THE ENGAGEMENT OF EMPLOYEES IN THE STRATEGY PROCESS AND FIRM PERFORMANCE: THE ROLE OF STRATEGIC GOALS AND ENVIRONMENT Linda F. Tegarden Virginia Tech Blacksburg, VA Yolanda Sarason Colorado State University Fort Collins, CO J. Stephen Childers Radford University Radford, VA Donald E. Hatfield Virginia Tech Blacksburg, VA Abstract De!>pite the call to engage employees in strategy making processes, empirical evidence that ties this engagement to financial performance has not been forth- coming. This study fills this gap by investigating whether involving employees in the strategy making process leads to a higher achievement ofstrategic goals and subsequently increased financial performance. Our findings suggest that the link between strategy making processes and financial performance may be underestimated unless strategic goals are included as a mediator. We also find environmental dynamism moderates the relationships we investigate. Under condi- tions of low dynamism, there is a stronger relationship between the engagement ofemployees and strategic goals related to innovation than under conditions of high dynamism. Conversely, strategic goals related to quality have a stronger relationship with engagement of employees under conditions ofhigh dynamism when compared to conditions oflow dynamism. The importance of involving employees throughout an organization in the strategy making process has been recognized by traditional academic research- ers (e.g. Burgelman, 1991; Floyd & Wooldridge, 2000; Hart, 1992), as well as consultants and practitioners (e.g. Hamel & Prahalad, 1994; Kaplan & Norton, 1996). One rationale behind the use of multi-level strategy processes is to provide employees with a better understanding of the company's strategy and build a 76 Journal ofBusiness Strategies Vol. 22, No.2 stronger commitment to achieving the goals in the implementation of the strategy (Floyd & Wooldridge, 2000; Hamel & Prahalad, 1994; Kaplan & Norton, 1996). A company's strategy can also be facilitated by employee involvement because knowledge and information relevant to strategy making is dispersed throughout the organization (Miller & Monge, 1986). Cognitive model theorists (Anthony, 1978; Frost, Wakely & Ruh, 1974) propose that when employees have more complete knowledge about their jobs and operations they can provide better information. Involvement in strategy making also allows employees to know more about implementation of decisions. For example, ideas that support new innovations often come from those employees that are in direct contact with the customer (Von Hippel, 1988). In order for innovation to be a source of competi- tive advantage for a company, there must be organizational processes that allow these ideas to be brought forward and integrated into the activities of the company (Burgelman,1983). Despite the call for more investigations into the relationship between strategy making process and firm performance (Chakravarthy & Doz, 1992; Huff& Reger, 1987), there has been relatively little empirical research on this link, especially when strategy making processes involve multiple levels of the organization (Floyd & Wooldridge, 2000; Rajagopalan, Rasheed, & Datta, 1993). While the intensity and use of employee involvement practices is on the increase, the stra- tegic impact of such practices has not been adequately investigated. This may be because most studies focus on narrow definitions of employee involvement (Ledford & Lawler, 1994). Our study uses a broad definition of employee in- volvement in strategy process to include information sharing, decision making, experimentation, understanding of company goals, and iterative strategy mak- ing that involves multiple levels of the organization. A broad definition captures the variety of involvement ranging from information sharing to actual decision making. We suggest that aspects of such involvement should relate to achieve- ment of strategic goals. The two types of strategic goals tested in this study are quality and innovation. It is suggested that both require employee involvement and empowerment to make decisions (Lawler, Mohrman, & Ledford, 1995; Burgelman, 1991). In addition, these goals often create a tension between striving to have organizational pro- cesses that ensure consistent quality and organizational processes that facilitate creativity and innovation (Brown & Eisenhardt, 1998; Jelinek & Schoonhoven, 1990; Miron, Erez, & Naveh, 2004). We also propose that multi-level strategy processes support the achievement of quality more than the achievement of in- novation, since achievement ofthese goals require different levels of cooperation and competition among employees within an organization. These differences in the strength of the path relationships