Intellectual Capital Bontis

[ 63 ]

Management Decision36/2 [1998] 63–76

© MCB University Press [ISSN 0025-1747]

Intellectual capital: an exploratory study that develops measures and models

Nick BontisNational Centre for Management Research and Development, Richard IveySchool of Business, University of Western Ontario, London, Ontario, Canada

This paper details an empiri-cal pilot study that exploresthe development of severalconceptual measures andmodels regarding intellectualcapital and its impact onbusiness performance. Theobjective of this pilot study isto explore the development ofitems and constructs throughprincipal components analy-sis and partial least squares(PLS). The final retained,subjective measures andoptimal structural specifica-tion show a valid, reliable,significant and substantivecausal link between dimen-sions of intellectual capitaland business performance.These results should helpboth academics and practi-tioners more readily under-stand the components ofintellectual capital and pro-vide insight into developingand increasing it within anorganization. Suggestions arethen made to advance andimprove this research pro-gramme.

All men by nature desire knowledge. Aristo-tle (384-322 BC), Greek philosopher. Meta-physics, Book 1, Chapter 1.

Whereas at one time the decisive factor ofproduction was the land, and later capital …today the decisive factor is increasingly manhimself, that is, his knowledge. Pope JohnPaul II (1991). Centesimus Annus.

Intellectual capital has been considered bymany, defined by some, understood by a selectfew, and formally valued by practically no one(Stewart, 1997; Sveiby, 1997). Therein lies oneof the greatest challenges facing businessleaders and academic researchers today andtomorrow. Recently, the job title of chiefknowledge officer (CKO) has been creepingup on annual reports and job advertisementswith ever-increasing frequency. Thesepathfinding individuals have been given theenviable task of channelling their organiza-tions’ intellectual capital as an essentialsource of competitive advantage. Knowledgeofficers are responsible for justifying thevalue of knowledge that is constantly beingdeveloped in their organizations (Nonaka andTakeuchi, 1995). This elusive intangible maynever be evaluated in the financial terms thatwe are currently accustomed to. However, itsstrategic impact is never in question. Fromthe capture, codification, and disseminationof information, through to the acquisition ofnew competencies via training and develop-ment, and on to the re-engineering of busi-ness processes, present and future success incompetition will be based less on the strategicallocation of physical and financial resourcesand more on the strategic management ofknowledge.

Intellectual capital research has primarilyevolved from the desires of practitioners(Bontis, 1996; Brooking, 1996; Darling, 1996;Edvinsson and Sullivan, 1996; Saint-Onge,1996). Consequently, recent developmentshave come largely in the form of popularpress articles in business magazines andnational newspapers. The challenge for acad-emics is to frame the phenomenon usingextant theories in order to develop a morerigorous conceptualization of this elusiveintangible. This paper coalesces many per-spectives from numerous fields of study in anattempt to raise the understanding and

importance of this phenomenon. The objec-tive here is to conceptualize and frame theexisting literature on intellectual capital as afoundation for further study.

This topic is important because intellectualcapital has been rarely studied orunderstood. In fact, managers and investorswoefully neglect intellectual inputs and out-puts, though these far outweigh the assetsthat appear on balance sheets (Stewart, 1991;1994). Handy (1989) suggests that the intellec-tual assets of a corporation are usually threeor four times tangible book value. He warnsthat no executive would leave his cash orfactory space idle, yet if CEOs are asked howmuch of the knowledge in their companies isused, they typically say only about 20 percent. The importance of this topic is alsoreflected in the increased importance of theprofessional services industry and the manynew knowledge-based firms that haverecently been launched.

This article is divided into five sections:1 review of concepts – a review of the recent

literature which includes definitions ofterms as well as a conceptual model;

2 research design – the methodologicalapproach utilized to administer the pilotstudy;

3 results – analysis of the measures andmodels;

4 discussion – highlights of the analysis,suggestions for future work, limitations ofthe research and the contribution it makesto academia and managers; and

5 conclusion – what managers can do next.

Review of concepts

Knowledge creation by business organiza-tions has been virtually neglected in manage-ment studies even though Nonaka andTakeuchi (1995) are convinced that thisprocess has been the most important sourceof international competitiveness for sometime. Drucker (1993) heralds the arrival of anew economy, referred to as the knowledgesociety. He claims that in this society, knowl-edge is not just another resource alongsidethe traditional factors of production – labour,capital, and land – but the only meaningful

Preparation of the articlewas assisted by the financialsupport received from anOntario Graduate Scholarshipand an Ivey Doctoral Grant.The author gratefullyacknowledges the sugges-tions and comments ofreviewers and of ProfessorsMary Crossan and John Hul-land. The author would alsolike to thank the inspirationalthinking of Hubert Saint-Onge, Leif Edvinsson andThomas Stewart. Previousversions of certain sections ofthis article may haveappeared elsewhere as aworking paper, conferenceproceeding or book chapter.

[ 64 ]

Nick BontisIntellectual capital: anexploratory study that develops measures and models


resource today. Because knowledge is sharedamong organizational members, it is con-nected to the firm’s history and experiences(Von Krogh et al., 1994) and soon becomes theultimate replacement of other resources(Toffler, 1990). This notion underpins a moregeneral idea that economies of the future willbe education-led (Young, 1995). What does thismean for managers? It means that the capac-ity to manage knowledge-based intellect isthe critical skill of this era (Quinn, 1992). It isup to symbolic analysts (Reich, 1991) who areequipped to identify and solve intellectualcapital issues, that will sustain the knowl-edge advantage for their own organizations.If there is one distinguishing feature of thenew economy that has developed as a result ofpowerful forces such as global competition, itis the ascendancy of intellectual capital. Ashift is clearly perceptible from a manufac-turing to a service-oriented economy: firmsthat are thriving in the new strategic environ-ment see themselves as learning organiza-tions pursuing the objective of continuousimprovement in their knowledge assets(Senge, 1990). Recently, there has been expo-nential growth in researching this area(Crossan and Guatto, 1996). Competitive,technological, and market pressures havemade continuous organizational learning acritical imperative in global strategy effec-tiveness (Osland and Yaprak, 1995). Organiza-tions that have been unable to enhance theirknowledge assets have failed to survive(Antal et al., 1994) and are left wonderingwhat the fuss is all about (Roos and vonKrogh, 1996).

The importance of this topic is alsoreflected in the growth of the professionalservices industry and the many new knowl-edge-based firms that have fuelled our econ-omy. Top MBA recruits no longer find asmany positions in manufacturing companiesas they did in the 1950s and 1960s. Nowadays,the career services offices of many businessschools report that most new graduatessecure positions with management consul-tants, accounting firms, investment banks,law firms, software developers and informa-tion brokers. The constant requirementfound in each of these positions is the impor-tance of intellectual capital.

