Interaction of Environmental Uncertainty, Organizational Reputation
and Management Control in the Hiring Process in Professional Service
Firms
Name: Ruonan Xie
Student number: 11331097
Thesis supervisor: Ms H. Kloosterman Date: June.25th, 2017
Word count: 14,984
MSc Accountancy & Control, specialization Control
2
Statement of Originality
This document is written by student Ruonan Xie who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3 Abstract
Service sector has been playing a dominant role in economy; however, the field of management control system (MCS) design in professional service firm (PSF) is relatively less explored. The attributes of PSFs cause problems due to the high human capitals and difficulty of measuring service quality. Firm reputation can serve as a quality guarantee to lessen the ambiguity in measuring service output while extensive hiring process can help to reduce the negativity brought by environmental uncertainty. This study draws conclusion from survey conducted from the period 2015 to 2017. Regression analyses were conducted using 414 questionnaires of professionals working in different industries within a broader PSF setting. Consistent with prediction, empirical evidence shows support that higher level of environmental uncertainty leads to more extensive use of personnel control in the hiring process. On the other hand, however, empirical results provide no evidence of interaction effect of firm reputation on the relationship between environmental uncertainty and personnel control. The paper contributes to the literature of less explored MCS mechanisms by investigating the interaction of environmental uncertainty, reputation and hiring process in a larger PSF context.
Key words: management control system; professional service firm; reputation; personnel control; environmental uncertainty; contingency theory
4 Table of Contents
1. Introduction ... 6
2. Literature Review ... 9
2.1 Professional Service Firms (PSFs)... 9
2.2 Firm Reputation in PSFs ... 11
2.3 Management Control System (MCS) and Personnel Control ... 12
2.4 Environmental Uncertainty in MCS and HRM ... 14
2.5 Hypothesis Development ... 16
2.6 Hypotheses Operationalization ... 18
3. Research Methodology ... 19
3.1 Sample and Data Collection... 19
3.2 Survey Demographics ... 20 3.3 Variable Measurement ... 21 3.3.1 Independent Variable... 21 3.3.2 Moderating Variable ... 23 3.3.3 Dependent Variable ... 25 3.3.4 Control Variables ... 27
3.4 Hypotheses Testing Models ... 29
4. Results ... 31 4.1 Descriptive Statistics ... 31 4.2 Main Findings ... 35 4.2.1 Hypothesis Testing H1 ... 35 4.2.2 Hypothesis Testing H2 ... 38 5. Conclusion ... 43 5.1 Discussion ... 43 5.2 Limitations ... 45
5.3 Future Research Directions ... 46
References ... 48
5 List of Tables and Figures
Table 1: Descriptive Statistics ... 21
Table 2: Factor analysis – Environment uncertainty ... 23
Table 3: Factor analysis – Reputation ... 24
Table 4: Factor analysis – Hiring in personnel control ... 26
Table 5: Descriptive Statistics ... 33
Table 6: Correlation Matrix ... 34
Table 7: Regression results of model 1a ... 35
Table 8: Regression results of model 1b ... 37
Table 9: Result of Mann Whitney U test ... 38
Table 10: Regression results of model 2a ... 39
Table 11: Regression results of model 2b ... 40
Figure 1: Overview of the hypotheses ... 18
6
1. Introduction
According to IBISWorld’s report published in 2014, service sector alone added to $2.5 trillion dollars of revenue in 2013. In a world of increasing growth of professional services in economy (Goodale et al., 2008), however, a large body of research pay more attention to the management controls in manufacture sector rather than service sector (Shields, 1997). The definition of the term ‘service’ in professional service firm (PSF) encompasses a vast diverse service group, from accounting firms and consulting firms to software development firms and health care institutions, just to name a few. The broad range of professional service makes the service sector harder to investigate, thus the subject of professional service firms and management control system (MCS) design in a broader context is relatively unexplored (von Nordenflycht, 2010). Additionally, the four characteristics of services – intangibility, variability, inseparability and perishability – contribute to the difficulty of measuring the output of services in PSFs (Reichheld and Sasser, 1990), resulting in less study in professional service research field. Together, the broad scope of industries and the characteristics of service lead to a knowledge gap in studying of professional service firms. This paper aims to study MCS in a boarder context of PSFs in respond to the call for further research in the field of MCS design and PSFs (Chenhall, 2003).
