The importance of brand equity to customer loyalty (PDF Download ...

Industrial Marketing Management 36 (2007) 109 – 120

Understanding small scale industrial user internet purchase and information management intentions: A test of two attitude models Kevin Celuch a,*, Stephen Goodwin b,1, Steven A. Taylor b,2 a

b

Department of Management and Marketing, School of Business, University of Southern Indiana, 8600 University Boulevard, Evansville, IN 47712, United States Department of Marketing, College of Business, Illinois State University, Normal, IL 61790, United States Received 12 September 2003; received in revised form 18 January 2005; accepted 9 August 2005 Available online 21 September 2005

Abstract The present research compares two attitudinal models—variants of the Theory of Planned Behavior (TOPB) in terms of understanding the determinants of industrial buyer intentions to use the internet. The first variant of the Theory of Planned Behavior examines a decomposed perceived behavioral control construct that consists of self-efficacy and perceived control. The second variant of the TOPB model adds past behavior. Data from small scale users of industrial equipment are used to explore the models using structural equation modeling. While both variants of the TOPB are comparable in terms of standard fit indices, the variant with past behavior added substantially to the variance explained for internet purchase intentions. These results hold implications for future theory, research, and management of information technology-related small scale industrial buyer motivation. D 2005 Elsevier Inc. All rights reserved. Keywords: Industrial user; Internet information management intentions; Internet purchase intentions; Attitude models

1. Introduction The internet is rapidly changing today’s business landscape. The total worldwide value of goods and services of business-to-business (B2B) eCommerce is predicted to reach $4.3 trillion by 2005 (IDC statistics in EMarketer’s report, 2001). Globally, the US will remain the largest region for B2B eCommerce increasing at a compound annual growth rate of 68% between 2001 and 2005. Western Europe and Asia will increase at compound annual growth rates of 91% and 109%, respectively, over this same period (IDC statistics in EMarketer’s report, 2001). Of relevance given the context of the present research, is that while US B2B spending is projected at $2.7 trillion for 2004, approximately 85% of US companies fall into the small- to medium-sized markets with companies with 100 or less employees accounting for 53% of economic activity (U.S. Bureau of Labor Statistics in * Corresponding author. Tel.: +1 812 461 5297; fax: +1 812 465 1044. E-mail address: [email protected] (K. Celuch). 1 Tel.: +1 309 438 2893; fax: +1 309 438 5510. 2 Tel.: +1 309 438 8772; fax: +1 309 438 5510. 0019-8501/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.indmarman.2005.08.004

Bousquin, 2000). Clearly, the need for large manufacturers to effectively engage the small- to medium-sized business sector in B2B activity is becoming increasingly important. There is strong consensus that eCommerce has the potential to change business (cf., Kalakota & Robinson, 1999). For example, manufacturers, distributors, and customers using the internet as a basis for communications and transactions are creating new platforms for competitive strategy. Indeed, Kalakota and Robinson (1999) describe how Web technology has served as a catalyst for customer relationship management (CRM). CRM strategy includes process reengineering, organizational change, and incentiveprogram change with the creation of customer value as the focal point. The overarching objective is to achieve integration among sales, marketing, and service through coordinated activity. More specific goals of CRM include: growing revenue through existing relationships, enhancing service through the use of integrated information, and facilitating consistent, replicable channel processes and procedures (Kalakota & Robinson, 1999). Potential benefits of CRM notwithstanding, industrial marketing managers are left with significant questions,

110

K. Celuch et al. / Industrial Marketing Management 36 (2007) 109 – 120

particularly when one considers that industrial manufacturers have relied heavily on traditional brick-and-mortar channel models, with an emphasis on developing and nurturing intermediary relations. Many industrial manufacturers utilize intermediaries such as dealerships for product/service distribution particularly for interaction with smaller scale end users such as farmers and light construction companies (Mirani, Moore, & Webster, 2001). While industrial manufacturers have invested heavily in developing relationships with channel members, they often lack (or have very limited) direct interaction with and knowledge of their end users. Clearly, the questions related to how industrial manufacturers can leverage the internet for direct interaction with end users in general, and small scale end users in particular, hold important customer relationship management implications. As such, an understanding of factors that influence small scale industrial user internet usage is an important step in the development of customer-focused strategy. Perspectives for understanding information technology usage have included: macroeconomic approaches (cf., Panko, 1991), firm level approaches examining relationships between information technology expenditures and firm performance, (cf., Banker, Kauffman, & Mahmood, 1993) and approaches examining determinants of usage at the individual level (cf., Davis, 1989; Davis, Bagozzi, & Warshaw, 1989; Taylor & Todd, 1995). We deem the latter approach particularly relevant given the nature and scope of the present research as small scale firms representing such sectors as agriculture and light construction are likely to consist of an individual buyer rather than a purchasing group as found in larger organizations. This perspective is consistent with viewpoints recognizing small firm (entrepreneurial) business decision making as usually the province of an individual decision maker (Sheth, Mittal, & Newman, 1999). Thus, within individual level approaches, attitudebased models, which focus on the identification of the determinants of behavioral intention (i.e., attitudes, subjective norms, perceived control), have been viewed as a useful means of understanding determinants of technology usage. This work is grounded in social psychological theoretical perspectives such as the Theory of Reasoned Action (Ajzen & Fishbein, 1980) and the Theory of Planned Behavior (Ajzen, 1985, 1991). A clear strength of this approach is that it combines well-grounded theory with practical, problem-relevant interventions. Recently, for example, investigations of technology-based self-service have utilized attitude-based models to understand intention and use of technology (Bobbitt & Dabholkar, 2001; Dabholkar & Bagozzi, 2002). The present study extends this line of research in several ways. We echo Bobbitt and Dabholkar’s (2001) admonition related to the need for more theory-based technology-related research. First, two attitudinal models are compared—two variants of The Theory of Planned Behavior (TOPB). Current literature in the area raises theory development issues with respect to variables that might further our understanding of the determinants of behavioral intentions and ultimately behavior.

