Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
International Marketing Involvement: The Construct, Dimensionality, and Measurement The growing importance of international marketing operations for ABSTRACT the survival and success of an increasing number of businesses underscores the need to understand their involvement in these activities. To this end, this article proposes an eclectic and multidimensional definition as well as a new measure of international marketing involvement where equity, administrative, and operational components represent the three distinct behavioral means that can be utilized by a business to perform foreign marketing activities. Based upon a field study conducted in the United States of 45 firms and 78 product market units, evidence is supportive of the internal consistency and construct validity of the proposed measure of international marketing involvement (IMI). The 1990s have been a watershed for most businesses worldwide where competing internationally has become a necessity rather than a matter of discretion. Consequently, there has been considerable interest among marketing scholars, practitioners, and policymakers in understanding the different facets of a business's international involvement in marketing programs. For marketing scholars, this interest centers on explaining and predicting the different degrees of international marketing involvement. For the marketing practitioner, the issues of concern relate to the assessment, selection, and implementation of different degrees of marketing involvement in serving foreign markets. For policymakers, the trade and balance of payments considerations call for efforts to promote, restrict, or otherwise influence the degree of firms' international marketing involvement so as to enhance national interests. Despite the growth and the widely recognized importance of international marketing involvement, surprisingly little attention has been focused on the systematic analysis of the meaning and measurement of the international marketing involvement concept itself. A cursory review of the literature suggests that international marketing involvement is a complex and, perhaps as a result, poorly understood phenomenon. The mostly imprecise and often contradictory terminology encountered in the literature underscores the concerns about Esra Gencturk Terry L. Childers Robert W. Ruekert Submitted October 1994 Revised April 1995 July 1995 Journal of International Marketing Vol. 3, No. 4, 1995, pp. 11-37 ISSN 1069-031X 11 the existing measures of this construct. The conspicuous absence of psychometric considerations in empirical studies further calls into question the reliability and validity of the existing measures. In the absence of a clearly delineated conceptual definition as well as reliable and valid measures of international marketing involvement, our ability to explain and predict this concept appears to be limited at best. In fact, Sullivan (1994, p. 338) argues that "theory testing remains ambiguous precisely because we cannot ascertain whether the acceptance or rejection of a hypothesis is the result of excessive error in measurement or the adequacy or inadequacy of prevailing theories." This limitation is critical given the importance of international marketing activities and the nontrivial level of investment and risk they entail. The purpose of this study is to overcome these limitations by developing a conceptual and operational definition of international marketing involvement (IMI) which can be incorporated into future research with reasonable assurance of its psychometric properties. To this end, we propose tbe following general yet inclusive constitutive definition of IMI: International marketing involvement refers to the degree to which a Product Market Unit engages in cross-national transfer of resources to support marketing activities in a host-country. Consistent witb the received conceptualizations, the definition proposed in this study posits that the distinguishing feature of IMI is the cross-national transfer of resources so as to serve one or more foreign markets (e.g., Bartels 1968; Cavusgil and Nevin 1981). The emphasis on cross-national performance of marketing activities also suggests that IMI refers to the behavior of a business and needs to be conceptualized as a continuous behavioral construct along a high-low involvement continuum. Furthermore, given the within firm and cross-country differences observed in international involvement behavior (e.g., Douglas and Craig 1983), the proposed definition explicitly acknowledges that IMI needs to be assessed in the context of a Product Market Unit's (PMU's) product, pricing, distribution, and promotion activities in a particular host-country. IMI is also a complex, abstract construct that is inherently ambiguous until it is separated into its component parts. Hence, the precise operationalization of the construct requires identification of the components that represent defining features of the underlying behavior. In this study, we build upon the constitutive definition of IMI and identify three theoretically defensible and distinct dimensions that distinguish 12 Esra Gencturk, Terry L. Childers, and Robert W. Ruekert among the behavioral means that can be deployed by a PMU in performing international marketing activities. In the section below, these dimensions are derived based on a review and synthesis of the various views on IMI. We then describe the measurement scale developed and the empirical research designed to assess its reliability and validity. After reporting the results, their implications for international marketing theory and practice are discussed. ° International involvement has been the subject of considerable research in various disciplines within the general field of international business and economics. Previous studies in the areas of international trade, industrial organization, and market imperfections provide a rich description of and an array of explanations for firms' international involvement behavior. The various definitions and measures of IMI used by these researchers in their quest to understand this phenomenon are discussed below by grouping them according to the attribute(s) emphasized. BACKGROUND AND LITERATURE REVIEW The importance of top management's orientation in shaping a firm's international involvement was originally proposed by Perlmutter (1969). He identified three orientations to describe companies with international involvement. They were: 1) ethnocentric (home country), 2) polycentric (host-country), and 3) geocentric (world) orientation. According to Perlmutter, these different orientations reflect increasing degrees of international involvement, respectively. Similar views on international involvement exist in the literature, each using global orientation as an indicator of a firm with high international involvement. For example, Kobrin (1994) suggests that a firm exploits its geographic scope and become highly involved in international operations when it is deeply committed to a geocentric mindset in its management and decision making. Managerial Orientation Despite the conceptual appeal of the managerial orientation criterion, most of the existing empirical work has concentrated on the geocentric orientation, rather than assessing the range of possible orientations. A notahle exception is a study hy Wind, Douglas, and Perlmutter (1973) which empirically explores each orientation in the context of 15 international marketing decisions. Their results suggest that while greater international involvement is associated with a geocentric orientation, each element of the marketing mix may reflect a different attitudinal orientation. This observed variability may explain the difficulty of using top management's orientation as an indicator of its involvement in international marketing. Several researchers have also shown that top management's orientation serves as an antecedent to the degree of international involvement rather International Marketing Involvement: The Construct, Dimensionality, and Measurement 13 than being its defining characteristic (e.g., Cavusgil 1984; Cavusgil and Nevin 1981). Hence, favorable orientation and deliberate interest in international operations appear to be instruments that allow a firm to mobilize its financial and managerial resources in serving host-country markets. Mode of Operations One of the most frequently discussed attributes of international involvement is the legal form of foreign operations which can range from ownership modes such as wholly owned foreign facilities to nonownership modes such as licensing and export marketing. Based