Our study found weak significance of IS infrastructure, top management support, IT expertise, resources, and organizational size on IT adoption of technology while formalization, centrali[r]
Trang 1A meta-analysis of relationships between organizational characteristics and IT
Department of Information Systems and Computing, Brunel University, Uxbridge, Middlesex UB8 3PH, UK
1 Introduction
Aninnovationcanbeanidea,product,program,ortechnology
thatis newtotheadoptingunit Itsadoption is a processthat
resultsin itsintroduction anduse that is newtotheadopting
personor organization.ITadoption presentspotentialadopters
withnewmeansofsolvingproblemandexploitingopportunities
Inthepasttwodecades,ISresearchershavefocusedparticularlyon
studyinginnovationintheadoptionofIT.Actualadoptioncanbe
initiatedbyeitheraresponsetoachangeintheenvironmentor
wheninnovation becomesa requirement forthe organization’s
operation or a belief by management that it will improve
organizational performance: an organization which possesses
thefinancialandtechnicalresourceshasastrongermotivationto
innovate
Studies addressing innovation adoption have often yielded
inconsistentandconflictingfindings;e.g.,RyeandKimberly[14]
statedthat inconsistencyhasbeena defining theme.Thusit is
almostimpossibletodrawfirmconclusionsontheeffectsofthe
factorsthatinfluenceITadoption
Identifyingthesefactorsisfundamentalinensuringsuccessful
adoptionandimplementationofIT.Wethereforeattemptedtofill
theknowledgegapbyvalidatingtheimportantdeterminantsof
organizationalITinnovationadoptioninanorganizationalcontext
Aggregating existing literature allows the validation of their findingsandclarificationof theinconsistenciesthatmightexist amongst primary studies [8] The key research question that motivatedourstudywastherefore:
Whatarethekeyorganizationalfactorsthatguidesuccessful adoptionandimplementationofITinnovationinorganizations?
2 Background 2.1 ITinnovationadoption Manypapershavebeen writtenonIT adoptionat boththe organizationalandindividuallevels.Organizationallevelstudies haveexaminedtheprocessofadoptionanddiffusionofIT[12] assuming thatIT improves the organization’s operational and strategic practices Similarly, many authors have examined a rangeoffactorsinfluencingITadoption.Four majorcategories are technological, organizational, environmental, and individual [1,5] In a technological context, perceived benefits, cost, complexityandcompatibilityarekeydeterminants.For organi-zationalcharacteristics,itssize,supportfromtopmanagement, existingresources,andITexpertiseareconsideredrelevant[6] Competitive pressure, demands from trading partners and customers, support from government, and environmental uncertainty have also been considered to be environmental factors[3,13].TheCEO’sknowledgeofITandhisorherattitude towards IT, amongst others have also been mentioned as importantfactors[4]
A R T I C L E I N F O
Article history:
Received 9 November 2010
Received in revised form 26 March 2012
Accepted 18 April 2012
Available online 24 May 2012
Keywords:
IT innovation adoption
IT implementation
Adoption and diffusion
Meta-analysis
Moderating effect
A B S T R A C T AdoptionofITinorganizationsisinfluencedbyawiderangeoffactorsintechnology,organization, environment,andindividuals.Researchershaveidentifiedseveralfactorsthateitherfacilitateorhinder innovationadoption.Studieshaveproducedinconsistentandcontradictoryoutcomes.Weperformeda meta-analysis of ten organizational factors to determine their relative impact and strength We aggregated their findings to determine the magnitude and direction of the relationship between organizationalfactorsandITinnovationadoption.Wefoundorganizationalreadinesstobethemost significantattributeandalsofoundamoderatelysignificantrelationshipbetweenITadoptionandIS departmentsize.OurstudyfoundweaksignificanceofISinfrastructure,topmanagementsupport,IT expertise, resources, and organizational size on IT adoption of technology while formalization, centralization,andproductchampionwerefoundtobeinsignificantattributes.Wealsoexaminedstage
ofinnovation,typeofinnovation,typeoforganization,andsizeoforganizationasmoderatorconditions affectingtherelationshipbetweentheorganizationalvariablesandITadoption
ß 2012ElsevierB.V.Allrightsreserved
* Corresponding author.
E-mail address: mumtaz.abdulhameed@brunel.ac.uk (M.A Hameed).
ContentslistsavailableatSciVerseScienceDirect Information & Management
j our na l ho me pa ge : w ww e l se v i e r com / l oca t e / i m
0378-7206/$ – see front matter ß 2012 Elsevier B.V All rights reserved.
