Export information acquisition modes: measure ... - CiteSeerX

Export information acquisition modes: measure development and validation Anne L. Souchon

Lecturer in Marketing, School of Business and Public Management, Victoria University of Wellington, Wellington, New Zealand and

Adamantios Diamantopoulos

Professor of Marketing and Business Research, The Business School, Loughborough University, Loughborough, UK

Export information acquisition 143 Received February 1998 Revised May 1998 September 1998 Accepted September 1998

Keywords Decision making, Export, Information, International marketing, Marketing research Abstract Export information acquisition has mostly been examined disparately as researchers have tended to focus on certain modes of information acquisition independently of others. Furthermore, past studies have typically employed single-item measures to operationalize information acquisition. The present study attempts to redress these deficiencies by considering a comprehensive set of export information acquisition modes and by developing psychometrically sound measures for each. The results show the adequacy of considering three broad export information acquisition modes (export marketing research, export assistance, and export market intelligence), each of which is operationalized by means of a multi-item scale. The latter are shown to be reliable and to possess content, convergent, discriminant, and nomological validity.

It has been pointed out that `ìnformation resources have a life cycle similar to that of other resources. The life cycle contains five phases: (1) planning, (2) acquisition, (3) stewardship, (4) exploitation, and (5) disposal'' (Weitzel, 1987, p. 10). The literature on export information has tended to focus on the first two stages, namely: (1) the identification of information needs and sources (which type(s) of information tend to be collected, and where from), and (2) the process of information acquisition (how such information will be collected). The authors would like to acknowledge Professor Constantine Katsikeas and two anonymous reviewers for helpful comments on a previous version of this paper.

International Marketing Review, Vol. 16 No. 2, 1999, pp. 143-168. # MCB University Press, 0265-1335

International Marketing Review 16,2 144

Yet, despite the interest that the topic of export information acquisition continues to raise (see Souchon and Diamantopoulos, 1996; Leonidou and Katsikeas 1997), very little attention has been paid to the simultaneous measurement of different information acquisition modes or the development of psychometrically sound scales of export information acquisition. The motivation for developing export-specific scales of information acquisition can be justified on several grounds. First, information is a powerful determinant of successful business decisions (Barabba, 1983; Barabba and Zaltman, 1991). Although use, rather than mere acquisition of information, directly affects the decisions made (Zaltman and Moorman, 1988; Hart and Diamantopoulos, 1993), information cannot logically be used unless it has previously been gathered (Weitzel, 1987). Second, exporters often cite lack of information as a major barrier to entering new foreign markets and an obstacle for expanding current export operations (e.g. Cavusgil and Naor, 1987; Katsikeas, 1994; Leonidou, 1995). Thus, information acquisition has been identified as a crucial activity for exporting firms (Cavusgil, 1984) in order to reduce uncertainty associated with export activities (Diamantopoulos et al., 1993; Koh et al., 1993; McAuley, 1993). Third, although effective use of export information has been shown to impact on export performance (Souchon and Diamantopoulos, 1997), acquisition of such information is a substantially more complex (and costly) activity than information collection in a domestic context (Loudon, 1975; Douglas and Craig, 1983; Grùnhaug and Graham, 1987). However, despite its importance and complex nature, export information acquisition ``has received disproportionately less empirical attention compared to other lines of enquiry in exporting'' (Leonidou and Katsikeas, 1997, p. 66). Fourth, export information acquisition can be undertaken by distinct modes (see Souchon and Diamantopoulos, 1996), for one of which (namely, export assistance) there is no domestic equivalent (Seringhaus, 1986/1987). Moreover, although there is evidence to suggest that the mode of acquisition has an impact on the extent to which and way in which export information is utilized (Sood and Adams, 1984; Garrett and Hart, 1993; Diamantopoulos and Souchon, 1996), there is no explicit analysis in the literature regarding exporters' usage of alternative acquisition modes. This is despite recent calls to establish ``the relative reliance of the firm on export marketing research versus export assistance versus export market intelligence'' (Souchon and Diamantopoulos, 1996, p. 66). Fifth, unlike in a domestic context (e.g. Sinkula and Hampton, 1988; Sinkula, 1990) there are no existing and psychometrically tested measures of export information acquisition. While several previous studies have investigated issues related to information collection by current and potential exporters (for a recent review, see Leonidou and Katsikeas, 1997), in most instances rather simplistic (single-item) measures have been used (often of the ``yes/no'' type); more important, ``seldom is any attempt made to test the measurement quality

of the indicators used. As in any research arena, attention should be paid to establishing the validity and reliability of the constructs used'' (Diamantopoulos et al., 1993, p. 12). In light of the above, the purpose of the present study is the construction of reliable and valid measures of distinct export information acquisition modes using rigorous measure development procedures (e.g. Churchill, 1979; DeVellis, 1991; Spector, 1992). The development of a comprehensive measuring instrument capturing information acquisition by exporters would be of value both as a management tool in its own right and as input to subsequent analyses linking information resources to export performance. Possible applications include diagnostic use (e.g. to what extent are alternative export information acquisition modes tapped by the firm?), evaluative use (e.g. given current reliance on certain modes, what is the perceived quality of the information generated?), and predictive use (e.g. what are the decision-making and performance implications of relying on alternative information acquisition modes?). In the section that follows, we briefly discuss a threefold classification of export information acquisition modes which is used to provide the conceptual underpinning of the investigation. We then provide preliminary support for the chosen classification by reporting on exploratory interviews with export managers. Next, we describe the data collection procedures for the main study and highlight issues related to questionnaire design and sample selection. The measure development procedure is elaborated next, addressing questions of dimensionality, reliability and validity. We conclude the paper with a descriptive analysis of the derived scales and some suggestions for further research. Information acquisition in exporting In a broad sense, `ìnformation acquisition'' can be defined as the generation of information relevant for decision making. It is an organizational-level construct, since ``the generation of information is considered to be the responsibility of all departments throughout the organization'' (Cadogan and Diamantopoulos, 1995, p. 43). The conceptual domain of export information acquisition is somewhat narrower in that it refers to the generation of information specific to export decision making (Cavusgil, 1984). Depending on the firm's resources, organizational structure, and stage of internationalization, generation of export information may be undertaken within a specialist unit (e.g. export department), be part of the activities of a marketing or sales department, and/or be (partly) out-sourced to independent firms (e.g. marketing research agencies) (Samiee and Walters, 1990; Schlegelmilch et al., 1993; Belich and Dubinsky, 1995; Diamantopoulos and Cadogan, 1996). Irrespective of the specific organisational arrangements for information acquisition, the construct needs to be sharply delineated from information use, the latter referring to ``whether information is actually taken into account when making decisions'' (Souchon and Diamantopoulos, 1996, p. 53).