on the market imperfections explanations of international investment (Hymer 1960, 1976; Kindleberger 1969; Rugman 1981), this literature considers equity presence in a foreign market as a hierarchical response to imperfections in international markets. Hence, the need to internalize exchanges to safeguard them against hazards of market imperfections is viewed as the principal reason for different institutional modes. Although acknowledged as a simplification, the prevailing view assumes that ovkrnership modes represent the highest level of international involvement since they entail the most extensive form of participation in and commitment to foreign operations (e.g., Anderson and Gatignon 1986; Erramilli and Rao 1993; Hill, Hwang, and Kim 1990). While this research stream has been instrumental in bringing much-needed theoretical rigor and analytical insight into the mode decision, the use of institutional ownership modes as an indicator of international involvement is not w^ithout its drawbacks. First, there is no empirical evidence in the literature which suggests that the aggregate shifts in legal claims from ownership to nonownership modes are in fact indicative of home-office managers' actual involvement in international marketing activities. In fact, research on international distribution channels suggests that international involvement tends to be issue- or task-specific and may best be captured by the vertical integration of specific activities (e.g., Anderson and Gatignon 1986; Davidson and McFetridge 1985; Klein, Frazier, and Roth 1990). In this context, a given task can be contracted out to external agents or suppliers (i.e., market contracting or nonequity) or it could be internalized and performed by a company's own employees (integrated or equity). Hence, a meaningful assessment of IMI requires addressing equity considerations in terms of vertical integration of marketing tasks rather than as discrete choices among institutional modes of operation. Control 14 While the prevailing view considers a business's legal ownership position and its control over foreign operations to be closely related, the need to distinguish between ownership and control considerations has been acknowledged by several Esra Gencturk, Terry L. Childers, and Robert W. Ruekert researchers (e.g., Anderson and Gatignon 1986; Erramilli and Rao 1993; Sohn 1994). In fact, the importance of control as a distinct consideration separate from the institutional mode decision is reflected in several conceptual definitions of IMI. Root (1984), for example, explicitly specifies control over the policies of foreign affiliates as the essence of international involvement. In examining the licensing agreements, Carstairs and Welch (1982) also provide empirical evidence suggesting that the control exercised can vary not only across different modes but also within a given mode as indicated by differences in the responsibilities assumed and carried out by the licensors. Consequently, it seems possible for a firm to control foreign operations without majority legal ownership or conversely to have little control over these operations despite majority or even complete legal ownership. That is, while institutional modes based on legal ownership claims give a firm legitimate authority to control foreign operations, it does not necessarily imply that this authority will be exercised (c.f., Gaski and Nevin 1985). Yet another view of international involvement is espoused by studies that consider the nonfinancial resources allocated to foreign operations (e.g., Baliga and Jaeger 1984; Cray 1984; Hill et al. 1990; Rosson and Ford 1982). These resources are generally discussed in terms of managerial time and effort expended on foreign activities. Cray (1984), for example, views international involvement as an effort directed at coordinating and supporting geographically dispersed activities. Coordination, as Cray uses the term, is distinguished by situating the foreign operations in a network of relationships with other parts of the firm. In a similar vein, Rosson and Ford (1982) use the term intensity to refer to the degree of contact and resource exchange between home and host-country members. The indicators of intensity used by these authors include number of letters and support materials exchanged, visits made, amount of effort expended, and training provided. Despite the variations in terminology, these studies reflect the importance and relevance of nonfinancial resource commitments in examining IMI. However, due to the prevalence of conceptual articles, the assessment of IMI in terms of deployment of these resources has been limited and inadequate. The definitions in this category emphasize financial performance such as profitability, foreign sales, or foreign assets as an indicator of international involvement. For example. Root (1984) considers a company whose foreign sales are 25 percent or more of total sales as "heavily involved" while Aggarwal (1979) measures international involvement not only by the percentage of foreign sales but also by the percentage of foreign income and assets. Effort Allocation Performance Outcomes International Marketing Involvement: The Construct, Dimensionality, and Measurement 15 While these findings lend some support to the relationship between overall performance of a firm and the extent of its international involvement, they all provide a simplistic and rather arbitrary distinction between high and low degrees of international involvement. The aggregate nature of firms' foreign sales or other financial performance data tends to disguise the underlying substantive differences in foreign operations and, hence, in international involvement behavior. Fluctuations in exchange rates as well as differences in accounting practices can also alter the reported figures substantially. Furthermore, as Mikhail and Shawky (1980) point out, it is not clear whether a firm's superior performance (even when detected) leads to, or is a by-product of, higher international involvement. Hence, in addition to the concerns about the direction of the relationship between performance and IMI, performance itself does not appear to be central to defining the IMI construct. TOWARD A MULTIDIMENSIONAL CONCEPTUALIZATION OF INTERNATIONAL MARKETING INVOLVEMENT As reflected by the various definitional attributes considered in the literature, IMI is a complex construct that is inherently ambiguous until it is separated into its component parts. In this study, IMI is hypothesized to have three distinct behavioral dimensions: 1) Equity Involvement is defined as the degree to which hostcountry marketing activities are performed by PMU employees: 2) Administrative Involvement is defined as the degree to which a PMU exerts control over host-country marketing activities. 3) Operational Involvement is defined as the degree to which a PMU spends time and effort on host-country marketing activities. These three dimensions correspond to the vertical integration, control, and effort allocation attributes, respectively. They are considered central to the IMI construct since they represent theoretically defensible and distinct behavioral means that can be used by a PMU when performing international marketing activities. The central premise of this study is that these three parameters are not interchangeable or synonymous but rather complementary and interdependent as each examines a different aspect of a business's involvement in international marketing activities. As such, their simultaneous assessment is deemed necessary to reflect the full scope of the phenomenon considered. Furthermore, given the observed variation in international involvement behavior across different marketing activities and foreign markets, each dimension incorporates involvement in product, pricing, promotion, and distribution 16 Esra Gencturk, Terry L. Childers, and Robert W. Ruekert distribution activities in a particular host-country. Each dimension is also viewed as a continuous variable where higher degrees of equity, administrative, and operational involvement reflect a higher degree of IMI. In the absence of a theoretical justification for attributing differential importance to any of these dimensions, they are weighted equally and assumed to serve complementary roles in reflecting the overall IMI of a PMU. Hence, over the conceptual range of IMI from highly involved units to relatively uninvolved ones, the three dimensions are expected to be positively related to each other. Any attempt to understand the IMI construct must also account for the relationships it is expected to have with other constructs. To this end, several