One issue relates to the decomposition of the perceived behavioral control construct into self-efficacy and perceived control. The second issue relates to the addition of past behavior (Conner & Armitage, 1998). The present research explores both of these theoretical extensions to the Theory of Planned Behavior. The authors recognize that an alternative framework –the Technology Acceptance Model (TAM) (cf., Taylor & Todd, 1995)– has received support in the literature. In the TAM, intention is explained by attitude toward usage as well as by direct and indirect effects of perceived ease of use and perceived usefulness. Note that the TOPB includes attitude, subjective norm, and perceived behavioral control components as determinants of intention. Within TOPB, the selfefficacy facet of perceived behavioral control captures the perceived ease of use construct in TAM. In addition, attitude items can reflect specific aspects of the perceived usefulness construct. Further, perceived control and subjective norm represented in the TOPB are not represented in TAM. We believe these constructs could add value to an understanding of internet usage as control and normative influence represent potentially relevant issues in the small scale industrial user context. This line of reasoning is consistent with concerns raised in the attitude literature regarding the insufficient consideration of the social context (Eagly & Chaiken, 1993). As acknowledged by Taylor and Todd (1995) in a comparison of the TAM and the TOPB in the context of utilization of a university computer resource center, the TOPB showed an improvement over TAM in explaining behavioral intention and concluded that the TOPB provided a more complete understanding of intention than did TAM. Thus, a conscious decision was made by the present authors to examine the more established theories owing to their potential to offer a richer theoretical understanding of psychological processes. We further extend research in the area by examining the above models for an important real world context that has received scant attention—small scale industrial end users. Prior related research has relied heavily on consumer and academic settings and has often employed student samples (cf., Eagly & Chaiken, 1993; Dabholkar & Bagozzi, 2002; Taylor & Todd, 1995). Beyond the theoretical dimension, recall that attitudebased approaches offer problem-relevant interventions. Thus, this study holds the potential of helping industrial manufacturers of equipment understand how to influence small scale end user internet usage. In addition, we assess models using structural equation modeling. As noted by Conner and Armitage (1998), given that models such as the Theory of Planned Behavior are based on assumptions regarding causal processes, there is a need to examine causal relationships among attitude model constructs. In the next section of the paper, the theoretical models which underlie this research are reviewed and we describe the specific models to be tested. Following, we provide an overview of the methodology of the study and then present the findings. The last section of the paper discusses results and addresses theoretical, research, and managerial implications.


1.1. Theoretical models 1.1.1. The theory of planned behavior The Theory of Planned Behavior was developed in response to a related existing model—The Theory of Reasoned Action (TORA) (Ajzen, 1988, 1991). Briefly, the Theory of Reasoned Action (Fishbein & Ajzen, 1975; Ajzen & Fishbein, 1980) places intention as the principal predictor of behavior. So conceived, the more one intends to engage in behavior, the more likely is the occurrence of the behavior. Determining intention are attitude and subjective norm. The attitudinal determinant of intention is defined as the overall evaluation of behavior. This overall evaluation, in turn, is composed of the salient beliefs: the perceived likelihood of particular consequences of the behavior occurring, weighted by an evaluation of the consequences. The subjective norm determinant of attitude is conceptualized as social pressure from significant others to perform or not perform the behavior. The subjective norm, in turn, is composed of normative beliefs: the perceived pressure from salient referents, weighted by the motivation to comply with the referents. The TORA has received support across a range of contexts (cf., Eagly & Chaiken, 1993; Sheppard, Hartwick, & Warshaw, 1988). A recognized limitation of the TORA is that it was developed to deal with behaviors that are completely under an individual’s volitional control (Ajzen, 1988; Fishbein, 1993). In response to the aforementioned limitation of the TORA, Ajzen (1988, 1991) proposed an extension—the Theory of Planned Behavior. The TOPB was designed to address behaviors not completely under volitional control. This is relevant to the current research as industrial buyer internet usage is not likely to be totally under volitional control particularly if small scale users have questions related to competency and control issues tied to information technology utilization. The model is identical to the TORA except that a perceived behavioral control construct has been added. Perceived behavioral control relates to the ease or difficulty of performing a behavior. Perceived behavioral control is affected by perceptions of access to necessary skills, resources and opportunities to perform a behavior, weighted by the perceived valence of each factor to facilitate or inhibit the behavior. In the TOPB, perceived behavioral control is viewed as determining intention as well as behavior directly. The TOPB has received strong support in predicting a wide range of intentions and behaviors (cf., Godin & Kok, 1996; Sutton, 1998). Recently, calls have been renewed for the exploration of additional constructs which might add value to the TOPB (Conner & Armitage, 1998). In fact, as noted by Conner and Armitage (1998), Ajzen (1991) conceived of the model as open to further extensions given the identification of important proximal variables. Conner and Armitage (1998) highlight the perceived behavioral control construct as ripe for reconsideration. Ajzen (1991) argued that perceived behavioral control and selfefficacy (perceived capabilities) (Bandura, 1982) are synonymous. As a result, some researchers utilizing the TOPB have used self-efficacy instead of perceived behavioral control.