Trang 22.2 OrganizationalcharacteristicsandITinnovationadoption
Wefocused ontheassociationbetween organizational
char-acteristicsandITadoption
Tsaoetal.[17]assessedtopmanagementsupport,
organiza-tionalreadiness,ITinvestmentandstaffresistancewithrespectto
organizational perspective in identifying success factors of
business-to-businesse-commerceadoptioninTaiwanese
compa-nies Similarly, Teo and Ranganathan [16] in discriminating
adoptersandnon-adoptersof business-to-businesse-commerce
inSingaporeanfirmsconsidered thedemographicprofileof the
organization, presence of a champion, formal plan, years of
e-commerce experience, expected and realised benefits from
e-commerce,managementsupport,andriskorientationas
organi-zationalfactors
2.3 Meta-analysis
Thefindingsofanumberofrelatedstudiesmaybeaggregated
to find an overall outcome Although such information is
sometimesqualitative, if it is quantitative,traditional methods
of aggregating data have focused on combining statistically
significant results, performing tests to verify factors affecting
theoutcome
Meta-analysisinvolvesseriesofproceduresforquantitatively
accumulatingeffectsizesacrossstudiesandanalyzingtheirresult
to reach overall conclusions Effect sizes can be expressed in
similar forms as correlation coefficients The meta-analysis
procedureshould,of course,include a waytocorrect sampling
errors,errorsofmeasurement,andrangeofvariance
2.4 Moderatorsforrelationshipsbetweenorganizational
characteristicsandITadoption
WefirstreviewedliteratureonITadoptionandorganizational
factorsinfluencingtheadoptionofIT.Thereviewshowedmixed
resultsinthefindingsasresearchhadbeenconductedindifferent
surroundings,sectors,groupsanddemographicconditions.Using
meta-analysisprocedures,itwaspossibletoexaminetheeffectsof
thesemoderatorsoftherelationshipbetweenorganizationalfactor
andITadoption
Weexaminedtheeffectoffourmoderators:stageofinnovation,
typeofinnovation,typeoforganization,andsizeoforganization
2.4.1 Stageofinnovationadoption
Theadoptionofinnovationinorganizationscanbeconsidered
tobeastage-basedprocesswhichmaybedividedintoanumber
of phases spanning from decision-making to its use for
organizational processing The pre-adoption stage identifies
theneedforaninnovation,gatheringknowledgeofthesolution
toreplacetheexisting.Adoption-decisionconsistsofevaluating
thetechnology,makingadecisiontoaccepttheinnovationand
allocation of resources to facilitate an environment for its
implementation Finally, the post-adoption stage includes the
acquisitionoftheinnovation,acceptanceofitbytheusersand
use of the innovation by the organization For our study, we
dividedtheliteratureintostudiesthatwereconductedforeither
initiation, adoption-decision, implementation, and those
con-ductedinallstages
2.4.2 Typeofinnovation
Innovation can be split into either: product or process
innovation Different factors influence the adoption of product
and process innovation and the degree to which innovation
impacts organizational performance We considered them as a
moderatingconditionontheadoptionofITinorganizations
2.4.3 Typeoforganization OrganizationsadoptITtoenhancethescopeoftheirproducts andservices.Almostallindustries,publicorgovernmentalutilize
IT According to their NAICS or SIC – UK, manufacturing organizationsareengagedinthetransformationofmaterialsor substancesintonewproductswhileserviceindustriesareengaged
inprovidingservicesforindividuals,businessesandgovernment establishmentsandother.FactorsinfluencingtheusageofITinthe manufacturingsectoraredifferentfromthoseintheservicesector [2].Wethereforemadeadistinctionbetweenthesetwotypesof organizations
2.4.4 Sizeoforganization Largeandsmallorganizationspossessdistinctcharacteristicsof theirownandarefundamentallydifferentinanumberofaspects
IT innovation adoption that influences larger organizational contextsmaynotbeappliedtosmallbusinesses.Inouranalysis,
we therefore adoptedsize ofthe organizationas a moderating factoranddividetheliteratureinto(a)studiesoflarge organiza-tions,(b)studiesforsmallorganizationsandSMEs,and(c)studies formixed-sizeorganizations
3 Researchmethod
Byusingmeta-analysis,weidentifiedthesignificant relation-shipbetweentheorganizationalcharacteristicsandITadoption Theaimofusingameta-analysisprocedureforthestudywasto evaluatefindingsofpaststudiesthathadexaminedorganizational attributes affectingIT adoption;thesewerethen aggregatedto obtainoverallconclusionsregardingthemagnitudeanddirection
of the relationships This allowed us to examine theeffects of differentmoderatorsinfluencingtherelationshipbetween organi-zationfactorsandITinnovationadoption
3.1 Studyselection
Wereviewedninety-twopublishedstudiesonITadoption.We searched IS Journals and Google Scholar with the key words innovation,adoption,diffusion,infusion,integration, implemen-tation,informationtechnology,informationsystemandITusageto obtainrelevantarticles.Thestudiesassessedwerefrom1990to
2009.Thestudyselectioncriteriaforthemeta-analysiswere: (a)ItwasanempiricalstudyonITinnovationadoption
(b)Thestudyexaminedtheorganizationalcontextofadoptionof innovationinitsempiricalevaluation
(c)Thedependentvariablesincludedinitiation,adoption-decision (adoption)orimplementation