To grasp the importance of why it is neces-sary to measure this phenomenon, we mustunderstand the concept of Tobin’s q from theaccounting and finance literature. This ratiomeasures the relationship between a com-pany’s market value and its replacementvalue (i.e., the cost of replacing its assets). Inother words, a company with a stock marketvalue of $100 million and a book value of $25million will have a Tobin’s q ratio of 4.00. The

ratio was developed by the Nobel Prize-win-ning economist James Tobin (White et al.,1994). In the long run, this ratio will tendtowards 1.00, but evidence shows that it candiffer significantly from 1.00 for very longperiods of time (Bodie et al., 1993). For exam-ple, companies in the software industry,where intellectual capital is abundant, tendto have a Tobin’s q ratio of 7.00, whereas firmsin the steel industry, noted for their largecapital assets, have a Tobin’s q ratio of nearly1.00. Intellectual capital valuation hasbecome an industry on its own. For example,the Royal Bank of Canada has launched asubsidiary business that concentrates exclu-sively on investing in knowledge-based indus-tries (Bontis, 1997).

Sveiby highlights intellectual capital valua-tion by citing a familiar example of a highTobin’s q:

Shares in Microsoft, the world’s largestcomputer software firm, changed hands atan average price of $70 during fiscal 1995 at atime when their so-called book value wasjust $7. In other words, for every $1 ofrecorded value the market saw $9 in addi-tional value for which there was no corre-sponding record in Microsoft’s balancesheet (Sveiby, 1997, p. 3).

There are numerous other examples thatmake the same case. The value of intellectualcapital in these firms has been cast as quasi-value by the invisible hand of the market.However, companies do not trade their intan-gible assets, so the value of items such asintellectual capital stocks or organizationallearning flows cannot be deduced from rou-tine market transactions like the value oftraditional tangible assets. Sometimes, thevalue of knowledge is attributed even withoutthe existence of any monetary transactions atall:

In August 1995 Netscape went public in oneof the most oversubscribed initial publicofferings in history. A company with negli-gible profits ended its first day of tradingwith a value of $2 billion – a value basedentirely on intangible assets (Sveiby, 1997, p. 114).

Another popular example of a knowledge-intensive organization that is internationallyknown for its products is Nike. However, Nikeis a shoemaker that makes no shoes – its workis research and development, design, market-ing, and distribution, almost all knowledge-based activities – but still has $334,000 in salesfor each employee (Stewart, 1997).

One of the purest examples of intellectualcapital valuation is in the consulting indus-try. McKinsey, one of the industry’s leaders,does not employ traditional marketing meth-ods; it sells by having clients come knocking

[ 65 ]



to purchase the best analytical knowledgeavailable (Nicou et al., 1994). McKinsey gener-ally sells its intellectual capital in teams offive, each led by a senior partner. Remarkably,clients are willing to pay for the transfer ofthis knowledge at an average annual rate of$500,000 per consultant (Sveiby, 1997).

Stewart (1997) defines intellectual capital as“the intellectual material – knowledge, infor-mation, intellectual property, experience –that can be put to use to create wealth.” Stewart goes on to identify several organiza-tions such as Skandia, Dow Chemical,Hughes Aircraft, and Canadian ImperialBank of Commerce that are in the process ofmanaging and developing this phenomenon.Stewart’s major contribution was in the defi-nition of intellectual capital and in the recog-nition of the difficulty to measure it. Theobjective of this pilot study is to explore thedevelopment of measures and models thatcould help both academics and practitionersmore readily understand the components ofintellectual capital and its impact on businessperformance.

Organizational learning, as described byChris Argyris at Harvard (1992) among others,has been thought of as the flow of knowledgein a firm; it follows then that intellectualcapital is the stock of knowledge in the firm(Dierickx and Cool, 1989). To marry the twoconcepts, it may be useful to consider intellec-tual capital as the stock unit of organizationallearning flows. However, intellectual capitalcannot necessarily be taught through educa-tion and training. The most precious knowl-edge in an organization often cannot bepassed on (Levitt, 1991).

Prior to continuing the conceptualization ofintellectual capital stocks, it may be helpful todefine what it is not. Intellectual capital doesnot include intellectual property. Intellectualproperty are assets that include copyrights,patents, semiconductor topography rights,and various design rights. They also includetrade and service marks. Undertaking anintellectual property audit is not a new idea.The following sections refer to the conceptu-alization of what intellectual capital is (seeFigure 1).

Human capitalFirst, the organization’s members possessindividual tacit knowledge (i.e. inarticulableskills necessary to perform their functions)(Nelson and Winter, 1982). In order to illus-trate the degree to which tacit knowledgecharacterizes the human capital of an organization, it is useful to conceive of theorganization as a productive process thatreceives tangible and informational inputsfrom the environment, produces tangible and

informational outputs that enter the environ-ment, and is characterized internally by aseries of flows among a network of nodes andties or links (see Figure 2).

A node represents the work performed –either pure decision making, innovative cre-ativity, improvisation (Crossan et al., 1996) orsome combination of the three – by a singlemember of the organization or by parallel,functionally equivalent members who do notinteract with one another as part of the pro-ductive process (see Figure 2). Thus, individ-ual tacit knowledge, when present, exists atthe nodes themselves. A tie or link is direc-tional in nature and represents a flow ofintermediate product or information from agiven node. Every node has at least one tie orlink originating from it, while multiple tiesoriginating from a single node imply that thetask performed at the node includes a deci-sion about where to direct the subsequentflow. Structural tacit knowledge, when pre-sent, implies that no member of the organiza-tion has an explicit overview of these ties orand consequently of the correspondingarrangement of nodes (see subsequent discus-sion on structural capital). Accordingly, aproductive process characterized by a sub-stantial degree of tacit knowledge is arrangedas a hodgepodge of nodes lacking any dis-cernible organizational logic.

Point A in Figure 2 represents the core ofhuman capital. Multiple nodes (human capi-tal units) attempt to align themselves in someform of recognizable pattern so that intellec-tual capital becomes more readilyinterpretable. This point represents the low-est level of difficulty for development as wellas the lowest level of externality from thecore of the organization.

Human capital has also been defined on anindividual level as the combination of thesefour factors:1 your genetic inheritance;2 your education;3 your experience; and4 your attitudes about life and business

(Hudson, 1993).

Human capital is important because it is asource of innovation and strategic renewal,whether it is from brainstorming in aresearch lab, daydreaming at the office,throwing out old files, re-engineering newprocesses, improving personal skills or devel-oping new leads in a sales rep’s little blackbook. The essence of human capital is thesheer intelligence of the organizational mem-ber. The scope of human capital is limited tothe knowledge node (i.e. internal to the mindof the employee). It can be measured(although it is difficult) as a function of

[ 66 ]



volume (i.e. a third degree measure encom-passing size, location and time). It is also thehardest of the three sub-domains of intellec-tual capital to codify.

The term human capital has also been usedby the American Nobel Prize-winning econo-mist Theodore W. Schultz (1981):

The decisive factors of production inimproving the welfare of poor people are notspace, energy, and cropland; the decisivefactors are the improvement in populationquality and advances in knowledge. Theseadvancements can be augmented by appro-priate investment in human capital.