111

However, Ajzen’s original conception reflects both control and skill/ability facets. In more recent work, Ajzen appears to orient perceived behavioral control to the easy/difficult dimension (Fishbein, 1993). Further, Bandura (1992) makes a theoretical distinction between control and self-efficacy. An individual may believe that task performance is under their control, however, they may question whether or not they posses the skills required for the given task. Indeed, empirical evidence also points to the notion of distinguishing between self-efficacy and perceived control (cf., Manstead & van Eekelen, 1998; Terry & O’Leary, 1995). Clearly, there is a need to further examine the self-efficacy –perceived control theoretical distinction, particularly in industrial marketing contexts. Conner and Armitage (1998) also note the argument that many behaviors are determined by past behavior. Indeed, the notion that habit or learned predispositions might influence intention and behavior in attitude-based models has been argued for years (cf., Eagly & Chaiken, 1993). However, issues have been raised about the addition of past behavior to a TOPB model which includes self-efficacy. Given self-efficacy relates to perceptions of an individual’s ability to perform a behavior in the future, questions exist regarding the added value of the inclusion of past behavior. Conner and Armitage (1998) present evidence from a review of studies that supports the contribution of past behavior to predictions of intentions and behavior over and above the contributions of TOPB variables. Thus, it would appear that the inclusion of past behavior might be usefully examined in tests of TOPB models. 1.2. Theoretical models of internet usage for industrial end users of heavy equipment Based on the preceding discussion, we offer two models adapted from the variants of the Theory of Planned Behavior. As noted previously we offer models to explain small scale industrial end user internet usage for the manufacturer contact point. Further, we develop models for two behavioral intention domains with respect to internet usage—purchase-related and information management-related. Our logic relates to the fact that internet usage encompasses a range of behaviors. We are focusing on information management versus purchase behavior as these are discrete behavioral domains requiring different levels of user commitment. An individual using the internet for information management purposes, that is, searching for, asking for, and receiving information is at a lower level in the ‘‘internet’’ hierarchy of effects than an individual using the internet to purchase equipment, parts, and/or service. Such differences may result in differential impact of model variables. Lastly, consistent with the TOPB (Ajzen, 1991), direct (nonmediated) effects from antecedents to intention are examined in this initial exploration of a small scale industrial user context. Models 1a and 1b are based on a variant of the TOPB (see Fig. 1a and b). In Model 1a, intention to use the internet for industrial equipment purchase is determined by the overall attitude toward internet purchasing and the subjective norm for using the internet. In addition, based on the rationale provided

112


a Attitude toward Internet Purchasing

Subjective Norm for Using the Internet Intention to Use the Internet for Purchase Internet Self-Efficacy

Perceived Control Related to Internet Usage

b Attitude toward Internet Information Management

Subjective Norm for Using the Internet Intention to Use the Internet for Information Management Internet Self-Efficacy


Fig. 1. Theory of planned behavior variant 1.

for exploring the influence of a decomposed perceived behavioral control construct, intention is also modeled as a function of internet self-efficacy and perceived control related

to internet usage constructs. In Model 1b, intention to use the internet for information management is determined by the overall attitude toward using the internet for information


a Attitude toward Internet Purchasing

Subjective Norm for Using the Internet

Intention to Use the Internet for Purchase

Internet Self-Efficacy


Past Behavior Related to Internet Purchase

b Attitude toward Internet Information Management

Subjective Norm for Using the Internet

Internet Self-Efficacy

Intention to Use the Internet for Information Management


Past Behavior Related to Internet Information Management Fig. 2. Theory of planned behavior variant 2.

113

114


management, the subjective norm for using the internet, internet self-efficacy, and perceived control related to internet usage. Thus, it is proposed that:

2. Method

H1a. Attitude toward internet purchasing will be positively related to intention to use the internet for purchase.

The sampling frame for the study derived from a purchased list provided by an industrial equipment manufacturer sponsoring the research. The list was compiled from the registrations of individuals who had purchased industrial moving equipment within the last 12 months. Purchase domains consisted of industry settings representing agriculture (crop or animal production) and light construction which typically utilize such equipment as small tractors and back hoes. Note that such equipment can come from a number of manufacturers both domestic and foreign. A total of 2380 mailed surveys were distributed to respondents at their business address with a $1 incentive and postage-paid return envelopes. The industrial manufacturer sponsoring the research was not identified in any way to respondents. Three hundred and three surveys were returned representing a response rate of 12.7%. Questionnaires were received from respondents representing virtually all of the 50 states with heavier representation from agricultural and industrial Midwestern states. The ‘‘typical’’ respondent was a male, 40 – 49 years old, with secondary education as their highest completed level of education, with more than 10 years industry experience, who own or work in firms of less than 50 employees. This profile is consistent with the population that buys the type of equipment (i.e., agricultural, light construction) that formed the basis of our sampling frame. We view this response rate as comparable to response rates typically found in general industrial sector research. In addition, discussion with managers at the firm whose grant enabled the research suggests that such response rates are typical for the specific industries examined. Further, the potential for nonresponse bias was assessed by comparing early and late respondents on several descriptive variables. No differences were found between groups for age, level of education, years associated with current employer, and years of industry experience. Thus, the lack of observed differences between early and late respondents suggests that differences between respondents and nonrespondents were not significant with these data.