In contingency literature, environmental uncertainty is a contingency variable and management control systems can be applied to reduce environmental uncertainty (Chenhall, 2003). Rastogi (2003) found that when there is high environmental uncertainty, firms are more inclined to set organizational strategies and controls to reduce the impact caused by unfavorable conditions. When firm face the environmental uncertain condition, Kren and Kerr (1993) found that this uncertainty calls for additional investment in MCS, and investment in MCS can be taken in the form of an increase in action controls. Herremans et al. (2011) did research on influence on result control in knowledge intensive firms and found that in high environmental uncertain situation firm focuses more on result controls. The papers mentioned above, however, have not yet investigated the relationship of environmental uncertainty and MCS from an input perspective, for instance, the personnel control. This lack of study is consistent with contemporary studies on MCS that researchers put more focus on bureaucratic mechanisms such as action controls and result controls (Ouchi, 1979; Jaeger and Baliga, 1985). However, professional service firms enjoy high human capitals and a professional
7
workforce which differ from non-PSFs (von Nordenflycht, 2010). Perrow (1986) proposed that in environmental uncertain condition, firms rely on a professional workforce will use professional control, similar to personnel control, to offset the uncertainty. Lippman and Rumelt (1982) stated that human capital is perceived as hard to reproduce because it is scarce and it owns specialized knowledge, thus it can serve as a sustainable advantage for PSFs if qualified people are recruited. When organizations find it difficult to align incentives by using output controls under environmental uncertain condition, it might be an effective alternative to align preferences through hiring process (Merchant, 1985; Prendergast, 2008). Yet no empirical result is provided to confirm that environmental uncertainty will result in more extensive hiring. Hence, it is interesting to investigate whether PSFs use more management control from an input perspective in uncertain environmental conditions.
The notion of corporate reputation has been receiving more attention from the management as well as stakeholders around the world (Fombrun, 2007). While the information asymmetry enlarges the service output ambiguity, reputation can reduce the impact brought by environmental uncertainty since reputation is regarded as a guarantee of service quality for customers (Greenwood et al., 2005). Other than regarded as a proof, reputation can serve as success factors for PSFs since reputable firms attract more qualified employees in hiring process (Cable and Turban, 2001). Firm reputation is likely to interact with environmental uncertainty and hiring process; the possible interaction is worth investigating in the PSF setting.
This paper contributes to study of MCS and PSF by using survey approach. Previous researchers analyzed the relationship of MCS and PSF by applying case studies and not public available database, for example a study on outsourcing relationship between two organizations (Langfield-Smith and Smith, 2003), and white collar incentives in US tech-based firms using a private database (Baik, 2016). Using case study approach and private database, however, limit the research scope to a single industry or single firm. For the service sector, it covers a wide range of industries and a case study only sheds light on a small segment of the service sector, making it hard to generalize to other settings. Thus, it is appealing to use survey approach to reach out to wider respondents and broader industries.
In this study, survey sample of 414 professionals working in PSFs from wide range of industries will be examined. The focus will be on environmental uncertainty and its effect on hiring process of MCS
8
design in the PSF context. Additionally, firm reputation will be introduced as a moderator and examined on the relationship between environmental uncertainty and personnel control. I believe this study of the hiring process of professional service firms in an uncertain environment and reputation as a moderator will bridge the knowledge gap in the study of PSF. This study seeks to answer the following research questions:
1. How does environmental uncertainty influence the hiring process in PSFs?
2. How does firm’s reputation play a role in hiring process in PSFs when there is environmental uncertainty?
The remainder of this paper is structured as followings. Literature review on PSFs, uncertainty, reputation and personnel control will be discussed and then hypotheses proposed in the next section. The third section addresses research methodology by providing with survey demographics and variable measurement. The results of linear regression analysis to test the hypotheses will be presented in the fourth section. Lastly, this paper will present conclusion and discussion and discuss the possible future research directions.
9
2. Literature Review
This section is structured as follows. First professional service firm will be illustrated. Then the importance of reputation in PSFs will be highlighted. Next, the framework of management control system will be summarized and personnel control of MCS will be zoomed in. Thereafter, environmental uncertainty and contingency theory will be introduced to better understand the conditions PSFs face. In the last part of this section the structure of theoretical frameworks will lead to the hypotheses proposed.
2.1 Professional Service Firms (PSFs)
What is Professional service firm (PSF)? To answer this question, it is important to classify what a professional service is. Professional service can be defined as actions that are intangible; actions one party provides to another party will not result in the change of ownership (Kolter, 1994). Other literature defined professional service in another way by asserting that service is not the same as supplying goods but rather to provide solutions to problems (Gadrey et al., 1995) while more recently von Nordenflychit (2010) argued that there is no universal definition of professional service from the literature.
From a broad perspective, PSF is a particular type of service firms it shares the characteristics that service firms have. Attributes of service firms give more insights into studying PSFs. Reichheld and Sasser (1990) concluded four distinct characteristics of services: intangibility, perishability, inseparability and variability. The outputs of service firms, as compared to non-service firms, are intangible. Due to the intangible characteristic of service firms, customers find it hard to recognize the difference of competence and quality of the services. The second characteristic service firms share with PSFs is perishability, which means that services cannot be stored as inventory and those not consumed are gone and cannot be recovered. For the inseparable characteristic, it implies that customer is the input of service, thus service or product offered by service firms cannot be separated from customer. The fourth characteristic variability refers to the outputs of service are usually not the same since the activities carried out can be very different, thus it is hard to set stand criteria to measure the results.