Export information acquisition 145


The current interest lies on alternative modes of information acquisition used by exporters, the latter referring to the mechanisms through which firms gather information relevant to export activities. Based on the literature, and following Souchon and Diamantopoulos (1996), three such modes can be distinguished based on a combination of: (1) the nature of the information provided (objective versus experiential), (2) the degree of formality/objectivity of the gathering process (i.e. formal versus informal), and (3) the location of the information (i.e. internal versus external). The first acquisition mode is export marketing research which refers to ``the research activities of firms carried out either in the home market or in foreign markets for the purpose of reducing uncertainty surrounding international marketing decisions'' (Cavusgil, 1984, p. 262). It is distinct from other types of export information acquisition modes, because it is formal, systematic and objective (Schlegelmilch et al., 1993). Export marketing research can be carried out internally (i.e. through an in-house unit/department) and/or externally (i.e. through a specialized market research agency/consultant); according to survey evidence, only about one in two exporters conducts export marketing research (Diamantopoulos et al., 1990; Schlegelmilch et al., 1993) and the sophistication of the research conducted lags behind that for the domestic market (Cavusgil, 1984; Diamantopoulos et al., 1991). The second information acquisition mode, export assistance, is provided by official bodies such as banks or governments, and encompasses ``first, standardized and customized market information and guidance on exporting and export marketing, and second, more comprehensive programs ranging from helping firms research specific foreign markets, market visits ± individual or with trade missions ± trade fairs, to actual market entry'' (Seringhaus, 1985, pp. 294-5). Export assistance is typically aimed at small and medium-sized companies and attempts to provide information which can be useful at different stages of the internationalization process (Diamantopoulos et al., 1993). This acquisition mode tends to be used less frequently either because the information obtained is considered ineffectual by exporters (e.g. Grùnhaug and Lorentzen, 1983) or because the latter's lack of awareness of assistance bodies and programs is low (e.g. Reid 1984). Finally, export market intelligence is an informal export information acquisition mode which includes approaching potential customers, distributors, and/or competitors, through attendance at international trade fairs and shows, or more directly through foreign visits (Kleinschmidt and Ross, 1984; Denis and Depelteau, 1985; Saunders and Jones, 1990). It is an ongoing information acquisition process, quite distinct from marketing research, reflecting `à set of procedures used by managers to obtain their everyday information about pertinent developments in the marketing environment''

(Kotler, 1991, p. 91). This form of direct, personal export information acquisition appears to be the information collection procedure which exporters favor (e.g. Reid, 1984; McAuley, 1993; Leonidou and Katsikeas, 1997). Collectively, export marketing research, export assistance and export market intelligence provide a comprehensive specification of potential information acquisition modes used by exporters, which is consistent with previous classifications found in the literature (e.g. Pointon, 1978; Sood and Adams, 1984; Seringhaus, 1988). However, a degree of overlap among the three modes is inevitable. For example, while participating in a seminar on exporting organized by an export assistance body, a decision maker may meet other delegates and thus informally obtain relevant information; the latter could be conceivably viewed as constituting export intelligence. Similarly, in the course of obtaining data for a marketing research project, a research agency may obtain secondary data prepared by an export assistance organization. While such overlap implies that the three export acquisition modes are not entirely mutually exclusive, it should be appreciated that ``there are fundamental differences among the three modes'' (Souchon and Diamantopoulos, 1996, p. 52). For example, compared to export intelligence, marketing research is a much more formalized, planned and focussed information acquisition mode (Douglas and Craig, 1983). Furthermore, ``the type of knowledge generated by export intelligence is largely experiential, unlike market research information which generates objective market knowledge'' (Diamantopoulos and Horncastle, 1997, p. 263). Similar arguments can be offered to support the distinct nature of export assistance versus the two other acquisition modes (see also Sood, 1981; Sood and Adams, 1984; Reid, 1984). It is important to note that not all firms may employ all three modes to generate export information. For example, some organizations may rely only on export market intelligence and/or export assistance because resource constraints and the longer payback periods associated with the use of formal export marketing research may make the latter unattractive (Douglas and Craig, 1983). Similarly, more experienced exporters tend to make less use of export marketing assistance services, having developed their own (i.e. internal) information gathering mechanisms (Seringhaus, 1986/1987; Young et al., 1989). It could also be the case that, depending on the particular type of export decision to be made (Goodman, 1993) or the particular stage of the decisionmaking process (Saunders and Jones, 1990), different emphasis is placed on export marketing research versus assistance versus intelligence[1]. These are currently issues on which very little is known as no empirical study has as yet directly investigated the relative influence of the three gathering approaches on export decision making. Preliminary research While the threefold classification of acquisition modes along the lines described above provides a conceptually appealing and convenient framework for studying export information, it is important that the chosen scheme is broadly



Table I. Export information classification: data display

consistent with executives' own perceptions of export information acquisition modes. To this end, a qualitative approach was initially adopted involving an experience survey of export managers. Specifically, 12 in-depth interviews were conducted with managers from UK exporting firms of different sizes, operating in different industries, and having different levels of export involvement and experience. Given the exploratory nature of this stage of the research, the small sample size was deemed suitable, not for the generalizability of the results but for gaining preliminary insights into the issues of interest and generating suitable items for the measure development procedure (Hart, 1987). Data analysis of the interview transcripts was conducted along the lines advocated by Miles and Huberman (1994) and involved the generation of in vivo codes, development of within-case displays, and construction of cross-case displays (meta-matrices). During the interviews, the export managers mentioned a variety of export information sources, which could be readily classified as export marketing research, export assistance, or export market intelligence (Table I)[2]. Moreover, the respondents drew sharp distinctions across the three information acquisition modes in terms of whether the nature of information generated tended to be objective or experiential (see Seringhaus, 1986/87), whether information gathering was undertaken in a formal vs informal manner (see Reid, 1984), and whether export information was generated internally (i.e. in-house) or externally (Pointon, 1978). Thus, the literature-based threefold classification was not found to be at odds with managers' own categorizations of alternative mechanisms of export information generation. Export marketing research tended to be considered reliable by the companies which used this source because it is a service they paid for. As a managing director said: ``we actually pay to have the surveys done annually, so we get pretty reliable data''; this assessment reflected the formal and systematic nature of this mode which was associated with specialist information providers. However, marketing research was considered less reliable than export market intelligence, because the information did not come from people who were actual players in the industry, but only observers. For Export marketing research Internal crash team Internal analysts Internal reports Test marketing External marketing research Independent agencies Marketing analysts Consultancy groups

Export market intelligence Agents/distributors Books/magazines/trade journals Dealer network/suppliers Competition Contacts/networking/traveling Related business Customers/market Engineering Exhibitions Material control Salespeople Company library

Export assistance Chamber of Commerce Conferences Local embassy

example, a sales director stated: `Ì think the information we build up informally is probably better than the information we get through formal market research ... When you approach the market from a very formal basis and do marketing research and come up with a set of results, they could vary quite dramatically from the information that you can gather from people who are really, really on the ground and very close to what is happening. What we found was that information gathered within the company was far more accurate.'' Many reservations were voiced with regard to the information provided by export assistance bodies. Of the 12 managers interviewed, seven reported not acquiring such information. A key reason for this was lack of awareness of what help the latter can provide; for instance, one manager admitted that ``we are not aware of all that is available through export assistance bodies ...''. Other companies had made use of assistance bodies in the past but did not rely on them any longer because government bodies were seen to provide `ìnformation that is so far behind, it is generally totally useless''. Assistance bodies were also accused of supplying information that is too general to suit the specific needs of exporters; for instance, a sales director said `Ì am not sure it is relevant to our particular market, because it is a business-to-business market and a very specialized field. Therefore, it is difficult for a generalist agency such as the DTI to give the specific information that we require.'' Export market intelligence was the unanimously favored way of acquiring information and, in the words of a managing director: ``the most reliable source of information is the company's own one''. The reason given for this was mainly because the information comes directly from the marketplace and, thus, is experiential in nature. That is why, although it is ``free'' information (for it generally comes from the company's own staff having/making contacts), it tends to be trusted more than any other form of export information collection method. Main survey A questionnaire was subsequently developed on the basis of the literature and the exploratory study. Each item within the questionnaire was drawn either from the literature or from expressions employed by respondents during the course of the qualitative study. Table II shows the constructs covered in the questionnaire, including those used for validation purposes. The questionnaire was pretested using a protocol approach with several export decision makers and via two separate mail pilots of 100 firms each. After implementing suggested revisions, the research instrument was administered to a sample of 1,327 UK exporters, drawn from the Dun & Bradstreet database. The specific respondents targeted were (in order of preference) export directors, sales/marketing directors, managing directors, and finance directors, depending upon the contact available in the Dun & Bradstreet database. In addition, since the contacted person may not always be the most knowledgeable on export-related issues, the recipient was asked to