variables that have generally been associated with IMI are also considered in this study. However, no research to date has examined the effects of these variables on all four elements of the marketing mix or across all three dimensions of IMI. Since the limited scope of past research does not allow a full consideration of a given variable's effect on each dimension of involvement in each type of marketing activity, the hypotheses considered below pertain only to the generally accepted measures of a few antecedent variables whose domain is very well defined and whose relationship to IMI behavior has been empirically established. TT] Related Variables Unit Size: The size of an organization (or subunit) has been used consistently in the extant literature to explain various aspects of its international involvement behavior. For the most part, the conceptual arguments and empirical findings emphasize the size-based advantages (e.g., financial and managerial resources) that enable larger firms to pursue higher degrees of equity involvement. Horst (1972), for example, suggests that larger firms may have an easier time financing the fixed cost of higher equity modes as they are often considered to be a better credit risk than small firms. Based on organizational behavior theory, it has also been argued that an increase in firm size contributes to complexity of the administrative task by compounding the problems of coordination, communication, and control. While larger organizations can enjoy size-based advantages, they also become more complex and differentiated, threatening to overload managers with too many decisions (Baliga and Jaeger 1984). Therefore, difficulty of monitoring increasingly specialized and differentiated roles along with greater administrative distance as well as pressures for administrative efficiency are expected to increase larger businesses' reluctance to exercise direct control over the activities conducted in host-country markets (e.g., Cray 1984; Rugman 1981). These arguments suggest that: International Marketing Involvement: The Construct, Dimensionality, and Measurement 17 Hi: The greater the size of a PMU: a. the greater will be its equity involvement; and b. the lower will be its administrative involvement. International Experience: International experience refers to knowledge accumulated through learning by doing whereby an understanding of international marketing is enriched by previous experience. While the relationship between international experience of a business and its international involvement has been examined extensively, these studies have produced mostly conflicting results. For example, while there is some evidence suggesting that international experience may not have any effect on international involvement (e.g., Kogut and Singh 1988), others have observed a positive (e.g., Gatignon and Anderson 1988), a negative (e.g., Davidson and McFetridge 1985), and a U-shaped relationship (Erramilli 1991). These results can be attributed not only to the different conceptualizations of international experience but also to the failure of previous studies to distinguish between different manifestations of international involvement. Accordingly, we rely mainly on the conceptual literature to posit that international experience will be positively related to equity involvement but negatively related to both administrative and operational involvement. Support for these expectations is provided in several streams of research. First, the importance of international experience in encouraging the use of integrated modes can be traced back to Johanson and Vahlne's (1977) work which views a firm's international involvement as an experiential process whereby degree of involvement is successively increased upon gradual acquisition, integration, and use of knowledge about foreign markets. According to this perspective, the accumulation of knowledge about foreign markets creates a highly specialized staff which reduces the need as well as the willingness to rely on local firms for market expertise and know how (e.g., Horst 1972; Johanson and Vahlne 1977; Rosson and Ford 1982). Based on market imperfection explanations of international involvement, it has also been argued that the tacit nature of experiential knowledge (Teece 1981) would favor internalization of host marketing activities, thereby increasing a firm's equity involvement. The difficulty of codifying experiential knowledge about foreign markets would, however, suggest that international experience cannot be disseminated effectively without the transfer of people who possess the relevant knowledge about foreign markets. As a result, greater international experience facilitates the development of a specialized staff organization in local markets which in turn requires less control and effort allocation by the parent unit. That is, internationally experienced PMUs may find it easier to internalize Esra Cencturk, Terry L. Childers, and Robert W. Ruekert foreign marketing activities and then to decentralize and delegate the decisions about these activities to specialized staff at lower levels in the hierarchy. Taken together, these considerations suggest that: H2: The greater the international experience of a PMU: a. the greater will be its equity involvement; b. the lower will be its administrative involvement; and c. the lower will be its operational involvement. Managerial Attitudes: An important consideration in the internationalization literature is senior management's attitudes toward international marketing activities. While there appears to be a conceptual consensus on the importance of managerial attitudes, only a few attempts have been made to empirically examine their impact on IMI behavior (e.g., Cavusgil and Nevin 1981; Wind et al. 1973; Kobrin 1994). A favorable attitude toward international marketing has generally been viewed as a prerequisite for higher degrees of equity and operational involvement. More specifically, deliberate and positive interest in international marketing is considered to be necessary for the deployment of financial and nonfinancial resources required for host-country marketing activities (e.g., Cavusgil 1984; Cavusgil and Nevin 1981). For example, senior management's negative attitude is expected to increase reliance on nonequity contractual arrangements in an attempt to externalize the fixed costs and the perceived risk generally associated with internalized exchanges. Similarly, negative attitudes and opinions about international marketing activities are likely to decrease the managerial time and effort expended on these activities. Accordingly, it is hypothesized that: H3: The more favorable senior management's attitudes toward international marketing activities: a. the greater will be the unit's equity involvement; and b. the greater will be its operational involvement. The procedures suggested by Nunnally (1978) were followed in generating the multiple item measurement scales used in this study. International Marketing Involvement: Given the eclectic conceptualization of IMI adopted in this study, each of its three components was measured across all four facets of the marketing mix concerning product, pricing, promotion, and distribution activities in foreign markets. The four subscales were constructed based on a sample of the most prominent observables suggested in the literature and the scale items were intentionally kept broad so as to apply to the different dimensions of RESEARCH DESIGN AND DATA COLLECTION Measures International Marketing Involvement: The Construct, Dimensionality, and Measurement 19 international involvement as well as to different products and industries. Care was also taken to ensure that all items conformed to their behavioral definition. The resulting 20 items, consisting of four sets of five-item marketing subscales, and the response formats used in measuring equity, administrative, and operational involvement, are presented in the Appendix. In addition to the aforementioned tripartite measure of IMI developed in this study, its traditional measure based on institutional mode of operations was included in order to address validity considerations. The alternative modes considered in operationalizing the mode of operations measure are provided in the Appendix. Unit Size: The size of a PMU was measured by three openended questions; each of which has been used in past research (Horst 1972; Erramilli and Rao 1993). The items concern the PMU's worldwide: 1) number of employees; 2) total assets (in dollars); and 3) dollar sales. International Experience: Based on measures used in past research (Erramilli 1991; Rosson and Ford 