Followingtheabovecriteria,weobtainedatotalof59empirical studiesfortheanalysis.Amongthese,41werepublishedbetween theyears2000and2009and18betweentheyears1990and1999
Assomeofstudiesconsideredmorethanonetypeofinnovation and different stages of innovation adoption, a total of 97 IT innovationadoptionrelationshipswereobtained
Thestudiesuseddifferentstatisticaltreatmentoranalysis.31 used correlation in their analysis; 5 were based on regression techniqueswhile5useddiscriminantanalysis.Descriptivestudies were conductedby 7 studies and 11 employedother forms of statistical evaluation We selectedstudiesthat usedcorrelation techniquesforourmeta-analysisandgatheredallorganizational factorsconsideredinthe59studies.Toperformthemeta-analysis,
wefilteredthestudiesthatprovidedcorrelationcoefficientsforthe relationshipbetweenorganizationalfactorsandITadoption.Thus
atleasttwocorrelationresultsforeachorganizationalfactorwere requiredtocarryouttheanalyticalprocedures.Studiesprovided
Trang 3meta-analysis
Table1 showsthetenorganizationalvariablesconsidered in
our study and the expected association with IT innovation
adoption based from the literature A positive association
facilitatedITadoptionwhileanegativeassociationinhibitedit
3.2 Coding
Before conducting the analysis, we coded dependent and
independent variables Adoption of IT was considered the
dependent variable and the organizational factors influencing
the adoption of IT was the independent variable Studies that
includedmorethan one innovationwerecoded separately and
treatedasindividualdata sets.Theindependent variables were
organizationalcharacteristicsthatinfluencedinitiation,
adoption-decision,andimplementationofanITinnovation
Thevariousstudiesuseddifferentnamestodescribesomeof
the independent variables Thus, in coding the independent
variable,wereferredtothecontextinwhichthevariables were
usedinthestudy.Table1alsoshowsthedifferentnamesusedin
thestudiestorefertotheindependentvariables
Information onthefourmoderatorswasalsocoded foreach
study.Fourmoderatorsandtheircategoriesweredefinedas(a)
stageofinnovation:initiation,adoption,implementation,mixed;(b)
typeofinnovation:product,process,mixed;(c)typeof
organiza-tion:manufacturing,service,mixed;and (d)sizeoforganization:
large,smallandmediumenterprise(SMEs),mixed
TheAppendixAshowsstudiesconsideredinouranalysis
3.3 Significancetestandcorrelationcoefficient
The resultsof therelationshipbetween organizational
char-acteristics and IT adoption were evaluated in terms of test of
significance.Statisticalsignificancedenotestheprobabilitythata
relationshipexistsbetween individualorganizational
character-isticsand IT adoption Thetest of significance verifies that the
observed value differs from the theorized value and statistical
significanceisdeterminedbyeffectsizeandsamplesize.Hence,
studieswiththesameeffectsizecouldhavedifferentstatistical
significance and thusaggregating the test of significance could
produceaconfusingoutcome
All59empiricalstudiesthatweusedinouranalysisprovided significant test results for differentindependent variables We analyzedtheaggregatedresultoftestofsignificancetoverifythe importanceofdifferentorganizationalfactors.Alsowedetermined thattheaggregatedsignificanttestshowedaninconsistencyacross thestudiesofITadoption
Valuesofthecorrelationcoefficientrangebetween 1and+1; values that fall 0 to 1 indicate a negative relationship A correlation coefficient of 0 denotes that the variable has no relationship Correlation coefficients are not precise but are generally classifiedas weak, moderate,or strong In ourstudy,
weassumed0–0.09tobeinsignificant,0.10–0.29tobeweakly significant, 0.30–0.49to bemoderately significant,0.5–0.69 tobe stronglysignificant,0.70–0.89tobeverystronglysignificantand0.9– 1.0nearlyperfect
3.4 Themetaanalysisprocedures
A sequence of procedures was used to aggregate statistical resultsfromindependent.Theproceduresinvolvedaccumulating effect sizes across studies, combining, and evaluating them to obtain an average effect size The statistic extracted from the studieswasthecorrelationcoefficientandweperformedfivebasic stepsinouranalysis
1Computethemeancorrelationcoefficientforthestudies
We first calculated the mean population correlation by converting each of the observed correlation values into populationcorrelationandaveragingthevalues.Thisprovided
aweightedmeancorrelationforeachobservedestimatebyits correspondingsamplesize.Tocalculatethemeanpopulation correlation,wethenmultipliedeachcorrelationcoefficientby itscorrespondingsamplesizeanddividedbytotalsamplesize This frequency weighted average gives a greater weight to results obtained from larger samples Averaging population correlationsacross studieseliminates the effectof sampling error.Thecorrelationcoefficientisnotnormallydistributedand itsvarianceisnotconstant.Fisher’sz-transformationisoften used to normalize the distribution and stabilize potential variance.Wecalculatedthisfromthemeancorrelationvalues and the z-values were then used to compute confidence internals
Table 1
Ten organization characteristics and their expected relationship to IT adoption.