for quality and innovation are proposed to vary with levels of environmental dynamism (Miller & Friesen, 1983). This study contributes to the understanding of the relationship between strategy processes and financial performance by examining information and communication benefits derived from use of multi-level strategy processes. We Fall 2005 Tegarden et at.: Engagement ofEmployees 77 extend this work with insights from cognitive and communication models that focus on information and communication benefits associated with participative management (Miller & Monge, 1986; Monge, Cozzens, & Contractor, 1992). We postulate that the relationship between multi-level strategy processes and firm financial performance is not a direct link. Rather, we suggest that involving multiple levels of the organization in strategy processes enables higher com- mitment and achievement of strategic goals because the quality of information used in decisions and communication of what strategies and goals are important increases with organization-wide participation in strategy processes. Moreover, it is the successful achievement of these strategic goals that leads to superior financial performance. Thus, we propose that strategic goals are an essential mediating variable in understanding the multi-level strategy process - financial performance link. Figure 1 outlines the model we propose and test. Figure 1 The Link Between Multi-Level Strategy Processes, Strategic Goals and Firm Performance and Moderating Role of Environment Strategy Making Processes Involving Multiple Levels of the Organization Achievement of ~ Strategic Goals ~r (Quality and Innovation) Firm Performance Environmental Dynamism We begin with a discussion of the rationale leading to our hypotheses. Sub- sequently, we describe the nature of our methodology and results. We end with a discussion of our conclusions and implications for research and managers. Literature Review and Hypotheses Strategy process research is concerned with how a firm's administrative sys- tems and decision processes influence its strategic positions (Chakravarthy & 002, 1992). There has been a call for more investigations of the links between firm performance and strategy processes in order that our understanding be more complete (Floyd & Wooldridge, 2000; Huff & Reger, 1987). Forthe link between strategy process and firm performance to be made, there is a need to explicate the relevant mediating and moderating variables (Rajagopalan et aI., 1993). We are 78 Journal of Business Strategies Vol. 22, No.2 proposing that strategic goals are important, yet overlooked, mediating variables in the strategy process - financial performance relationship. We will begin with a discussion of the extant work on strategy processes and firm performance. There is emerging empirical work that illustrates the potential competitive advantage from the use of multi-level strategy making. For example, Wooldridge and Floyd (1990) find that greater participation and involvement by middle level managers in strategy formation results in greater commitment and understanding of strategy, as well as improved economic performance. Monge et a1. (1992) finds that communication variables, like access to information throughout the organi- zation, increase the number of innovations in a firm. Powell (1992) shows that strategic planning processes, that are not available on the strategic factor market, can be sources of competitive advantage. Miller and Lee (200 I) give evidence that collaboration among employees toward decisions results in improved firm performance. Tegarden, Sarason, and Banbury (2003) give evidence that processes drawing upon all employees are more 1ikely to have a greater influence on strategic performance rather than financial performance. Finally, the use of participative management practices has been on the rise in organizations because most report success, especially with self-management teams (Lawler, Mohrman, & Ledford, 1992) and for more complex tasks (Ledford & Lawler, 1994). While these studies have begun to open the organizational black box and provide growing evidence that use of multi-level strategy processes can help explain firm performance, there is still a gap in our understanding of how and why involving employees in the strategy making processes improves performance. We offer that involving employees in strategy making results in higher achievement of strategic goals because of higher commitment to such goals and it is this achievement of strategic goals that leads to higher financial performance. In order to support this argument, we draw from participation and empowerment literatures to establish the mediating role of strategic goals with multi-level strategy processes and firm performance. There