Research Method Proceeding E Book 4A Turky

508 materials through a well-established courier company. We received 125 questionnaires representing 88.3 response rate. Sixteen respondents were excluded from further analyses because the respondents do not have a retail banking operations 66 eleven responses and do not answer all the required questions five responses. Our final sample consists of 109 banks. 3.2. Constructs and their measures Strategy Business strategy defines how a firm chooses to compete in its industry and tries to achieve a competitive advantage relative to its competitors Merchant and Van der Stede, 2007. Andrews 1980 argued that a clearly defined business strategy helps a firm allocates its resources to convert distinct competences into competitive advantage. Miles and Snow 1978 proposed three successful organizational strategies: defenders, prospectors, and analyzers. 67 Defenders focus on their niche market and emphasize on high product quality and services James Hatten, 1995. Defenders tend to have a narrow product lines and are less involved in product or market development Langfield-Smith, 1997. The critical success factors for defenders are stable product and services, high product quality and services on existing products and low prices Miles and Snow, 1978. Prospectors strive to take advantage of market opportunities by producing new products and services and are rewarded by their ability to charge premium prices for their innovative products and services. The critical success factors for prospectors are innovative products and services, broad range of products and services and quick response to changing business environment. Analyzers are characterized by their ability to take advantage of the strengths of both the defenders and the prospectors. The key success factors for analyzers are ability to 66 These respondents work either work for foreign banks or joint-venture that are not allowed to have retail businesses. 67 The forth type of strategy is reactor. However, since reactor is considered an unsuccessful type of strategy see for example Shortell and Zajac, 1990; Langfield-Smith, 1997, it is not discussed and used in this paper. 509 adopt innovation, in-depth analyses for innovative products and services before adopting them, adoption of innovative products with lower prices achieved through efficiency and close monit oring of competitors‘ activities. We use a multi-item scale approach to measure business strategy see section A1 of Appendix 1. This approach has been used successfully to measure types of business strategy in previous studies e.g., Segev, 1987; Conant Mokwa Varadajan, 1990. Some authors e.g., Segev 1987; McDaniel Kolari, 1987; McKee, Vanarajan Pride, 1989 argued that the multi-item approach is superior to self-typing method since the multi-item approach has detail questions that tend to lead the respondents to choose a particular type of strategy that is closely represent their firms‘ actual strategy. This approach has been acknowledged as an appropriate method when conducting strategy research Snow Hambrick, 1980; Huber Power, 1985 and has been frequently used in previous studies Snow Hrebiniak, 1980; Smith, Guthrie Chen, 1986; Segev 1987; McDaniel Kolari, 1987. The questionnaire contained a list of firms‘ characteristics and the respondents were asked to indicate the characteristics that best described their firm on a scale of 1 the characteristic does not suit my firm at all to 6 the characteristic suits my firm to a very high degree. The respondent were informed that there is no ―good‖ or ―bad‖ characteristic and they were asked to indicate their actual situation. The characteristics represent εiles and Snow‘s 1978 strategic typology of defender, prospector, and analyzer. The terms ―defender‖, ―prospector‖, and ―analyzer‖, however, were not used on the questionnaire. Rat her, each description was replaced by the terms ―Type 1‖, ―Type β‖ and ―Type γ‖. There are eighteen items used to measure strategy. The first six items in the questionnaires refers to the key success factors of defenders. Items seven to thirteen represent the key success factors of prospectors. The last five items in the questionnaire indicate the key success factors of analyzers. Following the approach 510 introduced by Conant, εokwa, Varadarajan 1987, we use the ―majority-rule‖ decision structure to categorize firms into their strategic archetypes. Firms are classified as defenders, prospectors, or analyzers based on the highest average score of the three types of strategies. Based on this procedure, 64 respondents are classified as defender, 28 respondents are classified as prospectors and 17 respondents are classified as analyzers. .Management control systems MCS MCS are put in place to ensure that employees only engage in value maximizing activities. MCS ensure that employees understand and consistently work hard to accomplish what are expected of them, they implement the firm‘s intended strategy, and they are capable of performing their jobs Merchant and Van der Stede, 2007. MCS includes both formal and informal systems. Formal control systems include rules, standard operating procedures, manuals, and budgeting systems. Informal control systems are not deliberately designed but are important to achieve superior performance. Informal control systems include work ethics, management style, and o rganizational culture. Both formal and informal control systems influence employees‘ behavior and consequently affect the degree to which goal congruence can be achieved. We adapt the MCS constructs used in previous studies e.g., Selto, Renner Young, 1995; Van de Ven and Ferry, 1980; Gresov, 1989; Drazin Van de Ven, 1985. There are six variables used for the MCS: performance evaluation, compensation, communication, conflict resolution, commitment, and product and market policy. Table 1 provides the definition of these variables. We use a multi-item scale approach to measure the control system construct see section A2 of Appendix 1. Respondents were asked to