10
The characteristics of service firms provide academic support to further define the characteristics of the particular type of service firms - professional service firms. Von Nordenflychit (2010) defines three attributes of PSFs as knowledge intensity, low capital intensity and professional workforce. The first one is knowledge intensity. PSFs are knowledge intensive firms since outputs of professionals working in PSFs encompass specific knowledge. PSF possesses high educated human capitals and this type of organizations is often referred to as ‘intellect industry’. The outputs of this type of firms rely heavily on knowledge input, which is embedded into services (Scott, 1998). Thus, for PSF, the intellectual input is important because of high possession of human capitals. The second one is low capital intensity, which refers to tangible and intangible assets which are non-human assets in PSFs. Assets such as facilities and inventories are not largely applied in service production in PSFs and this results in decreasing demand for investments and thus provides related opportunities for professional service firms since the needs to protect investors are decreasing. The third attribute of PSF is concluded as the professional workforce. Two features can be drawn from the professional workforce characteristic; professionals own the knowledge; professionals create particular norms to define their code of ethics and to define proper behaviors at work.
The attributes of PSFs mentioned above can lead to problems, however. Auzair and Langfield-Smith (2005) argue that the attributes of professional service firms create challenge for firms such as the needs to attract and retain both customers and employees, need for autonomy within the organization and reliance on informal controls. Von Nordenflycht (2010) concludes two problems in his paper. The first one is that talents are hard to retain in PSFs. On the one hand, professionals own the knowledge and the skill sets they have is scarce in the market. Since knowledge is transferrable and employees cannot be stored and are mobile, professionals thus enjoy stronger bargaining power with PSFs. On the other hand, the intangibility and ambiguity attributes of service outputs made customers rely on the employees from the professional workforce to deliver satisfying outcomes. The bargaining power of professionals exposes PSF to losing talents at any moment. The second one is that quality of services is difficult for customers to evaluate. The evaluation problem lasts even after the service is delivered. This is not uncommon since customers do not know whether services provided are the cause of certain subsequent consequences. The author concludes the problem as ‘opaque quality’ of professional services.
11 2.2 Firm Reputation in PSFs
The potential problems of losing knowledgeable employees and the ambiguous service quality create challenges for PSFs. However, previous literature suggested that the problems can be alleviated by firm reputation. Von Nordenflycht (2010) stated that firms can design management control mechanisms or pay attention to mechanisms such as firm reputation to handle the problems result from ambiguous service quality. Greenwood et al. (2005) mention that because customers are dependent on professional workforce of PSFs, firm reputation serves as proof of competence of workforce and thus a social guarantee of service quality.
Building firm reputation as a way to alleviate problems result from PSF attributes leads to the discussion of what reputation is. According to Fombrun (1996), reputation is ‘a perceptual representation of a company’s past actions and future prospects that describes the firm’s overall appeal to all of its key constituents when compared with other leading rivals.’ From a global perspective, the concept of corporate reputation has been receiving more attention from the management as well as stakeholders. In another Fombrun’s paper (2007) he stated that reputation reflects firms’ historical performance, and reputation could be utilized to forecast performance and actions conducted in the future.
Due to the high human capital and output ambiguity, firm reputation is considered especially crucial for PSF (Greenwood et al., 2005). Additionally, firm reputation can bring professional service firms benefits. Firstly, for customers, reputation plays a signaling effect of the services PSFs provide and helps to decrease customers’ purchase risk. Because of the intangibility and ambiguity of service outputs, it is difficult for customers to judge a PSF’s service quality or compared to other PSFs based on the service quality. Fombrun (2000) mentions in his paper that reputation can generate favorable public opinion and create a business-friendly environment for PSFs. Podolny (1993) is convinced that reputation is an important indicator to attract customers for PSF because reputable firm signals a guarantee of service. Paper from Rao et al. (2001) further linked firm reputation to uncertain environment by stating that in environmental uncertain condition, firm reputation can serve as social guarantee to signal the quality PSF provides to customers.
Secondly, reputation helps employers to attract talents and future employees to seek for firms they want to stay with. For firms, reputation can serve as a sustainable competitive advantage for reputable PSFs. Cable and Turban (2001) suggest that a firm's reputation has an impact on hiring
12
qualified employees and can be a critical success factor for organizations. Human capital is the most valuable intangible resource in professional service firms and reputation enables firms the ability to attract high quality candidates and thus bring benefits to PSFs (Greenwood et al., 2005). For employees, firm reputation helps them find jobs and companies they prefer, which in turn benefits professional service firms by obtaining qualified professionals. Firm reputation shows its value through the signaling effect when observing and measuring outputs are typically difficult (Hirshleifer, Hsu and Li, 2013). Job seekers turn to firm reputation as a signaling indicator in the hiring market (Kreps and Wilson, 1982) since output of PSF is characterized as intangible and opaque.
