تاثیر بازارگرایی بر نوآوری محصول و عملکرد کسب و کار
|کد مقاله||سال انتشار||تعداد صفحات مقاله انگلیسی||ترجمه فارسی|
|19066||2003||22 صفحه PDF||سفارش دهید|
نسخه انگلیسی مقاله همین الان قابل دانلود است.
هزینه ترجمه مقاله بر اساس تعداد کلمات مقاله انگلیسی محاسبه می شود.
این مقاله تقریباً شامل 13828 کلمه می باشد.
هزینه ترجمه مقاله توسط مترجمان با تجربه، طبق جدول زیر محاسبه می شود:
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : International Journal of Research in Marketing, Volume 20, Issue 4, December 2003, Pages 355–376
The paper reports a study of the impact of market orientation on business performance. The use of product innovativeness is proposed as a mediator of the influence of market orientation on business performance. Product innovativeness is defined along two dimensions: use of new-to-the-firm and use of new-to-the-market products. Business performance was represented by relative price premium, sales growth, capacity utilization, and profitability. The findings provide support for the positive influence of market orientation on both dimensions of product innovativeness. However, only use of new-to-the-market products turns out to be a positive contributor to business performance.
In recent years there has been an increased focus on the relationship between market orientation and business performance. Studies of this issue have generally demonstrated that market orientation has a positive impact on firm performance (e.g., Jaworski & Kohli, 1993, Narver & Slater, 1990, Pelham & Wilson, 1996, Slater & Narver, 1994, Slater & Narver, 2000a and Slater & Narver, 2000b. However, few studies have investigated the potential mediators of the market orientation/performance relationship (for exceptions, see Atuahene-Gima, 1996 and Han et al., 1998). Such research is needed to understand how market orientation influences the different competitive parameters of the firm (e.g., price, advertising, distribution, etc.), and, in turn, how market orientation indirectly affects business performance. A firm's ability to employ new and successful product innovations is an important competitive weapon. Indeed, Jaworski and Kohli (1996) and Varadarajan and Jayachandran (1999) have argued that innovation has been inappropriately absent in models of market orientation. They suggest that future studies of market orientation should include the effects of market orientation on innovativeness of goods and services, in order to learn more about how market orientation works, and how it may be beneficial as a strategic firm capability. Consequently, this study explicitly includes the use of product innovation as a mediator of the relationship between market orientation and firm performance. More specifically, it explores the influence of market orientation on the use of the two dimensions of innovativeness that have been suggested by scholars of market orientation Atuahene-Gima, 1995, Atuahene-Gima, 1996, Danneels & Kleinschmidt, 2001, Jaworski & Kohli, 1996 and Lukas & Ferrell, 2000. These two dimensions are: new-to-the-firm products and new-to-the-market products Booz, Allen & Hamilton, 1982, Danneels & Kleinschmidt, 2001, Kleinschmidt & Cooper, 1991 and Olson et al., 1995. There is little research that conclusively demonstrates the effects of using new-to-the-market and new-to-the-firm product strategies (for exceptions, see Atuahene-Gima, 1995, Atuahene-Gima, 1996, Cooper, 1994 and Lukas & Ferrell, 2000). Thus, the goal of this research is to explore these effects on business performance.