H1b. Attitude toward internet information management will be positively related to intention to use the internet for information management. H2a. Subjective norm for using the internet will be positively related to intention to use the internet for purchase. H2b. Subjective norm for using the internet will be positively related to intention to use the internet for information management. H3a. Internet self-efficacy will be positively related to intention to use the internet for purchase. H3b. Internet self-efficacy will be positively related to intention to use the internet for information management. H4a. Perceived control related to using the internet will be positively related to intention to use the internet for purchase. H4b. Perceived control related to using the internet will be positively related to intention to use the internet for information management. Models 2a and 2b are also based on a variant of the TOPB (see Fig. 2a and b). In Model 2a, as in Model 1a, intention to use the internet for industrial purchase is determined by the overall attitude toward internet purchasing, the subjective norm for using the internet, internet self-efficacy, and perceived control related to internet usage. In addition, the value of the past behavior construct is examined by explicitly adding past behavior related to internet purchase to the model. Similarly, in Model 2b, intention to use the internet for information management is determined by the overall attitude toward using the internet for information management, the subjective norm for using the internet, internet self-efficacy, perceived control related to internet usage, and past behavior related to managing information. Thus, based on the variant of the TOPB model presented earlier, the two immediately preceding models add a past behavior construct to the explanation of behavioral intention. Questions relating to the strength of prior behavior (i.e., the strength of representation in respondents behavioral repertoire), particularly for the industrial context examined in the present research, might influence whether this variable proves a viable addition to the models tested. Thus, beyond testing the posited relationships specified for Models 1a and 1b we add hypotheses related to past behavior. Therefore: H5a. Past behavior related to internet purchase will be positively related to intention to use the internet for purchase. H5b. Past behavior related to internet purchase will be positively related to intention to use the internet for information management.

2.1. Sample and procedure

2.2. Measures Measures employed in the questionnaire consisted of scales developed specifically for TOPB model constructs adapted for the industrial context studied. As noted in the attitude literature, compatibility of measures with the context studied enhances model predictive power (cf., Eagly & Chaiken, 1993). Scale items were developed based on a literature review, extensive focus-group research with members of the target population, coupled with lengthy discussions with organizational managers. Early drafts of the questionnaire were reviewed by industry representatives and pretested for readability, understandability, and comprehensiveness. The final questionnaire included measures of industrial end user perceptions, oriented


115

manufacturers encouraging internet use for communication and purchases and the competitive pressure to use the internet for communication with and purchasing from manufacturers. Again, consistent with the TOPB, all matched likelihood and motivation to comply items were multiplied together and summed to form overall subjective norm.

toward manufacturer contact, related to intention to use the internet for industrial purchase, intention to use the internet for information management, attitude toward internet purchasing, attitude toward using the internet for information management, the subjective norm for using the internet, internet self-efficacy, perceived control related to internet usage, past behavior related to internet purchase, past behavior related to managing information, and demographic descriptors. In this study we employed multi-item scales for both the exogenous variables (attitudes, subjective norms, self efficacy, perceived control, and past behavior) and for the endogenous variables (intentions). Please refer to Appendix A for the measures used in the current research.

2.2.4. Internet self-efficacy, perceived control, and past behavior measures An important objective of this research was to explore a decomposed perceived behavioral control construct in an industrial context. As such, self-efficacy and perceived control measures consistent with the work of Terry and O’Leary (1995) and Bandura (1992) were created. Internet self-efficacy consisted of 3 six-point items relating to perceptions of difficulty in using and confidence in ability using the internet for manufacturer communication and transactions. Perceived control related to internet usage was assessed via 3 six-point items relating to perceptions tied to the degree of personal choice in using the internet for communication and transactions with manufacturers. Past behavior related to internet purchase consisted of four items on which respondents indicated whether they had used the internet for equipment and parts purchases during the past 12 months. Past behavior related to managing information consisted of six items on which respondents indicated whether they had used the internet for searching for and asking for information, and receiving and responding to requests for information from manufacturers during the past 12 months. Table 1 provides the correlations among the constructs used in the study.

2.2.1. Intention measures Intention to use the internet for industrial equipment purchase consisted of two six-point items, one relating to purchasing equipment and the other relating to the purchase of parts over the next 12 months. Intention to use the internet for information management was assessed via 3 six-point items relating to reviewing, asking for, and responding to requests for information from manufacturers over the next 12 months. 2.2.2. Attitude measures Attitude toward internet purchasing consisted of 10 sixpoint items, with respondents providing perceptions of likelihood and importance relating to internet usage enhancing respondent ability to make better decisions regarding purchasing equipment, parts, services, and increasing security concerns. Attitude toward using the internet for information management was assessed via 8 six-point items with respondents providing perceptions of likelihood and importance relating to internet usage enhancing respondent ability to compare offerings, access information, request information, respond to requests, as well as increasing costs of transactions, decreasing the value of personal relationships, and increasing the volume of unsolicited messages. Consistent with the TOPB, all matched likelihood and importance items were multiplied together and summed to form overall attitude.

3. Results The primary objective of the present research was to test two related attitude-based models that could be used to explain behavioral intentions related to industrial internet usage. Structural equation modeling was employed for model evaluation using SPSS 10.0 and LISREL 8.51. Structural equation modeling of multiple endogenous variables requires assessment of measurement model fit. Consistent with the suggestion of Hair, Anderson, Tatham, and Black (1998), we investigated whether all variables used for analyses were significantly related to their specified constructs, which we found to be true; and reliability estimates and variance extracted measures (where appropriate) were calculated for each construct. Table 2 presents these results and demonstrates that our construct measures generally exceeded the accepted reliability standard of a > 0.7 and certainly compare favorably with results reported in related research. The one exception to these findings related to the reliability coefficient for the past behavior measure. Given this measure

2.2.3. Subjective norm measures The subjective norm for using the internet consisted of 4 six-point items, with respondents providing perceptions of likelihood and motivation to comply relating to important Table 1 Correlations among the study constructs

Purchase intentions Info. man. intentions Att. toward purchase Att. toward info. man. Subjective norm Self-efficacy Perceived control Past behavior-purchase Past behavior-info. man.