Table II. Questionnaire content

Construct Export marketing research (EMR) Undertaken in-house Purchased on a subscription basis Commissioned to independent agencies Percentage of all export information collected through EMR (validation item)

Illustrative source Sinkula Sinkula Sinkula Sinkula

(1990) (1990) (1990) (1990)

Export assistance (EA) Government departments Trade associations Export association/club Chamber of Commerce Confederation of British Industry Banks Embassies Trade missions Conferences/seminars on exporting Percentage of all export information collected through EA (validation item)

Crick et al. (1994) Diamantopoulos et al. (1991) Bodur and Cavusgil (1985) Exploratory study McAuley (1993) Walters (1983) Crick et al. (1994) Seringhaus and Mayer (1988) Grùnhaug and Lorentzen (1983) Sinkula (1990)

Export market intelligence (EMI) Trade journals/magazines/books Contacts at trade fairs/exhibitions Overseas staff Export customers Export agents/distributors Percentage of all information collected through EMI (validation item)

Walters (1983) Hart et al. (1994) Keegan (1974) Koh (1991) Bodur (1994) Sinkula (1990

Information source awareness Self-rated awareness of export information sources (five-point scale)

Seringhaus (1986)

Company size Total number of employees

Hart et al. (1994)

Export experience Number of years exporting

Bodur (1994)

Export dependence Export sales to total sales (%)

Diamantopoulos and Schlegelmilch (1994)

Export specificity Presence of an export department

Samiee and Walters (1990)

pass the questionnaire on to whoever was deemed to be in the best position to provide the requested information[3]. Two weeks after the initial mailing, follow-up letters were sent out to a random subsample of 300 non-respondents. A total of 198 usable responses was obtained and deemed an adequate sample size for measure development purposes (Spector, 1992). After adjusting for noneligible firms (e.g. firms no longer in business, firms no longer exporting), the effective response rate came to 24 percent (Wiseman and Billington, 1984). The

replies were then analyzed for non-response bias using a time trend procedure (Armstrong and Overton, 1977) which involved splitting the sample between early and late respondents; a series of t-tests showed no significant differences in the variables of interest between the two groups. In addition, to obtain firsthand information about reasons for non-response, 100 randomly selected nonrespondents were contacted by telephone; this showed that the main reason was ineligibility (e.g. the firm no longer exported) or that time pressure or company policy prevented participation in any survey. None of these reasons suggests that non-response reflects characteristics which directly and differentially affect responses to the substantive issues under study (Lesley, 1972). Thus, again, no evidence indicating the presence of non-response bias was obtained. Measure development The first step in measure development was to decide on the nature of the measurement model to be used to link the construct(s) involved to empirical indicators[4]. Specifically, a choice had to be made between using ``cause'' (formative) or `èffect'' (reflective) indicators to represent the construct(s) of interest (Bollen and Lennox, 1991). In this context, ``cause indicators are observed variables that are assumed to cause a latent variable. For effect indicators the latent variable causes the observed variable'' (Bollen, 1989, p. 65). Distinguishing between the two types of indicators is important because the measurement procedure is different in each case. With cause indicators, one seeks to form an index, and ``reliability in the internal consistency sense and construct validity in terms of convergent and discriminant validity are not meaningful when indexes are formed'' (Bagozzi, 1994, p. 33). In contrast, with effect indicators, one seeks to develop a scale for which ``reflective indicators provide a way to test hypotheses about the properties (e.g. reliability, single factoredness) and construct validity of the measures'' (Bagozzi, 1994, p. 332). In the present case, and after much deliberation, it was decided to use effect indicators to represent each export information acquisition mode. Four key considerations led to this decision. First, recalling that the concept of information acquisition refers to the generation of information, reference to the general marketing literature revealed that information generation has been operationalized by means of effect rather than cause indicators (e.g. Jaworski and Kohli, 1993; Kohli et al., 1993). Second, previous multi-item measures of information acquisition in a domestic setting have taken the form of scales rather than indexes and have been shown to be highly reliable (e.g. Sinkula, 1990)[5]. Third, similar efforts in an export context have also approached measure development using effect indicators with good results (e.g. Belich and Dubinsky, 1995). Finally, and most important, the exploratory interviews with export managers showed that the various sources under each acquisition mode were viewed in the same light yet differently from sources under the other modes; put another way, respondents



seemed to draw broad distinctions between the three modes but not within each mode. In light of these considerations, it was decided to develop three separate scales of export information acquisition, one for each acquisition mode. The psychometric properties of the three scales were subsequently assessed using established measurement development procedures (e.g. Churchill, 1979; Carmines and Zeller, 1979; DeVellis, 1991; Spector, 1992; Nunnally and Bernstein, 1994). Dimensionality was first examined, as it has been argued that the most critical and basic assumption of (classic) measurement theory is that a set of items captures just one underlying construct (Hattie, 1985). Following Carmines and Zeller (1979), a two-stage strategy was followed in dimensionality assessment. First, a principal component analysis was run on all items comprising the three modes of export information acquisition to provide initial support for the chosen threefold classification. Subsequently, principal axis factoring was employed on the items comprising each individual mode of information acquisition; in this analysis, and consistent with the use of effect indicators (see above), only shared variance of each set of items was considered[6]. The second step in constructing the measures involved the assessment of reliability (internal consistency). For each acquisition mode, an item analysis was performed, the aim being to identify the items which form an internally consistent scale, and eliminate those which do not (Churchill, 1979; DeVellis, 1991). The procedure followed was to correlate each item with the sum of the remaining items (item-whole correlations) and with every other item (inter-item correlations). Spector (1992) describes two criteria for retaining items within the scale: (1) to set a specific number of items which should be contained within a scale, and (2) to set a minimum item-whole correlation coefficient beyond which the items should be excluded from the scale. In the present study, only items with statistically significant (at p < 0.05) itemwhole correlations were to be included within each scale. Following the item analysis, Cronbach's (1951) measure of internal consistency (alpha) was computed for the three scales. Once dimensionality and reliability had been assessed, scale validation was undertaken. The three scales (export marketing research, export assistance, and export market intelligence information acquisition) were considered to be content valid since: (1) all items pertaining to each scale were derived from the literature, (2) the construct domains had been clarified during the qualitative interviews, and (3) the domains were clearly defined in the introduction in the questionnaire in order to enable the respondents to relate easily to the specific constructs (see Carmines and Zeller, 1979).

Convergent and discriminant validity were assessed via a variation of the multitrait-multimethod (MTMM) matrix (Campbell and Fiske, 1959), while nomological validity was established by linking the scales to other theoretically relevant constructs uncovered via a literature review (Peter, 1981). Dimensionality and reliability When a set of items is supposed to measure more than one construct (as is the case when all indicators for the three export information acquisition modes are considered concurrently), a principal component analysis (PCA) should fulfil the following conditions: ``(1) the number of statistically meaningful components should equal the number of hypothesized phenomena; (2) after rotation, specific items should have higher factor loadings on the hypothesized relevant component than on other components; and (3) components extracted subsequent to the number of hypothesized components should be statistically unimportant and substantively uninterpretable'' (Carmines and Zeller, 1979, p. 24).


