egic i and pe o mance. A ou ine-based
iew. In e na ional Jou nal o Ope a ions & P oduc ion Managemen , 24 (2), 171-91.
Leong, G. K., Snyde , D., Wa d, P., 1990. Resea ch in he p ocess and con en o manu ac u ing s a egy.
OMEGA, 18(2), 109—122.
Lu , J., Shields M.D., 2003. Mapping managemen accoun ing: g aphics and guidelines o heo y-consis en
empi ical esea ch. Accoun ing, O ganisa ions and Socie y, 28, 169–249
Maidique, M.A., Pa ch, B.J., 1988. Co po a e S a egy and Technology Policy. In: Tushman, M.L., Moo e,
W.L., (Eds.). Readings in Managemen o Inno a ion (2nd Ed.). Camb idge, MA: Ballinge , 236–248.
Maie , F., Sch oede , R., 2001. Compe i i e P oduc and P ocess Technology. In: Sch oede , R., Flynn, B.,
(Eds.). High Pe o mance Manu ac u ing-Global Pe spec i es. New Yo k: John Wiley & Sons, Inc., 74-114.
Ma ucheck, A., Pannesi, R., Ande son, C., 1990. An explo a o y s udy o he manu ac u ing s a egy p ocess in
p ac ice. Jou nal o Ope a ions Managemen , 9 (1), 101–23.
Ma sui, Y., 2002. Con ibu ion o manu ac u ing depa men s o echnology de elopmen : An empi ical analysis
o machine y, elec ical and elec onics, and au omobile plan s in Japan. In e na ional Jou nal o P oduc ion
Economics, 80, 185–97.
McKel ey, B., 1982. O ganisa ional Sys ema ics: Taxonomy, E olu ion, Classi ica ion. Be keley, CA:
Uni e si y o Cali o nia P ess.

33
McKone, K. E., Sch oede , R.G., 2002. A plan ’s echnology emphasis and app oach. A con ex ual iew.
In e na ional Jou nal o Ope a ions & P oduc ion Managemen , 22 (7), 772-92.
Meilich O., 2006. Bi a ia e Models o Fi in Con ingency Theo y. C i ique and a Polynomial Reg ession
Al e na i e. O ganisa ional Resea ch Me hods, 9 (2), 161-93
Meye , M. W., Zucke , L. G., 1989. Pe manen ly ailing o ganiza ions, Sage: Newbu y Pa k, CA.
Milg om, P., Robe s, J., 1995. Complemen a i ies and i : s a egy, s uc u e and o ganisa ional change in
manu ac u ing. Jou nal o Accoun ing and Economics, 19 (2-3), 179–208.
Mo i a, M., Sakakiba a S., 1994a. Linkage as a Key o Excellence o Managemen , Pa 1. Managemen 21
(Japan Managemen Associa ion), 4 (8), 48-52.
Mo i a, M., Sakakiba a S., 1994b. Linkage as a Key o Excellence o Managemen , Pa 2. Managemen 21
(Japan Managemen Associa ion), 4 (9), 44-48.
Mo i a, M., Flynn, E.J., 1997. The Linking among Managemen Sys ems, P ac ices and Beha io in Success ul
Manu ac u ing S a egy. In e na ional Jou nal o Ope a ions and Managemen , 17 (10), 967-93.
Mo se, E.V., 1977. The e ec s o o maliza ion o ganiza ional adop ion beha io . O ganiza ion and
Adminis a i e
Sciences, 8, 107-122.
Nassimbeni, G, 1996. Fac o s Unde lying Ope a ional JIT Pu chasing P ac ices. In e na ional Jou nal o
P oduc ion Economics 42 (3), 275-88.
Nunnally, J.C., 1967. Psychome ic Theo y. New Yo k: McG aw-Hill.
Nunnally, J.C., 1978. Psychome ic Theo y. 2nd ed., New Yo k: McG aw-Hill.
Olhage , J., 1993. Manu ac u ing lexibili y and p o i abili y. In e na ional Jou nal o P oduc ion Economics, 30-
31, 67-78.
Pa ke , A., 2000. Impac on he O ganisa ional Pe o mance o he S a egy-Technology Policy In e ac ion.
Jou nal o Business Resea ch, 47, 55–64.
Pa hasa hy, R., Se hi, S., 1993. Rela ing s a egy and s uc u e o lexible au oma ion: a es o i and
pe o mance implica ions. S a egic Managemen Jou nal, 14, 529–49.
Pennings, J.M., 1992. S uc u al con ingency heo y: a eapp aisal. In: S aw, B., Cummings, I.L., (Eds.).
Resea ch in O ganiza ional Beha io . G eenwich, CT: JAI P ess, 267-309.
34
Po e , M.E., 1983. The Technological Dimension o Compe i i e S a egy. In: Rosenbloom, R.S., (Ed).
Resea ch on Technological Inno a ion, Managemen , and Policy. G eenwich, CT: Jai., 1–33.
Po e , M.E., 1985. Compe i i e Ad an age—C ea ing and Sus aining Supe io Pe o mance. New Yo k: F ee
P ess.
P ahalad, C.K., Hamel, G., 1990. The Co e Compe ence o he Co po a ion. Ha a d Business Re iew, 79–91.