is a plethora ofliterature that focuses on the outcome effects of employee involvement in organizations. Multi-level strategy processes encompass the no- tions that involvement of employees and managers is beneficial to organization outcomes. Both motivation and communication are central to the idea that em- ployee involvement produces higher satisfaction and productivity. The emphasis is on outcomes that impact strategic position rather than financial performance. Each set ofliterature is summarized below as to the links between greater involvement of employees and managers and its impact on organizational performance. Involvement as a subject has a long history, dating back to McGregor's (1960) idea that workers can contribute to governing their own situations. McGregor's specification of Theory X and Theory Y focus on beliefs about human nature and how they impact work to be done. Likert (1961) described management models that went beyond the traditional command and control structure. The work of McGregor (1960) and Likert (1961) encouraged investigations on participative management during the 1960s. Participative management "recognized the good- Fall 2005 Tegarden et al..' Engagement ofEmployees 79 will, interest, talent, and needs ofemployees, and encouraged open communication and cooperation between management and employees" (Forrester, 2000: 67). During the 1980s, Lawler and his colleagues (Bowen & Lawler, 1992; Lawler, 1986, 1992; Lawler et aI., 1995) extended this idea by identifying levels of involvement from low to high. The low end of involvement is idea generation, suggestions that employees give to managers on ways to improve operations. A higher level of involvement is giving employees the ability to determine the methods they employ on the job. The highest level of involvement is giving lower level managers and employees the ability to influence decisions beyond one's own job that affect the business at large. The major reason for involving employees in decision making is that it will improve organizational effectiveness. The empowerment literature focuses on empowerment as a motivational con- struct that affects individual and organizational outcomes (Rudolph & Peluchette, 1993). Empowerment comprises individual cognitions and perceptions that constitute feelings of behavioral and psychological investment in work (Conger & Kanungo, 1988; Spreitzer, 1995, 1996; Zimmerman, 1990). These include meaning, self-determination, competence and impact. It describes "people's belief in their capabilities to mobilize the motivation, cognitive resources, and courses of action needed to exercise control over given events" (Ozer & Bandura, 1990, p. 472). Empowerment also relates to a belief in self-efficacy and an expecta- tion that effort will lead to performance (Koberg, Boss, Senjem, & Goodman, 1999). "Empowerment implies the freedom and the ability to make decisions and commitments, not just to suggest them or be part of making them" (For- rester, 2000, p. 67). It is recognized that organizations are characterized with an increasingly blurred distinction between managers and workers with a growing reliance on horizontal structures and peer networks (Kantor, 1989; Pfeffer, 1994). The importance of empowering employees is also discussed in the entrepreneur- ship literature (Kuratko, Hornsby, & Corso, 1996; Willard, Krueger, & Fesser, 1992). When the founder shifts from creation to exploitation, employees should be empowered with the ability to create and facilitate adaptation and change. The founder's shift to a more managerial role requires an increase in employees' roles in identifying and adapting to change. As such, empowerment has received increasing attention. Empowerment has been shown to affect managerial and organizational effectiveness (Spreitzer, 1995). Perceptions of empowerment can enhance the value of work for individuals, increase job satisfaction and contrib- ute to work productivity and success (Eylon & Au, 1996; Fulford & Enz, 1995; Spreitzer, 1995). For example, Koberg, et al. (1999) found that empowerment perceptions were associated with increased job satisfaction and work producti v- ity/effectiveness. The important point here is that outcomes studied in relationship to empowerment include worker effectiveness/productivity and job satisfaction, not a direct link to financial performance. Another stream of research is the participation literature which focuses on participation's influence on satisfaction and productivity. The primary foundations ofthis literature are communication and affective models that predict the positive 80 Journal olBusiness Strategies Vol. 22, No.2 relationship between participation and organizational outcomes, like innovation, productivity and job satisfaction (Miller & Monge, 1986; Monge et aI., 1992; Wagner, 1994). Communication models assume that participation in decision making enhances the flow and use of important information in organizations. Theorists (Anthony, 1978; Frost, et aI., 1974) propose that workers typically have more complete knowledge of their work than management; hence if work- ers participate in decision making, decisions will be made with better pools of information. In addition, it is suggested that if employees participate in decision making, they will know more about implementing work procedures after decisions have been made (Maier, 1963; Melcher, 1976). In essence, the role of information and communication is central to the positive relationship between participation and organizational outcomes. Some support for this assertion is found in Miller & Monge's (1986) meta-analysis. Stronger support is shown in the Monge, et al. (1992) study of participation and organizational innovation. In addition to the communication predictors, participation literature also stresses the motivational predictors or affective models of participation, which are the foundations of the human relations school of management (Blake & Mouton, 1970; Ritchie & Miles, 1970) as well as the theory supporting empowerment. The relationship between participation and organizational outcomes is indirect. Participation will enhance organization effectiveness through intervening moti- vational processes: participation fulfills needs, fulfilled needs lead to satisfaction, satisfaction strengthens motivation, and increased motivation improves worker's productivity. A small, but positive relationship is supported by the literature (Cot- ton, Vollrath, Legnick-Hall, & Froggatt, 1990; Wagner, 1994), yet most studies focus on trivial changes in complex systems (Ledford & Lawler, 1994). In any event, again, participation studies focus on its relationship to organizational outcomes and effectiveness and not financial performance. Engaging employees at multiple levels ofthe organization in the strategy mak- ing process results in greater commitment to strategic goals (Locke, Latham, & Erez, 1988), as well as the potential for improved decision making, either through use of more relevant information or higher satisfaction, which in turn leads to the greater motivation to achieve goals. While we recognize that commitment to strategic goals and achievement of strategic goals are two different constructs, we follow the logic of Lock, et al. (1988, p. 24) that commitment to goals can be inferred from performance. As they state, "(w)hile performance cannot be a catch-all measure of commitment, since performance can be caused by other factors such as ability, judicious use of inference from performance seems both theoretically and empirically justified." Consistent with this insight, Salancik (1977) indicates that behavior or action is the ultimate proof of commitment and thus, by implication, the most accurate measure of it. Thus, multi-level strategy making processes are directed toward the achievement of organizational goals, such as innovation, efficiency, and quality (what we call strategic goals). Stated more formally: Fall 2005 Tegarden et al.: Engagement of Employees 81 Hypothesis Ia: The use olmulti-level strategy processes is positive(v related to the achievement ofstrategic goals. The literature on multi-level strategy process has traditionally focused upon commitment to goals or a strategy rather than financial performance. Conven- tional strategy models always focus on strategy (a stream of actions) that can lead to superior strategic performance. When multi-level actions are "directed" by a commonly agreed on set of strategic goals, the result is higher financial performance. As such, we propose that: Hypothesis I b: The achievement ofstrategic goals mediate the relationship of multi-level strategy processes andfinan- cial performance. Type of Strategic Goal We investigate two types of strategic goals, goals that focus on the achieve- ment of quality and goals that focus on the achievement of innovation. Quality focuses on striving to be more effective and efficient, while innovation focuses on developing new products and services. We propose that the degree of internal cooperation and internal competition required to achieve these strategic goals differs. This difference impacts the effectiveness ofmulti-level strategy processes on firm performance. Striving for quality requires organization members to collaborate and coordinate their decisions and actions (Dean & Bowen, 1994; Powell, 1995). To achieve higher quality, organization members must coordinate their actions across func- tional and hierarchy boundaries (Floyd & Wooldridge, 2000). For example, the quality of a product should