1

2

3

4

– 0.578 0.381 0.367 0.430 0.486 0.036 0.383 0.363

– 0.416 0.450 0.439 0.531 0.036 0.244 0.389

– 0.775 0.528 0.313 0.082 0.177 0.369

– 0.568 0.398 0.105 0.171 0.347

5

6

7

8

9

– 0.359

–

– 0.413 0. 070 0.296 0.240

– 0.212 0.313 0.359

– 0.018 0.023

116


Table 2 Reliability analyses for model variables

4. Discussion

Variable

Coefficient Construct Construct SEM a SEM variance reliability extracted

Attitude toward info. management Attitude toward purchasing Subjective norm Self-efficacy Perceived control Previous behavior info. management Previous behavior purchase Purchase intention Information management intention

0.8155 0.9235 0.8625 0.7552 0.7477 0.7259 0.6110 0.8700 0.8994

The intent of the present research was to compare two attitudinal models—variants of The Theory of Planned Behavior in terms of understanding the determinants of small scale industrial buyer intentions to use the internet. The first variant of the TOPB examined a decomposed perceived behavioral control construct that consisted of self-efficacy and perceived control. The second variant of the TOPB model added past behavior. Data from industrial end users were used to explore the models using structural equation modeling. While both variants of the TOPB model performed well and were comparable in terms of standard fit indices, the variant with past behavior, added substantially to the variance explained for purchase intentions. Implications of these findings are discussed in detail below. From the standpoint of theory testing, this study sheds additional light on the applicability of attitude-based models and addresses questions regarding possible conceptual extensions. First, based on the explained variances observed in the present research, the TOPB appears to be a viable conceptual framework from which to understand buyer technology-related motivation in business-to-business settings. The proportion of variances accounted for in the current study compare favorably with variances explained in reviews of research employing attitude-based models (cf., Godin & Kok, 1996; Sutton, 1998). This research also provides evidence relevant to the argument for distinguishing between self-efficacy and perceived control constructs. Recall that behavior control is at the heart of the TOPB in that it was formulated to explain behavior not totally under an individual’s volitional control. This study found internet self-efficacy to be the strongest predictor of purchase as well as information management intentions for manufacturer contact. In contrast, perceived control was not found to have impact. Thus, findings with respect to the strong influence of self-efficacy on intentions in the present industrial context mirror results reported in consumer, self-help, and academic settings (Conner & Armitage, 1998). With respect to the addition of a past behavior construct to the TOPB model, in the present study, past behavior did substantially enhance the model related to internet purchase intentions. It would appear that, for the context examined in the

0.91 0.92

0.83 0.72

utilized a ‘‘check all that apply’’ format a lower reliability estimate is not surprising. We next assessed the validity of our measures. Given that measures were derived from previous qualitative and quantitative research followed by consultation with industry personnel, we are confident that face and content validity have been established. Furthermore, the calculated variance-extracted scores for the endogenous variables in Table 2 exceed the 50% recommended criteria for model constructs (Raines-Eudy, 2000). Thus, we find evidence in support of the validity of the measures used in the research. As noted previously, structural equation modeling was employed for model evaluation. We tested the models using item parcels with items summed (in the case of attitudes and subjective norms, related items were multiplied and summed) to form unique construct parcels. Bandalos and Finney (2001) note that the use of item parcels has become a common practice in structural equation modeling in recent years. Little, Cunningham, Shahar, and Widaman (2002) most recently examine the issue of parceling and reach the same general conclusions as Bandalos and Finney (2001). In particular, they conclude that the use of parceling can be warranted in social science research. We thus combined measure items into unique global exogenous parcels, and considered their contributions to the dependent variables. The results of testing the competing models using structural equation analysis can be found in Table 3. There were no warning messages from the LISREL software or negative error covariances noted. These results suggest that both models fit the data well and are relatively comparable in terms of RMSEA, CFI, and SRMR indices. However, the TOPB variant that includes past behavior adds substantially to the variance explained for purchase intentions (an 8% increase) while the increase in variance explained is perhaps negligible for information management intentions (a 2% increase). Of interest is the finding that internet self-efficacy is consistently the strongest predictor of usage intentions. In the case of purchase intentions, past behavior is also a strong predictor of intentions. Attitude toward usage is also a relatively strong predictor of intentions. Further, in all instances, perceived control is not found to be a significant predictor of internet usage intentions. Table 3 Results of model assessment using structural equation modeling (N = 256) Dependent variable

Equationa,b

R2

RMSEA

CFI

SRMR

Purchase intentions

Eq. (1): 0.21*Att. toward Purchase + 0.18*Sub. Norm + 0.42*Self-Eff. NS*Per. Control Eq. (2): 0.20*Att. toward Purchase + 0.13*Sub. Norm + 0.30*Self-Eff. NS*Per. Control + 0.30*Previous Behavior Eq. (1): 0.29*Att. toward Info. Management + 0.11*Sub. Norm + 0.38*Self-Eff. NS*Per. Control Eq. (2)c: 0.24*Att. toward Info. Management + 0.12*Sub. Norm + 0.32*Self-Eff. + NS*Per. Control + 0.18*Past Behavior

