The Antecedents and Consequence of Trust for Structural Equation Modeling - Term Paper Example

The antecedents and consequence of trust for SMEs Executive summary This analysis is designed to capture the effect of five independent variables - reputation, skill, information exchange, power and flexibility- on trust and the influence of trust on long-term orientation. In doing so, we compare two model specifications using structural equation modelling. In the first one, trust is an observed variable; in the second one trust is a latent variable. Data were collected from 229 SMEs operating in the UK electrical and electronic sector. The results show that reputation, information exchange, and flexibility have positive effect on trust, while power has a negative effect. Skill was not found to impact trust. In addition, trust has a significant positive impact on long-term orientation of the relationship among SMEs. Table of Content Introduction 3 Description of structural equation modelling and its applicability to the field of marketing 4 Explanation of the variables and their inter-relationships 6 Design and purpose of the analysis 8 Explanation and debate of the final solution 9 References 14 Introduction The origins of the structural equation modelling (SEM) have its roots in three disciplines: sociology, psychology and economics. In marketing SEM starts its application in November 1982 in the issue published by the Journal of Marketing Research (Bollen 1989). SEM grows out of and serves purposes similar to multiple regression, but in a more powerful way which takes into account the modeling of interactions, nonlinearities, correlated independents, measurement error, correlated error terms, multiple latent independents each measured by multiple indicators, and one or more latent dependents also each with multiple indicators. SEM may be used as a more powerful alternative to multiple regression, path analysis, factor analysis, time series analysis, and analysis of covariance. That is, these procedures may be seen as special cases of SEM, or, to put it another way, SEM is an extension of the general linear model (GLM) of which multiple regression is a part. In this analysis we apply SEM to examine the influence of reputation, flexibility, information exchange, power and skill on trust and consequently on long-term orientation. As long-term orientation is a central theme in marketing currently, it is crucial to know what are the variables that help explain successful long-term relationship building. The analysis is structured as follows. First, we give a brief description of SEM and its importance for marketing research, then we provide the bases for the interrelationship between the variables used in the model; this is followed by description of the design of the analysis and finally, discussion of the result is provided. Description of structural equation modelling and its applicability to the field of marketing SEM is usually viewed as a confirmatory rather than exploratory procedure, using one of three approaches: 1. Strictly confirmatory approach: A model is tested using SEM goodness-of-fit tests to determine if the pattern of variances and covariances in the data is consistent with a structural (path) model specified by the researcher. However as other unexamined models may fit the data as well or better, an accepted model is only a not-disconfirmed model. 2. Alternative models approach: One may test two or more causal models to determine which has the best fit. There are many goodness-of-fit measures, reflecting different considerations, and usually three or four are reported by the researcher. Although desirable in principle, this AM approach runs into the real-world problem that in most specific research topic areas, the researcher does not find in the literature two well-developed alternative models to test. 3. Model development approach: In practice, much SEM research combines confirmatory and exploratory purposes: a model is tested using SEM procedures, found to be deficient, and an alternative model is then tested based on changes suggested by SEM modification indexes. This is the most common approach found in the literature. The problem with the model development approach is that models confirmed in this manner are post-hoc ones which may not be stable (may not fit new data, having been created based on the uniqueness of an initial dataset). Researchers may attempt to overcome this problem by using a cross-validation strategy under which the model is developed using a calibration data sample and then confirmed using an independent validation sample. Regardless of approach, SEM cannot itself draw causal arrows in models or resolve causal ambiguities. Theoretical insight and judgment by the researcher is still of utmost importance. According to Bollen (1989) the general SEM can be viewed in terms of three components: (1) path analysis, (2) conceptual synthesis of latent variable and a measurement model, and (3) general estimation procedures. In relation to the path analysis, it has three aspects to be considered: the path diagram, the equation relating correlations or covariances to parameters and the decomposition of effects. The first aspect, the path diagram is a pictorial representation of a system of simultaneous equations. It shows the relation between all variables, including disturbances and errors. The second aspect, the correlations of variables to the model parameters, is equivalent to covariance structure equations. The third aspect of path analysis provides a means to distinguish direct, indirect and total effect of one variable on another. The direct effects are those not mediated by any other variable; the indirect effects operate through at least one intervening variable, and the total effect is the sum of direct and all indirect effects. In addition to path analysis, the conceptual synthesis of latent variable and measurement models is essential to contemporary structural equation techniques. In particular, measurement is a process by which a concept is linked to one or more latent variables, and these are linked to observed variables. The concept can vary from one that is highly abstract, such as economic development, expectations, intelligence or trust, to one that is more concrete, such as age, sex, or race. One or several latent variables may be needed to represent the concept. The observed variables can be responses to questionnaire items, census figures, or any other observable characteristic. The last aspect, discussed by Bollen, is general estimation procedures. Bock and Borgman (1966) proposed an analysis of covariance structures to estimate the component of variance due to latent variables in multinormal observed variables. Bentler (1983) suggested estimators that treat higher-order product moments of the observed variables. He demonstrated that these moments can help identify model parameters that are not identified by the covariances and the gains in efficiency that may result. Finally, the developments in SEM would not be complete without mentioning the computer software that has emerged. The three most widely used software packages in SEM are LISREL, AMOS and EQS. Steenkamp and Baumgartner (2000) reflect on the role of SEM in marketing modelling and decision making by taking a process view on marketing science. They argue that SEM is especially useful for theory testing and knowledge building, which are key inputs for valid marketing models. However, they