be tracked throughout the value chain of a firm, from inputs to manufacturing through sales and customer service. Feedback and learning associated with the achievement of higher quality requires that the organization provides cross-functional communication, not only within the organization but also with suppliers and customers (Hart, 1992). Decisions reflect cooperation and tight coordination between related entities to achieve high quality. Employees are empowered at multiple levels to make decisions that impact quality, but they must be orchestrated within a tight, cohesive system. A strong commitment to quality by all employees is necessary to achieve a goal of quality (Bowles, 1992; Gabor, 1990; Port & Smith, 1995). Multi-level strategy processes include the aspects of learning and adaptation required to achieve greater quality. The relationship between multi-level strategy processes and the processes required to achieve quality are highly complementary in nature, as both require cooperation and coordination in the development of strategy and the implementation of quality. Greater achievement ofthe strategic goal of innovation also requires multiple levels of involvement in strategy processes. As others have advocated (Dam- anpour, 1991, Kantor, 1988), it is important that individuals are encouraged to continually commit to product and process innovation in order for creativity not 82 Journal of Business Strategies Vol. 22, No.2 to be strangled. While innovation as a strategic goal requires employees at all levels ofthe organization to commit to new product development, there is a lesser need for systematic coordination across all levels and functions of the organiza- tion to achieve it. In contrast to those firms in pursuit of quality, firms striving for innovation use teams (usually multi-functional) that compete for resources (Burgelman, 1983). Thus, the innovation process is rooted more in competition rather than cooperation. Because the overall coordination required to successfully implement innovation is lower, the effectiveness ofmulti-level strategy processes on innovation goals will be lower relative to quality goals. To summarize, high achievement of innovation requires a team-level goal com- mitment, more than an organizational-level goal commitment. In addition, for innovation, coordination is confined to fewer individuals in the organization and there is competition between teams for resources. In contrast, achievement of quality requires greater coordination across the entire organization. As such, the use of multi-level strategy processes is more effective when the goal requires broader coordination and communication across a greater number of organiza- tional members. In the case of innovation, organization members need only to communicate with the specific team members involved in a specific innovation project. Since multi-level processes facilitate cooperation and commitment, the fit between quality and multi-level processes is higher compared to the fit with inno- vation and multi-level processes. Therefore, the link between multi-level strategy processes and firm performance will differ. We hypothesize that: Hypothesis 2: The indirect relationship between multi-level pro- cesses andfirm performance is stronger when the strategic goal is quality rather than when the strategic goal is innovation. Environmental Dynamism as a Moderator In order to test the strength of our predictions and argument, we investigate the role of environmental dynamism on the differences between quality and in- novation. Environmental dynamism is defined as both technological change and unpredictable instability (Miller & Friesen, 1983; Powell, 1996) and has been an important variable in investigations of strategy processes and firm performance (Hart & Banbury, 1994; Tegarden et al., 2003). While investigations have yielded mixed results (Miller & Cardinal, 1994; Rajagopalan et al., 1993), there is grow- ing evidence that strategy processes such as strategic planning (Brews & Hunt, 1999) and rational decision making (Goll & Rasheed, 1997) are more effective in dynamic environments. We predict that the role of environmental dynamism will depend on whether the strategic goal is quality or if the strategic goal is innova- tion. We propose that environmental dynamism will positively impact the multi-lev- el strategy process when the strategic goal is quality. To be successful in dynamic environments, firms must adapt to unpredictable external changes. Firms that strive for high quality have more connections with external actors because quality Fall 2005 Tegarden et al.: Engagement of Employees 83 requires extensive collaboration with both suppliers and customers. In addition, organization members are more connected