نتیجه گیری انگلیسی
The two-step approach suggested by Anderson and Gerbing (1988) was applied to validate the measures used to test this model. To prepare the data for such a test, a special procedure was used for the validation of the market orientation scale. The scale for each of the dimensions of market orientation is considered to be formative since the more of the activities the organization performs, the more market oriented it becomes (please see Diamantopoulos & Winklhofer, 2001 for an examination of the nature of formative indicators). Thus, the indicators determine the dimension rather than the reverse. The firm may find the different activities in the scale to be interchangeable. Accordingly, for a given level of each of the dimensions, the firm may use different combinations of activities. Some companies can make use of formal data gathering systems while others may prefer informal ways (e.g., customer visits) to gather market information. Additionally, the measures may be nonlinearly related to the latent construct indicating that the different activities may be of different difficulty. The items within each of the three dimensions do not have to be highly correlated to satisfy the requirement of being valid indicants of market orientation (cf., Bollen & Lennox, 1991). Thus, omitting indicants based on classical test theory (cf., the requirement of unidimensionality) may be excluding a part of the construct. A principal-component analysis was conducted to identify the facets for each of the three dimensions of market orientation. Each principal component represents a facet of the dimension. For dimensions with more than one facet, each facet was weighted equally within the parcel to secure that no particular facet dominates the dimension. Accordingly, the value for each dimension (i.e., parcel) is a function of unique information (i.e., facets) and not a result of the number of items that represent each facet. Technically, the scores for each facet were computed by equally weighting and adding the corresponding items scores divided by the number of items representing the facet. The same procedure was applied for computing the parcels for the dimensions, where the scores for each dimension were computed by equally weighting and adding the corresponding facet scores divided by the number of facets representing the dimension. Since the indicants are parceled for each of the dimensions, the unidimensionality and reliability of the items are unknown Anderson & Gerbing, 1988 and Bollen & Lennox, 1991. On the other hand, the measures for each dimension of market orientation have a satisfactory face validity (Jaworski & Kohli, 1993) and parcels have better reliability (Bollen & Lennox, 1991) and normality than single items. Validation of the full measurement model was done with the three dimensions of market orientation treated as reflective indicants, the other measures in the model, and with the two business strategy variables. The analysis was done by using structural equation modeling (LISREL 8.5) to test the measurement model's ability to reproduce the observed variation–covariation matrix for the measures. The initial measurement models resulted in exclusion of two items for differentiation strategy and one item for overall cost leadership due to cross-loadings. Thus, overall cost leadership strategy is measured with a single indicator in this study, while two indicators represent differentiation strategy. The final measurement model fits well to the data. All fit indices are above the suggested cut-off values for satisfactory fit. The chi-square is 62.34 with 47 degrees of freedom and significant at 6.6%. The RMSEA-value is 0.033, which is below 0.05, the cut-off for close fit (p=0.91). The NNFI and CFI values are 0.97 and 0.98, respectively, which are above the 0.95 requirements suggested by Hu and Bentler (1999). Accordingly, the remaining measures meet the requirement of a well-fitting measurement model and thus will be applied in the structural analysis. The reliability information of the final measurement model is presented in Table 1 using standardized coefficients. The average variance extracted is considered satisfactory for all constructs, meeting the suggested cut-off value of 0.50 suggested by Bagozzi and Yi (1988). The composite reliability should be above 0.6 (Bagozzi & Yi, 1988) and is satisfactory for all constructs.To test the structural model the three dimensions of market orientation were multiplied to develop a new composite measure of market orientation as a multiplicative index. The error term of the new measure was set to 1 minus its composite reliability estimate (see Table 1) and multiplied with the variance of the composite measure (Anderson & Gerbing, 1988). The other single-item measures were not assigned any random measurement error due to lack of available reliability information. The estimated correlation matrix used for test of the structural model is shown in Table 2.To test the hypothesized model we included differentiation strategy and overall cost leadership as control variables to avoid potential spurious or masked effects in the model. The controls were treated as exogenous variables free to covary with the other latent variables in the model. The results from test of the linear effects in the model are shown in Table 3.The