also argue that SEM has decision making potential. SEM is a powerful and versatile technique that can be fruitfully used by modelers to account for measurement error and / or assess the robustness of the results to measurement error, to test substantive hypotheses in cross-sectional models containing varying sets of constructs and indicators, non-linear effects, and heterogeneous relationships, and to analyse longitudinal data sets to investigate change processes and the drivers of this change. Judgemental parameter values can in principal be blended with hard data. Moreover, and more generally, SEM 's focus on the relations between constructs and their operationalizations pushes the marketing modeler to think more critically about the conceptual and empirical quality of their observed measures. Explanation of the variables and their inter-relationships The antecedents and consequences of trust (See figure 1) have been subject of intensive research in the marketing literature recently (e.g. De Ruyter et al. 2001, Geyskens et al. 1998, Ivens 2005). Inspired by interpersonal research, many studies define trust as "the extent to which a firm believes that its exchange partner is honest and/or benevolent" (Anderson and Narus 1990; Dwyer and Oh 1987), or some variant thereof. Trust in the partner's honesty is a member's belief that one's partner is reliable, stands by its word, fulfils promised role obligations, and is sincere. Trust in partner's benevolence is a channel member's belief that its partner is genuinely interested in one's interests or welfare and is motivated to seek joint gains. A benevolent partner subordinates immediate self-interest for long range group gain and will not take unexpected actions that would have a negative impact on the firm. Amazingly, the studies identified and examined over sixty constructs as antecedents and consequences of trust (See Geyskens et al. 1998 for a review of research on antecedents and consequences of Trust). Among the most common used antecedents have been Reputation (Anderson and Weitz 1989; Ganesan 1994), Skill/Expertise (Grosby et al. 1990), Information Exchange/Communication (Anderson and Weitz 1989; Anderson and Narus 1990; Morgan and Hunt 1994), and Power (Anderson and Weitz 1989; Busch and Wilson 1976). Recently, additional variable has received attention as an antecedent of trust that is Flexibility (e.g. Ivens 2005). Based on the research mentioned above, reputation, information exchange/ communication and skill/expertise of partner have been found to influence positively Trust. Power asymmetry, on the other hand, has negative effect according to Anderson and Weitz (1989). However, referent power of partner and expert power of partner influence positively Trust (Busch and Wilson 1976). In addition, flexibility has positive effect on Trust (Ivens 2005). Long-term orientation has been found the most frequent consequence of Trust (Andaleeb 1991; Anderson and Weitz 1989; De Ruyter et al. 2001; Ganesan 1994; Geyskens et al. 1998; Morgan and Hunt 1994). Figure 1: Theoretical model Design and purpose of the analysis In order to reveal the relationship of the above described five independent variables (reputation, skill, information exchange, power and flexibility) to trust and consequently to long-term orientation, data were collected from 229 SMEs operating in the UK electrical and electronic sector and were subjected to covariance based SEM. The decreasing emphasis on short-term transactions and the increased focus on long-term relationships is an important trend in marketing nowadays. However, researchers claim that trust building and the antecedent of trust are vital for a successful long-term relationship. That is why this analysis is designed to capture the effect of the five independent variables on trust and the influence of trust on long-term orientation. In doing so, we compare two model specifications. In the first one, trust is an observed variable; in the second one trust is a latent variable composed of reputation, skill, information exchange, power and flexibility. The rational behind this procedure is to see which model explains better the long-term orientation. As trust has enjoyed a lot of research efforts, it has also received numerous definitions. Therefore, we are interested to know whether reputation, skill, information exchange, power and flexibility can explain more forming part of the notion of trust or being its antecedents. If the second is true, and if we find a significant positive relationship between trust and long-term orientation, companies that strive to achieve long-term orientation should focus on controlling the five antecedent of trust discussed here. The analysis follows the logical sequence of Bollen (1989); first, we designed the path diagrams - one without a latent variable and another with trust as a latent variable; then we go to the estimation procedure, and finally we compare the two models and discuss the results and implications. AMOS software has been used for this analysis. Explanation and debate of the final solution As it can be observed from the results given below, both models have a good global fitness (Chi-square = 47.426 for model 1 and Chi-square = 74.067 for model 2 with probability level = .000 for both models). However, model 1 is better. The noncentrality parameter (NCP) and the expected cross-validation index (ECVI) are smaller for model 1, which also give reason to choose this model as the better one. Besides, the incremental fitness measures normed fit index (NFI), incremental fit index (IFI) and comparative fit index (CFI) have values bigger than 0.9 for model 1, which is not the case for model 2. All things considered, we can conclude that model 1 represents better the relationships among our variables. Trust has a significant positive relationship with long-term orientation in both models. In addition, the first model confirms the results of other researches that have found positive effect of reputation, information exchange/communication and flexibility on trust. However, skill does not appear as having a significative relationship with trust in our analysis. Importantly, power has a negative effect, which calls for attention to power imbalances in building trust and long-term orientation. The covariances also show negative relationship of power with reputation, flexibility, information exchange and skill (although for information exchange it is not significative). As reputation, information exchange/communication and flexibility have significant positive relationship with trust, it is important for managers to pay attention to the issues of power when building a long-term orientation. Another interesting result of our analysis is that we found skill to has no relationship with trust in the first model, but in the second model skill has significant positive relation with trust. This is important in terms of the operationalization of the variable trust, i.e. the belief of the partner having been skilled can be included as a part of the notion/operationalization of trust, and not as an antecedent. Finally, this analysis confirms the importance of trust for long-term orientation and brings to light some useful additional considerations as regard the most important antecedent of trust. Model 1 Result (Default model) Chi-square = 47.426 Degrees of freedom = 5 Probability level = .000 Regression Weights: (Group number 1 - Default model) Estimate S.E. C.R. P Label TRUST Read More