to ensure that coordination and com- munication transpires. Because of the greater number oflinkages implemented in an organization when quality is a priority, information regarding changes in the environment will transmit to a greater number of organization members faster. Thus: Hypothesis 3a: The strength ofthe relationship between multi- level strategic processes and the strategic goal of quality is greater when environmental dynamism is high compared to when environmental dynamism is low. We suggest that the relationship between multi-level strategy processes and the strategic goal of innovation will be negatively impacted by environmental dynamism. In contrast to the implementation mechanisms in place for quality, the number of communication and coordination links among employees is lower with the goal of innovation, resulting in less information being relayed about the changing environment. While a single project team can respond quickly to changes, the entire organization will take longer to absorb and determine the strategic impact of external changes. The use of multi-level strategy processes with processes that support innovation are more likely to generate multiple inter- pretations ofthe nature and impact of change because employees are often tied to individual projects with individual goals and reward structures. This is also more likely to result in greater negotiation to reach agreement about the company's future strategy. Moreover, we suggest that under conditions of low dynamism it is even more important for firms to continually innovate to stay ahead of their industry's competitors. Thus we predict effectiveness will decline with the use of multi-level strategy processes with the goal of innovation in more dynamic environments. We hypothesize that: Hypothesis 3b: The strength ofthe relationship between multi- level strategic processes and the strategic goal of innovation is greater when environmental dynamism is low compared to when environmental dynamism is high. Method OUf sample was obtained from the 1996 CorpTech Directory of Technology Companies. This data set ofU.S. firms was published by the Corporate Technology Information Services. The range of these technology industries includes advance materials, biotechnology, defense, environmental, manufacturing equipment, transportation, and chemicals. The relationships being tested are not time depen- dent. Following guidelines for strategy research, there is no reason for suspecting that the age of the data influenced the results in this study (Robins, 2004). 84 Journal of Business Strategies Vol. 22, No.2 We randomly selected 2,000 organizations from among the firms provided by the CorpTech Directory and mailed surveys to the chief executives of these firms. Research indicates that top administrators provide reliable information about basic environmental and organizational characteristics of their organiza- tions (Miller & Friesen, 1983). A total 01'377 surveys were returned for a 19% response rate. This response rate is not atypical for research using CEOs as respondents (Milliken, 1990). Non-responding firms did not differ significantly from responding firms in the proportion of privately owned firms, number of employees, sales revenue, annual percentage growth in number of employees, or year of formation. We eliminated returned surveys with incomplete information, leaving the data from 335 surveys for use in our statistical analysis. Most (80%) are private firms and over half of the firms report annual sales revenues of less than $2.5 million and employ fewer than 25 workers. Less than 2% employ over 2,500 workers and about 3% have sales of over $500 million. Approximately 8% of the firms in the sample have been in business for 25 or more years, approximately 36% for between 10 and 24 years, and the remaining firms have been in business for less than 10 years. Measures. The scales used to construct the variables in this investigation have been standardized and validated by other researchers. The questions were pre- sented on a five-point Likert scale. Variables are measured using averages across multiple items. Multi-Level Strategy Processes. Multi-Level Strategy Processes is measured using the average of five items developed by Hart and Banbury (1994). These strategy processes reflect engagement of employees at all levels of the organiza- tion. Items in this measure include the questions: "Strategy is made on an iterative basis, involving managers, staff and executives in an on-going dialogue." and "Most people in this company have input into the decisions that effect them." The alpha coefficient for this measure is 0.74. Strategic Goal- Quality. Quality as a strategic goal is measured using the aver- age of three items developed by Hart and