As Table III shows, the PCA conducted on the entire set of acquisition items meets these requirements. There are three components which are easily interpretable as representing:

Items

P1

Government departments Chambers of Commerce Conferences Trade missions Embassies Trade associations Export associations Banks Contacts with or visits to export customers Contacts with or visits to agents/distributors Company's own staff abroad Personal contacts made at trade fairs/exhibitions Export marketing research commissioned to outside agencies Export marketing research information purchased on a subscription basis Export marketing research conducted in-house Confederation of British Industry Trade journals/magazines/books

0.77 0.73 0.67 0.62 0.61 0.57 0.51 0.50

Eigenvalues Variance explained (%)

Loadings P2 P3

0.83 0.78 0.69 0.63

0.74 0.71 0.65 0.54

4.50 2.26 26.46 13.32

1.51 8.89

Table III. PCA of export information acquisition items


(1) export assistance (P1); (2) export market intelligence (P2); and (3) export marketing research (P3)[7]. Moreover, out of 17 items, no fewer than 15 (88 percent) load highly on the appropriate information mode. Having established the viability of classifying export information acquisition into the three proposed modes, the dimensionality of the items comprising each mode was further investigated by means of principal axis factoring (PAF). In this context, Hair et al. (1992, p. 237) recommend that if one is testing a hypothesis about the number of factors underlying a set of items, then the analyst should ``simply instruct the computer to stop the analysis when the desired number of factors has been extracted''. Evidence to suggest unidimensionality would be provided if all the items load significantly on a single factor; for sample sizes of 200 (like the current sample) loadings of 0.18 or greater can be considered significant at the 1 percent level (Hair et al., 1992). Tables IV, V, and VI show the results for the PAF analyses for the three export information acquisition modes (as well as reliability coefficients and descriptive statistics for each scale). The loadings for all items are significant on a single factor solution across all three modes. Moreover, the magnitudes of the loadings are substantial (all but one are greater than 0.40). Thus the unidimensionality of each of these scales is supported. Regarding reliability, for the export marketing research scale, computation of Cronbach's (1951) alpha yielded a value of 0.64, falling somewhat short of Nunnally's (1978) widely-used guideline of 0.70 as a minimum acceptable level. However, the formula for coefficient alpha shows that as the number of items increases, so does the coefficient. The scale in question is composed of only three items which would explain a coefficient alpha slightly below the

Factor loadings

Itemwhole correlation

Export marketing research information purchased on a subscription basis Export marketing research commissioned to outside agencies Export marketing research conducted in-house

0.78 0.77 0.75

0.44 0.48 0.46

Eigenvalue

1.77

Items

Table IV. Dimensionality and reliability of the export marketing research scale

Mean 6.15

Standard deviation

Summary statistics Average inter-item correlation

Alpha

Number of cases

2.46

0.39

0.64

198

ItemFactor whole loadings correlations

Items Government departments Trade missions Conferences Export associations Trade associations Chambers of Commerce Embassies Banks Confederation of British Industry

0.75 0.55 0.64 0.50 0.52 0.57 0.62 0.43 0.47

Eigenvalue

3.56

Mean

Standard deviation


Alpha

18.77

5.39

0.32

0.80

0.66 0.49 0.56 0.46 0.47 0.50 0.55 0.38 0.43

Number of cases 198


Table V. Dimensionality and reliability of the export assistance scale

generally accepted yardstick. Using the Spearman-Brown prophecy formula, the number of items necessary to obtain a coefficient alpha of 0.70 was subsequently calculated (Nunnally and Bernstein, 1994); this resulted in four items, which signifies that only one additional item to the export marketing research scale would have ensured a coefficient alpha of 0.70 if the additional item were to have been of the same quality as the original three items[8]. Analysis of the export assistance items uncovered a minimum item-whole correlation coefficient of 0.38. The scale attained an alpha value of 0.80, well

Factor loadings

Items Contacts with or visits to export customers Contacts with or visits to export agents/distributors Company's own staff abroad Personal contacts made at trade fairs/exhibitions Trade journals/magazines/books

0.83 0.74 0.62 0.47 0.29

Eigenvalue

2.43

Mean

Standard deviation


18.44

4.02

0.34

Alpha 0.72

Itemwhole correlation 0.66 0.58 0.50 0.43 0.26

Number of cases 198

Table VI. Dimensionality and reliability of the export market intelligence scale


above the threshold level for reliable scales advocated by Nunnally (1978). As a result, all nine items were retained within the scale. When assessing the reliability of the five export intelligence items, a minimum item-whole correlation coefficient of 0.26 was obtained, and alpha was found to be 0.72. While the item with the low item-whole correlation (trade journals/magazines/ books) could have been eliminated with little loss in reliability, it was retained for conceptual reasons, i.e. because it was the only item capturing non-personal sources of export information[9]. Convergent and discriminant validity A strict MTMM approach (Campbell and Fiske, 1959) could not be used to assess the convergent and discriminant validity of the three export information scales, as alternative (i.e. ``maximally-different'') methods of measurement for the constructs involved were not available[10]. Thus, a ``second best'' approach had to be followed whereby the scales reflecting the three export information acquisition modes were correlated with the percentages of export information collected via each of the three modes (as derived from a constant sum technique ± see Table II). Figure 1 shows the resulting matrix, with the three modes reflecting the ``trait'' and the two measures (scale scores and percentages) reflecting the ``method'' components respectively. The validity diagonal (between the two dotted triangles) shows that both the export marketing research (EMR) and export assistance (EA) scales present good evidence of convergent validity. The heterotrait-monomethod (the triangles drawn in solid lines) and heterotrait-heteromethod (the triangles in dotted lines) triangles also show evidence of discriminant validity for these two scales. However, the matrix highlights that the export market intelligence

Figure 1. Convergent and discriminant validity of export information acquisition scales

EA

.365 .000

EMI

.280 .000

.219 .001

% EMR

.717 .000

.299 .000

.150 .020

% EA

.131 .036

.473 .000

–.042 .281

.130 .037

% EMI

–.523 .000

–.409 .000

.030 .342

–.689 .000

–.612 .000

%EMR

%EA

EMR

EA

EMI

(EMI) scale does not correlate significantly with the proportion of export information acquired via export market intelligence. The case of matrices not fully meeting the Campbell and Fiske (1959) criteria is common and the interpretation of the degree of validity exhibited by the scales can be difficult (Peter, 1981). One reason for the present finding may well be found in the dissimilarity of the two variables intended to capture export market intelligence acquisition. Specifically, the percentage of export information collected via export market intelligence need not equate the frequency with which this particular mode is consulted; the non-significant result may reflect an attempted validation of a trait with a dissimilar trait rather than from poor scale definition. For example, it could be the case that export market intelligence sources are consulted frequently (indeed, this often occurs in the course of day-to-day business), but relatively little useful information is actually gathered that way. In contrast, the very purpose of conducting export marketing research or consulting export assistance sources is to generate information and, thus, a much closer correspondence would be expected between the frequency of these activities and the quantity of information generated. While there are positive significant relationships between the three scales of export information acquisition (see top left triangle in Figure 1), the moderate size of the relevant correlations implies that this should not be interpreted as indicating that export information acquisition constitutes a single latent variable. Indeed, the appropriateness of considering three separate acquisition modes was highlighted in the exploratory study and supported by the earlier PCA analysis (see Table III); the hypotheses set out in the next section also indicate that different antecedent factors impinge upon different modes. Note also that, in relative terms, firms which primarily rely on export market intelligence as a mode of information acquisition tend to rely less on export marketing research and export assistance (see pattern of correlations in bottom right triangle of Figure 1). Nomological validity The export literature suggests a number of antecedent factors to, and outcomes of, export information acquisition (for a review, see Leonidou and Katsikeas, 1997). In some instances, however, the literature is divided and some linkages remain controversial. For example, the link between collection of export marketing research information and export experience is ambiguous; while some authors have observed a positive relationship (e.g. Cavusgil, 1984), other researchers found a negative relationship (e.g. Sood and Adams, 1984), and a third group of studies uncovered no differences in export marketing research acquisition based on the number of years a firm has been exporting (e.g. Koh et al., 1993). Assessing the scale's nomological validity based on ambiguous theory would be pointless as disconfirming evidence could be attributed to shortcomings of the theory rather than those of the scale (see Peter, 1981). Therefore, in what follows, only relationships which are firmly based on