fit of the structural model is satisfactory and the model shows a great ability to reproduce the observed variation–covariation matrix of the sample. All fit indices are clearly above the suggested cut-off values for satisfactory fit. The chi-square is 34.94 with 35 degrees of freedom and is significant with p=0.47. The RMSEA-value is 0.00, which is below 0.05, the cut-off for close fit (p=0.99). The NNFI and CFI values are both 1.00, and above the 0.95 requirements (Hu & Bentler, 1999). The market orientation has a significant positive impact both on the use of new-to-the-market (γ11=0.33, p<0.01) and use of new-to-the-firm products (γ21=0.33, p<0.01). These findings support hypotheses 1 and 2. Moreover, market orientation was hypothesized to have greater impact on the use of new-to-the-market products than the use of new-to-the-firm products ( H3). The testing of H3 was done by examining the difference of the fit between two forms of the model. In the first form of the model the effects of market orientation on product innovation were set equal (i.e., γ11=γ21). In the second, these effects were allowed to be estimated individually (see Anderson & Gerbing, 1988). The chi-square difference between the models was not significant, rejecting H3. The use of new-to-the-market products was posited to affect both relative price premium, sales growth, and capacity utilization positively. The findings support H4 (β31=0.17, p<0.01), H6 (β41=0.12, p<0.01), and H8 (β51=0.12, p<0.01). The effects of the use of new-to-firm products on relative price premium, sales growth, and capacity utilization are not significant. Accordingly, H5, H7 and H9 were not supported in this study. Relative price premium, sales growth, and capacity utilization were all hypothesized to have a positive impact on profitability H11, H12 and H13. The effect of relative price premium on profitability is not significant, while the effect of sales growth on profitability is positive and significant (β64=0.24, p<0.01). Furthermore, the effect of capacity utilization on profitability is positive and significant (β65=0.30, p<0.01). H10 claims that there are positive interaction effects from the use of new-to-the-market and the use of new-to-the-firm products. Since the two constructs are measured with single items, and the error terms are fixed to zero, the interaction effect was tested by simply computing a new variable in the multiplicative index of the two innovativeness variables. Each of the two innovativeness variables that were used to compute the index were standardized (with mean 0 and variance 1) due to different distribution properties of the two variables and to avoid multicollinearity with the two original innovativeness variables in the model (Jaccard, Turrisi, & Wan, 1990). The new variable was entered into the model and was used as an exogenous variable together with use of new-to-the-market products, use of new-to-the-firm products, market orientation and business strategy. An extract of the analysis is reported in Table 4. The results show no significant interaction effects between use of new-to-the-market products and use of new-to-the-firm products on relative price premium, sales growth and capacity utilization. Accordingly, H10 is not supported.The model that has been developed and tested in the paper reflects specific choices regarding measurement and structure. Although a model fits well to data, other models might fit equally well or better, and a more rigorous test is to assess the relative fit of the hypothesized model Jöreskog, 1993 and Lambe et al., 2002. Therefore, the hypothesized model is compared with three rival models. First, the rival models are described and the results presented (see Table 5). Second, the formal model comparison is made collectively for the three rival models.Rival Model 1 holds that the effects of market orientation are not restricted to being mediated by product innovativeness. Many of the studies of market orientation have included profitability, overall performance, sales growth and other performance indicators without controlling for possible mediating variables that is done in the hypothesized model (e.g., Jaworski & Kohli, 1993, Matsuno et al., 2002, Narver & Slater, 1990 and Slater & Narver, 1994). Accordingly, the rival model includes additional paths from market orientation to relative price premium, capacity utilization, sales growth, and profitability to test whether product innovativeness mediates the effects of market orientation. The rival model has a very good model fit, and shows that market orientation has a positive impact on capacity utilization (p<0.05). Capacity utilization includes decisions about capacity planning as well as filling the capacity (e.g., van Dierdonck, 1998). Accordingly, it may be possible to argue that market orientation affects how well the capacity matches the demand in the market in addition to that use of new-to-the-market products affects how the capacity is being utilized. However, market orientation does not have direct effects on relative price premium, sales growth and profitability, controlling for product innovativeness (and capacity utilization). Rival Model 2 deals with market orientation as latent variable with the three dimensions (i.e., generation, dissemination, and responsiveness) as reflective composite