association Organizational size Number of employees within the organization or total sales revenue Business size Positive
IT expertise Prior experience of IT in term knowledge of individuals and within
the organization
Technology competence, Technical capability, IT knowledge,
IT sophistication, Employees IT knowledge, Existence of IS department, Knowledge of IT in company, IT maturity, Education
Positive
Top management
support
Extent of commitment of resource and support from the top management to the innovation
Resources Amount of financial, technical and human resources for the adoption
process
IS department size Existing IT function and dedicate IT personal within the organization IT function size Positive
IS infrastructure Availability of IT resources within the organization for the adoption IT resources, IT sophistication Positive Formalization The extent of the use of rules and formal procedures within the
organization
IS structure, Technology strategy, Organizational objective consensus
Negative Centralization Level of centralization of decision making in organization Organizational structure, Decision
making pattern
Negative Organizational
readiness
Level of awareness, resources, commitment and governance for adoption Technical competence, IT maturity,
Education
Positive Product champion Existence of high level individual to promote the innovation within
the organization
Innovation champion, Technology leader Positive
Trang 4The observed variance is explained by variations due to
populationcorrelationandsamplecorrelationsproducedbythe
samplingerror,whichaddstothevarianceofcorrelationsacross
studies.Variationduetopopulationcorrelationcanbeobtained
byeliminatingvariationduetosamplingerror
3Calculatethesamplingerrorvariance
Wefirstderivedtheeffectofvariancebysamplingerrorby
usingthemeanpopulationcorrelationandaveragesamplesize
Bysubtractingsamplingerrorvariancefromthevarianceinthe
sample correlation (observed variance), the variance due to
populationcorrelationwasobtained
4Computethepercentageofobservedvarianceexplainedbythe
samplingerrorvariance
Toaccountformoderatoreffectoftheindividual
organiza-tional attributes, we calculated the percentage of observed
variance explained by sampling error variance If the
percentage of the observed variation is mostly due to
samplingerrorvariance, a moderatoreffectcan beassumed
tobeminimal If thepercentage obtained is not sufficiently
high, a substantial amount of observed variance is due to
variation in population correlations This indicates that
the study requires the examination of moderator effect
For every organizational attribute that showed a sampling
error variance of less than 60% of observed variance,
we introducedfour moderating conditions and performeda
meta-analysis
5Computethe95%confidenceintervalusingthemeancorrelation
Finally,tofindthesignificanceoftheindependentvariable
inITadoption,wecomputeda 95%confidenceintervalusing
thevalues obtained fromz-transformation of mean
correla-tion The confidence intervals cannot be computed directly
usingmeancorrelationcoefficientduetovarianceinsample
size of individual studies Use of Fisher’s z-transformation
valuemakesitpossibletocalculatethesevaluesindirectly.The
relationship between independent variable and IT adoption
was regarded as statistically significant if the confidence
interval didnot include zero If the 95%confidence interval
was in the range 0–1, itindicated a positive association;if
the interval fell between 0 and 1, it implied a negative
association
3.4.1 Meta-analysisprocedureandmoderators
Ourreviewof studiesonfactorsaffectingtheadoptionof IT
showed mixed findings It was therefore necessary to explore
differentmoderatorsthat mayhave influencedthe relationship
betweenorganizationalfactorandITadoptions.Toexaminetheir
effect,studiesweredividedintosubgroupsofthesemoderators
andmeta-analysesweremadeforeachsubgrouptodeterminethe
strengthoftherelationshipbetweenorganizationalfactorsandIT
adoptioninthatcategory
4 Results 4.1 Significanttestresults
Ofthe59studies,atotalof97innovationadoptionrelationships with organizational factors were assessed Six relationships considered the initiation stage of adoption, 58 relationships examined theadoption-decision stage,27 verified relationships
attheimplementationstageandsixassessedmixedstages Table2showstheaggregatedsignificancetestresultsforallour independent variables It shows that, for each organizational attribute, the number of studies found to be significant or in agreement withthehypothesis and thenumberoffoundtobe insignificantorindisagreementwiththehypothesis
Thuswefoundthat,exceptforformalization,allorganizational variableshadarelationshipwithITadoption
The aggregated test of significance did not provide the magnitudeanddirectionoftheeachattributeinITadoption.In addition,theresultsdonotprovideanymechanismfor generaliz-ing and identifying the impact of different organizational attributes for IT adoption However, significance tests demon-stratedinconsistencyoffindingsinthestudies
4.2 Meta-analysissummaryoffindings
Weconductedmeta-analysisproceduresfortenorganizational characteristicsthat influenced theadoptionof ITand extracted studiesthatprovidedcorrelationvalues.31studieswereusedto carry out the meta-analysis procedures; fromthese,57 sets of correlationvalueswithITinnovationadoptionrelationshipwere obtained
4.2.1 Overallfindings Table3showstheresultsofthemeta-analysisofrelationships betweenthetenorganizationalfactorsandITadoption
Themeta-analysisresultsconfirmedtherelationshipbetween organizational attributes and ITinnovation adoptionexcept for formalization,centralizationandproductchampion.Mean corre-lation results showed that the strongest relationship with innovationadoptionwasorganizationalreadiness.Theimpactof organizationalreadinessonITadoptionwasevidentfromthe12 studiesonthevariable,ofwhich10weresignificant.However,this meta-analysisresultneededtobeconsideredwithcautionsince onlytwoinnovationadoptionrelationshipswereused.Thestudy was not able to perform moderator effects for organizational readinesspartlyduetolackofinnovationadoptionrelationships andbecausethesamplingerrorvarianceofthevariablewasfound
tobemorethan60%
OurresultsalsosuggestthatISdepartmentsizehadamoderate effect and that IS infrastructure, top management support, IT expertise, resources and organizational size had only weak significancewithITadoption.Wehadexpectedthatorganizational
Table 2
Aggregated significance test result.