established theory are examined to ascertain the nomological validity of the three export information acquisition scales. Note also that the aim is not to develop a comprehensive model of the antecedents (or outcomes) of export information acquisition but simply to test a few theory-driven hypotheses as part of measure validation (a much more modest aim); in this context, it is an `èrroneous conclusion that only formal, fully developed theories are relevant to construct validation .... What is required is that one should be able to state several theoretically derived hypotheses involving the particular construct'' (Carmines and Zeller, 1979, p. 24). Awareness of export assistance sources of information has long been recognized as a prerequisite for collection of export assistance information (Seringhaus, 1985, 1986; Diamantopoulos et al., 1993). The exploratory study also highlighted the importance of export assistance awareness on the collection of export assistance information. Thus, H1: Export information source awareness is expected to be positively related to export assistance acquisition. The amount of resources available to larger firms signifies the possibility of acquiring more `èxpensive'' export information from specialized marketing research firms (Douglas and Craig, 1983). Larger firms are also more likely to possess in-house marketing research capabilities (Cavusgil, 1984; Schlegelmilch et al., 1993) and employ more sophisticated methods for gathering information than small and medium-sized companies (Pointon, 1978). Indeed, marketing research is a sophisticated information gathering mechanism (Douglas and Craig, 1982) which is mostly employed by large firms (Cavusgil, 1984), whereas smaller firms are unlikely to collect export marketing research information largely due to lack of resources (Reid, 1984; Samiee and Walters, 1990; Crick et al., 1994). Therefore, H2: Company size is expected to be positively related to export marketing research acquisition. The literature is quite clear as to the effect of export experience on collection of export assistance information. Czinkota (1982) notes that the more export experience companies possess, the lower their belief in export assistance. Cavusgil (1984) also found that the greater the export experience of companies, the less they rely on export assistance information sources. Furthermore, most firms tend to rely, to some extent, on export market intelligence as a source of information (Seringhaus, 1988), the most common being personal contacts with customers, suppliers etc. (McAuley, 1993; Leonidou and Katsikeas, 1997). Importantly, it takes time to develop these contacts and networks from which to acquire export information (Crick et al., 1994). Therefore, H3: Export experience is expected to be negatively related to export assistance, and H4: positively related to export market intelligence acquisition.

Past studies relating the impact of export dependence on the acquisition of export marketing research information provide evidence that the greater the export dependence, the more resources companies devote to formal research (e.g. Cavusgil, 1984; Sood and Adams, 1984; Koh, 1991). A high level of export sales compared to domestic sales is an indication of the importance of export activities to the company (Diamantopoulos and Inglis, 1988). The more dependent companies are on exporting, the more critical export information is to them, since a mistake would have more damaging repercussions on the company's performance (Axinn, 1988). Therefore, given the high uncertainty and risks associated with foreign markets (Young et al., 1989), acquisition of export information represents a critical activity for firms which are highly dependent on exporting (Koh et al., 1993). Such firms attempt to reduce the risks associated with foreign markets by seeking international information (Cavusgil, 1984; Sood and Adams, 1984). Thus, H5: Export dependence is expected to be positively related to export marketing research, and H6: Export market intelligence acquisition.


As Table VII shows, the vast majority of the postulated relationships turned out to be positive and significant, providing support for the nomological validity of the measures; only one linkage (H3) was found to be contrary to expectations, indicating a positive link between export experience and export assistance acquisition[11]. Further nomological validition was undertaken by the examination of the relationships between acquisition modes and export specificity. It has been argued that the presence of a separate export department within an organization tends to reflect the company's commitment to its exporting operations (Samiee and Walters, 1990). Since it has also been found that ``the more committed the firm was to export markets, the more aggressive it became in export marketing and the more sophisticated the research methods it employed'' (Cavusgil, 1984, p. 270), it is likely that the reliance on export marketing research as a mode of export information acquisition will increase with the export specificity of the firm (i.e. whether or not the organization has a separate export department handling its export business). In addition, the Source awareness EMR EA EMI

n/a (H1) Sig.

n/a

0.206 0.002

Firm size (H2) Sig.

n/a n/a

0.271 0.000

Export experience n/a (H3) Sig. (H4) Sig.

Notes: n/a = not applicable; Sig. = one-tailed significance level

0.108 0.065 0.279 0.000

Export dependence (H5) Sig. (H6) Sig.

n/a

0.178 0.006 0.425 0.000

Table VII. Nomological validation of export information acquisition scales


presence of a separate export department within the company is likely to be positively related to the exporters' awareness of export information sources. Companies which are knowledgeable about exporting will be in a better position to determine what information to collect and where to collect it from than less knowledgeable exporters (Seringhaus and Rosson, 1990). Given the fact that lack of awareness of export assistance sources is a key determinant of export assistance's lack of use (Seringhaus, 1986), the presence of a separate export department is likely to enhance collection of information from export assistance bodies. Finally, the literature also suggests that companies which conduct their export business through an export department tend to meet more frequently with their export customers (Koh, 1991); such direct visits fall under what has been defined earlier as export market intelligence (see Table II). As a result: H7: Export specificity is expected to be positively related to export marketing research; H8: Export assistance; and H9: Export market intelligence acquisition. Three independent-sample t-tests were computed in order to test the above relationships. The results are all significant and consistent with expectations, providing further nomological support for the proposed scales (Table VIII). Scale description Table IX provides summary measures of location and variability for each scale; to make the descriptive statistics between the three scales directly comparable, total scores were divided by the number of items comprising the corresponding scale. Almost 44 percent (87 firms) of the total sample of exporters averaged a score of 4 or higher for export market intelligence information acquisition. In contrast, only 2 percent (four companies) of the respondents averaged a score equal to or greater than 4 on the export marketing research scale, and only a

Sample size

Mean

Sig-t

Acquisition of export marketing research information (H10) No export department Export department

125 73

1.9 2.3

0.001

Acquisition of export assistance information (H11) No export department Export department

125 73

2.0 2.3

0.002

125 73

3.5 4.0

0.000

Table VIII. Further nomological validation of the export Acquisition of export market intelligence information (H12) information acquisition No export department Export department scales

single company averaged a score over 4 on the export assistance scale! An interesting but not altogether surprising finding is the small spread of the respondents around the small mean for the export assistance scale. These results support Denis and Depelteau's (1985) contention that the two main types of information sources which exporters tend to rely upon are: (1) intelligence; and (2) marketing research.


Institutional sources tend to be disregarded by the majority of exporters, probably reflecting the fact that export assistance is often considered not specific enough to provide useful data. To assess the (mean rank) differences of the export marketing research, export assistance, and export market intelligence scales, a Friedman two-way ANOVA test was conducted. As can be seen from Table X, the probability of obtaining the mean rank differences through sampling error is virtually nil, reflecting significant differences between the frequency with which each export information acquisition mode is consulted. It can thus be concluded that, although the three scales are significantly and positively correlated (see Figure 1 earlier), exporters tend to consult export market intelligence sources much more frequently than export marketing research or export assistance sources. This finding corroborates past research on export information acquisition which has dealt with information sources on a single-item basis (e.g. McAuley, 1993). Conclusion and future research directions The results of the present study support the categorization of export information acquisition along three distinct modes, namely: (1) export marketing research, (2) export assistance, and (3) export market intelligence.