indicate whether they use a particular MCS and to rate the degree of importance attached to each system used on a 6-point Likert-type scale 1 = negligence; and 6 = significantly very important. [Insert Table 1 here] 511 Performance We investigate the performance implication of strategy-control system misfits using both financial and non-financial measures that are critical for banks to thrive and success. The financial measures consist of four variables: return on assets ROA, return on equity ROE, non-performing loan NPL and net interest margin NIM. The non-financial measures consist of two variables: customer satisfaction and employee satisfaction. ROA is the ratio of profit to total assets. ROE is the ratio of profit to total equity. NPL is the ratio of non-performing loan to total credit. NIM is the difference between interest revenues and interest expenses. Customer satisfaction is measured based on the length of time a customer stays with the bank. Employee satisfaction is measured based on the average length of time that employees work for the bank. The abbreviated version of the questionnaires on how we measure the firm performance is shown in section Aγ of appendix 1. Respondents were asked to indicate their banks‘ performance relative to their leading competitors. Responses were given on a 6-point Likert-type scale 1 = significantly below average; and 6 = significantly above average. Strategy-MCS misfit The strategy-structure misfit is measured based on the degree of departure of the observed configurations from the ―ideal‖ configurations for a given type of strategy. Van de Ven and Drazin 1985 proposed that there are two steps to measure a misfit using a system approach. First, an ideal profile is generated from high performing organizations. Second, the sampled organizations‘ configurations are compared to the ―ideal‖ profile using the Euclidean distance measures. Following Venkatraman and Prescot 1990, we measure a misfit based on the weighted Euclidean distance of an organization from the ideal profile for a specific type of strategy. Specifically, the misfit is measured using the following formula: MISFIT =    1 2 j cj sj j X X b 1 512 Where, X sj = the score for the organization in the study sample for the j th variable; X cj = the mean score for the ―ideal‖ profile along the j th variable; b j = standardized beta weight of the OLS regression equation for the j th variable in the specific type of strategy; j = 1, n where n is the number of control systems variables in the specific type of strategy. The measure of a misfit is based on the approach first introduced by Van de Ven Drazin 1985, but has been modified by Venkatraman Prescott 1990 to consider variables and their relative weights that are critically related to performance or those that are not critically related to performance for a given type of strategy. 4. Data Analysis and Result 4.1. Analytical Procedure There are three stages used to analyze the data and test the hypotheses. In the first stage, we perform separate analyzes for each type of strategy. These analyzes include: 1 run a separate OLS regression for each performance measure financial and non-financial measures on the six control systems i.e., performance evaluation, compensation, communication, conflict resolution, commitment, and product and market policy to obtain the standardized coefficients for the MCS variables that are significantly related to performance and MCS variables that are not significantly related to performance; 2 rank-order the firms with respect to their financial non-financial performance; 3 consider the top 10 percent firms as the ―ideal‖ group and the remaining 90 percent firms as the ―sample‖ group. In the second stage, we measure the degree of strategy-MCS misfits for both the significant variables MISFIT_C and the non-significant variables MISFIT_NC for each type of strategy. This procedure is performed separately for financial and non financial performance measures and include: 1 determine the ―ideal‖ control systems 513 for each type of strategy based on the control systems of the high performing firms i.e., top 10 percent; β measure the deviation from the ―ideal‖ profile by comparing the sample firms‘ εCS to the εCS of the ―ideal‖ profile; γ calculate the misfit for each firm by summing the products of the deviation score of each MCS with its corresponding standardized coefficient as shown in equation 1. In the third stage, we correlate the misfit and performance for financial and non financial measures. We run separate analyses for each type of strategy. We then use z-tests to investigate whether the correlation coefficients between the MISFIT_C and performance is significantly different from those coefficients between the MISFIT_NC and performance. The analytical procedures are presented in figure 1. [Insert Figure 1 here]

4.2. Statistical Analyses and Results

We begin our analyses by assessing the reliability of the constructs used in this study. The eleven constructs and their inter-item reliability are shown in Table 2. The results indicate that the reliability of the constructs is within the acceptable range Nunnally, 1967 with the minimum of 0.70 and the maximum of 0.94. [Insert Table 2 here] To investigate whether the multi-item questionnaires have dimensions that are consistent with the proposed construct used in this study, we perform a principal component factor analysis 68 . As shown in Table 3, defender has one factor 66.25 of the variance explained, prospector has one factor 63.34 of the variance explained, analyzer has one factor 60.05 of the variance explained, performance evaluation has one factor 84.36 of the variance explained, compensation has one factor 75.97 of the variance explained, communication has one factor 75.69 of the variance explained, conflict resolution has one factor 75.69 of the variance 68 To obtain a clear pattern of loadings, we use Varimax with Kaiser Normalization rotation method. In applying this procedure, factors with Eigenvalues greater than 1.00 were retained.