Structural capitalThe organization itself embodies structuraltacit knowledge, which exists in:

the myriads of relationships that enable theorganization to function in a coordinatedway [but] are reasonably understood by [atmost] the participants in the relationshipand a few others…” This means that “theorganization is … accomplishing its aims byfollowing rules that are not known as suchto most of the participants in the organiza-tion” (Winter, 1987, p. 171).

This construct deals with the mechanismsand structures of the organization that canhelp support employees in their quest foroptimum intellectual performance and there-fore overall business performance. An indi-vidual can have a high level of intellect, but ifthe organization has poor systems and proce-dures by which to track his or her actions, theoverall intellectual capital will not reach itsfullest potential.

An organization with strong structuralcapital will have a supportive culture thatallows individuals to try things, to fail, tolearn, and to try again. If the culture undulypenalizes failure, its success will be mini-mal. Structuring intellectual assets withinformation systems can turn individualknow-how into group property (Nicolini,1993). It is the concept of structural capitalthat allows intellectual capital to be mea-sured and developed in an organization. Ineffect, without structural capital, intellec-tual capital would just be human capital.This construct therefore contains elementsof efficiency, transaction times, proceduralinnovativeness and access to information forcodification into knowledge. It also supportselements of cost minimization and profitmaximization per employee. Structuralcapital is the critical link that allows intel-lectual capital to be measured at an organi-zational level.

Point B in Figure 2 illustrates the struc-tural ties or links of human capital nodesthat are required to transform human capi-tal into structural capital. The arrowswithin structural capital represent the focusof intellectual capital development from thenodes into the organization’s core. Theessence of structural capital is the knowl-edge embedded within the routines of anorganization. Its scope lies internal to thefirm but external to the human capitalnodes. It can be measured (although it isdifficult) as a function of efficiency (i.e., anoutput function per some temporal unit).Organizational processes (such as thosefound in structural capital) can eventuallybe codified.

Increase inLevel of

Externalityfrom the

Organization’sCore

customerrelationships

marketingchannels

A - human capitalB - structural capitalC - customer capitalIncrease in

Level ofDifficulty to

Develop

node link

C

B

A

Figure 2Discriminating intellectual capital sub-domains

Figure 1Conceptualization of intellectual capital

IntellectualCapital

StructuralCapital

HumanCapital

CustomerCapital

2nd Order

1st Order

Essence

Scope

Parameters

human intellectinternal withinemployee node

volumeappropriateness

high

organizational routinesinternal

organizational linksefficiency

accessibility

medium

market relationshipsexternal

organizational links

longevity volume

highestCodificationDifficulty

[ 67 ]



Customer capitalKnowledge of marketing channels and cus-tomer relationships is the main theme ofcustomer capital. Frustrated managers oftendo not recognize that they can tap into awealth of knowledge from their own clients.After all, understanding what customerswant in a product or a service better thananyone else is what makes someone a busi-ness leader as opposed to a follower.

Customer capital represents the potentialan organization has due to ex-firm intangi-bles. These intangibles include the knowledgeembedded in customers, suppliers, the gov-ernment or related industry associations.Point C in Figure 2 illustrates that customercapital is the most difficult of the three sub-domains to develop since it is the most exter-nal to the organization’s core. The arrowsrepresent the knowledge that must flow fromexternal to the organization (i.e. its environ-ment) into the organization’s core by way oflinked nodes. The essence of customer capitalis knowledge embedded in relationshipsexternal to the firm. Its scope lies external tothe firm and external to the human capitalnodes. It can be measured (although it isdifficult) as a function of longevity (i.e. cus-tomer capital becomes more valuable as timegoes on). Owing to its external nature, knowl-edge embedded in customer capital is themost difficult to codify.

One manifestation of customer capital thatcan be leveraged from customers is oftenreferred to as “market orientation”. There isno consensus on a definition of market orien-tation, but two recent definitions have becomewidely accepted. The first is from Kohli andJaworski (1990), who define market orienta-tion as the organization-wide generation ofmarket intelligence pertaining to current andfuture needs of customers, dissemination ofintelligence horizontally and verticallywithin the organization, and organization-wide action or responsiveness to market intel-ligence. Similar definitions are found in Dengand Dart (1994) and Lichtenthal and Wilson(1992). The second is from Narver and Slater(1990), who define market orientation as a one-dimension construct consisting of threebehavioural components and two decisioncriteria – customer orientation, competitororientation, inter-functional co-ordination, along-term focus, and a profit objective. Withclose parallels to Kohli and Jaworski (1990),Narver and Slater (1990) include the genera-tion and dissemination of market intelligenceas well as managerial action.

Intellectual capitalThe term intellectual capital was first pub-lished by John Kenneth Galbraith in 1969

(Feiwal, 1975). He believed that intellectualcapital meant more than just “intellect aspure intellect” but rather incorporated adegree of “intellectual action”. In that sense,intellectual capital is not only a static intan-gible asset per se, but an ideological process; ameans to an end.

Some readers may wonder whether or notintellectual capital is just the effective dissec-tion of information overload. To clarify thispoint, an examination of the differencebetween “information” and “knowledge”would be helpful. Simply put, information isthe raw material, and knowledge is the fin-ished product. When a manager receives aprintout from a computer detailing lastweeks cost per transaction figures, he or sheis reading information. The implicationsderived from the trends or underlying issuesfrom those data are what is referred to asknowledge. Intellectual capital is thereforethe pursuit of effective use of knowledge asopposed to information.

Research design

A survey was designed that taps into theintellectual capital constructs as well as busi-ness performance within the context of theconceptual model. Many of the TDM (totaldesign method) recommendations suggestedby Dillman (1978) were adopted. A copy of thecover letter and questionnaire can berequested from the author. The questionnairewas administered to one section of MBA stu-dents at the Ivey School of Business in theUniversity of Western Ontario. The question-naire was designed in an easy to read bookletformat with a total of eight pages. The coverletter was on the first page and it introducedthe concept of intellectual capital. There wasno incentive to fill out the survey for the stu-dents and it was completely optional. Sixty-four students took approximately ten minutesto complete the questionnaire.

Since this study concentrates on the firmlevel of analysis, each respondent wasrequired to answer the questionnaire as arepresentative of the organization theyworked for prior to entering the MBA pro-gramme. In effect, each respondent acted as aproxy for their organization. Some studentsreturned the questionnaires unansweredbecause they worked for governmental officesand felt that they could offer absolutely nofeedback on the topic. Others were uncom-fortable in filling out the survey because theyfelt that they did not know enough about thefirm and were not in a position high enoughto fill out the questionnaire adequately.