1982), international experience is measured in this study with the following five open-ended questions: 1) the number of different countries in which the PMU's products are currently sold; 2) the number of different foreign countries in which a PMU's products have ever been sold; 3) the number of years since the first foreign sale of a PMU to any foreign country; 4) number of foreign countries where the PMU has a majority-owned subsidiary; and 5) number of majority-owned subsidiaries which currently sell PMU's products. Managerial Attitudes: The managerial attitude scale used in this study was based on the measure used by Cavusgil and Nevin (1981), and Cavusgil (1984). The statements comprising the 6-point Likert scale concerned the degree to which a PMU's senior management: 1) expects international marketing activities to have a positive effect on the unit's profits; 2) expects international marketing activities to have a positive effect on the unit's growth; 3) considers international marketing activities as a waste ofthe unit's resources; 4) considers the unit's domestic activities to be more important than its international marketing activities; 5) has no intention of increasing the unit's international marketing activities; 6) actively explores international market opportunities; and 7) frequently travels abroad. 1— Research Setting 20 The psychometric properties of the scales were tested between late 1987 and 1990 in a field study using cross sectional data from U.S. based international firms located in a Midwestern Esra Gencturk, Terry L. Childers, and Robert W. Ruekert state. To assess the potential biasing effect of this research setting, two preliminary analyses were conducted on measures of international marketing activity for which comparable data were available at the national as well as state level. The results of these analyses suggest that the firms in this research setting do not differ from their national counterparts in terms of either percent of export-related shipments (t=1.04, p=.31) or percent of export-related employment (t=1.36, p=.18). The selection of sample units was accomplished in three stages. First, the International Firm Directory (IFD) published by the state's Trade Office was used to select a random sample of 130 firms. The cbief executive officer named in the directory for each firm was sent a letter introducing the study and a participation request form. After a second wave of letters and request forms, completed participation forms were received from 97firms.Since six mailings bad undeliverable addresses, the returned participation forms represent 78 percent of the 124 net mailings. Sample and Survey Instrument Of tbe 97 firms tbat responded to tbe introductory letter, 58 agreed to take part in tbe study and were contacted in tbe second stage of the sampling process to identify the PMUs of eacb firm. In tbis study, a PMU was operationally defined as tbe smallest set of products wbicb were grouped together for international strategic planning and reporting purposes. In tbe tbird stage, eacb unit manager was asked to identify tbe respondents wbo would be most knowledgeable about the unit's foreign marketing activities. While tbe managers were encouraged to designate as many informants as possible, the concentrated nature of PMU and foreign country knowledge in most firms rendered systematic collection of multiple informant data unfeasible. As a result, only 158 informants were identified across tbe 89 PMUs of tbe 58 firms tbat bad agreed to participate in tbis study. After content validation of tbe measurement scales by five individuals wbo were familiar witb tbe research topic, tbe questionnaire was pretested extensively in personal interviews witb tbree informants from two PMUs. In the final pretest, tbe questionnaire was administered by mail to 10 informants across five PMUs to test tbe data collection procedures. Overall, no difficulties were observed and no major cbanges were made eitber in tbe content or tbe format of tbe instrument. Tbe data for the main study were collected by a mail survey directed to tbe remaining 145 informants at the 82 PMUs. In Pretest of the Instrument Data Collection International Marketing Involvement: The Construct, Dimensionality, and Measurement 21 tbe first section of tbe survey, informants were asked to selfselect tbe PMU and tbe foreign country tbat tbey bad tbe most knowledge about and tben respond to tbe survey questions in reference to their particular selection, unless asked to do otberwise. This procedure enabled an unambiguous identification of tbe unit of analysis and ensured tbat eacb respondent could truly assume tbe role of an informant by providing information about tbe PMU and tbe foreign country tbat s/be believed to be tbe most qualified and knowledgeable about. Tbe follow-up procedure involved tbree mailings yielding usable responses from 78 PMUs in 42 firms. Even tbougb multiple informant data were obtained for 11 PMUs, tbese data bad to be treated as independent observations in subsequent analyses since they addressed PMUs' marketing activities in different foreign countries. Accordingly, the results reported in tbis article are based on informant data generated by the 91 respondents participating in tbe main study. PRELIMINARY ANALYSES Assessment of Sample Firms Assessment of Sample PMUs and Key Informants 22 Prior to assessing tbe psycbometric properties of tbe survey instrument, several preliminary tests were performed to assess tbe appropriateness of tbe sample firms, PMUs, and tbe informants. Tbree analyses were conducted to assess tbe appropriateness of sample firms based on tbe IFD data on dollar sales, number of employees, and firm age. In the first analysis, no significant differences were observed between tbe firms tbat agreed to participate in this study and those that declined to do so in terms of their dollar sales volume (t= .99, p=.32), employees (t= 1.04, p=.3O), or age (t=.2O, p=.8O). In tbe second preliminary analysis, tbe firms responding to eitber tbe first or tbe second wave of tbe participation request mailout were compared, and no significant differences were observed between tbe dollar sales (t=1.48, p=.15), employees (t=.92, p=.36) or age (t=.28, p=.78) of the two groups. Survey nonresponse bias was assessed in the tbird preliminary analysis wbere the results revealed that the 7 nonrespondent firms closely resembled tbe 45 firms witb completed questionnaires witb respect to their sales volume (t=.64, p=.53), employment (t=.96, p=.34), and age (t=1.37, p=.18). Since tbe data on tbe key descriptive parameters of population PMUs were not available, sample variance in relevant PMU cbaracteristics was used to assess tbe appropriateness of tbe PMUs included in tbe main study. Tbe sample statistics on tbe tbree cbaracteristics of 78 PMUs are depicted at tbe top portion of Table 1. Tbe observed sample beterogeneity was conEsra Cencturk, Terry L. Childers, and Robert W. Ruekert sidered desirable for estimating the validity of the proposed scale, since the degree of international marketing involvement of a PMU was hypothesized to be a function of these variables. Characteristics of PMUs Number of countries where PMU's products are currently sold: Worldwide employees: Foreign sales as a % of worldwide sales Mean Range 1 - 115 20 - 10,000 1-90 33 1,089 24 Tahle 1 Characteristics of the Sample PMUs and Informants Characteristics of Informants 2-38 4-41 4-23 Years of international experience 8 Years with PMU selected 9 Years with PMU in the foreign country selected 6 Title Number of Informants (absolute frequency) President/CEO/Chairman Vice President Director PMU/Area Manager Other (e.g., Country Specialist, Brand Manager) 6 15 18 45 7 The appropriateness of respondents as key informants assessed by examining the self-report information provided by each informant in the last section of the survey. While there is no comparable objective benchmark to judge the relative magnitude of informant attributes, the sample statistics on four informant characteristics reported in the lower section of Table 1 provide preliminary evidence for the quality and appropriateness of study informants. As the first step in the measure evaluation process, the reliability of each multi-item scale was estimated based upon domain sampling theory which relates the measurement error directly to the internal consistency of the items included in an instrument (Nunnally 1978). Accordingly, the items considered in each scale were evaluated based on item-to-total correlations to identify and delete ill-fitting items. The pre- and post-purification coefficient alpha estimates as well as the