Organizational factors No of studies No of innovation Significant Not significant % Significance
Trang 5adoption However, the amount of resources that has to be
committedforadoptingIThasbecomeminimal
The mean correlationresultsof formalization, centralization
andproductchampionshowedinsignificancefortheirrelationship
withITadoption.Onlytwoinnovationadoptionrelationshipswere
foundtoperformthemeta-analysisofproductchampion.More
than70%ofthestudiesthatinvestigatedproductchampionfound
ittobeasignificantattributeinITadoption
The results of95%confidence intervalfound anassociation
withITadoptionexceptforcentralizationandproduct
champi-on For formalization, most past studies suggest a negative
association; our meta-analysis found it to be positive This
impliesthatformalizationfacilitatesadoptionbyfollowingrules
and use of formal procedures Formalization permits better
decentralizationofISdevelopmentandsmoothermanagement
of IS projects The result for explained variance (EXP VAR)
showed that except for organizational readiness and product
champion,allfactorshavesamplingerrorvariancelessthan60%
oftheobservedvariance
5 Discussion
5.1 Organizationalsize
Organizationalsizehasbeenfrequentlyexaminedinthestudy
of organizational innovation adoption However, the impact of
organizationalsizeonITadoptionhasbeenfoundtobemixed;in somestudiesitwasfoundtobeanimportantattributewhilein others it was found it to be insignificant Availability of slack resourcesinlargerorganizationsfacilitatesinnovationadoption [19] but the flexible organizational structure and centralized decision-making in smaller organizations assists innovation adoption[20]
However, most researchhas hypothesized that larger orga-nizationstendedtoadoptITmorerapidlythansmallorganizations 5.1.1 Effectofmoderatoreffectonorganizationalsize
Table4illustratestheresultsofthemeta-analysisofmoderator effects on the relationship between organizational size and adoptionofIT
For all the subgroups categorized by stage of innovation (initiation, adoption, implementation and mixed), their mean correlation and 95% confidence intervals showed a significant (>0.10)andpositiveassociationbetweenorganizationalsizeand
IT adoption The results for mean correlation suggested that organizational sizewasa more significantattribute for process innovationthanproductinnovation,presumablybecauseprocess innovationinvolves replacingtheentire systemorwork proce-dure;smallorganizationswillhavelessavailableresources.Sizeof organization was a better determinant of IT adoption in manufacturingorganizationsthanserviceorganizations
Itisimportanttonotetheweaksignificance(correlationvalue between0.10and0.29)ofsizetoITadoptioninmostmoderating
Table 3
Meta-analysis results of organizational factors.
INN STD, no of innovation studied; SAM SIZ, sample size; MEN COR, mean correlation; ZTR VAL, z-transformation; OBS VAR, observed variance; SAM EVA, sampling error variance; EXP VAR, explain variance; COF INT, 95% confidence interval.
Table 4
Meta-analysis result of organizational size.
Organizational size
Stage of innovation
Type of innovation
Type of organization
Size of organization
INN STD, no of innovation studied; SAM SIZ, sample size; MEN COR, mean correlation; ZTR VAL, z-transformation; OBS VAR, observed variance; SAM EVA, sampling error variance; EXP VAR, explain variance; COF INT, 95% confidence interval.
Trang 6realisedthebenefitofITandhavebeenutilizingitintheiroperations
Theresultoforganizationalsizefromthemeta-analysisbyLeeand
Xia[9]weremoresignificantthanourresults.Inourstudy,however,
weincludedmorerecentstudiesandthismighthaveinfluencedthe
overallresult
5.2 ITexpertise
In an organization, knowledge of IT is a major needin the
adoptionofnewtechnologies.Anorganizationwithknowledgeof
anewinnovationcanadoptaninnovationeffortlesslyandretain
knowledgeforinnovationadoption[10]
ITexpertiseisakeydeterminantoforganizationalinnovation
adoption In our study, IT expertise was examined by 18
researchers.The numberofrelationships betweenIT expertise
andIT adoption found were 32.25 studies concluded thatIT
expertisepossesseda significantrelationship withITadoption
whilesevenstudiesdisagreed.WeconcludethatITexpertiseis
one of the major factorsfacilitating innovation adoption The
resultof95%confidenceconfirmedtheassociationbetweenIT
expertiseandITadoption.Meancorrelationofthemeta-analysis confirmed a weakly significant relationship (correlation value between0.10and0.29)betweenITexpertiseandITinnovation adoption
5.2.1 FindingsofmoderatoreffectonITexpertise The meta-analysis results are shown in Table 5 The mean correlation and 95% confidence interval verified that all four assumed moderators influenced the relationship between IT expertise and IT adoption This result is consistent with past literature[15].ITexpertisewasabetterdeterminantforprocess innovationthanproductinnovation
Oneoftheimportantfindingsofthemoderatoreffectonthe relationship between IT expertise and IT adoption was its significance (strong significance – correlation value between
0.5 and0.69) forsmallorganizations LackofITexpertiseand insufficient knowledge ofthe benefitsof innovation inhibitsmall businesses from adopting IT Small organizations usually seek assistancefromITconsultingfirmsorITvendors.Usinganexternal source,ITcanbeadoptedinsmallorganizations;however,itmight struggletosustainitsuse
Table 6
Meta-analysis result of top management support.
Top management support
Stage of innovation
Type of innovation
Type of organization
Size of organization
INN STD, no of innovation studied; SAM SIZ, sample size; MEN COR, mean correlation; ZTR VAL, z-transformation; OBS VAR, observed variance; SAM EVA, sampling error
Table 5
Meta-analysis result of IT expertise.