[ 68 ]



In designing the questionnaire, a 7-pointLikert scale (strongly disagree to stronglyagree) for each item with medium-length (16to 24 words) questions was used, as suggestedby Andrews (1984). A total of 63 items,designed to tap into four constructs (threeconstructs relating to intellectual capital plusperformance), were included in the question-naire. The items included in the survey weredeveloped from concepts that were accentu-ated during the literature review phase of thestudy. Since this is an exploratory pilot study,no previous instruments were replicated. SeeAppendix 1 for a summary of the items thatwere developed and used for each construct.

The results were coded in SPSS for Win-dows. The following items were reversecoded: human capital (H5R, H13R, H14R,H15R, H19R), customer capital (C13R, C15R),and structural capital (S13R, S16R). Of thetotal 4,032 data cells (63 items * 64 observa-tions), less than 2.5 per cent had missingvalues which were assigned to the means ofeach variable.

The following statistical tests were exe-cuted:• Kolmogorov-Smirnov test for normality;• Cronbach’s alpha test for reliability;• principal components analysis with VARI-

MAX rotation , and• Partial least squares (PLS).

Results

The 64 observations represented a variety oforganizations in numerous industries. Table Ihighlights the profile of the data with somedescriptive statistics.

Respondents were promised organizationalanonymity. These industries and the numberof times they were each represented show awide cross-section of businesses accountedfor by the data: financial services (7), chemi-cal (4), insurance (4), computers and software(3), and courier services (2).

Kolmogorov-Smirnov TestThe results show that most items spanned thewhole range of possible responses except for:C6, C13R (customer capital), H1, H14R(human capital), S5, S6, S7, S12, S16R (struc-tural capital), and P1, P5, P8 (performance).

The Kolmogorov-Smirnov test for normalitywas used to see whether the responses had anormal curve about the mean. Just over halfof the items (33 out of 63) were considered tohave normal distributions. However, theassumption of normality is not a major issuefor structural modelling. In fact, PLS isrobust enough to not require normal data(Barclay et al., 1995).

Feedback from respondents highlightedcertain items that were difficult to interpretand thus rejected. For example, C4 (our mar-ket share is the highest in the industry) wasdifficult to answer on a “strongly disagree” or“strongly agree” scale from 1 to 7. If you hadthe third highest market share in your indus-try, where would you mark your response? Itwould be difficult to decide in this casebecause respondents interpret the questiondifferently.

Cronbach’s AlphaTo test the reliability of the measures, Cron-bach’s alpha was used as suggested by Nun-nally (1978). This calculation should be thefirst measure one calculates to assess thequality of the instrument (Churchill, 1979). Asatisfactory level of reliability depends onhow a measure is being used. In the earlystages of research on predictor tests orhypothesized measures of a construct (as isthe case with this exploratory pilot study)instruments that have reliabilities of 0.7 orgreater will suffice (Nunnally, 1978). Thereliabilities for each of the four constructs isfine since the alpha values for each aregreater than 0.85.

Principal components analysisFactor analysis is a multivariate statisticalmethod whose primary purpose is data reduc-tion and summarization (Hair et al., 1987). Byusing factor analysis, a factor loading foreach item and its corresponding constructwas determined. In order to verify that theitems tapped into their stipulated constructs,a principal components analysis with a VARI-MAX rotation was executed. The items wereforced into three factors and the output wassorted and ranked based on a 0.5 loading cut-off. Typically, loadings of 0.5 or greater areconsidered very significant (Hair et al., 1987).

The VARIMAX rotation was used because itcentres on simplifying the columns of thefactor matrix. With the VARIMAX rotationalapproach, there tends to be some high load-ings (i.e. closer to 1) and some loadings near 0in each column of the matrix. The logic isthat interpretation is easiest when the vari-able-factor correlations are either closer to 1,thus indicating a clear association between

Table IProfile of data

Mean Std. dev. Minimum MaximumItem $ $ $ $

Sales ($million) 588.15 931.46 1.00 4,000.00Employees (#) 8,731 28,489 8 180,000

[ 69 ]



the variable and the factor, or 0 indicating aclear lack of association (Hair et al., 1987).

Only the items that loaded on their corre-sponding factors at levels of 0.5 or greaterwere retained for the rest of the analysis.These items are highlighted in the last col-umn. Items were not retained because they• did not load on any factor with a value of 0.5

or greater;• loaded on the wrong factor; or• had cross-loadings on two factors.

Items S12, S7, S8, and S13R were not retainedbecause they did not load on their appropri-ate factor and also cross-loaded on Factor 3 ata loading of less than 0.5. The three factorshad Eigenvalues and percentage of varianceexplained of 13.735 (25.9 per cent), 7.634 (14.4per cent) and 3.289 (6.2 per cent) respectivelywith a total cumulative variance explained of46.5 per cent.

Partial least squaresPartial least squares (PLS) allows theresearcher to test a model within its nomolog-ical network. Constructs derive their mean-ing from their underlying measures as well astheir antecedent and consequent constructsgiving a researcher the benefit of examiningthe constructs in an overall theoretical con-text.

The objective in PLS is to maximize theexplanation variance. Thus, R2 and the signif-icance of relationships among constructs aremeasures indicative of how well a model isperforming. The conceptual core of PLS is aniterative combination of principal compo-nents analysis relating measures toconstructs, and path analysis permitting theconstruction of a system of constructs. Thehypothesizing of relationships between mea-sures and constructs, and constructs andother constructs is guided by theory. Theestimation of the parameters representingthe measurement and path relationships isaccomplished using ordinary least squares(OLS) techniques.

The first step in PLS is for the researcher toexplicitly specify both the structural modeland the construct-to-measures relationshipsin the measurement model. The exogenousconstructs are consistent with the idea ofindependent variables (antecedents). Simi-larly, the endogenous constructs are consis-tent with the idea of dependent variables(consequents).

The constructs can be specified as “forma-tive” indicators or “reflective” indicators.Formative indicators imply a construct thatis expressed as a function of the items (theitems form or cause the construct). Reflectiveindicators imply a construct where the

observable items are expressed as a functionof the construct (the items reflect or are mani-festations of the construct). One looks to the-ory to decide on which type of epistemic orconstruct-to-measure relationship to specify.In this case, all constructs were “reflective”indicators. Once specified, the measurementand structural parameters are estimatedusing an iterative process of OLS, simple andmultiple regressions. The process continuesuntil the differences in the component scoresconverge within certain criteria.

One of the key benefits of using PLS as astructural modelling technique is that it maywork with smaller samples. In general, themost complex regression will involve:1 the indicators on the most complex forma-

tive construct; or2 the largest number of antecedent

constructs leading to an endogenous con-struct.

Sample size requirements become at least 10times the number of predictors from 1 or 2,whichever is greater (Barclay et al., 1995). Inthis study, the sample size of 64 is highenough for PLS. There were no formativeindicators so it is the second requirementthat must be met. The largest number ofantecedent constructs leading to an endoge-nous construct is three (3 * 10 = 30 < 64).