Thirdly, from a long-run perspective, reputation creates more profits for firms. Reputable PSF can charge more service fees because their brand recognition is high (Beatty 1989), thereby reputation helps PSFs achieve better financial performance (Fombrun et al., 2000). Wilson (1985) proposed that a desired reputation can bring excess returns for firms because the competitive advantages reputations bring can inhibit the mobility of competitors. In addition, because of the quality of service in PSFs is opaque, customers tend to stay with the current service providers they have experience with and are reluctant to change to other PSFs since they are uncertain about the service quality of other service providers (Greenwood et al., 2005). Therefore, reputation helps firms achieve more profits compared with firms with the same order of talents but are less reputable.
Summarizing, firm reputation is an important influencing factor to investigate in PSF research field because 1) customers see reputation as a proof of good service quality; 2) reputation helps firms attract more talented employees and help employees seek for companies to stay with; 3) profits firms in the long-term.
2.3 Management Control System (MCS) and Personnel Control
Management control refers to the processes by which management ensures that employees carry out the firm’s objectives and strategies. Management control systems are applied to help in achieving desired behaviors and outcomes in organizations (Simons, 1994; Chenhall, 2003). Several widely-used management control system frameworks can be found in management accounting literature such as Ouchi (1979), Simon’s four levers of controls (1994), Ferreira and Otley (2009), and the most recent one from Merchant and van der Stede (2012). These MCS frameworks are not
13
completely independent of each other; together they provide a clear and broad outline in better understanding of how management control systems work.
Merchant (2012) developed a framework which describes three different types of MCS mechanisms: social controls, action controls and result controls. In his framework social controls can be further divided into cultural control and personnel control. The personnel control is fundamentally the same as input control or clan control as in Ouchi’s paper (1979) but Merchant elaborates social control in a more detailed way by mentioning selection, placement, and training of employees in personnel control. As for cultural control, it refers to code of conduct, shared values and beliefs for employees; the use of social controls can serve action controls and result controls better in the MCS mechanisms. Action controls can be associated with bureaucratic mechanisms in Ouchi’s framework (1979), focusing on prescribing and controlling behavior through monitoring activities. Employees are given clear protocols to follow in order to achieve targets. Action controls are applied to ensure employees’ behaviors are aligned with organizational objectives when they conduct tasks. The last MCS mechanism is result control, similar to Ouchi’s market mechanism (1979), which mainly looks at the performance measures and incentive systems. Result controls set targets for employees and measure employees’ output through performance indicators; incentive systems are used to reward employees’ performance when targets are reached.
Personnel control is worthy of more attention in professional service firms. First of all, other than non professional service firms, PSFs have characteristics that other firms do not have, which are highly knowledge intensive, low non-human assets employed and possession of professional workforce. For PSF, the most critical resource is their employees (Hitt et al., 2001) and labor intensity is usually higher than capital intensity in PSFs. However, this brings high mobility to PSFs due to the higher bargaining power employees enjoy (von Nordenflychit, 2010). To deliver professional services, complex knowledge and personal judgment are required (Larsson and Bowen, 1989). The uniqueness of professional service has made PSFs dependent on its professional workforce. Since PSFs need more talented employees, the input of PSFs, personnel controls are needed to secure a high quality labor. Therefore, selection in personnel control becomes more important for PSFs as compared to non-PSFs. Secondly, it appears that actions controls and result controls can be difficult to apply and therefore personnel controls become more favorable in PSF. The characteristics of service are that service is intangible and variable (Reichheld and Sasser, 1990),
14
therefore making it harder to set standards to evaluate the quality service output. Studies from Merchant (2012) and Brivot (2011) argue that when output is hard to measure, control mechanisms that are more formal, such as action controls and result controls, tend to be less effective and even counter-productive. Hence, more input controls, such as personnel controls are favorable when action controls and result controls are less effective. Last but not least, the hiring process, as one component of personnel control, interacts with firm’s reputation in PSF setting. Jones (1996) stated that hiring would be influenced by firm reputation. Other academic papers (Becker and Gerhart, 1996; Dess and Shaw, 2001) proposed that reputable companies put more focus on obtaining better quality of professionals.
Based on the highly professional workforce of PSF and ambiguous performance measures and outputs, control mechanisms such as action controls and results controls are less effective. Firm’s reputation as discussed in previous sections, interacts with the hiring process of personnel control in professional service firms. Therefore, the personnel control deserves more focus in PSF than non-PSF. This paper chooses hiring process of personnel control from the framework of Merchant and van der Stede (2012) to look into the relationship of MCS and PSF.
2.4 Environmental Uncertainty in MCS and HRM
In management control, contingency theory holds that no universally best management control system can be found to apply to every situation and organization (Burns and Stalker, 1961). Contingency theory claims that control systems must be aligned with organizational characteristics (Fisher, 1995). For PSFs, if the designed management control system aligns with organizational objectives, it can lead to better performance (King and Clarkson, 2015). Environment has been explored frequently as influencing factor in organizational practices (Child, 1972). In more recent research, Chenhall (2003) confirmed that environment is one of the most frequently discussed variables in contingency-based research and linked environment to uncertainty as a contingency variable. Uncertainty encompasses two aspects - environment and technology in management control research.