descriptive summary statistics for each multi-item scale are presented in Table 2. PSYCHOMETRIC ANALYSES— MEASURE VALIDATION Internal Consistency International Marketing Involvement: The Construct, Dimensionality, and Measurement 23 Table 2 Internal Consistency Estimates and Descriptive Statistics for the Multi-Item Measures Full Scale Reliability (# of items) Worldwide size 92 (3 items) International experience 83 (5 items) Managerial attitudes .72 (7 items) Purified Scale Reliability (# of items) Mean S.D. .21 2.93 .85 (4 items) .52 3.67 .75 (6 items) 26.75 4.63 .87 (4 items) 29.86 26.80 24.65 17.95 5.60 8.69 7.22 7.36 .93 (4 items) 26.83 21.06 20.39 14.54 6.25 9.83 8.28 7.51 .88 (4 items) 21.04 16.48 16.68 11.28 8.14 8.46 7.51 6.28 Equity involvement Product Price Promotion Distribution .87 (5 .91 ( .85 ( .83 ( items) " ) " ) " ) Administrative involvement Product .87 (5 items) Price .94 ( " ) Promotion .92 ( " ) Distribution .91 ( • " ) Operational involvement Product Price Promotion Distribution .92 (5 .93 ( .91 ( .87 ( items) " ) " ) " ) International Marketing Involvement Measure: The proposed measure of IMI is expected to display a hierarchical factorial structure where the four marketing components represent the separate dimensions subsumed under the more general equity, administrative, and operational involvement dimensions. Since such factorially complex constructs require a separate internally consistent scale for each dimension, reliability assessment was performed separately for each of the four marketing subscales embedded under each involvement dimensions. Based on the item-to-total correlations, the items in the respective four subscales appeared to be relatively equivalent in tapping the domain of their specific marketing activity, except for item 5 in the distribution subscale. Accordingly, the distribution subscales for equity, administrative, and operational involvement were purified by removing item five from further consideration. The pre- and post- purification coefficient alphas in the range of .85 to .94 obtained across the 12 market- 24 Esra Cencturk, Terry L. Childers, and Robert W. Ruekert ing subscales provide strong evidence for the internal consistency of each. Antecedent Variables: Of the three variables hypothesized to influence a PMU's IMI behavior, the item scores of size and international experience scales were standardized through a Z-score transformation because they were both measured by open-ended scale items involving different units of measurement. Accordingly, internal homogeneity of these two scales was assessed based on the corresponding standardized item scores. In the case of PMU size measure, scale purification was not considered necessary because the three scale items exhibited consistently high correlations with the standardized total score. On the other hand, item 5 in the international experience scale and item 7 in the managerial orientation scale were deleted because of their substantially lower correlations with the total score of their respective scales. The purified scales were subjected to a factor analysis to assess the unidimensionality and construct validity of each set of congeneric items. Since the items in a unidimensional scale represent repeated attempts to measure the underlying construct, they are expected to load strongly on one factor to satisfy the requirements of convergent validity while loading weakly on all other factors to meet discriminant validity requirements. While the extraction of the factors was guided primarily by the a priori specification of the number of factors, the highest derived primary and secondary item-factor loadings of ± .40 and above were used in interpreting the instruments' factorial structure. In the interest of space, only a summary of the factor analytic results is provided."• Factorial Composition International Marketing Involvement: Given the complex measurement structure of the IMI scale proposed in this study, the substructure of each dimension was subjected to a common factor analysis.2 Since varimax and oblique rotations yielded similar factor patterns, the results of the varimax solution are used in assessing the factor structure of the IMI dimensions. Four factors with eigenvalues greater than one emerged from the factor analyses, confirming the a priori specification of the four dimensional factorial structure for each component of IMI. The amount of variance extracted in the four factor solutions for the three involvement dimensions was: equity—74 percent; administrative—78 percent; and operational—78 percent. Furthermore, in all the analyses, the items in each marketing subscale had their highest loading on the same factor, and each factor clearly corresponded to a different element of the marketing mix. For the equity involvement dimension, the average loading for the distribution component items was .74, International Marketing Involvement: The Construct, Dimensionality, and Measurement 25 while for product, pricing, and promotion the average loadings were .75, .73, and .68, respectively. In the case of administrative involvement dimension, the average loading for distribution was .71 and for product, pricing, and promotion the average loadings were .74, .76, and .71, respectively. The operational involvement dimension demonstrated comparable loadings with average loadings of .79, .76, .76, and .70 for distribution, product, pricing, and promotion respectively. The pattern of derived factor loadings, hence, reaffirms the unidimensionality of the marketing subscales and provides at least preliminary evidence for their convergent and discriminant validity. Antecedent Variables: To assess the internal structure of the measures used for the antecedent variables, the 13 purified items across the three independent scales were subjected to a principal components analysis with varimax rotation (see Note 2). A three-factor solution was derived accounting for 64 percent of the variance. As expected, the highest loadings on a factor were exhibited only by those items included in the same measure. The items for the size measure had an average loading of .90, for international experience average loadings were .80, and for the managerial orientation scale the average was .66. Consequently, the derived pattern of factor loadings confirms the hypothesized unidimensionality of each scales and provides evidence for their convergent and discriminant validityT; MTMM Examination of IMI Inter-Correlational Properties 26 In view of the encouraging factor analytic result, a more rigoro^^ assessment of construct validity was undertaken by examining the inter-correlation matrix of the IMI subscales. To this end, each marketing mix variable was represented by a composite score based on an unweighted linear sum of the purified item scores in the corresponding marketing subscale. The pairwise correlations among the four marketing mix measures both within and across the three types of international marketing involvement are displayed in Table 3. These correlations can be construed as a quasi multitrait-multimeasure matrix where the equity, administrative, and operational involvement dimensions constitute the traits, and the four marketing subscales constitute the measures. However, since all of these marketing measures rely on self-report data based on a 7-point Likert-type scale, they do not meet Campbell and Fiske's (1959) requirement of distinctly different and independent methods of measuring the same variable. Therefore, the matrix of correlations reported in Table 3 should be evaluated with caution. In this matrix, the previously reported internal consistency coefficients are displayed in parentheses along the main diagonal while the italicized entries identify the validity diagonal of each heterotrait-heteromeasure matrix. Esra Gencturk, Terry L. Childers, and Robert W. Ruekert Pricing Product £7 Al 01 El Al Promotion 01 El Al 01 Distribution El Al 01 Product El (.87) Al .29 .87 01 .03 .40 (.92) Pricing El Al 01 .32 .09 .04 .10 .38 .30 .06 (.01) .34 .39 (.94) .58 .17 .64 (.93) Promotion El .22 Al -.03 01 -.06 .03 .42 .32 .08 .34 .62 .71 .27 .13 .38 .68. .48 .12 (.85) .45 .41 (.92) .73 .11 .57 (.91) Distribution El .24 -.08 Al .07 .43 01 .03 .26 .06 .37 .48 .68 .26 .15 .30 .75 .45 .14 .54 .58 Note: El: Al: OI: .64 .31 .12 .21 .68 .43 Table 3 Within and Cross-construct Pairwise Correlations between the Marketing Subscales .07 (.87) .45 .41 (.93) .61 .33 .71 (.88) Entries in the diagonal are sub.scale reliabilitie.s. Equity Involvement Administration Involvement Operational Involvement Substantively, the correlations in the validity diagonal are all significant at the .01 