items. This corresponds to the market orientation construct used in validation of the measurement model (Table 1). Each of the dimensions reflects the market orientation construct, and thus, are three indicants of the construct. This view is consistent with a recent scale validation study (Matsuno et al., 2000). The use of Rival Model 1 enables us to test whether a multiplicative index is superior to a construct with three reflective dimensions. Although acceptable, the fit indexes of Rival Model 2 (see Table 5) are not as good as those of the hypothesized model. The rival model does not show any differences in structure (cf., modification indices) but it has larger residuals in general. The effects of market orientation on two product innovativeness variables are different in the rival model while they are equal in the hypothesized model. Although both effects of market orientation on product innovativeness are lower than in the hypothesized model, market orientation has a relatively greater effect on use of new-to-the-market products than on use of new-to-the-firm products in this rival model. The weaker fit of Rival Model 2 indicates greater uncertainty about the model's estimates due to the use of market orientation with three reflective indicants rather than with a multiplicative index. Accordingly, the parameters of the rival model are less stable and reliable than those of the hypothesized model. Rival Model 3 holds each of the dimensions of market orientation as separate independent variables.2 This model allows a test of whether there are different effects of the different dimensions on the variables market orientation is hypothesized to impact. According to Kohli et al. (1993), the three dimensions of market orientation may or may not be correlated. This indicates that the different dimensions may have different effects, and thus, may be more effectively treated as three separate constructs. For instance, Ottum and Moore (1997) found that market information use had a positive impact on new product success, while generation and dissemination did not have any effect. The rival model only provides support for generation dimension of market orientation having a positive impact on both of the two innovativeness variables. The fit indexes of Rival Model 3 are acceptable, but not as good as those of the hypothesized model. The model comparisons are done in two steps. First, Rival Model 1 is a less restrictive version of the hypothesized model, and thus, the hypothesized model can be seen as a nested model of Rival Model 1 Anderson & Gerbing, 1988 and Jöreskog, 1993. A more flexible model will always fit better in absolute terms (i.e., chi-square) than a more restrictive model. The comparison is then a test of whether the Rival Model 1 with additional paths adds significantly to model fit compared to the more restrictive hypothesized model. If the chi-square difference exceeds α=0.05 this will lead to accepting the more flexible Rival Model 1 (Jöreskog, 1993). The model comparison (see Table 6) shows that the chi-square difference is not significant (p=0.26), and thus, the hypothesized model is preferred. The lack of direct effects of market orientation in the rival model supports the general position that market orientation is most effectively studied through the use of mediating variables that tap how market orientation works.Second, Rival Models 2 and 3 have different exogenous variables (i.e., measures of market orientation) than the hypothesized model. Accordingly, the models are non-nested and the models are compared to the hypothesized model using the Second-Order Akaike Information Criterion (CAIC) Burnham & Anderson, 1998 and Jöreskog, 1993. 3 The model with the smallest CAIC is the best approximation for the information in the data (i.e., chi-square adjusted for the degrees of freedom), relative to other models considered. The hypothesized model has smaller CAIC value than the two rival models (Table 6). To assess the size of the CAIC differences (Δi) between the hypothesized model and the two rivals, Akaike weights can be calculated (Burnham & Anderson, 1998). The Akaike weight (wi) is the relative likelihood of each model being the best model out of a set of evaluated models. The relative weight is expressed as exp(−1/2Δi) divided by the sum of exp(−1/2Δi) for each of the three models compared. The weights (wR2 and wR3) for Rival Models 2 and 3 are both less than 0.001 (see Table 6) and the weight for the hypothesized model is 0.999. This implies that Rival Models 2 and 3 are very unlikely to be equally good or better than the hypothesized model. In contrast, the hypothesized model is likely to be the best approximating model compared to the two rival models, and thus, is supported by the analysis. The hypothesized model has a better fit and approximation than its rivals with market orientation as latent variable with the three dimensions as well as when each of the three dimensions of market orientation appear as separate constructs. The model comparison supports that a multiplicative index is more effective and explains more of the variance in product innovativeness. Given the good fit of the hypothesized model, this may provide support for the position that market orientation dimensions have to operate in concert to provide positive effects for the company.