The retained items from the previous testswere used in PLS to test their loadings withina nomological network. Nine structural com-binations were examined using differentrelative positions for the intellectual capitalconstructs. The nine models represent differ-ent combinations of the intellectual capitalconstructs leading into performance. The R2

figures represent the predictive power withinthose constructs as explained by the mea-sures that represent the precedingconstructs. The path loadings represent thecausal links from one construct to the other.

The previous analysis was used to deter-mine which of the retained items (from theoriginal principal components analysis) werenow going to be kept for further investiga-tion. These remaining items were then placedin a structural configuration which yieldedthe highest original R2 for performance at56.0 per cent (which is very high relativelyspeaking for such a construct). The selectedmodel was then tested using PLS once moreand the statistical highlights are illustratedin Appendix 2.

Tests for individual item reliability, inter-nal consistency and discriminant validitywere completed for the selected model. The R2

or predictive power in the endogenous con-structs were as follows: customer capital =24.53 per cent, structural capital = 24.89 per

[ 70 ]



cent and performance = 56.02 per cent. Indi-vidual item reliability is assessed by examin-ing the loadings, or simple correlations, of themeasures with their respective construct. Arule of thumb is to accept items with loadingsof 0.7 or more, which implies more sharedvariance between the construct and its mea-sures than error variance (Carmines andZeller, 1979). All lambdas (or loadings) wereover the 0.7 threshold.

Internal consistency was verified since allof the items loaded at 0.7 or greater on theircorresponding constructs. Internal consis-tency was tested using the Fornell and Lar-cker (1981) measure. Discriminant validitywas tested using the correlation matrix ofconstructs. The diagonal of this matrix is thesquare root of the average variance extracted.For adequate discriminant validity, the diago-nal elements should be significantly greaterthan the off-diagonal elements in the corresponding rows and columns as was thecase for the selected model (see Appendix 2).

To assess the statistical significance of thepath coefficients, which are standardizedbetas, a jackknife analysis was performedusing a program developed by Fornell andBarclay (1983). The use of jackknifing, asopposed to traditional t-tests, allows the test-ing of the significance of parameter estimatesfrom data which are not assumed to be multi-variate normal. In this case, 32 subsampleswere created by removing two cases from thetotal data set. PLS estimates the parametersof each subsample and “pseudovalues” arecalculated by applying the jackknife formula.Four of the five paths proved to be significantat the p-value < 0.001 level. The one path fromcustomer capital to structural capital was notsignificant. Interestingly, this was also theonly path to have a negative coefficient andwas the least substantive of them all.

Path analysis can be used to calculate thetotal direct, indirect and spurious effects foreach endogenous construct. Table II summa-rizes the results for each path highlighted inAppendix 2:

Discussion

PLS analyses the measurement model andstructural model concurrently. Model fit isdependent upon the integrity of the data aswell as the strength of the theory. In the caseof my model, the integrity of the data was fine

and all but one of the paths proved to be sig-nificant. The strong contribution of PLS inexploratory work is that principal compo-nents analysis and path analysis are incorpo-rated into an a priori theoretical and mea-surement model, and thus the parameters areestimated in this specific context.

There are numerous improvements thatcan be made from this pilot study for futureresearch. First of all, the use of a conveniencesample (MBAs) is a strong criticism againstthese data because of the appropriateness andrepresentativeness of the respondents. Someof the MBAs mentioned that they had forgot-ten or were not currently close enough to theorganization to respond accurately to some ofthe questions. Others thought that they werenot in high enough positions to respondthoughtfully.

To improve on this study it would be benefi-cial in the future to elicit responses directlyfrom a wide variety of organizations thatinclude both manufacturing and serviceindustries. By examining these two differenttypes of organizations, one would hope to finda relatively larger concentration of intellec-tual capital in the professional servicesindustry (i.e. organizations such as softwaredevelopers, research laboratories and lawfirms).

The objective of the study thus far has beento determine which items effectively capturethe constructs of human capital, structuralcapital, customer capital, and performance.This was done by examining their loadingsusing a variety of structural model specifica-tions. It was also noted that certain paths (i.e.the ones leading into customer capital) wereneither substantive nor significant). To solvethis dilemma, it may be useful in future stud-ies to utilize model specifications that do notrequire paths into structural capital. Twoexamples of this possibility using the currentpilot study data are depicted in Appendix 3.

The Diamond specification is the optimalmodel encountered in the pilot study. All ofthe paths are substantive and significant andthe R2 of performance is high. This modelalso makes intuitive sense. A brilliant busi-ness school graduate that is recruited into anorganization as a product manager symbol-izes the human capital that starts off thismodel. With the advent of a supportive cul-ture (structural capital) and market research(customer capital), the new employee canlaunch a very successful product (perfor-mance).

Although the Simplistic specification con-jectures that the three components of intellec-tual capital lead into performance directly, itdoes not account for the interrelationshipsamong the three. It is for this reason that it is

1

2

2

2

2

2

( )

( ) ( )

( )

λ

λ ε

λλ ε

yi

yi i

yi

yi i

Var

Var

∑∑∑

∑∑∑

+

+

[ 71 ]



not supported even though the R2 value wasstill high for the performance construct.Given the literature review of intellectualcapital, the three constructs that make upthis phenomenon are known to affect eachother. In other words, an intellectualemployee (human capital) is practically use-less without the supportive structure of anorganization (structural capital) that canutilize and nurture his or her skills. This mayaccount for the unsubstantive and insignifi-cant path from human capital toperformance.

There is an important implication hiddenbehind Appendix 3 for managers. What thetwo different model specifications are sayingis that there must exist a constant interplayamong human, structural and customer capi-tal in order for an organization to leverage offits knowledge base. Isolated stocks of knowl-edge that reside in the employees’ minds thatare never codified into organizational knowl-edge will never positively affect businessperformance. In other words, it is not enoughfor an organization to hire and promote thebrightest individuals it can find. An organiza-tion must also support and nurture brightindividuals into sharing their human capitalthrough organizational learning. Unlikenormal inventory that can be found in tradi-tional manufacturing settings, individualknowledge stocks that reside in human capi-tal become obsolete. This obsolescence is notnecessarily due to outdated knowledge. Thereis a behavioural explanation instead. Humanbeings become unmotivated when they feelthey are not being utilized or challenged.That is why a stock of human capital willdeteriorate if not constantly supported andnurtured.

The results of this research programmeshould be very beneficial to both academicsand practitioners. Academics in the policyand accounting areas have traditionally beenvery interested in how intangible assetsreflect on the performance of firms. The pilotstudy thus far has shown that intellectual

capital has a significant and substantiveimpact on performance. Future research mayshow that this causal link may be more sub-stantive in certain specific industries. Also,future research may show that organizationswith predominant home country profiles maybe more in tune with intellectual capital andits effect on performance. Cross-referencingthe intellectual capital data with a variety ofinternational respondents and Hofstede’s(1978) cultural dimensions may highlightsome interesting relationships in this case.