Various definitions can be found on ‘uncertainty’ in management accounting literature. Argote (1982) describes uncertainty as ‘the absence of complete information about an organizational phenomenon,
15
which in turn leads to an inability to predict its outcome’. Chenhall (2003) states that uncertainty raises from a lack of information and uncertainty can lead to difficulty in making contingency plans. More recent research defines uncertainty as organizations having difficulty in predicting the future because of the dynamic conditions and incomplete information (Germain et al., 2008). Since this study concentrates on discussion of the environmental aspect of uncertainty, I define environmental uncertainty as the dynamic change of industrial intensity, increasing the difficulty of translating actions into desire output, therefore intensifies the difficulty of output prediction and measurement. In MCS research field, previous research has shown that environmental uncertainty has an impact on MCS design and organizational outcomes. The link between environmental uncertainty and contingency theory can be seen from Burns and Stalker’s paper (1961). They investigated the effects of environmental uncertainty on organization structure in a study of Scottish defense electronics industry and found that regardless of organizational structures, organizations respond to both low environmental uncertainty condition and high environmental uncertainty condition effectively. Their paper is perhaps one of the earliest literatures that describe the link between environmental uncertainty and contingency in management control. More recent research showed that when firms are operating in an uncertain environment, the use of traditional MCS is not effective and this could lead to undesired decision making and consequently undesired outcomes (Eldridge et al., 2013). Contingency theory encourages designing appropriate MCS based on contingency factors (Chenall, 2003). Given the fact that environmental uncertainty is a contingent variable, MCS design in professional service firms should take environmental uncertainty into consideration. Perrow (1986) used the concept of professional control, similar to Merchant’s personnel control mechanism, to stress that professionals rely more on self-control and social controls. Perrow stated that professional control can be used to cope with uncertain conditions for firms which need expertise to complete tasks. When information is absent or when desirable performance is not clear, action controls and result controls appear to be less effective in these conditions (Brivot, 2011). Chenhall (2003) stated that environmental uncertainty makes it harder for employees to understand how to conduct tasks and turn actions into favorable outputs. Employees are the most critical resource and it adds value to firms (Barney et al., 2011), therefore it is important for PSFs to recruit qualified employees who have the knowledge and are flexible in tasks. Hence, from MCS research findings, personnel control is likely to be an alternative for organizations in uncertain environment conditions.
16
Support of using more personnel control to combat environmental uncertain conditions can also be found in human resource management (HRM) papers. Discussion of the importance of human capital to overcome the impact brought by environmental uncertainty can be found in HRM literature. When firms face a high environmental uncertain condition, firms will try to find a buffer, for example through ensuring the input resources, to offset the negative influence (Ghosh et al., 2009). Barney (2011) is convinced of the importance of people factor as company resource by stating that human capital can add value to firms and consequently generate a firm's core competence. Ghosh et al. (2009) stated that employees with specified knowledge are more flexible at work and less influenced by environmental uncertainty because these employees can find solutions to problems effectively when compared with other employees who do not require specified knowledge at work. Baron and Kreps (1999) are convinced that firms can transfer the pressure resulting from environmental uncertainty to employees as they are the ones who conduct the task and that the skill sets and expertise employees have help firms to stay efficient. Consequently the educated professionals need fewer action controls such as guidance and monitoring activities from employees when they face environmental uncertainty. Therefore, inputs of professional services, which are the service professionals, are valuable for PSFs when affected by uncertain environment. The professional workforce can deal with environmental uncertainty with more flexibility and therefore help firms to offset unfavorable situations. This results in firms' demand to hire more talents as buffer to lower the undesired impact of environmental uncertainty.
In light of the above mentioned literature from MCS and HRM literature, contingency theory and professional human capitals can thus be served as ground theories in explaining that need to use more extensive hiring processes accordingly in environmental uncertain conditions.
2.5 Hypothesis Development
Employees and customers are important inputs for PSFs (Auzair and Langfield-Smith, 2005). For professional service firms, the knowledge intensive attribute and the professional workforce attribute cause the possibility of losing talents and the difficulty in telling service quality (von Nordenflychit, 2010). The nature of PSF exposes PSF to uncertain environment. As a contingency variable, uncertainty encompasses environmental and technological dimensions (Chenall, 2003). Drawing
17
from literature, environmental uncertainty is an important influencing factor of MCS design. Literature shows that environmental uncertainty influences management outcomes and suggest personnel control can be used to reduce the environmental uncertainty. Ouchi (1979) argues that personnel control might be the most appropriate strategy under conditions of incomplete information about the task and ambiguous standards of desirable performance. Brivot (2011) found that when information is absent or when outputs are hard to measure, action controls and result controls appear to be less effective. Ghosh et al. (2009) suggested that environmental uncertainty triggers firms to apply more personnel control as a buffer. Despite the importance of hiring process in management control systems (Campbell, 2012), past studies focused more on bureaucratic mechanisms such as action controls and result controls in organizational controls thereby personnel control has not yet been addressed. Based on the literature review, a possible relationship between environmental uncertainty and hiring process from MCS can be expected, leading to the first hypothesis:
H1: Firms facing higher environmental uncertain conditions use more extensive personnel controls than firms that are confronted with less environmental uncertain conditions.