Banbury ( 1994). This measure included answers to questions focusing on whether the respondents' organization produced high quality of product (service) or if quality was a focus in the technical product! service design and development. The alpha coefficient for this measure is 0.61. Strategic Goal-Innovation. Innovation as a strategic goal is measured using the average of three items developed by Hart and Banbury (1994). This mea- sure included answers to rating the company's performance over the last three years on activities such as the extent of product/service changes over the past five years and the number of new products/services in the next year. The alpha coefficient is 0.66. Strategic Goals Combined. The items to measure the strategic goal of quality and the strategic goal of innovation were combined for an overall average of strategic goals. The alpha coefficient for this measure is 0.70. Fall 2005 Tegarden et al.: Engagement of Employees 85 Financial Performance. Financial Performance is measured as an average of a five-item scale developed by Powell (1995). Respondents are asked questions that reflect the extent that revenue, growth, or financial performance has exceeded their competitors' or has been judged outstanding by the respondent in the last three years. Although the measures obtained are subjective, past research has validated the use of such measures for performance (Lawrence & Lorsch, 1967), and is appropriate for samples such as ours in which most (80%) of the firms are privately owned. The alpha coefficient for this measure is 0.91. Environmental Dynamism. Environmental Dynamism is measured as an average of five items from a scale developed by Powell (1996) with questions related to the frequency of change in the environment and the predictability of this change (Miller & Friesen, 1983). Respondents were asked the extent to which they agreed with statements such as "Demand in our industry has been growing rapidly in the past 3 years" and "Our industry is more unstable than most, changing more quickly and unpredictably." The alpha coefficient for this measure is 0.73. Typically, alpha coefficients of multi-item measures should fall within a range of 0.70 and 0.90 for narrow constructs, and 0.55 to 0.70 for moderately broad constructs (Van de Ven & Ferry, 1979). All of the coefficient alphas in this study are in these ranges and are consistent with those used in other studies using these variables (Hart & Banbury, 1994; Tegarden et aI., 2003). Results Table 1 reports the descriptive statistics and correlations for the variables. Re- sults of the correlations show that financial performance is not significantly relat- ed to multi-level strategy processes or environmental dynamism. However, multi- level strategy processes is significantly related to the innovation strategic goal, quality strategic goal and the combined strategic goal. As would be expected, strategic goals are related to each other. Interestingly, environmental dynamism is significantly related to each strategic goal, as well as the goals combined. Hypothesis 1a states that the use ofmulti-level strategy processes is positively related to the achievement of strategic goals. To test this relationship we used regressions, with combined quality and innovation, quality alone, and innova- tion alone as the dependent variables and multi-level strategy processes as the independent variable. Table 2 indicates that multi-level strategic processes are significantly related to the combination of quality and innovation as the strategic goal (F = 88.52, p < 0.001), quality as the strategic goal (F = 80.06, p < .001) and innovation as the strategic goal (F = 40.64, P < .001). Thus, Hypothesis la is supported. 86 Journal o./Business Strategies Vol. 22, No.2 Table 1 Means, Standard Deviations and Correlations (n = 335) Std. Variables Mean Dev. 1 2 3 4 5 6 1. Multi-Level Strategy Processes 3.85 .58 1.00 2. Quality Strategic Goal 4.08 .60 .44*** 1.00 3. Innovation Strategic Goal 3.57 .73 .33*** .37*** 1.00 4. Combined Quality & Innovation 3.82 .55 .46*** .79*** .86*** 1.00 5. Financial Performance 3.14 .96 .08 .19*** .26*** .28*** 1.00 6. Environmental Dynamism 3.61 .77 .26*** .27*** .28*** .33*** .07 1.00 One-tailed test: * p < .05 , ** P < .01, *** p < .001 Table 2 Regression Model Estimates to Test Hypotheses 1 a Regression Model For Paths Modell: DV: Combined Quality & Innovation Goals IV: Constant IV: Multi-Level Strategy Processes R2 F Model 2: DV: Quality Strategic Goal IV: Constant IV: Multi-Level Strategy Processes R2 F Model 3: DV: Innovation Strategic Goal IV: Constant IV: Multi-Level Strategy Processes R2 F Estimates 2.14*** .44*** .21 88.52*** 2.33*** .46*** .19 80.06*** 1.96*** .42*** .11 40.64*** One-tailed test: * p < .05, ** P < .01, *** p < .001 Fall 2005 Tegarden et al.: Engagement of Employees 87 Hypothesis 1b states that the achievement of strategic goals mediate the relationship of multi-level strategy processes and financial pefformance. To investigate this relationship we employed the test for mediation developed by Feedman and Schatzkin (1992)1. This test is consistent with the recommendation for the investigation of a mediating relationship in strategy research outlined by Venkatraman, (1989). That is, mediation specifies the existence of a significant intervening mechanism between an antecedent variable and the consequent vari- able and is carried out within a path-analytic framework. Moreover, the mediator variable should account for a significant proportion of the relationship between the predictor and the criterion (Venkatraman, 1989, p. 429). We test Hypothesis I b with a series of regression models (See Table 3), contrasting three strategic goal mediators with the unadjusted regression (multi- level strategic processes on financial performance). First, we tested whether the combined strategic goals of quality and innovation mediated the relationship between multi-level strategic processes and performance (Model I v. Model 2). Subsequently, we test for the mediating eHect of the quality strategic goal (Model I v. Model 3) and then the innovation strategic goal (Modell v. Model 4). The results indicate a significant mediating relationship for the combined strategic goal (t = -5.12, P < .001), as well as a significant mediating relationship for the quality goal (t = -3.20, P < 0.001) and the innovation goal (t = -4.67, P < .001). This represents a complete mediation model, rather than partial mediation model, because the relationship between multi-level strategic processes and firm perfor- mance is not significant without the inclusion of strategic goals (Venkatraman, 1989). We find positive support for Hypothesis 1b with all three models, indicating support that strategic goals mediate the relationship between multi-level strategy processes and firm performance. Hypothesis 2 states that the relationship between multi-level processes and firm performance is stronger when the strategic goal is quality rather than when the strategic goal is innovation. Hypothesis 2 is tested using path analysis, a recommended statistical tool when examining theorized pathways (Asher, 1983). Path analysis uses the un standardized beta coefficients from the regression equa- tions in Table 4. Hypothesis 2 would be supported if the indirect effects using quality (Indirect Path Model A) are greater than those focusing on innovation (In- direct Path Model B). Hypothesis 2 was not supported as the indirect pathways of quality (.46 * .31 = .14) are less than the indirect effect pathways of innovation (.42 * .35 = .15). Moreover, the quality pathway accounts for 49% of the indirect effect and the innovation pathway accounts for 51 % of the effect (Trevino & Youngblood, 1990)2, further evidence that our hypothesis is not supported. We test Hypotheses 3a and 3b by comparing the regressions of multi-level strategy processes on the strategic goals of quality (Hypothesis 3a) and innova- tion (Hypothesis 3b) under conditions of high dynamism and low dynamism. Hy- pothesis 3a states that the strength of the relationship between multi-level stra- tegic processes and the strategic goal of quality is greater when environmental dynamism is high compared to when environmental dynamism is low. Responses 88 Journal ofBusiness Strategies Table 3 Regression Model Estimates Regression Model Estimates to Test Hypotheses Ib Vol. 22, No.2 Model I: Unadjusted Regression for MLSP on Financial Performance IV: Constant IV: Multi-Level Strategy Processes ('t') --~ --- --- -- Standard Error for MLSP Parameter Est. (0,.) R2 F 2.65*** .13 "- - ~ .- - ~ - .09 .01 1.90 Model 2: Adjusted for Combined Quality & Innovation Goals IV: Constant IV: Multi-Level Strategy Processes ('t) IV: Combined Quality & Innovation Goals . --~ --~ --- Standard Error for MLSP Parameter Est.(o,l) R2 F Mediation test (Model I v. Modell)! Model 3: Adjusted for Quality Strategic Goal IV: Constant IV: Multi-Level Strategy Processes ('1) IV: Quality Strategic Goal - - - - Standard Error for MLSP Parameter Est. (U'2) R2 F Mediation test (Model Iv. ModeI3)! Model 4: Adjusted for Innovation Strategic Goal IV: Constant IV: Multi-Level Strategy Processes (t) IV: Innovation Strategic Goal - - - - - - - - - - - - - Standard Error for MLSP Parameter Est. (a'3) R2 F Mediation test (Model I v. Model 4) I I Freedman and Schatzkin test: t N-2= T -r' 1.52*** -.11 .53*** .10 .08 14.18*** -5.12*** 1.93*** -.02 .31*** .10 .04 6.10*** -3.20*** 1.98*** -.02 .35*** .09 .07 12.02*** -4.65*** ~a 2 , + a 2 t' - 2al a f ~1- p2 !