level, and range from .22 to .73. The degree of agreement observed between multiple attempts at measuring the same trait with different measures provides strong evidence for the convergent validity of the four measures of equity, administrative, and operational international marketing involvement. Preliminary support for discriminant validity is also provided, since each validity estimate exceeds all the cross-dimension correlations between different measures. However, as would be expected, a few of the validity correlations are lower than the corresponding cross dimension correlations between the same measures since the three dimensions share a moderate amount of variance partly because of common measure variance and also due to true variation among related constructs. Even in the presence of common measure variation, the results are encouraging since all the cross-dimension correlations for International Marketing Involvement: The Construct, Dimensionality, and Measurement 27 the same measure are lower than the corresponding internal consistency estimates. Furthermore, a relatively consistent pattern is observed in both the magnitude and order of crossdimension correlations. That is, across most of the heterotrait entries, the lowest correlation is observed between the equity and operational involvement dimensions and the highest correlation is displayed between the administrative and operational involvement. Based on the criteria established by Campbell and Fiske (1959), the above results provide additional evidence for, and reaffirm the convergent and discriminant validity of, the four marketing subscales. These results, taken together, make it possible to examine the psychometric qualities of the entire scale used for the three dimensions of IMI. Accordingly, a separate composite score was calculated for equity, administrative, and operational dimensions as a linear unweighted sum of the corresponding item scores across the four marketing subscales. The internal consistency of each composite scale was then estimated by the formula suggested by Nunnally (1978, pp. 246-54) for calculating the reliability of linear combinations. The resulting internal consistency estimates for the equity, administrative and operational involvement scales equaled .95, .97, and .97, respectively. The validity of these three composite measures was assessed by evaluating their pairwise correlations and also their correlation with the traditional mode of operations measure. The top portion of Table 4 displays these correlations where the entries in parentheses represent the previously reported linear combination reliability estimates. Table 4 Pairwise Product Moment Correlations between Measures and Antecedents of International Marketing Involvement Equity Administrative Operational Measures Equity Administrative Operational Mode of operations (.946)* .328"= .361 '^ (.967) .551 «= -.065 (.969) -.078 .201 ^ -.298^ -.139 .157 a -.298 •= -.269 f^ .255 "J Antecedents Worldwide size (n=52) International experience (n=90} Managerial attitudes (n=91) -.032 .021 .237^ * = Entries in the diagonal are linear combination reliability estimates, a = p<.10 D<.10 b = p<.05 c = p<.01 28 Esra Gencturk, Terry L Childers, and Robert W. Ruekert The pairwise correlations between equity, administrative, and operational involvement measures support our a priori expectations in that they are all significant at the .05 level or below, indicating adequate convergence. However, the magnitude of these correlations is not so large as to suggest the treatment of these three measures as a unidimensional indicator of the same underlying trait. In fact, support for their discriminant validity is provided by reliability estimates that are considerably higher than each measure's correlation with any of the other scales. Additional support for the convergent validity of the equity, administrative, and operational involvement scales was obtained from their correlation with the traditional mode-ofoperations measure which addresses the institutional mode choices. These correlations provide a relatively stringent validation test for two reasons. First, the traditional mode measure involves a single item objective identification of legal ownership structure of foreign operations. As such, it differs from the task-specific, multidimensional measure of international marketing involvement proposed in this study, thereby better satisfying the maximally different method requirement for convergent validity. Second, unlike the tripartite measure of IMI proposed in this study, the mode of operations measure reflects a rather narrow conceptualization of international marketing involvement since it focuses primarily on the legal ownership aspect of foreign marketing activities. ===== Additional Evidence of IMI Convergent Validity ' The traditional mode of operations measure is also expected to display a differential association with each of the three dimensions of IMI. More specifically, we expect the mode of operation measure to have the strongest association with equity dimension of IMI given that both measures are concerned with responses to market imperfections in foreign markets. Yet, equity and mode measures address these responses differently because they operate at two distinct levels. Equity involvement, as used in this study, refers to the extent to which marketing activities are purchased in the market as opposed to performed internally hy company employees. This is a related but separate consideration from mode of foreign operations decisions which is concerned with the legal ownership claims of the parent company in a foreign country. The pairwise correlations of the mode of operations measure reported in Tahle 4 substantiate these expectations. More specifically, support for convergent validity is provided hy the relatively high and significant correlations hetween mode of operations and the equity involvement measure. Similarly, the low and insignificant correlation of mode measure with the measures of administrative as well as operational involvement International Marketing Involvement: The Construct, Dimensionality, and Measurement 29 provide evidence of discriminant validity. These findings also attest to the narrowness of the construct's domain as captured by the traditional mode of operations measure. While the mode of operations criterion has been the most frequently used indicator of international involvement, its usefulness as a single measure of a multidimensional activity appears to be overestimated and indeed potentially misleading. ==^====s========i Correlates of IMI Based on the relatively strong internal consistency and validity evidence provided by the aforementioned analyses, the proposed IMI measure was evaluated further to ascertain its value in producing observable results consistent with a priori expectations. To this end, the composite scores for the three involvement measures were correlated with the measures of the antecedent variables to test the substantive hypotheses considered in this study. In general, the correlational results depicted in the lower portion of Table 4 are encouraging, supporting four of the seven bypotbeses. Furtbermore, two otbers are in tbe expected direction but fail to reacb significance at tbe .05 level. Specifically, tbe observed positive association between a PMU's size and its equity involvement is consistent witb Hla but is significant only at p<.10. On tbe otber band, tbe significant negative relationsbip a PMU's size bas witb administrative involvement confirms Hlb supporting tbe contention tbat home office managers of larger PMUs are less likely to exercise control over foreign marketing activities. While no relationsbip was bypotbesized, tbe results also reveal tbat a PMU's size is not related to tbe operational involvement of its managers in international marketing activities. Tbe results concerning a PMU's international experience reveal tbat its positive relationsbip witb PMU's equity involvement is consistent witb tbe literature yet not significant at .05. In contrast, international experience of a PMU is strongly and negatively related to its administrative and operational involvement, confirming botb H2b and c. As sucb, tbese results provide support for tbe role internationally experienced staff plays in enabling bome office managers to reduce tbe time and effort tbey spend on foreign marketing activities as well as tbe control they exercise over tbem. A finding tbat was not anticipated is tbe insignificant relationsbip between managerial attitudes and equity involvement. Wbile no relationsbip was bypotbesized, managerial attitudes toward international marketing are not found to be related to a PMU's administrative involvement eitber. Tbese insignificant relationsbips are interesting especially in view of tbe attention given to managerial attitudes in tbe extant literature. One pos- 30 Esra Gencturk, Terry L. Childers, and Robert W. Ruekert sible explanation for these results is that a more complex relationship exists where the impact of managerial attitudes is moderated hy variahles not considered in this study. It is also conceivable that managerial attitudes toward international marketing may no longer he an important or relevant factor in today's increasingly glohal environment where almost every manager knows the inevitability of international involvement.