EMR EA EMI

Mean

Standard deviation

Median

Mode

2.051 2.085 3.689

0.820 0.599 0.805

2.000 2.111 3.80

1.000 2.222 4.000

EMR

Mean ranks EA

EMI

Chi-square

DF

Sig.

N

1.46

1.60

2.94

267.180

2.0

0.000

198

Table IX. Scale summary statistics

Table X. Friedman two-way ANOVA for export information acquisition modes


Three multi-item scales of these modes were developed and shown to be psychometrically sound in terms of unidimensionality, internal consistency, and validity. Nevertheless, the present study is but a first attempt at empirical measurement of export information acquisition modes using multi-item scales. Replication of the scales is clearly necessary in order to establish the stability of their psychometric properties and generalizability to other settings (e.g. in different countries or between industrial and consumer goods companies). The three scales provide a platform from which to initiate empirical assessment of the antecedents and outcomes of export information acquisition. Regarding the former, it has been pointed out that ``very few attempts have been made to investigate the effect of organizational factors on firms' export behavior pertaining to information sources'' (Leonidou and Katsikeas, 1997, p. 66). While some of these factors were investigated in the present study in the context of measure validation, the development of an integrated framework encompassing a variety of organizational (and environmental) antecedents is clearly a major avenue for future research. With regard to the outcomes of information acquisition, the links to information use and organizational learning are obvious issues for future study. According to Goldstein and Zack (1989), information supply (i.e. acquisition) is followed by information use which, in turn, precedes knowledge acquisition. Parallel to this conceptualization, Fiol and Lyles (1985, p. 811) describe learning as ``the development of insights, knowledge, and associations between past actions, the effectiveness of those actions and future action''. Thus acquisition of export information has an impact on export learning (Johanson and Vahlne, 1977) via the process of information use (Sinkula 1994; Souchon and Diamantopoulos, 1996). Moreover, acquisition of export information can be expected to have an (indirect) impact on export performance, mediated by information use (Souchon and Diamantopoulos, 1997; Diamantopoulos and Souchon, 1998); this is an issue of obvious importance to practitioners but one on which there is very little empirical evidence. The two proposed streams of study: (1) comprehensive assessment of antecedent factors, and (2) evaluation of ultimate impact on export performance are clearly complementary and have important managerial implications: identification of the acquisition mode(s) providing the ``best'' information for decision making (i.e. that which ultimately results in increased export success) in conjunction with the examination of the factors affecting such acquisition, should result in solid, empirically-based guidelines on how to best manage information acquisition in an export context. Notes 1. This is not to say that information in general and export information in particular are only used for decision-making purposes. For example, information can be utilized to broaden the general managerial knowledge base without being specifically considered in the

2.

3.

4.

5. 6. 7.

8.

9. 10.

11.

making of a particular decision (e.g. Rich, 1977). This is commonly referred to as ``conceptual'' use of information, while use of information in the making of specific decisions is termed `ìnstrumental'' use (Caplan et al., 1975). Information can also be used to acquire power in the workplace (e.g. Piercy, 1983), to keep information providers content (e.g. Menon and Varadarajan, 1992), or to justify decisions made prior to the acquisition of information (e.g. Feldman and March, 1981). An extensive discussion of alernative uses of information in a marketing context can be found in Menon and Varadarajan (1992), while an analysis of information use in an export context is provided by Souchon and Diamantopoulos (1996, 1997) and Diamantopoulos and Souchon (1996, 1998). Note that Table I is constructed directly from the in vivo quotes of the respondents in the exploratory study. As such, it contains some (trivial) redundancy (e.g. `èxternal marketing research'' and `ìndependent agencies'' both clearly refer to commissioned marketing research). We eliminated this redundancy when developing items for the main survey instrument (see Table II). In other instances ± namely in the case of export assistance ± we added items (drawn from the literature) to ensure that the domain of the construct was appropriately captured. These procedures are entirely in line with the recommendations of the measure development literature (e.g. Churchill, 1979; DeVellis, 1991; Spector, 1992). The choice of export decision makers (whether they be export managers, marketing managers, or managing directors) as informants was justified on the basis that these individuals are likely to be the most knowledgeable about the firm's export activities; selection of respondents, therefore, was not based on the assumption that they would be the only persons acquiring export information within their organization. This step is more often than not bypassed in measure development studies, not least because ``traditional measures of reliability and the examination of the correlation matrix of indicators are so ingrained that researchers have failed to realise that these are not appropriate under all situations'' (Bollen and Lennox, 1991, p. 312). This implies that the scale items are strongly intercorrelated, a key requirement with effect indicators (see Bollen, 1989; Bollen and Lennox, 1991). The choice of the common factor model was based on the fact that ``the primary objective is to identify the latent dimensions or constructs represented in the original variables'' (Hair et al., 1992, p. 231). Given this overall picture, the fact that one item fails to load highly (4 0.40) on any component (journals/magazines/books) and another loads on the wrong component (the CBI) is most probably a statistical aberration rather than indicative of an alternative factorial structure. Interestingly, Slater (1995) reports that about half of the papers published in leading marketing journals (e.g. Journal of Marketing, Journal of Marketing Research) which employed multi-item scales do not meet the Nunnally (1978) criterion: he states that the ``criterion in use'' tends to be closer to 0.60 than to 0.70. Moreover, the low values obtained could reflect sampling variability and, given its conceptual importance, it is best to retain this item until independent replications of the proposed scale show otherwise. Convergent validity requires that a measure of a given construct correlates positively with other measures of the same construct; discriminant validity entails that a measure of a construct does not correlate highly with measures from which it is supposed to differ (see Peter, 1981). Whether the observed positive relationship is only the result of a sampling artifact or whether it reflects the fact that experienced UK exporters behave differently from their US counterparts with respect to export assistance (since the hypothesized negative relationship was based predominantly on US evidence) is an interesting question itself but clearly outside the scope of the present paper.