For accounting researchers, intellectualcapital may prove to be an important item ofdisclosure in the future (especially for profes-sional services firms whose knowledge assetsare not currently reflected in today’s account-ing procedures). Churchill’s (1979) final sug-gestion in creating better measures is to“develop norms”. Once accepted items formeasuring intellectual capital are selected,organizations might be assessed by theirrelative positioning on each characteristic.Since respondents in this study participatedanonymously, the relative positioning ofactual firms was not reported.

By making periodic assessments of keyintellectual capital components, their poten-tial sale value to an outsider, and any measur-able trends in these values can offer a newperspective. Another interesting calculationfor accounting and finance academics is toexamine what the companies actually didwith their intellectual capital. For example,one might calculate a firm’s “exploitationratio” comparing the value of its intellectualcapital with its actual relative performance.This would suggest how effective the organi-zation has been in making the causal linkfrom intellectual capital to performance.

Once managers realize the importance ofmeasuring and developing their intellectualcapital, they will invariably want to increaseit since it positively affects firm performance.In recognizing the key to intellectual capitaldevelopment, Professor Neil Postman (1985)of New York University believes that the mostimportant thing one learns is always

Table IIResults for path analysis

Path Direct Path Indirect Path Total Effect Spurious Total Sumfrom ➜ to Correlation r (1) (2) (1) + (2) (3) (1)+(2)+(3)

Human ➜ Customer 0.499 0.499 0 0.499 0 0.499Human ➜ Structural 0.492 0.524 (0.499)(–0.065) 0.492 0 0.492Customer ➜ Structural 0.197 –0.065 0 –0.067 (0.499)(0.524) 0.197Customer ➜ Performance 0.639 0.560 (–0.065)(0.398) 0.534 (0.499)(0.524)(0.398) 0.639Structural ➜ Performance 0.508 0.398 0 0.398 (–0.065)(0.560) + 0.508

(0.524)(0.499)(0.560)

[ 72 ]



something about how one learns. This notionis similar to the idea of deutero learning asput forward by Argyris, one of the most pro-lific writers on organizational learning.Argyris and Schon (1978) identified threetypes of learning, single loop, double loop anddeutero learning. Through case study analy-sis they examined which of the three types oflearning was most prevalent in business.They concluded that most businesses followsingle loop learning which merely detectsand corrects problems as soon as possible sothat the organization can continue with theirregular activities. Double loop learning onthe other hand not only involves the detectionand correction phase of problem resolution,but also attempts to modify underlyingnorms, policies and objectives. Deutero learn-ing, the most advanced of the three, involvesunderstanding the whole process or learninghow to learn. Although this concept is intu-itively appealing, managers have yet to find apractical means to adopt the deutero learningprocess and will therefore continue to strug-gle to develop intellectual capital.

Conclusion

The management of intellect lies at the heartof value in the current “knowledge era” ofbusiness. Unfortunately, methods of measur-ing and evaluating intellectual capital havebeen slow to develop. There is an extremelylimited literature on the study and manage-ment of intellectual capital. This is partly dueto the privacy that accompanies most organi-zations and their discussion on intellectualcapital. Continued research of this phenome-non should show that organizations with ahigh level of intellectual capital will be thosein which the value-added service of the firmcomes from deep professional knowledge,organizational learning, and protection andsecurity of information. Managers, analystsand researchers should also be wary of look-ing for a formula of intellectual capital. Bydefinition, the tacitness of intellectual capitalmay not allow analysts to ever measure itusing economic variables. A warning must besent out to those accountants and financialanalysts who are asking the question, “Howmuch is my intellectual capital worth?” Aformula may never exist. That is not to saythat metric development is a waste of time.Longitudinal examination of metrics as wellas benchmarking against industry norms canhelp managers in examining their own intel-lectual capital. In this case, examining theprocesses underlying intellectual capitaldevelopment may be of more importance thanever finding out what it is all worth.

Managers who are interested in strategi-cally managing their intellectual capital fortheir own organizations should follow thesesteps (Bontis, 1996):1 Conduct an initial intellectual capital

audit. Such an examination may consistof a survey design and administrationusing Likert-type scales in order to get asnapshot of the benchmark level of intel-lectual capital in existence. Some firmslike Skandia (1994, 1995a, 1995b, 1996a,1996b) use their own metrics of intellec-tual capital. However, each firm is differ-ent and must thrive in the context of itsown industry. Each organization shoulddesign their metrics for their own strate-gic purposes.

2 Make knowledge management a require-ment for evaluation purposes for eachemployee – assign personal targets tointellectual capital development. Forexample, companies can have eachemployee aim to learn something that theorganization currently does not know.

3 Formally define the role of knowledge inyour business and in your industry – findand secure the greatest resources of intel-lectual capital inside and outside yourfirm from places like industry associa-tions, academia, customers, suppliers,and the government.

4 Recruit and hire a leader responsible forthe intellectual capital development ofyour organization. This person musthave an integrated background in humanresources, strategy and informationtechnology.

5 Classify your intellectual portfolio byproducing a knowledge map of your orga-nization – determine in which people andsystems knowledge resides. For example,create a central database in which allcompetitive intelligence information canbe accumulated and accessed.

6 Utilize information systems and sharingtools that aid in knowledge exchange andcodifying such as groupware technology,videoconferencing, Intranets, corporateuniversities and storytelling amongstemployees.

7 Send employees to conferences and tradeshows and have them spy. Do not pay fortheir travel expenses unless they sharewhat they learned with the rest of theorganization when they return.

8 Consistently conduct intellectual capitalaudits to re-evaluate the organization’sknowledge accumulation – use monetaryvalues if at all possible, but do not beafraid to develop customized indices andmetrics.

[ 73 ]



9 Identify gaps to be filled or holes to beplugged based on weaknesses relative tocompetitors, customers, suppliers andbest practices.

10 Assemble the organization’s new knowl-edge portfolio in an intellectual capitaladdendum to the annual report.

As with the human body’s muscles, intellec-tual capital suffers from, if you don’t use it,you lose it (Cohen et al., 1993). There is anincreasing emphasis on survival of the fittestin international competitiveness. In order tostay alive, organizations must win the inter-national organizational learning race (Hampden-Turner, 1992). Future research inthis area may want to tap into comparisons ofknowledge management characteristics bypersonality type with the use of the Myers-Briggs type indicator (Wiele, 1993). Also,researchers could correlate knowledge man-agement with diversity or different leader-ship traits (Boehnke et al., 1997) and deter-mine if in fact there is a relationship betweenthe interorganizational learning of diversegroups or international teams and overallbusiness performance.

The purpose of this study was to explore theideas and concepts that have been publishedthus far on intellectual capital as well as pushthese forward through empirical analysis.The measures and models developed havebeen proven to be valid and reliable in addi-tion to significant and substantive. This phe-nomenon should interest both academics andbusiness practitioners alike because thedevelopment and management of intellectualcapital will require more dedication andeffort in the future relative to the traditionaltasks of monitoring and deploying the physi-cal and capital assets of an organization.Ideally, the shift of thinking in the future willbe from shorter-term product focus strategiesto longer-term human, structural and cus-tomer capital focus strategies.