^ The results, however, support the importance of managerial orientation for operational involvement in international marketing where a significant positive association is observed, confirming H3c. As such, this study provides support for the contention that favorable managerial attitude is important in making home office managers commit time and effort to foreign marketing activities. In summary, the pattern of correlational evidence supports the broader conceptualization of IMI advocated in this study. The ohserved differences in the direction and magnitude of these correlations underscore the need to examine equity, administrative, and operational involvement as separate behavioral strategies that can be deployed by a PMU in performing international marketing activities. The primary thrust of this study was to develop a comprehensive definition of IMI and to empirically evaluate the psychometric properties of the tripartite measure proposed in this study. The results offer a number of implications for both managers involved with international marketing as well as researchers interested in furthering our theoretical understanding of this important area. However, there are also a numher of limitations of this research that should be explicitly recognized. Ideally, the purified and validated scales should be applied to new data to rule out errors caused by factors external to the instrument or to reduce the possibility of attributing initial findings to chance. Hence, in the absence of a diligent crossvalidation procedure, the IMI scale cannot be offered as a definitive measurement scale. The results of this study merely suggest that the proposed IMI scale is a reasonable approximation of the underlying construct, and that the evidence is encouraging enough to treat it as a reliahle and valid measure in our study context. However, these results should be considered tentative until they are replicated and confirmed by future research. DISCUSSION Limitations International Marketing Involvement: The Construct, Dimensionality, and Measurement 31 Managerial Implications An important implication of this work for international marketing practice comes from the simultaneous consideration of three distinct yet complementary behavioral means that can be deployed by a business in performing international marketing activities. This eclectic approach and the choices it highlights are particularly important since most published material in this area has focused primarily on the institutional mode of foreign operations as an indicator of international marketing involvement. Our results highlight the fact that businesses can and do use means other than direct ownership to increase their involvement in performing marketing activities in a foreign country. The results also confirm the growing reliance of businesses on contractual rather than ownership-based alliances, and attest to tbe increasing importance of administrative and operational involvement in representing a business's involvement in foreign marketing activities. By drawing attention to different types of international involvement and assessing them simultaneously in the context of a single study, the findings suggest that businesses can still be actively involved in foreign marketing activities even though internal or external constraints prevent them from using the traditionally advocated legal ownership route. A second interesting finding from tbis study is tbe lack of direct relationsbip between institutional mode of operations in a foreign market and tbe level of administrative and operational involvement of a business in tbat market. As some managers migbt expect, institutional modes based on legal ownersbip claims were not related to differences in tbe degree of administrative control exercised over bost-country marketing activities or to tbe amount of time and effort a business spends on these activities. Tbe implications of tbis finding is that control is probably determined by a bost of factors and institutional mode of foreign operations is not a particularly good predictor of ultimate managerial control exercised over the foreign operations. Surprisingly, the same general finding was also evident in the relationsbip between mode of entry and amount of effort expended on tbe foreign operation. Theoretical Implications 32 Tbeoretically, support gained for the robustness and generality of relationships predicted from past research as well as the unanticipated relationships provide a basis for future research in the area of international marketing involvement. The results suggest that studies which rely on global estimates of IMI or those that focus on only one type of involvement, such as equity involvement, run the risk of misrepresenting the way businesses conduct marketing activities in foreign markets. Wbile sucb limited measures are easier to use, conclusions on the causes or effects of IMI based on such measures should be viewed as extremely tentative. This is especially relevant since Esra Gencturk, Terry L Ghilders, and Robert W. Ruekert the results of this study suggest that different dimensions of this construct may be influenced by different explanatory factors. Thus, the importance of using a broader definition of the IMI construct as well as measuring and modeling each of its components cannot be overstated for both diagnostic and theoretical purposes. As mentioned earlier, the lack of a strong relationship between institutional mode choices and either administrative or operational involvement poses a number of interesting questions for theory development and testing. While the prevailing view assumes that control and legal ownership claims are strongly and positively related, the results of this study suggest that institutional mode of operations in a foreign market is not significantly related to control exercised over the various types of marketing activities performed in that market. Hence, it appears that the rather global measure of institutional modes may not be sensitive to how particular foreign marketing functions get executed. In this respect, this study's explicit recognition that marketing activities represent a broad set of functions, each of which could theoretically he performed hy either external agents or company employees, hecomes important in capturing equity aspects of IMI hehavior. ACKNOWLEDGMENT This article was partially supported by a grant from the University of Minnesota Curtis L. Carlson School of Management. The authors would like to thank Orville C. Walker, Ir. and Kenneth Roering for their contrihutions to this research project, and gratefully acknowledge the many helpful suggestions provided hy Tamer Cavusgil and the anonymous //M reviewers. As a first attempt at developing and testing a tripartite conceptualization of the IMI construct, this study focused exclusively on relatively well established correlates of IMI. It should he noted, however, that there are other antecedent variahles that influence the nature as well as the magnitude of a PMU's involvement in international marketing activities. Hence, a natural extension of this study is to expand the scope of explanatory variahles in order to yield a broader explanation of a PMU's international marketing involvement behavior. In this respect, particular attention should he focused on identification of external and additional internal determinants of IMI. Finally, future research also needs to explore how well IMI serves as a predictor of a PMU's performance. This could be accomplished in a numher of ways. First, the relationship hetween the three dimensions of IMI could be examined for a single performance outcome, such as profitahility. Similarly, a single dimension of marketing involvement could he related to alternative performance outcomes. Given the importance of these activities, such questions deserve more attention than we have provided to date. It is hoped that the precise operationalization of the IMI scale and its current application provide a solid foundation for future efforts designed to better understand a business's involvement in international marketing activities. THE AUTHORS Esra Gencturk is assistant professor of marketing and international business at the University of Texas at Austin. Terry L. Childers is professor of marketing at the University of Minnesota. Robert W. Ruekert is associate professor of marketing at the University of Minnesota. International Marketing Involvement: The Construct, Dimensionality, and Measurement 33 NOTES i. summary tables depicting the results of all the factor analyses performed are available from the authors upon request. 