References Armstrong, J.S. and Overton, T.S. (1977), `Èstimating nonresponse bias in mail surveys'', Journal of Marketing Research, No. 14, August, pp. 396-402. Axinn, C.N. (1988) `Èxport performance: do managerial perceptions make a difference?'', International Marketing Review, No. 5, Summer, pp. 61-71. Bagozzi, R.P. (1994), ``Structural equation models in marketing research: basic principles'', in Bagozzi, R.P. (Ed.), Principles of Marketing Research, Blackwell Publishers, Cambridge, MA. Barabba, V.P. (1983), ``Making use of methodologies developed in academia: lessons from one practitioner's experience'', in Kilman, R. et al. (Eds), Producing Useful Knowledge for Organizations, Praeger, New York, NY. Barabba, V.P. and Zaltman, G. (1991), Hearing the Voice of the Market ± Competitive Advantage through Creative Use, Harvard Business School Press, Boston, MA. Belich, T.J. and Dubinsky, A.J. (1995), ``Factors related to information acquisition in exporting organizations'', Journal of Business Research, No. 33, pp. 1-11. Bodur, M. (1994), ``Foreign market indicators, structural resources, and marketing strategies as determinants of export performance'', in Cavusgil, S.T. (Ed.) Advances in International Marketing, No. 6, JAI Press, Greenwich, CT, pp. 183-205. Bodur, M. and Cavusgil, S.T. (1985), `Èxport market research orientations of Turkish firms'', European Journal of Marketing, Vol. 9 No. 2, pp. 5-16. Bollen, K.A. (1989), Structural Equations with Latent Variables, John Wiley & Sons, New York, NY. Bollen, K. and Lennox, R. (1991), ``Conventional wisdom on measurement: a structural equation perspective'', Psychological Bulletin, Vol. 110, pp. 345-70. Cadogan, J.W. and Diamantopoulos, A. (1995), ``Narver and Slater, Kohli and Jaworski and the market orientation construct: integration and internationalization'', Journal of Strategic Marketing, Vol. 3 No. 1, pp. 41-60. Campbell, D.T. and Fiske, D.W. (1959), ``Convergent and discriminant validation by the multitrait-multimethod matrix'', Psychological Bulletin, Vol. 56 No. 2, pp. 81-105. Caplan, N., Morrison, A. and Stambaugh, R.J. (1975), The Use of Social Science Knowledge in Policy Decisions at the National Level, Institute for Social Research, Ann Arbor, MI. Carmines, E.G. and Zeller, R.A. (1979), Reliability and Validity Assessment, Sage, London. Cavusgil, S.T. (1984), `Ìnternational marketing research: insights into company practices'', Research in Marketing, Vol. 7, pp. 261-88. Cavusgil, S.T. and Naor, J. (1987), ``Firm and management characteristics as discriminators of export marketing activity'', Journal of Business Research, Vol. 15, pp. 221-35. Churchill, G.A. (1979), `À paradigm for developing better measures of market constructs'', Journal of Marketing Research, Vol. 16, February, pp. 64-73. Crick, D., Jones, M. and Hart, S. (1994), `Ìnternational marketing research activities of UK exporters: an exploratory study'', Journal of Euromarketing, Vol. 3 No. 2, pp. 7-26. Cronbach, L.J. (1951), ``Coefficient alpha and the internal structure of tests'', Psychometrika, Vol. 16 No. 3, September, pp. 297-334. Czinkota, M.R. (1982), `Àn evaluation of the effectiveness of US export promotion efforts'', in Czinkota, M.R., and Tesar, G. (Eds) Export Policy: A Global Assessment, Praeger, New York, NY, pp. 63-71. DeVellis, R.F. (1991), Scale Development: Theory and Applications, Sage, London.

Denis, J.E. and Depelteau, D. (1985), ``Market knowledge, diversification and export expansion'', Journal of International Business Studies, Fall, pp. 77-89. Diamantopoulos, A. and Cadogan, J.W. (1996), `Ìnternationalizing the market orientation construct: an in-depth interview approach'', Journal of Strategic Marketing, Vol. 4, pp. 23-52. Diamantopoulos, A. and Horncastle, S. (1997), `Ùse of export marketing research by industrial firms: an application and extension of Deshpande and Zaltman's model'', International Business Review, Vol. 6 No. 3, pp. 245-70. Diamantopoulos, A. and Inglis, K. (1988), `Ìdentifying differences between high- and lowinvolvement exporters'', International Marketing Review, Vol. 5 No. 2, pp. 52-60. Diamantopoulos, A. and Schlegelmilch, B.B. (1994), ``Linking export manpower to export performance: a canonical regression analysis of Europe and US data'', in Cavusgil. S.T. (Ed.), Advances in International Marketing, JAI Press Inc., Greenwich, CT, pp. 161-81. Diamantopoulos, A. and Souchon, A. (1988), `Ìnformation utilisation by exporting firms: conceptualisation, measurement, and impact on export performance'', in Information and Management, Gabler, Wiesbaden. Diamantopoulos, A. and Souchon, A. (1996), `Ìnstrumental, conceptual and symbolic use of export information: an exploratory study of UK firms'', in Cavusgil, S.T. (Ed.), Advances in International Marketing, Vol. 8, JAI Press, Greenwich, CT, pp. 117-44. Diamantopoulos, A., Schlegelmilch, B.B. and Allpress, C. (1990), `Èxport marketing research in practice: a comparison of users and non-users'', Journal of Marketing Management, Vol. 6 No. 3, pp. 257-73. Diamantopoulos, A., Schlegelmilch, B.B. and Tse, K. (1993) `Ùnderstanding the role of export marketing assistance: empirical evidence and research needs'', European Journal of Marketing, Vol. 27 No. 4, pp. 5-18. Diamantopoulos, A., Schlegelmilch, B.B., Tse, K. and Allpress, C. (1991) `Èxport marketing research: a comparative analysis of Finnish and British firms'', Proceedings of the Annual Conference of the Academy of Marketing Science, Fort Lauderdale, FL. Douglas, S.P. and Craig, C.S. (1982), ``Marketing research in the international environment'', in Walter and Murray (Eds), Handbook of International Business, John Wiley & Sons, New York, NY, Chapter 30. Douglas, S.P. and Craig, C.S. (1983), International Marketing Research, Prentice-Hall, Englewood Cliffs, NJ. Feldman, M.S. and March, J.G. (1981) `Ìnformation in organizations as signal and symbol'', Administrative Science Quarterly, Vol. 26, pp. 171-86. Fiol, C.M. and Lyles, M.A. (1985), `Òrganizational learning'', Academy of Management Review, Vol. 10 No. 4, pp. 803-13. Garrett, K. and Hart, S. (1993), `Àn investigation of market research activities of multinational SBUs'', in Chiar, J. and Sureda, J. (Eds), Marketing for the New Europe: Dealing with Complexity, Proceedings of the 22nd Annual Conference of the European Marketing Academy, Vol. 1, pp. 519-42. Goldstein, D.K. and Zack, M.H. (1989), ``The impact of marketing information supply on product managers: an organizational information processing perspective'', Office: Technology and People, Vol. 4 No. 4 , pp. 313-36. Goodman S.K. (1993), `Ìnformation needs for management decision making'', Records Management Quarterly, Vol. 27 No. 4, pp. 12-23. Grùnhaug, K. and Graham, J.L. (1987), `Ìnternational marketing research revisited'', in Cavusgil, S.T. (Ed.), Advances in International Marketing, Vol. 2, JAI Press, Greenwich, CT, pp. 121-37.