IT expertise
Stage of innovation
Type of innovation
Type of organization
Size of organization
INN STD, no of innovation studied; SAM SIZ, sample size; MEN COR, mean correlation; ZTR VAL, z-transformation; OBS VAR, observed variance; SAM EVA, sampling error variance; EXP VAR, explain variance; COF INT, 95% confidence interval.
Trang 75.3 Topmanagementsupport
Top management support is one of the critical factors
influencingITimplementations.Thereisevidencewhichsuggests
thattopmanagementsupportispositivelyrelatedtotheadoption
ofnewtechnologiesinorganizations
InreviewingstudiesofITadoption,wefound24studiesthat
examinedtopmanagementsupportwith31setsofITinnovation
relationships;24oftheseshowedapositiverelationship,while7
showedaninsignificantassociationwithITinnovationadoption
ToverifytheeffectoftopmanagementsupportonITadoption,we
considered 10 studies with 13 relationships A 95% confidence
intervalverifiedtheassociationbetweentopmanagementsupport
and IT adoption Mean correlationresults of themeta-analysis
showedthat topmanagementsupporthad onlyaweak
signifi-canceonITadoption
5.3.1 Findingsofmoderatoreffectontopmanagementsupport
Table 6 shows the result of the moderator effects on the
relationshipbetweentopmanagementsupportandITadoption
The mean correlation and 95% confidence interval of all four
moderatorsshowedittobesignificantwithapositiveassociation
betweentopmanagementsupportandITadoption
Onenotableresultwasthattopmanagementsupportwasmore
significantforlargerorganizationsthansmallorganizations
5.4 Resources
Organizational slack resources may be a fundamental
ingredient for innovation adoption Lack of technological
infrastructureand ITknowledge can bea major barrier forIT
adoption[18]
In our study, we used 11 studies that considered the
influence of resources on innovation adoption 23 sets were
explored,outofwhich14werefoundtobesignificant.However,
thesignificancetestresultswereinconclusive.Atotalof19sets
ofcorrelationrelationshipswereusedtoperformmeta-analysis
onresourceswithITadoption.Theserelationshipsweredrawn
from 8 studies The 95% confidence interval confirmed the
associationbetweenresourcesandITadoption.However,results
ofmeancorrelationofthemeta-analysisshowedthatresources
availableinanorganizationhadonlyaweaksignificanceonIT
adoption
5.4.1 Findingsofmoderatoreffectonresources Table 7 illustrates the results of the meta-analysis of the moderator effect on the relationship between resources and adoptionofIT.Thestageofinnovationwasfoundtobeasignificant moderator for the relationship The mean correlation result suggested that resources were a better determinant of the implementation stage of adoption than the other two stages Results found weak significance between resources and IT adoptionforbothproductandprocessinnovation
Therearetwofindingsabouttheresultsofmoderatoreffects
ontherelationshipbetweenresourcesandITadoption.First,the significanceofresourcesfortheimplementationstageofadoption compared to initiation and adoption stages The literature suggeststhatasuccessfulimplementationrequiresasubstantial financial investment and competent human resources [11] Second,themeta-analysis results verified aweakly significant relationship between resources and IT adoption for small organizationsbutaninsignificantrelationshipforlarge organiza-tionsinITadoption
5.5 ISdepartmentsize
ISdepartmentsizeislikelytohaveasignificantimpactinthe adoptionofITinnovation.AlargerISdepartmentsizemeansthat theorganizationcanundertakeinnovationadoption
Fourstudiesempiricallyexaminedtherelationshipbetween
ISdepartmentsize andITadoption.Inthese, theeffectsof 15 relationships were reported Of these, 14 had significant associations and onlyone was found tohave no significance
A set of12 ITadoption relationships from threestudies were considered for meta-analysis The results showed that IS department size was a significant predictor of innovation adoptionofIT
5.5.1 FindingsofmoderatoreffectonISdepartmentsize Table8illustratesthemeta-analysisresultsofthemoderator effects on the relationship between IS department size and adoptionofIT
The stage of innovation wasa significant moderator of the relationship between IS department size and IT adoption The meancorrelationshowedthatISdepartmentsizewassignificant for the implementation stage of IT adoption For successful implementationofIT,organizationsrequireITexpertise
Table 7
Meta-analysis result of resources.
Resources
Stage of innovation
Type of innovation
Type of organization
Size of organization
INN STD, no of innovation studied; SAM SIZ, sample size; MEN COR, mean correlation; ZTR VAL, z-transformation; OBS VAR, observed variance; SAM EVA, sampling error
Trang 85.6 Formalization
In an IT context, formalization can be considered as the IS
structure/technologystrategywithintheorganization.Outof59
studies,nineassessedformalization.17relationshipswerestudied
offormalizationandITadoptionandfivefoundthatformalization
wassignificantwhile12 studiesfoundno relevance.Our
meta-analysis procedure for formalization considered 10 sets of
formalizationITadoptionrelationshipsfromfourstudies
Themeancorrelationresultofthemeta-analysisrevealedthat
formalizationhadnosignificanteffectontheadoptionofIT.The
95% confidence interval resultsof the meta-analysis showed a
positiveassociationthoughtheorysuggestsanegativeone
5.6.1 Findingsofmoderatoreffectonformalization
Table9showstheresultsofthemeta-analysisofthemoderator
effectontherelationshipbetweenformalizationandadoptionof
IT
The 95% confidence interval verified a positive association
between adoption and implementation stages of IT adoption
Although the meta-analysis results for formalization and IT innovation adoption werefound to be insignificant, the meta-analysismoderatorofthestageofinnovationconfirmedaweak significance relationship with the implementation stage and insignificancefortheadoptionstage
The meancorrelationand 95% confidenceinterval resultsof meta-analysisfoundthatformalizationwaspositivelyassociated withweaksignificancewiththeadoptionofproductinnovation butnosignificancewasfoundwithprocessinnovation
5.7 Centralization Centralizationof anorganizationhasa negativerelationship withinitiationandadoption,buthasapositiverelationshipwith implementation Centralization was considered in six of our reviewedliteraturestudieswith16ITadoptionrelationships.The aggregatedsignificanttestresultsofcentralizationshowedeight significant and eight insignificant For the meta-analysis, nine centralization-ITadoptionrelationshipsweregatheredfromthree studieswhichperformedcorrelationanalysis.The95%confidence
Table 8
Meta-analysis result of IS department size.