The high human capital nature of professional service firms makes employees the most critical resource for firms (Hitt et al., 2001). Besides the employee input component, service output ambiguity and intangibility is another important attribute of PSF. These two characteristics of professional service firms, however, can lead to problems (von Nordenflychit, 2010). Firm reputation helps to deal with the PSF problems for the signaling effect reputation plays. Empirical results of a survey of 100 firms found a positive effect reputation has on performance of PSFs (Greenwood et al., 2005). Reputation provides benefits to the stakeholders in PSFs. Customers, another important input of PSF, see reputation as a proof of good service quality; employers use this signaling effect to attract more future employees while future employees turn to firm reputation as signal of ‘goodness’ to seek for companies to stay with (Wilson, 1985; Rao et al., 2001; Greenwood et al., 2005). As mention in previous sections, the nature of PSF exposes PSF to a more uncertain environment. The quality of outputs is even harder to measure in uncertain environment, and this leads more importance of developing and maintaining firm reputation (Greenword et al., 2005). Reo et al. (2001) linked the reputation to the environmental uncertainty by stating that in environmental uncertain
18
condition, firm reputation can serve as social guarantee to signal the quality PSF provides to customers. Reputation is not only important for PSF because of the uncertain environment PSF is in but also important because it impacts hiring in PSF. Jones (1996) stated that hiring would be influenced by reputation. Other academic papers (Becker and Gerhart, 1996; Dess and Shaw, 2001) proposed that reputable companies put more focus on obtaining better quality of professionals. Therefore, reputation is valued in PSF since the service output is hard to measure and reputation impacts hiring process. Although reputation is especially important for PSFs, no study of reputation can be found in the PSF literature. The discussion of previous academic papers leads to a logical proposal that reputable PSFs invest more in obtaining qualified professionals in uncertain environment. Therefore, this leads to the second hypothesis:
H2: The environmental uncertainty conditions firms are facing together with firm’s reputation will lead to more extensive use of personnel controls.
2.6 Hypotheses Operationalization
In the previous section, hypotheses are proposed. Environmental uncertainty is the independent variable; personnel control is the dependent variable, and firm reputation as the moderator. The dependent variable personnel control is measured by hiring process. This paper uses hiring process as a lens to look into the MCS design. Below is the figure represents the two hypotheses:
19
3. Research Methodology
3.1 Sample and Data Collection
This study applies an empirical approach to test the hypotheses of the interaction of environmental uncertainty, reputation and the hiring process. Data is collected by sending out questionnaires online through Qualtrics as part of a larger PSF survey project by Faculty of Economics and Business at University of Amsterdam. Acquiring the data needed for this research topic is difficult due to the low availability in collecting field data. Professional service firms cover a wide range of industries and this has made collecting surveys through personal efforts unrealistic. Thus, joining a survey project can help to solve the problems by using a joint dataset. Since this questionnaire is designed for a target group of people, it cannot be spread out randomly. So participants in this survey project help to contribute to reliable survey information through their connections.
The PSF project mainly looks into factors that influence professional service firms and their management control systems on an individual level. The questionnaire was designed to cover broad range of topics, for example performance, tight and loose control and four types of controls. The survey was conducted online from 2015 to 2017 with 5 survey collection deadlines throughout the time period.
The design of this questionnaire is for individuals who work in PSFs, thus a few criteria must be met to be eligible to fill in the survey. First, the respondent should be working in a professional service firm, thus non-profit firms, such as government organizations are not taken into consideration. Secondly, individuals should be working in medium to large size professional service firms (more than 50 employees), regardless of their nationalities and working locations. Thirdly, respondents should have at least three years of working experience but a maximum of ten years. This criterion is designed to ensure that respondents have acquire the necessary experience to perform their jobs since employees in the learning phase of their jobs are often subject to various control systems and they might respond differently to controls. As for setting a ceiling for working experience this is because the survey is aimed to analyze how individuals experience MCS but rather than individuals who design MCS. When individuals have been working longer it is probable they get to a higher level that will be part of MCS design in larger organizations. Lastly, the questionnaire is written in English,
20
thus surveyed individuals are expected to have a good command of business English in order to fill out the surveys.
3.2 Survey Demographics
Questionnaires were sent out online to individuals working in medium and large PSFs worldwide. Two pre-tests were carried out to ensure the reliability of the questionnaire. The first pre-test was conducted after the survey design and 20 people were asked to sort the items provided with two sheets of paper. The first sheet of paper included definitions for the eight control constructs, which were implicit and explicit results control, behavior control and personnel control. Then subjects were provided with the second paper with 52 statements and were asked to match definition to each statement. The items that were sorted wrong most were removed from the questionnaire. The second pre-test was conducted with another group of 20 individuals to do the entire survey online to assess the quality. The questionnaire was then improved with the feedback provided.