Grùnhaug, K. and Lorentzen, T. (1983), `Èxploring the impact of governmental export subsidies'', European Journal of Marketing, Vol. 17 No. 2, pp. 5-12. Hair, J.F., Anderson, R.E., Tatham, R.L. and Black, W.L. (1992), Multivariate Data Analysis with Readings (3rd ed.), Macmillan, New York, NY. Hart, S. (1987), ``The use of the survey in industrial market research'', Journal of Marketing Management, Vol. 3 No. 1, pp. 25-38. Hart, S. and Diamantopoulos, A. (1993), ``Marketing research activity and company performance: evidence from manufacturing industry'', European Journal of Marketing, Vol. 27 No. 5, pp. 54-72. Hart, S., Webb, J.R. and Jones, M.V. (1994), `Èxport marketing research and the effect of export experience in industrial SMEs'', International Marketing Review, Vol. 11 No. 6, pp. 4-22. Hattie, J. (1985), ``Methodology review: assessing unidimensionality of tests and items'', Applied Psychological Measurement, Vol. 9 No. 2, pp. 139-64. Jaworski, B.J. and Kohli, A.J. (1993), ``Market orientation: antecedents and consequences'', Journal of Marketing, Vol. 57, pp. 53-70. Johanson, J. and Vahlne, J. (1977), `Ìnternationalization process of the firm ± a model of knowledge development and increasing foreign market commitments'', Journal of International Business Studies, Vol. 8, pp. 23-32. Katsikeas, C.S. (1994), ``Perceived export problems and export involvement: the case of Greek export manufacturers'', Journal of Global Marketing, Vol. 7 No. 4, pp. 29-57. Keegan, W.J. (1974), ``Multinational scanning: a study of the information sources utilized by headquarters executives in multinational companies'', Administrative Science Quarterly, Vol. 9, pp. 411-21. Kim, J. and Meuller, C.W. (1978a), Introduction to Factor Analysis: What it is and How to Do it, Sage Publications, Beverly Hills, CA. Kim, J. and Mueller, C.W. (1978b), Factor Analysis: Statistical Methods and Practical Issues, Sage Publications, Beverly Hills, CA. Kleinschmidt, E.J. and Ross, R.E. (1984), `Èxport performance and foreign market information: relationships for small high-technology firms'', Canadian Journal of Small Business and Entrepreneurship, Vol. 2 No. 4, pp. 8-23. Koh, A.C. (1991), `Àn evaluation of international marketing research planning in United States export firms'', Journal of Global Marketing, Vol. 4 No. 3, pp. 7-25. Koh, A.C., Chow, J. and Smittivate, S. (1993), ``The practice of international marketing research by Thai exporters'', Journal of Global Marketing, Vol. 7 No. 2, pp. 7-26. Kohli, A.J., Jaworski, B.J. and Kumar, A. (1993), ``MARKOR: a measure of market orientation'', Journal of Marketing Research, Vol. XXX, pp. 467-77. Kotler, P. (1991), Marketing Management: Analysis, Planning and Control, Prentice-Hall, Englewood Cliffs, NJ. Leonidou, L.C. (1995), `Èxport barriers: non-exporters' perceptions'', International Marketing Review, Vol. 12 No. 1, pp. 4-25. Leonidou, L.C. and Katsikeas, C.S. (1997), `Èxport information sources: the role of organizations and international influences'', Journal of Strategic Marketing,Vol. 5 No. 2, pp. 65-88. Lesley, L.L. (1972), `Àre high response rates essential to valid surveys?'', Social Science Research, Vol. 1, pp. 323-34. Loudon, D.L. (1975), ``The influence of environmental variables on the use of marketing research'', Management International Review, Vol. 15 No. 2-3, pp. 95-111. McAuley, A. (1993), ``The perceived usefulness of export information sources'', European Journal of Marketing, pp. 52-64.

Menon, A. and Varadarajan, R. (1992), `À model of marketing knowledge use within firms'', Journal of Marketing, Vol. 56, pp. 53-71. Miles, M.B. and Huberman, A.M. (1994), An Expanded Sourcebook: Qualitative Data Analysis, Sage Publications, Beverly Hills, CA. Nunnally, J.C. (1978), Psychometric Theory (2nd ed.), McGraw-Hill Book Company, New York, NY. Nunnally, J.C. and Bernstein, I.H. (1994), Psychometric Theory (3rd ed.), McGraw-Hill Book Company, New York, NY. Peter, J.P. (1981), ``Construct validity: a review of basic issues and marketing practices'', Journal of Marketing Research, Vol. 18, pp. 133-45. Piercy, N. (1983), `À social psychology of marketing information ± learning to cope with the corporate battleground'', Journal of the Market Research Society, Vol. 25 No. 2, pp. 103-19. Pointon, T. (1978), ``Measuring the gains from government export promotion'', European Journal of Marketing, Vol. 12 No. 6, pp. 451-62. Reid, S.D. (1984), `Ìnformation acquisition and export entry decisions in small firms'', Journal of Business Research, Vol. 12, pp. 141-57. Rich, R.F. (1977), `Ùses of social science information by federal bureaucrats: knowledge for action versus knowledge for understanding'', in Weiss, C.H. (Ed.) Using Social Research in Public Policy Making, D.C. Heath and Company, Lexington Books, Lexington, MA. Samiee, S. and Walters, P.G.P. (1990), `Ìnfluence of firm size on export planning and performance'', Journal of Business Research, Vol. 2, pp. 235-48. Saunders, C. and Jones, J.W. (1990), ``Temporal sequences in information acquisition for decision making: a focus on source and medium'', Academy of Management Review, Vol. 15 No. 1, pp. 29-46. Schlegelmilch, B.B., Diamantopoulos, A. and Tse, K. (1993), ``Determinants of export marketing research usage: testing some hypotheses on UK exporters'', in Baker, M.J. (Ed.), Perspectives on Marketing Management III, John Wiley & Sons, Inc., New York, NY. Seringhaus, R.F.H. (1985), ``How do major industrial countries support firms' international efforts?'', ASAC Conference, UniversiteÂ du QueÂbec aÁ MontreÂal. Seringhaus, R.F.H. (1986), `Èmpirical investigation of awareness, use and impact of export marketing support by government in manufacturing firms'', Contemporary Research in Marketing, Proceedings of the 25th Annual Conference of the European Marketing Academy, Vol. 1. Seringhaus, R.F.H. (1986/1987), ``The role of information assistance in small firms' export involvement'', International Small Business Journal, Vol. 5 No. 2, pp. 26-36. Seringhaus, R.F.H. (1988), `Èxport knowledge, strategy and performance'', Developments in Marketing Science, Vol. 10, pp. 97-101. Seringhaus, R.F.H. and Mayer, C. (1988), ``Different approaches to foreign market entry between users and non-users of trade missions'', European Journal of Marketing, Vol. 22 No. 10, pp. 7-18. Seringhaus, R.F.H. and Rosson, P.J. (1990), Government Export Promotion: A Global Perspective, Routledge, London. Sharma, S. (1996), Applied Multivariate Techniques, John Wiley & Sons Inc., New York, NY. Sinkula, J.M. (1990), ``Perceived characteristics, organizational factors, and the utilization of external market research suppliers'', Journal of Business Research, August, pp. 1-17. Sinkula, J.M. (1994) ``Market information processing and organizational learning'', Journal of Marketing, Vol. 58, pp. 35-45.



Sinkula, J.M. and Hampton, R.D. (1988), ``Centralization and information acquisition by in-house market research departments'', Journal of Business Research, Vol. 16, June, pp. 337-49. Slater, S.F. (1995), `Ìssues in conducting marketing strategy research'', Journal of Strategic Marketing,Vol. 3, pp. 257-70. Sood, J.H. (1981), ``The international market information model of export behaviour: an empirical investigation based on the size of the exporting firm'', Proceedings of the European Academy of Advanced Research in Marketing, Copenhagen. Sood, J.H. and Adams, P. (1984), ``Model of management learning styles as a predictor of export behavior and performance'', Journal of Business Research, Vol. 12, pp. 169-82. Souchon, A.L. and Diamantopoulos, A. (1996), `À conceptual framework of export marketing information use: key issues and research propositions'', Journal of International Marketing, Vol. 4 No. 3, pp. 49-71. Souchon, A.L. and Diamantopoulos, A. (1997), `Ùse and non-use of export information: some preliminary insights into antecedents and impact on export performance'', Journal of Marketing Management, Vol. 13 No. 1-3, pp. 135-51. Spector, P.E. (1992), Summated Rating Scale Construction: An Introduction, Quantitative Applications in the Social Sciences, Series No. 07-082, Sage, Newbury Park, CA. Walters, P. (1983), `Èxport information sources ± a study of their usage and utility'', International Marketing Review, Winter, pp. 34-43. Weitzel, J.R. (1987), ``Strategic information management: targeting information for organizational performance'', Information Management Review, Vol. 3 No. 1, pp. 9-19. Wiseman, F. and Billington, M. (1984), ``Comment on a standard definition of response rates'', Journal of Marketing Research, Vol. 21, August, pp. 336-8. Young, S., Hamill, J., Wheeler, C. and Davies, J.R. (1989), International Market Entry and Development: Strategies and Management, Harvester Wheatsheaf, Prentice-Hall. Zaltman, G. and Moorman, C. (1988), ``The importance of personal trust in the use of research'', Journal of Advertising Research, Vol. 28 No. 3, pp. 16-24.