0.40

0.000

1.00

0.0063

0.48

0.027

1.00

0.011

0.40

0.000

1.00

0.0093

0.42

0.000

1.00

0.009

Information management intentions

a b c

Please note that all equation coefficients are fully standardized. Please note that Eqs. (1) and (2) are model variants of the Theory of Planned Behavior. We dropped the intention item relating to responding to requests for information from this analysis.


present research, the inclusion of past behavior in models examining internet purchase behavior would be worthwhile. A potential interpretation relates to the fact that using the internet for purchase rather than for information acquisition is further along on the ‘‘hierarchy of commitment’’ and thus is better explained with the additional variable of past behavior. Findings of this research related to efficacy and past behavior are also consistent with findings of related research examining factors that contribute to website visitation. Ilfeld and Winer (2002) conclude that internet advertising should focus primarily on inducing website visitation rather than on impression building. They note that website loyalty and equity appear to develop primarily from experience and usage. These findings are consistent with the findings of the present research which point to the strength of efficacy and past behavior in determining internet purchase intentions. While Ilfeld and Winer (2002) employ persuasive hierarchy of effects models to examine the affect of advertising on website awareness, traffic, and brand equity, the present research employs attitude models to explore internet information management and purchase intentions within the broader tableau of CRM. A fusion of these two approaches could prove interesting by adding depth to future research in the area. Future research might want to build on the models examined in the current study by exploring additional industries and measures/variables. As noted previously, much of the research examining attitude-based models has examined consumer and academic contexts and/or relied heavily on student samples. The present research provides support for the viability of using attitude-based models to understand the small scale industrial end user. With respect to measurement issues, given the centrality of the notion of control in the TOPB, an examination of alternative measures of perceived control may prove fruitful in future research. Skinner (1995) delineates a more fine grained approach to assessing control beliefs. In much the same way that belief-based measures of attitude have been found to contribute to intention in ways that differ from simpler attitude measures (Nucifora, Gallois, & Kashima, 1993), this approach may allow for a more detailed assessment and may better capture the construct and enhance explanatory power. Further, the addition of actual behavior would allow for the examination of direct as well as indirect effects (working though intention) of perceived control and efficacy. Moving beyond a subjective norm to the inclusion of a behavioral norm might also prove useful. A behavioral norm addresses what significant others are perceived to do. Alternatively, a subjective norm addresses what a significant other is perceived to think an actor should do. Adding such a distinction could broaden the concept of normative influence in TOPB models for industrial contexts. Future research might also examine potential mediated relationships. Although not suggested by the TOPB, alternative models (e.g., TAM) suggest the potential for mediated relationships. In addition, potential moderators can also be explored. The present data, given the nature of the sample consisting of small scale industrial users, provided minimal variability on

117

demographic variables which could moderate relationships in other contexts. 4.1. Managerial implications How can industrial manufacturers leverage the internet for direct interaction with end users in general, and small scale end users in particular, as part of a customer relationship management (CRM) initiative? As noted earlier, an understanding of factors that influence small scale industrial user internet usage is an important step in the development of customer-focused strategy. From a practical standpoint, the TOPB provides an industrial manufacturer focal points from which to affect end user behavior. Based on the strong impact of internet self-efficacy in models, efforts aimed at enhancing end user self-efficacy would help strengthen their internet use intentions. Social cognitive theory (Bandura, 1986, 1997) accords beliefs of personal efficacy a central role in the regulation of motivation and performance. The work of Bandura et al. has demonstrated that in addition to knowledge and skills, competent behavior in any situation requires specific selfbeliefs of efficacy (i.e., judgments of what one can do with the knowledge/skills one possesses to meet situational demands). As noted by Wood and Bandura (1989) there is a difference between: (a) possessing skills and (b) being able to use them consistently well under various circumstances. Task-specific self-beliefs of efficacy help account for why an individual possessing the requisite skills performs suboptimally in a given task context and why two individuals at the same skill level perform differently in the same task situation. Within this view, perceived self-efficacy is a dynamic and generative cognition. That is, efficacy judgments may change over time given new experiences and include a mobilization component involving the orchestration of behavior to fit changing contexts (Gist & Mitchell, 1992). Bandura (1997) details sources of experiences which can contribute to an individual’s efficacy perceptions. Two of the most powerful sources of efficacy are direct and vicarious experience. Thus, the provision of direct experience through manufacturer sponsored training programs initially related to information management would likely enhance efficacy perceptions related to internet usage. Such programs aimed at information management intentions could be one component aimed at achieving the CRM goal of enhancing service through integrated information. In addition, efficacy enhancement interventions focused on directly observing others engaging in positive industrial-related experience with internet usage (both information management and purchase) could also prove beneficial in achieving the CRM goal of revenue growth through existing relationships. Recall that attitude toward usage was also a relatively strong predictor of intentions. Thus, the use of persuasive communication interventions (i.e., direct mail, business advertising) coupled with incentive programs by manufacturers which aim at enhancing the perception of potential benefits and weakening perceptions of potential costs could