P im ose, P.L.., 1992. E alua ing he in oduc ion o JIT. In e na ional Jou nal o P oduc ion Economics, 27, 9–
22.
Robbins, S. P. 1990. O ganiza ion heo y: s uc u e, design, and applica ions (3 d ed.). Englewood Cli s, NJ:
P en ice-Hall Inc.
Sakakiba a, S., Flynn, B.B., Sch oede , R.G., Mo is, W.T., 1997. The Impac o Jus -In-Time Manu ac u ing
and i s In as uc u e on Manu ac u ing Pe o mance. Managemen Science, 43 (9), 1246-57.
Sch oede , R.G., Flynn, B., (Eds.), 2001. High Pe o mance Manu ac u ing-Global Pe spec i es. New Yo k:
John Wiley & Sons, Inc.
Simons, R., 1987. Accoun ing Con ol Sys ems and Business S a egy: an Empi ical Analysis. Accoun ing,
O ganisa ions and Socie y, 12, 357-74.
Skinne , W., 1969. Manu ac u ing - Missing link in co po a e s a egy. Ha a d Business Re iew, May-June,
136-45.
S obaugh, R., Telesio, P., 1983. Ma ch manu ac u ing policies and p oduc s a egy. Ha a d Business Re iew,
61(2), 113—20.
Umana h, N., 2003. The concep o con ingency beyond ‘‘I depends’’: illus a ions om IS esea ch s eam.
In o ma ion & Managemen , 40, 551–62
Umana h, N., Kyu. K., 1992. Task-S uc u e Rela ionship o In o ma ion Sys ems De elopmen Subuni : A
Cong uence pe spec i e. Decision Sciences, 23 (4), 819-38.
Van de Ven, A.H., D azin, R., 1985. The concep o i in con ingency heo y. In: S aw, B.M., Cummings, L.L.,
(Eds.). Resea ch in O ganisa ional Beha io . G eenwich, CT: JAI P ess.
Venka aman, N., P esco , J.E., 1990. En i onmen -s a egy coalignmen : an empi ical es o i s pe o mance
implica ions. S a egic Managemen Jou nal, 11 (1), 1–23.
Zah a, S., Co in, J., 1993. Business S a egy, Technology Policy, and Fi m Pe o mance. S a egic Managemen
Jou nal, 14, 451–78.
35
a. Selec ion: no pe o mance a ia ion
b. In e ac ion: pe o mance a ia ion
Figu e 1. Fi s. mis i
2a. All plan s
2b. Low le el MP2 plan s
2c. High le el MP2 plan s
Figu e 2. Rela ionship in a Selec ion Fi
(Adap ed om Ge din and G e e, 2004)
a. Bidi ec ional
b. Unidi ec ional MS-T
c. Unidi ec ional T-MS
Figu e 3. MS-Technology Rela ionship in a Selec ion Fi
MS T
MS T
MS T
36
Table 1. MS and T co ela ions
Fi s canonical pai
Canonical Co ela ion
0.7711
Likelihood a io
0.3803
Signi icance
0.0000
Redundancy index: MS
0.4361
Redundancy index: T
0.3117
Co ela ions be ween manu ac u ing s a egy p ac ices and canonical
a iable o echnology ela ed p ac ices (canonical c oss-loadings)
An icipa ion o New Technologies (MS1)
0.665
Fo mal S a egic Planning (MS2)
0.627
Manu ac u ing - Business S a egy Linkage (MS3)
0.669
Co ela ions be ween echnology p ac ices and canonical a iable o
MS ela ed p ac ices (canonical c oss-loadings)
In e unc ional Design E o s (T1)
0.490
E ec i e P ocess Implemen a ion (T2)
0.743
Technology supplie In ol emen (T3)
0.329
Table 2. MS se as p edic o
a. P edic o eg ession: signi icance on
mul i a ia e es s
b. MS o T: es s o
be ween-subjec s e ec s
MS1***
GL: 0.210, 0.790, 0.266, 0.266
F 6.478
MS2*
GL: 0.096, 0.904, 0.106, 0.106
F 2.573
MS3*
GL: 0.086, 0.914, 0.094, 0.094
F 2.297
MS1
MS2
MS3
T1
***
F 7.922
F 0.444
**
F 3.818
T2
***
F 16.863
*
F 2.640
**
F 5.633
T3
F 1.301
F 1.027
F 0.305
GL: Pillai’s T ace, Wilks' Lambda, Ho elling's T ace and Roy's La ges Roo espec i ely; * P ≤ 0.1, **P ≤ 0.05, *** P ≤ 0.01
37
Table 3. Technology se as p edic o
a. P edic o eg ession: signi icance on
mul i a ia e es s
b. T o MS: es s o be ween-subjec s
e ec s
T1
GL: 0.053, 0.947, 0.057, 0. 057
F 1.375
T2***
GL: 0.459 , 0.541, 0.847 , 0.847
F 20.609
T3**
GL: 0.107, 0.893, 0.120, 0.120
F 2.919
MS1
MS2
MS3
T1
F 0.388
F 2.770
F 0.074
T2
F 25.372
***
F 39.187
***
F 34.857
***
T3
F 1.962
F 7.658
***
F 4.249
**
GL:Pillai’s T ace, Wilks' Lambda, Ho elling's T ace and Roy's La ges Roo espec i ely; **P ≤ 0.05, *** P ≤ 0.01