2. The principal factor analysis was deemed more appropriate than a maximum likelihood confirmatory factor analysis for two reasons. The first reason concerns the preliminary nature of this study in proposing a tripartite conceptualization of international involvement across all four aspects of marketing activities. Hence, a priori specification of measurement parameters is considered to be premature in the absence of a sufficiently detailed theory about the relations among the three dimensions and the corresponding indicators. The second reason is a methodological one concerning the adequacy of sample size, especially since the robustness of the maximum likelihood estimation procedure against small sample sizes remains an unresolved issue. 3. The authors thank an anonymous reviewer for suggesting this explanation. REFERENCES 34 Aggarwal, Raj. "Multinationality and Stock Market Evaluation: An Empirical Study of U.S. Companies and Markets." Management International Review 19 (1979): 5-21. Anderson, Erin M., and Hubert Catignon. "Modes of Foreign Entry: A Transaction Cost Analysis and Propositions." Journal of International Business Studies 17 (Fall 1986): 1-26. Baliga, B. R., and Alfred Jaeger. "Multinational Corporations: Control Systems and Delegation Issues." Journal of International Business Studies 3 (Fall 1984): 25-40. Bartels, Robert. "Are Domestic and International Marketing Dissimilar?" Journal of Marketing 32 (July 1968): 56-61. Campbell, Donald T, and Donald Fiske. "Convergent and Discriminant Validation by the Multitrait-Multimethod Matrix." Psychological Bulletin 56 (March 1959): 100-122. Carstairs, R. T., and L. S. Welch. "Licensing and the Internationalization of Smaller Companies: Some Australian Evidence." Management International Review 22 (1982): 33-44. Cavusgil, Tamer S. "Organizational Characteristics Associated with Export Activity." Journal of Management Studies 21 (January 1984): 3-22. Cavusgil, Tamer S., and John R. Nevin. "Internal Determinants of Export Marketing Behavior: An Empirical Investigation." Journal of Marketing Research 18 (February 1981): 114-19. Cray, David. "Control and Coordination in Multinational Corporations." Journal of International Business Studies 15 (Fall 1984): 85-98. Esra Gencturk, Terry L. Childers, and Robert W. Ruekert Davidson, William H., and D. G. McFetridge. "Key Characteristics in the Choice of International Technology Transfer Mode." Journal of International Business Studies 16 (Summer 1985): 521. Douglas, Susan, and C. Samuel Craig. "Examining Performance of U.S. Multinationals in Foreign Markets." Journal of International Studies (Winter 1983): 51-62. Erramilli, Krishna M. "The Experience Factor in Foreign Market Entry Behavior of Service Firms." Journal of International Business Studies 22, no. 3 (1991): 479-501. Erramilli, Krishna, and C. P. Rao. "International Entry Mode Choice: A Modified Transaction Cost Analysis Approach." Journal of Marketing 57 (July 1993): 19-38. Gaski, John F., and John R. Nevin. "The Differential Effects of Exercised and Unexercised Power Sources in a Marketing Channel."/ourna7 of Marketing Research 22 (May 1985): 130142. Gatignon, Huhert, and Erin Anderson. "The Multinational Corporation's Degree of Control Over Foreign Subsidiaries: An Empirical Test of a Transaction Cost Explanation." Journal of Law and Economics (Fall 1988): 305-336. Hill, Charles W. L., Peter Hwang, and W. Chan Kim. "An Eclectic Theory of the Choice of International Entry Mode." Strategic Management Journal 11, no. 20 (1990): 117-28. Horst, Thomas. "Firm and Industry Determinants of the Decision to Invest Abroad: An Empirical Study." The Review of Economics and Statistics 54 (August 1972): 258-66. Hymer, S. The International Operations of National Firms: A Study of Direct Investment. Cambridge, MA: MIT Press, 1960, 1976 Johanson, J., and J. Vahlne. "The Internationalization Process of the Firm: A Model of Knowledge Development and Increasing Foreign Commitments." Journal of International Business Studies 8 (Spring/Summer 1977): 23-32. Kindleberger, Charles P. "The Theory of Direct Investment." In International Financial Management, edited by Donald L. Lessard, 19-29. Boston: Warren, Gorham and Lamont, 1969. Klein, Saul, Gary Frazier and V. J. Roth. "A Transaction Cost Model of Channel Integration in International Markets." Journal of Marketing Research 27 (May 1990): 196-208. Kobrin, Stephen J. "Is There a Relationship Between Geocentric Mind-Set and Multinational Strategy?" Journal of International Business Studies 25, no. 3 (1994): 493-511. Kogut, Bruce, and Harbir Singh. "The Effect of National Culture on the Choice of Entry Mode." Journal of International Business Studies 19 (Fall 1988): 411-32. Mikhail, A. D., and H. A. Shawky. "Investment Performance of U.S.-Based Multinational Corporations: Comments and a Perspective on International Diversification of Real Assets: A Reply." Journal of International Business Studies 11 (Spring/Summer 1980): 105-106. International Marketing Involvement: The Gonstruct, Dimensionality, and Measurement 35 Nunnally, Jum C. Psychometric Theory. New York: McGraw-Hill Book Company, 1978. Perlmutter, Howard V. "The Tortuous Evolution of the Multinational Corporation." Columbia Journal of World Business 4 (January/February 1969): 9-18. Root, Franklin R. International Tirade and Investment. Cincinnati: South-West Publishing Co., 1984. Rosson, Philip J., and I. David Ford. "Manufacturer-Overseas Distributor Relations and Export Performance." Journal of International Business Studies 13 (Fall 1982): 57-72. Rugman, Alan M. Inside the Multinationals: The Economics of Internal Markets.New York: Columbia University Press, 1981. Sohn, Jung Hoon Derick. "Social Knowledge as a Control System: A Proposition and Evidence from the Japanese FDI Behavior." Journal of International Business Studies 25 (1994): 295-324. Sullivan, Daniel. "Measuring the Degree of Internationalization of a Firm." Journal of International Business Studies 25, no. 1 (1994): 325-42. Teece, David J. "The Multinational Enterprise: Market Failure and Market Power Considerations." Sloan Management Review 23 (Spring 1981): 3-17. Wind, Yoram, Susan P. Douglas, and Howard V. Perlmutter. "Guidelines for Developing International Marketing Strategies." Journal of Marketing 37 (April 1973): 14-23. 36 Esra Gencturk, Terry L. Ghilders, and Robert W. Ruekert APPENDIX Measures of International Marketing Involvement Scale Mode of Operations Equity Involvement Scale Items Response Anchors Internally developed wholly owned subsidiary Acquired wholly owned subsidiary Joint Venture Licensing Direct exporting Indirect exporting via merchant distributors Indirect exporting via commission agents Product Subscale 1. Product Design Specifications 2. Package Design Speciflcations 3. Process Design Specifications 4. Brand Name Selection 5. Development of New Products Pricing Subscale 1. End User Price Specification 2. Provision of Credit 3. List price Specification 4. Dealer Allowance 5. Price Discounts Promotion Subscale 1. Sales Force Training 2. Advertising 3. Non-advertising Promotion 4. Design of Promotional Materials 5. Personal Selling Distribution Subscale 1. Warehousing 2. Inventory Management 3. Local Transportation 4. Order Processing 5. Specification of Territorial Coverage l=Always Performed by Company Employees 7=Never Performed by Company Employees Administrative Involvement *t l=Home-office Managers Have Complete Control 7=Home-Office Managers Have No Control Operational Involvement t l=Home-Office Managers Spend Very Little Time and Effort 7=Home-Office Managers Spend A Lot of Time and Effort * In analyzing the data, the equity and administrative involvement scales were reverse coded such that higher values represent more of the trait. t The marketing subscale items for Administrative and Operational Involvement are the same as those for Equity Involvement. International Marketing Involvement: The Construct, Dimensionality, and Measurement 37