118


prove useful in strengthening internet usage intentions. Potential benefits identified in the present research which could be addressed by the interventions would relate to improved information access, improved ability to compare and select offerings, improved ability to request, receive, and respond to information, and improved decision making with respect to purchase. Potential costs identified in this research which could be addressed by the interventions include concerns related to security, costs, unsolicited messages, and personal relationships. In this way, the use of persuasive communication interventions along with incentive programs that positively impact internet intentions could also play a role in CRM efforts aimed at enhancing service through information as well as growing revenue. Note that the use of persuasive communication interventions exclusively, while enhancing attitudes toward internet usage among industrial end users, is not likely to contribute greatly to efficacy enhancement (Bandura, 1986). Subjective norms and perceived control related to internet usage were found to exhibit more limited effects on intentions. Nevertheless, as part of a CRM program, manufacturer communication addressing company imperatives and competitive pressure appear to be relevant for encouraging internet usage. In conclusion, understanding information technology-related motivation in industrial settings will continue to be a significant topic for marketing researchers and practitioners. It is hoped that this theory-based approach related to information management and purchase intentions will contribute to future empirical efforts aimed at increasing our understanding of internet usage by small scale industrial buyers. Appendix A A.1. Measures Intention to use the internet for purchase: To what extent do you intend to use the Internet to purchase equipment? To what extent do you intend to use the Internet to purchase parts? (Definitely will not use 1 2 3 4 5 6 Definitely will use) Intention to use the internet for information management: To what extent do you intend to use the Internet to review information sent from equipment Manufacturers? To what extent do you intend to use the Internet to ask for information from equipment Manufacturers? To what extent do you intend to use the Internet to respond to requests for information from equipment Manufacturers? (Definitely will not use 1 2 3 4 5 6 Definitely will use) Attitude toward internet purchasing: Using the internet will significantly improve your ability to make better decisions regarding . . . purchasing new equipment. . . . purchasing used equipment.

. . . purchasing parts that would enable customization of your own equipment. . . . purchasing replacement parts. . . . renting machines or parts. . . . purchasing services related to: repairs. . . . purchasing services related to: maintenance. . . . purchasing services related to: diagnostic information. . . .purchasing services related to: equipment financing (including the possibility of leasing). Using the internet will significantly increase your concern about security (e.g., making on-line orders using a credit card). (Likelihood: Very unlikely 1 2 3 4 5 6 Very likely; Importance: Of little or no importance 1 2 3 4 5 6 Of utmost importance) Attitude toward using the internet for information management: Using the internet will significantly improve your ability to compare different manufacturers’ equipment offerings in terms of specifications. Using the internet will significantly improve your ability to compare different manufacturers’ equipment offerings in terms of prices. Using the internet will significantly improve your ability to request and receive specific information directly from Equipment Manufacturers. Using the internet will significantly improve your ability to make better decisions regarding selecting which Manufacturer’s brand to purchase. Using the internet will significantly improve your ability to respond to requests from Equipment Manufacturers. Using the internet will significantly increase costs (e.g., monetary, time) associated with Manufacturer transactions due to Internet training requirements of personnel in your company. Using the internet will significantly decrease the value of personal relationships with Manufacturers. Using the internet will significantly increase the volume of unsolicited messages from Manufacturers. (Likelihood: Very unlikely 1 2 3 4 5 6 Very likely; Importance: Of little or no importance 1 2 3 4 5 6 Of utmost importance) Subjective norm for using the internet: Over the course of the next 12 months: Equipment Manufacturers who are especially important to you and your company will encourage you to regularly use the Internet for business communications. Equipment Manufacturers who are especially important to you and your company will encourage you to regularly use the Internet for making on-line purchases. You will be under severe competitive pressure to use the Internet for business communications with Manufacturers. You will be under severe competitive pressure to use the Internet for making on-line purchases from Manufacturers. (Likelihood: Very unlikely 1 2 3 4 5 6 Very likely; Motivation to comply: Do not want to comply 1 2 3 4 5 6 Want to comply very much)Internet self-efficacy:


I would not have any difficulty using the Internet for making transactions (e.g., on-line purchase orders) with one or more Equipment Manufacturers. I am very confident in my ability to use the Internet to search for equipment and/or parts. (Strongly disagree 1 2 3 4 5 6 Strongly agree) For me to use the Internet for communicating with one or more Equipment Manufacturers would be: (Very difficult 1 2 3 4 5 6 Very easy) Perceived control related to internet usage: I feel it’s completely my choice as to whether or not I use the Internet for Manufacturer communication. I feel it’s completely my choice as to whether or not I use the Internet for Manufacturer transactions (e.g., making on-line purchases). (Strongly disagree 1 2 3 4 5 6 Strongly agree) How much choice do you feel you will have over whether you use the Internet for any kind of contact with Manufacturers? (No personal choice 1 2 3 4 5 6 Complete personal choice) Past behavior related to internet purchase (Please check only those that apply). I have used the Internet during the past 12 months . . .to place an on-line purchase order for new Equipment. . . .to place an on-line purchase order for used Equipment. . . .to place an on-line purchase order for new Parts. . . .to place an on-line purchase order for used Parts. Past behavior related to information management: (Please check only those that apply). I have used the Internet during the past 12 months . . .to search for information about new Equipment. . . .to search for information about used Equipment. . . .to search for information about new Parts. . . .to search for information about used Parts. . . .to ask for specific information from at least one Equipment Manufacturer. . . .to receive and then respond to requests for information from at least one Equipment Manufacturer. References Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl, & J. Beckman (Eds.), Action control: From cognition to behavior. Berlin’ Springer-Verlag. Ajzen, I. (1988). Attitudes, personality, and behavior. Chicago’ Dorsey Press. Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179 – 211. Ajzen, I., & Fishbein (1980). Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ’ Prentice Hall. Bandalos, D. L., & Finney, S. J. (2001). Item parceling issues in structural equation modeling. In G. A. Marcoulides, & R. E. Schumacker (Eds.), New developments and techniques in structural equation modeling. London’ Lawrence Erlbaum Associates. Bandura, A. (1982). Self-efficacy mechanisms in human agency. American Psychologist, 37, 122 – 147.