Finally, all business leaders should beappreciative of the power knowledge manage-ment can have on business performance. Thestudy of intellectual capital produces atremendous amount of energy, energy thatcan take companies far beyond their currentvision (Ward, 1996). It requires people torethink their attitudes on intangible assetsand to start recognizing that measuring andstrategically managing knowledge may makethe difference between mediocrity and excel-lence.

ReferencesAndrews, F.M. (1984), “Construct validity and

error components of survey measures: a

structural modeling approach”, Public Opin-ion Quarterly, Vol. 48, pp. 409-42.

Antal, A.B., Dierkes, M. and Hahner, K. (1994),“Business in society: perceptions and princi-ples in organizational learning”, Journal ofGeneral Management, Vol. 20 No. 2, pp. 55-77.

Argyris, C. (1992), On Organizational Learning,Blackwell, Cambridge, MA.

Argyris, C. and Schön, D. (1978), OrganizationalLearning: A Theory of Action Perspective,Addison-Wesley, Reading, MA.

Barclay D., Higgins, C. and Thompson, R. (1995),“The partial least squares (PLS) approach tocausal modeling”, Technology Studies, Vol. 2No. 2.

Bodie, Z., Kane, A. and Marcus, A.J. (1993), Invest-ments, Irwin, New York, NY.

Boehnke, K., DiStefano, A., DiStefano, J. andBontis, N. (1997), “Leadership for exceptionalperformance”, Business Quarterly, Summer.

Bontis, N. (1996), “There’s a price on your head:managing intellectual capital strategically”,Business Quarterly, Summer.

Bontis, N. (1997), “Royal Bank invests in knowl-edge-based industries”, Knowledge Inc., Vol. 2No. 8.

Brooking, A. (1996), Intellectual Capital: CoreAsset for the Third Millennium Enterprise,International Thomson Business Press, London.

Carmines, E. and Zeller, R. (1979), Reliability andValidity Assessment. Sage Paper Series onQuantitative Applications No. 07-017. Sage,Beverly Hills, CA.

Churchill, G.A. Jr (1979), “A paradigm for develop-ing better measures of marketing constructs”,Journal of Marketing Research, Vol. 16, pp. 64-73.

Cohen, G., Kiss, G. and Le Voi, M. (1993), Memory –Current Issues, Open University, Buckingham.

Crossan, M. and Guatto, T. (1996), “Organizationallearning research profile”, Journal of Organi-zational Change Management, Vol. 9 No. 1.

Crossan, M., White, R.E., Lane, H.W. and Klus, L.(1996), “The improvising organization: whereplanning meets opportunity”, OrganizationDynamics, Vol. 24 No. 4, pp. 20-34.

Darling, M. (1996), “Building the knowledge orga-nization”, Business Quarterly, Winter.

Deng, S. and Dart, J. (1994), “Measuring marketorientation: a multi-factor, multi-itemapproach”, Journal of Marketing Manage-ment, Vol. 10, pp. 725-42.

Dierickx, I. and Cool, K. (1989), “Asset stock accu-mulation and sustainability of competitiveadvantage”, Management Science.

Dillman, D.A. (1978), Mail and Telephone Surveys:The Total Design Method, Wiley & Sons, NewYork, NY.

Drucker, P.F. (1993), Post-Capitalist Society, Butter-worth Heinemann, Oxford.

Edvinsson, L. and Sullivan, P. (1996), “Developinga model for managing intellectual capital”,European Management Journal, Vol. 14 No. 4.

[ 74 ]



Feiwal, G.R. (1975), The Intellectual Capital ofMichal Kalecki: A Study in Economic Theoryand Policy, The University of Tennessee Press,Knoxville, TN.

Fornell, C. and Barclay, D. (1983), Jackknifing: ASupplement to Lohmoller’s LVPLS Program,University of Michigan, Ann Arbor, MI.

Fornell, C. and Larcker, D. (1981), “Evaluatingstructural equation models with unobserv-able variable and measurement error”, Jour-nal of Marketing Research, Vol. 18, pp. 39-50.

Hair, J., Rolph, A. and Tatham, R. (1987), Multi-variate Data Analysis, 2nd ed., Macmillan,New York, NY.

Hampden-Turner, C. (1992), Creating CorporateCulture: From Discord to Harmony, Addison-Wesley, Reading, MA.

Handy, C.B. (1989), The Age of Unreason, ArrowBooks Ltd, London.

Hofstede, G. (1978), “Value systems in 40 coun-tries”, Proceedings of the 4th InternationalCongress of the Association for Cross-CulturalPsychology.

Hudson, W. (1993), Intellectual Capital: How toBuild it, Enhance it, Use it, John Wiley & Sons,New York, NY.

Kohli, A.K. and Jaworski, B.J. (1990), “Marketorientation: the construct, research proposi-tions, and managerial implications”, Journalof Marketing, Vol. 54, pp. 1-18.

Levitt, T. (1991), Marketing Imagination, The FreePress, New York, NY.

Lichtenthal, J.D. and Wilson, D.T. (1992), “Becom-ing market oriented”, Journal of BusinessResearch, Vol. 24, pp. 191-207.

Narver, J.C. and Slater, S.F. (1990), “The effect of amarket orientation on business profitability”,Journal of Marketing, October, pp. 20-35.

Nelson, R.R. and Winter, S.G. (1982), An Evolution-ary Theory of Economic Change, BelknapPress, Cambridge, MA.

Nicolini, D. (1993), “Apprendimento organizzativoe pubblica amministrazione locale”,Autonomie Locali e Servizi Sociali, Vol. 16 No. 2.

Nicou, M., Ribbing, C. and Ading, E. (1994), SellYour Knowledge, Kogan Page, London.

Nonaka, I. and Takeuchi, H. (1995), The Knowl-edge-Creating Company, Oxford UniversityPress, New York, NY.

Nunnally, J.C. (1978), Psychometric Theory, 2nded., McGraw-Hill, New York, NY.

Osland, G.E. and Yaprak, A. (1995), “Learningthrough strategic alliances: processes andfactors that enhance marketing effective-ness”, European Journal of Marketing, Vol. 29,pp. 52-66.

Postman, N. (1985), Amusing Ourselves to Death:Public Discourse in the Age of Show Business,Viking, New York, NY.

Prusak, L. (1996), “The knowledge advantage”,Strategy & Leadership, March/April.

Quinn, J.B. (1992), Intelligent Enterprise, FreePress, New York, NY.

Reich, R.B. (1991), The Work of Nations, Alfred A.Knopf, New York, NY.

Roos, J. and von Krogh, G. (1996), “The epistemo-logical challenge: managing knowledge andintellectual capital”, European ManagementJournal, Vol. 14 No. 4.