IS department size
Stage of innovation
Type of innovation
Type of organization
Manufacturing 0
Size of organization
INN STD, no of innovation studied; SAM SIZ, sample size; MEN COR, mean correlation; ZTR VAL, z-transformation; OBS VAR, observed variance; SAM EVA, sampling error variance; EXP VAR, explain variance; COF INT, 95% confidence interval.
Table 9
Meta-analysis result of formalization.
Formalization
Stage of innovation
Type of innovation
Type of organization
Size of organization
INN STD, no of innovation studied; SAM SIZ, sample size; MEN COR, mean correlation; ZTR VAL, z-transformation; OBS VAR, observed variance; SAM EVA, sampling error
Trang 9interval of meta-analysis showed that centralization had no
significanceintheadoptionofIT
5.7.1 Findingsofmoderatoreffectoncentralization
Theresultsofmeancorrelationand95%confidenceintervalof
allcategories of four moderatorfoundno associationwith the
relationshipbetweencentralizationandITadoption.The
impor-tantconclusionfromthiswasthatcentralizedstructure neither
inhibitsnorfacilitatesITadoption.Allstudiesconsideredin the
meta-analysisofcentralizationandITadoptioninTable10were
performedforlargeorganizations
5.8 Organizationalreadiness
Organizationalreadinessisthedegreetowhichanorganization
hasthe awareness, resources, commitment and governance to
adoptIT[7].Manystudieshaveaddressedorganizationalreadiness
intermsofavailabilityoffinancialandtechnologicalresourcesin
organizations.12setsoforganizationalreadinessandIT
innova-tionadoptionrelationshipswerereviewedinourstudy.Combined
resultsfromsignificancetestsfoundapositiveassociationwith10
relationships of organizational readinessand IT adoption while twofoundotherwise.Significancetestresultsconfirmed organi-zationalreadinesstobeamajorfactorindeterminingtheadoption
ofIT
The mean correlation results of meta-analysis showed that organizationalreadinesshadastrongsignificanceforadoptionof
IT.However,themeta-analysissamplesizewasverysmallwith onlytwostudies.Thesamplingerrorforthisvariablewasnotless than60%oftheobservedvariance
5.9 ISinfrastructure Organizations with well established IS infrastructure were morelikelytoadoptITinnovation
Onlyninestudies consideredIS infrastructure effectin our study 16 innovation adoption and IS infrastructure relation-ships were considered in these nine, with 12 showing IS infrastructure to have significant effect on IT adoption while fourdidnot.Forthemeta-analysis,wegathered10relationships from five studies that related IS infrastructure to innovation adoptionofIT.Themeancorrelationresultsofthemeta-analysis
Table 11
Meta-analysis result of IS infrastructure.
IS infrastructure
Stage of innovation
Type of innovation
Type of organization
Manufacturing 0
Size of organization
INN STD, no of innovation studied; SAM SIZ, sample size; MEN COR, mean correlation; ZTR VAL, z-transformation; OBS VAR, observed variance; SAM EVA, sampling error
Table 10
Meta-analysis result of centralization.
Centralization
Stage of innovation
Type of innovation
Mixed
Type of organization
Manufacturing
Size of organization
SMEs
Mixed
INN STD, no of innovation studied; SAM SIZ, sample size; MEN COR, mean correlation; ZTR VAL, z-transformation; OBS VAR, observed variance; SAM EVA, sampling error variance; EXP VAR, explain variance; COF INT, 95% confidence interval.