119

Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ’ Prentice Hall. Bandura, A. (1992). On rectifying the comparative autonomy of perceived control: Comments on Fcognates of personal control_. Applied and Preventive Psychology, 1, 121 – 126. Bandura, A. (1997). Self-efficacy: The exercise of control. New York’ W.H. Freeman and Company. Banker, R. D., Kauffman, R. J., & Mahmood, M. A. (1993). Strategic information technology management: Perspectives on organizational growth and competitive advantage. Harrisburg, PA’ Idea Group Publishing. Bobbitt, L. M., & Dabholkar, P. A. (2001). Integrating attitudinal theories to understand and predict use of technology-based self-service: The internet as an illustration. International Journal of Service Industry Management, 12(5), 423 – 450. Bousquin, J. (2000). Macro to Micro: B2B Shifts Focus, http://thestreet.com. Posted Aug. 8, 2000. Conner, M., & Armitage, C. J. (1998). Extending the theory of planned behavior: A review and avenues for further research. Journal of Applied Social Psychology, 28, 1429 – 1464. Dabholkar, P. A., & Bagozzi, R. P. (2002). An attitudinal model of technology-based self-service: Moderating effects of consumer traits and situational factors. Journal of the Academy of Marketing Science, 30(3), 184 – 201. Davis, F. D. (1989, Sept.). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13, 319 – 339. Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1989). User acceptance of computer technology: A comparison of two theoretical models. Management Science, 35(8), 982 – 1003. Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Fort Worth, TX’ Harcourt Brace and Company. eMarketer’s report. (2001). E-Commerce Trade and B2B Exchanges, http://www.clikz.com, Posted fourth quarter. Fishbein, M. (1993). Introduction. In D. J. Terry, C. Gallois, & M. McCamish (Eds.), The theory of reasoned action: Its application to AIDS-preventive behavior. Oxford, UK’ Pergamon. Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Reading, MA’ Addison-Wesley. Gist, M., & Mitchell, T. (1992). Self-efficacy: A theoretical analysis of its determinants and malleability. Academy of Management Review, 17(2), 183 – 211. Godin, G., & Kok, G. (1996). The theory of planned behavior: A review of its applications to health-related behaviors. American Journal of Health Promotion, 11, 87 – 98. Hair, J. F. Jr., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate data analysis (5th edition). Upper Saddle River, NJ’ Prentice Hall. Ilfeld, J. S., & Winer, R. S. (2002, September/October). Generating website traffic. Journal of Advertising Research, 49 – 61. Kalakota, R., & Robinson, M. (1999). E-business: Roadmap for success. Reading, MA’ Addison-Wesley. Little, T. D., Cunningham, W. A., Shakar, G., & Widaman, K. F. (2002). To parcel or not to parcel: Exploring the question, weighing the merits. Structural Equation Modeling, 9(2), 151 – 173. Manstead, A. S. R., & van Eekelen, S. A. M. (1998). Distinguishing between perceived behavioral control and self-efficacy in the domain of academic achievement intentions and behaviors. Journal of Applied Social Psychology, 28, 1375 – 1392. Mirani, R., Moore, D., & Webster, J. A. (2001). Emerging technologies for enhancing supplier-reseller partnerships. Industrial Marketing Management, 30, 101 – 114. Nucifora, J., Gallois, C., & Kashima, Y. (1993). Influences on condom use among undergraduates: Testing the theories of reasoned action and planned behavior. In D. J. Terry, C. Gallois, M. McCamish (Eds.), The theory of reasoned action: Its application to AIDS-preventive behavior. Oxford, UK’ Pergamon. Panko, R. R. (1991). Is office productivity stagnant. MIS Quarterly, 15(2), 191 – 204.

120


Raines-Eudy, R. (2000). Teacher’s corner: Using structural equation modeling to test for differential reliability and validity: An empirical demonstration. Structural Equation Modeling, 7(1), 124 – 141. Sheppard, B., Hartwick, J., & Warshaw, P. (1988). The theory of reasoned action: A meta analysis of past research with recommendations for modifications and future research. Journal of Consumer Research, 15(4), 325 – 343. Sheth, J. N., Mittal, B., & Newman, B. I. (1999). Customer behavior: Consumer behavior and beyond. Fort Worth, TX’ The Dryden Press. Skinner, E. A. (1995). Perceived control, motivation, and coping. Thousand Oaks, CA’ Sage Publications. Sutton, S. (1998). Predicting and explaining intentions and behavior: How well are we doing? Journal of Applied Social Psychology, 28, 1317 – 1338. Taylor, S., & Todd, P. A. (1995). Understanding information technology usage: A test of competing models. Information Systems Research, 6(2), 144 – 176. Terry, D. J., & O’Leary, J. E. (1995). The theory of planned behavior: The effects of perceived behavioral control and self-efficacy. British Journal of Social Psychology, 34, 199 – 220. Wood, R., & Bandura, A. (1989). Social cognitive theory of organizational management. Academy of Management Review, 14(3), 361 – 384.

Kevin Celuch (PhD, Syracuse University) is a Professor of Marketing and holder of the Blair Chair in Business Science at the University of Southern Indiana. His research interests include advertising, product manual, consumer behavior, salesperson, market information, and organizational partneringrelated research. Stephen Goodwin (PhD, University of Iowa) is a Professor of Marketing at Illinois State University. His primary research interests focus on psychological aspects of consumer choice behavior, communication-effectiveness of alternative message appeals, and E-marketing. Steven A. Taylor is a Professor of Marketing at Illinois State University. He received his degree from Florida State University. His areas of interest relate to services marketing, relationship marketing, and e-business with an emphasis on the conceptualization and operationalization of the service quality, satisfaction, and value constructs.