Saint-Onge, H. (1996), “Tacit knowledge: the key tothe strategic alignment of intellectual capi-tal”, Strategy & Leadership, April.

Schultz, T.W. (1981), Investing in People: the Eco-nomics of Population Quality, University ofCalifornia, Berkeley and Los Angeles, CA.

Senge, P.M. (1990), The Fifth Discipline: The Artand Practice of the Learning Organisation,Doubleday Currency, New York, NY.

Skandia (1994), “Visualizing intellectual capital inSkandia”, A supplement to Skandia’s 1994Annual Report, Sweden.

Skandia (1995a), “Renewal and development:intellectual capital”, A supplement to Skan-dia’s 1995 Interim Annual Report, Sweden.

Skandia (1995b), “Value-creating processes: intel-lectual capital”, A supplement to Skandia’s1995 Annual Report, Sweden.

Skandia (1996a). “Power of innovation: intellec-tual capital”, A supplement to Skandia’s 1996Interim Annual Report, Sweden.

Skandia (1996b), “Customer value”, a supplementto Skandia’s 1996 Annual Report, Sweden.

Stewart, T.A. (1991), “Brainpower: how intellec-tual capital is becoming America’s mostvaluable asset”, Fortune, 3 June, pp. 44-60.

Stewart, T.A. (1994), “Your company’s most valu-able asset: intellectual capital”, Fortune, 3 October, pp. 68-74.

Stewart, T.A. (1997), Intellectual Capital: The NewWealth of Organizations, Doubleday/Cur-rency, New York, NY.

Sveiby, K.E. (1997), The New OrganizationalWealth: Managing and Measuring Knowledge-Based Assets, Berrett-Koehler, New York, NY.

Toffler, A. (1990), Powershift: Knowledge, Wealthand Violence at the Edge of the 21st Century,Bantam Books, New York, NY.

Von Krogh, G., Roos, J. and Slocum, K. (1994), “Anessay on corporate epistemology”, StrategicManagement Journal, Vol. 15.

Ward, A. (1996), “Lessons learned on the knowl-edge highways and byways”, Strategy & Lead-ership, March/April.

White, G.I., Sondhi, A.C. and Fried, D. (1994), TheAnalysis and Use of Financial Statements,John Wiley & Sons, New York, NY.

Wiele, B. (1993), “Competing from the neck up”,Performance & Instruction, March.

Winter, S.G. (1987), “Knowledge and competenceas strategic assets”, in Teece, D.J. (Ed.), TheCompetitive Challenge: Strategies of IndustrialInnovation and Renewal, Ballinger, Cambridge, MA, pp. 159-84.

Young, M. (1995), “Post-compulsory education andtraining for learning society”, Australian andNew Zealand Journal of Vocational Educa-tional Research.

[ 75 ]



Appendix 1. Summary of survey items (excerpts from questionnaire)

Human capitalH1 competence ideal level H11 employees perform their bestH2 succession training program H12 recruitment program comprehensiveH3 planners on schedule H13R big trouble if individuals leftH4 employees cooperate in teams H14R rarely think actions throughH5R no internal relationships H15R do without thinkingH6 come up with new ideas H16 individuals learn from othersH7 upgrade employees’ skills H17 employees voice opinionsH8 employees are bright H18 get the most out of employeesH9 employees are best in industry H19R bring down to others’ levelH10 employees are satisfied H20 employees give it their all

Customer capitalC1 customers generally satisfied C10 meet with customersC2 reduce time to resolve problem C11 customer info disseminatedC3 market share improving C12 understand target marketsC4 market share is highest C13R do not care what customer wantsC5 longevity of relationships C14 capitalize on customers’ wantsC6 value added service C15R launch what customers don’t wantC7 customers are loyal C16 confident of future with customerC8 customers increasingly select us C17 feedback with customerC9 firm is market-oriented

Structural capitalS1 lowest cost per transaction S9 develops most ideas in industryS2 improving cost per revenue $ S10 firm is efficientS3 increase revenue per employee S11 systems allow easy info accessS4 revenue per employee is best S12 procedures support innovationS5 transaction time decreasing S13R firm is bureaucratic nightmareS6 transaction time is best S14 not too far removed from each otherS7 implement new ideas S15 atmosphere is supportiveS8 supports development of ideas S16R do not share knowledge

PerformanceP1 industry leadership P6 after-tax return on assetsP2 future outlook P7 after-tax return on salesP3 profit P8 overall response to competitionP4 profit growth P9 success rate in new product launchP5 sales growth P10 overall business performance

Note: R – reverse coded items

Appendix 2. Statistical highlights on selected model specification

Table AI

Number of Internal consistency Discriminant validity Ritems (Fornell and Larcker) correlation of constructs squared (%)

Human 7 0.9194 0.936Customer 7 0.9228 0.499 0.940 24.43Structural 7 0.9258 0.492 0.197 0.943 24.89Performance 9 0.9535 0.509 0.639 0.508 0.967 56.02

Human capital H8 H15R H9 H20 H18 H6 H110.8055 0.7855 0.8392 0.8556 0.7006 0.7059 0.8091

Customer capital C14 C1 C16 C9 C6 C5 C80.8189 0.7924 0.8365 0.7297 0.8205 0.7879 0.7706

Structural capital S10 S2 S6 S5 S1 S3 S40.8215 0.8431 0.8030 0.7368 0.7873 0.8058 0.8021

Performance P2 P3 P4 P5 P6 P7 P8 P9 P100.7823 0.8209 0.8978 0.7977 0.8514 0.8348 0.8054 0.7406 0.9591

[ 76 ]



StructuralCapital

R2 = 24.3%

CustomerCapital

R2 = 24.7%

HumanCapital

PerformanceR2 = 56.0%

0.493(22.06)

***

0.400(19.86)

***

0.497(20.26)

***

0.559(33.84)

***

DIAMOND SPECIFICATION

StructuralCapital

0.065(4.62)

SIMPLISTIC SPECIFICATION

PerformanceR2 = 56.9%

HumanCapital

CustomerCapital

0.373(12.48)

***

0.538(24.03)

***

Note: top number is path, t-values in brackets, *** significant at p-value < 0.001

Appendix 3. Further statistical highlights on selected model specifications

Application questions

1 Will “intellectual capital” start appearingon balance sheets as an asset with mone-tary value? How would you start to calcu-ate the intellectual capital asset base ofyour organization?

2 Does high intellectual capital suggestbusiness success?

HumanCapital

0.499(20.5032)

***

CustomerCapital

R2 = 24.53%

StructuralCapital

R2 = 24.89%PerformanceR2 = 56.02%–0.065

(–1.1792)

0.560(32.6709)

***

0.398(20.4885)

***

0.524(16.3878)

***

Note: top number is path, t-values in brackets, *** significant at p-value < 0.001

Figure A1

Figure A2

Intellectual Capital Bontis

Documents

Transcript of Intellectual Capital Bontis