Trang 10showed that IS infrastructure had a weak significance for IT
adoption
5.9.1 FindingsofmoderatoreffectonISinfrastructure
Table11illustratesthemeta-analysisresultsofthemoderator
effectontherelationshipbetweenISinfrastructureandadoption
ofIT
Theresultsofmeancorrelationand95%confidenceintervalof
implementation sub-category of stages of innovation showed
moderate significance and positive association between IS
infrastructureandITadoption.Oneimportantaspectofthisresult
wasthesignificanceof ISinfrastructurefortheimplementation
stagecomparedtothatoftheadoptionstage.Theadoptionstage
showed a weak significance between IS infrastructure and IT
adoption
5.10 Productchampion
The presence of a product champion is critical to the
introductionofnewtechnologiesinorganizationsitinfluences
all stages of innovation adoption In the initiation stage, the
productchampionwillpersuademanagementtoacquire
tech-nologyandcreateawarenessoftheinnovationinthe
organiza-tion In the adoption andimplementation stages, the product
championfacilitatesuseracceptancebyprovidingvarioustypes
oftraining
Weaccumulatedfivestudiesthatdealtwiththeinfluenceof
productchampionon innovationadoption.Fromthese,asetof
seven product champion and IT adoption relationships were
obtained Aggregated test of significance showed that they
included five significant and two insignificant relationships
betweenproductchampionandITadoption
The 95% confidence interval results showed that product
champion had no in the adoption of IT Also, the mean
correlation was found to be insignificant for the relationship
betweenproductchampionandITadoption.Wedidnotperform
moderator effects for product champion since the evaluation
showed that sampling error was not less than 60% of the
observedvariance
6 Implicationsoftheresearch
We foundthatthemost significantorganizationalfactor for
adoption of IT was organizational readiness followed by IS
departmentsize,IS infrastructure, top management support, IT
expertise,resource,andorganizationalsize.Wedidnotfindthat
centralizationorproductchampionwerefactorsdeterminingIT
adoption.Mostpaststudieshadsuggestedthatformalizationhada
negativeassociationwithITadoption;however,wefoundthatit
facilitatedITadoption
We conducted tests for moderator effects of individual
organizational factors except for organizational readiness and
productchampion:samplingerrorvarianceofthesetwowasfound
tobemorethan60%oftheobservedvarianceandthusmoderator
effectscouldnotbeperformedforthem
Wefoundthatprimarystudiesrarelyexaminedtheinitiation
stagesofinnovationadoptionofIT.MoreITadoptionrelationships
between organizational attributesand theinitiationstage of IT
adoptionarerequiredtoperformmeta-analysismoderatoreffect
Theinitiationstagewastestedfororganizationalsizefactorwitha
small sample and was found to be weakly significant All
organizational factorsweresignificantin theadoption decision
andimplementationstages.Organizationalsize,ITexpertise,top
managementsupport,ISinfrastructure,resources,and
formaliza-tionweremoresignificantfortheimplementationstagethanfor
theadoptiondecisionstages
WealsofoundthatITexpertise,topmanagementsupport,and resourcesweresignificantin theadoption ofboth product and processinnovationofITadoption.Additionally,organizationsize wassignificantforprocessinnovationwhileISinfrastructurewere significantforproductinnovation.ITexpertise,topmanagement support,andresourceswerekeydeterminantsforsmall organiza-tions; top management support and resources were deciding factorsforlargeorganizations
Our analysis allowed us to assess the current state of IT adoption in organizations and, in particular, organizational attributesinfluencingtheprocess.Weidentifiedthekey organiza-tionalattributesinfluencingtheadoptionofITinorganizations.In addition,wecategorizeddifferentfactorsinfluencingtheadoption
of IT in organizations under different conditions Managers involvedinITadoptionneedtotakeaccountofthesedeterminants duringtheinnovationprocessandorganizationsshouldfocuson relevantattributesbasedontheconditionsinwhichinnovation adoptionbecomeseffective
The usecorrelationcoefficientsallowedthe combinationof smallandnon-significanteffectstoshowanoverallviewofthe relationshipbetweenorganizationalattributesandITinnovation adoption The findings of individual studies on IT innovation adoption hadproducedcontradictory outcomes,largelydueto statisticalerrorandmeasurementvariations.Also,differencein the interpretation of tests of significance contributed to inconsistency By aggregating observed correlation coefficient and examining for moderator effects, the meta-analysis pre-sented overcame these drawbacks and explained past incon-sistencies
The findings of themeta-analysis allowed us to drawmore definitiveconclusionsontherelationshipbetweenorganizational factorsandITinnovationadoptionandallowedtheidentification
of relationships that would not necessarily be apparent from individualstudies
Anumberoflimitations needtobeconsidered.Weincluded onlystudiesthatperformedcorrelationanalysis;ofthefifty-nine studiesthat examined theinfluenceoforganizational variables, onlythirty-oneperformed it.Forsomevariables,thenumberof datasetsavailablewasinadequatetoperformthemeta-analysis Similarly,thestudycouldnotevaluatethemoderatorfactoreffect for some variables due to lack of data We used studies that providedcorrelationvaluesfortherelationshipbetween organi-zationalattributesandITadoption
Anotherlimitationwasthatthemajorityofthestudieswere intended for large organizations and only a few examined smaller ones Furthermore most meta-analyses experience publicationbias.However,wedidsearchcarefullyinanattempt
to find all literature that examined factors influencing the adoptionofIT
7 Conclusions Our findingshaveconsiderablesignificanceinunderstanding thedeterminantofITadoptionintermsoforganizationalcontext The study provides researchers and practitioners with a set of factorsthataffecttheadoptionofITinorganizations.Resultsserve
as a guideline for practitioners to identify and address the facilitatingandinhibitingissuesintheorganizationalcontextin theprocessofITadoption.Managersneedtoconsidertheseissues whenembarkingonITadoption
In our study and meta-analysis we identified gaps in understanding theattributesofIT.We wereonlyabletoverify organizational size factorswithtwo datasets for theinitiation stageofinnovationadoption;wecouldnotverifyother organiza-tionalfactors.Alldatausedinthisstudycanbemadeavailable uponrequesttotheleadauthor