The PSF survey project questionnaire collection was closed in February 2017. After the closure of the survey collection, the total recorded responses amounted to 612. Through a data cleaning process, surveys that are not fully completed, respondents lacking working experience, and some other outliers such as did not read definitions and remarks were excluded from the analysis. Of the total 612 entries, 198 entries were deducted from the dataset and this gives total entries of 414, amounting to a usable rate of 67.6%. Among the valid questionnaires, the occupation of respondents spread in different fields. Most respondents (16%) work as accountants, 33 respondents (8%) work in physician practices; following by consulting management (29 respondents, 7%), consulting IT (25 respondents, 6%) and engineering (25 respondents, 6%). Eighteen percent of survey respondents give ‘other’ as they did not find relevant choices of occupation matches with their jobs. For the gender of survey respondents, number of female respondent counts for 147 (35.6%), while number of male respondent counts for 266 (64.4%). As for highest education level, three scales are designed to classify respondents’ level of education with 1 as the lowest and 3 as the highest. For the education level, 40.8% (169 individuals) of respondents have obtained Bachelor’s degree, 44.4% (184 individuals) of respondents with Master degree, and the rest 14.7% (61 individuals) achieve PhD or other equivalent degree. From our respondents’ education level, approximately 60%
21
individuals have obtained at least a master degree, which are the subjects we are looking for as PSF is known for high human intensity. In the years of experience in field and organization, descriptive statistics have shown that on average respondents have 7.47 years in a specific field and approximately 6.21 years in an organization; with a median of 7 years and 6 years respectively. For key demographic information, a statistic summary of respondents can be seen in Table 1.
Table 1: Descriptive Statistics
Items # Sample Min Max Mean Median Std.
deviation Age 414 22 63 35.72 34.00 8.586 Education* 414 1 3 1.74 2.00 .699 Experience in field** 414 1 11 7.47 7.00 2.908 Experience in organization** 413 1 11 6.21 6.00 3.153 *
. Bachelor degree or lower=1; Master degree=2; PhD or other professional doctorate degree=3 **
. Less than 1 year=1, 1 year=2, … 10 or more=11
3.3 Variable Measurement
From the previous section, after survey screening, 414 observations will be processed and analyzed. According to Hair et al. (1995), a survey with more than a hundred samples should be sufficient for analysis. For this study, although the questionnaire has passed two pre-tests, whether items are grouped as one factor or a defined number of factors are unknown. Thus, in this section, independent, mediating and dependent variables will be introduced, following by exploratory factor analysis (EFA) and reliability analysis for each variable before going to the next step analysis.
3.3.1 Independent Variable
The independent variable in this paper is the environmental uncertainty. For this variable originally 6 items were designed and questions on uncertainty were asked on three aspects: intensity, innovation and predictability of industry. The constructs are designed to ask respondents about predictability of
22
the business environment the professional service firm faces on a five-point Likert Scale. Questions involved intensity of price competition and competition for talents, for instance. For the intensity items designed, 1 is defined as of negligible intensity whereas 5 as extremely intense. A higher score refers to a more dynamic and uncertain external environment; on the contrary, a lower score represents for an external environment with less uncertainty.
The descriptions were adapted from several academic papers. Gordon and Narayanan (1984) analyzed the correlation between perceived environmental uncertainty and the degree of organic organization structures. In Child's paper (1972), he defined uncertainty as one of the two characteristics of environment, which can be expressed as "environmental variability", or in simpler terms, the degree of change. He proposed that to measure this construct, market characteristics such as competition in the market should be taken into consideration. In the survey, we use a subjective measure from our respondents by using perceived uncertainty. Applying perceived uncertainty instead of objective measures is supported by previous papers. Leifer and Huber (1977) are convinced that to study environment factors, what people think provides more insights into organizational behavior. They claim that perceptual information is more relevant than archival data. The Keyser-Meyer Olkin of the environmental uncertainty variable is .657, and the Bartlett’s test of Sphericity shows P value lower than .001, therefore the data is suitable for factor analysis. From the factor analysis a screen plot is drawn and two components can be found from six items. Two components explained 56.13% for the variance. Items have to load well when compared with the benchmark of .32 recommended by Tabachnick and Fidell (2001). For the innovativeness item, stated as “How many new products and/or services have been marketed during the past 5 years by your industry” the loading factor (.302) is lower than the benchmark. Therefore, this item is removed from the analysis. In addition, for the environmental uncertainty variable, the questionnaire was originally constructed to combine three components. However, from the first factor analysis, results show that innovativeness item and two other items referring to predictability of the industry are grouped together as one component. A reliability test is carried out to see if the remaining five items are valid, and this gives value of Cronbach’s alpha of .594. While if the three items (included the deleted one) grouped together are removed, factor analysis shows a result of one component with loading factors all above .60 (.855, .624, .815), which indicates items remained load well. When another reliability test is conducted, a much higher Cronbach’s alpha can be seen ( =.657 . The
23
variance explained is 59.49%. Given the above analysis, two items stated as “How could you describe the tastes and preferences of your clients” and “How could you classify the market activities of other firms in the industry” will no longer be considered in the variable analysis. According to Nunnally (1978), Cronbach’s alpha value of around .50 to .60 is considered to be acceptable for exploratory research; the Cronbach’s alpha result for the second loading analysis is above the lowest acceptable level for this study. The values of the items will be summated into a composite score and also a mean score for further analysis.
Table 2: Factor analysis – Environment uncertainty
First Loadings Analysis Second Loadings Analysis
Items Component Component
1
1 2
Price competition .847 .855
Competition for manpower .561 .624
Bidding for new contracts/clients .836 .815
New products/services numbers .302
Predictability of tastes and preferences .689
Predictability of market activities .503
Variance explained (%) Cronbach’s alpha 56.13% .59 59.49% .66 3.3.2 Moderating Variable
The moderating variable in this paper is reputation of a professional service firm. Four items were constructed for this variable on a five-point Likert Scale and questionnaire was designed to compose them in one general factor. Questions for this variable can be seen in Appendix. Individuals are
24
required to answer questions about perception of their companies. In this construct, a 5 indicates ‘strongly agree’ whereas 1 indicates ‘strongly disagree’. Thus, a higher score means respondents perceive their firms more reputable and vice versa.
The measurement for reputation is mainly adapted from Combs and Ketchen (1999). Jones (1996) thinks that hiring would be influenced by reputation, which means the perception of employees of whether the company is a suitable place to work for or not affects the selection procedure. Thus, reputation is applied as a moderating variable to test if it positively correlates with environmental uncertainty and thus contributes to more extensive use of personnel control.
The Keyser-Meyer Olkin and the Bertlett’s test of Sphericity for this variable are .750 and p < .000 respectively, which indicate reputation construct is suitable for factor analysis. In the next step of factor analysis, a screen plot is created and of the four items only one component is found and the total variance explained is 62.972%. This matches with the survey design, which aims to collect a general reputation perception of a firm.
To test the internal validity for this variable, reliability analysis is done in SPSS and this results in a Cronbach’s alpha of .799. This is above the upper limit of acceptability for exploratory research, which is usually approximately .70 (Nunnally, 1978). The data set passed the internal validity test and the scores for the four questions are computed as sum and as mean for next step analysis.
Table 3: Factor analysis – Reputation
Items Loadings
Perceived to provide good value for the price .872
Well respected in its field .701
Strong brand name recognition .862
Strong reputation for consistent quality and service .723
Variance explained (%) Cronbach’s alpha
62.97 .80
25 3.3.3 Dependent Variable
The dependent variable in the research is personnel control and this paper investigates the hiring process in the personnel control. The construct to measure personnel control a new construct developed from the definition of personnel control by Merchant and van der Stede (2012) and Ouchi (1979). For this construct, survey respondents need to answer 8 questions designed in a five-point Likert Scale on personnel control in their organizations. Questions asked all relate to the hiring process in respondent’s organizations. These questions are constructed with most questions designed to correspond with the rest, which means a higher score indicates more use of personnel control while a lower score indicates less extensive use of personnel control. There are two items, however, were designed to ask in an opposite way which means a lower score indicates more extensive use of personnel control and vice versa. The answers from these two items are reverse coded. For 5 which indicates ‘strongly agree’ and 1 indicates ‘strongly disagree’; they are coded as new variable in SPSS setting new value of ‘1’ for a ‘5’ while ‘5’ for a ‘1’. By recoding into new variables, the two reverse-coded items are aligned with the rest of the items in terms of degree of personnel control measured. The reason behind this is to see if respondents can do the survey and indicate answers in the same positive or negative degree without actually finding out the differences in constructs (Bryman, 2012). The two items are:
There seems to be little consistency in the type of professional that gets hired for my job. (Q3_7) The competence of employees within my job title varies greatly. (Q3_11)
The results for Keyser Meyer Olkin (.681) and the Bartlett’s test of Sphericity (p <.001) proves that data base is suitable for factor analysis. Three components are found from the screen plot and the variance explained is 62.25%. A follow up analysis was done to see if items load well with the three components. However, from the component matrix it can be seen only first five items group together as one component which represents for explicit personnel control items. While two reverse coded items share the same component being a new factor, these two items were originally designed to be grouped with two other implicit personnel control items. A reliability analysis is also conducted and result shows a Cronbach’s alpha of .666.
Another factor loading analysis is carried out to see the results if taken out the four not aligned items how to rest items load. Keyser Meyer Olkin (.733) and the Bartlett’s test of Sphericity (p <.001)
26
show it is suitable for factor analysis. This time the screen plot shows one component among the rest items and variable explained for 57.63%. Cronbach’s alpha verified the validity of .745, which improves the reliability of the personnel control variable and it is above the limit of acceptability for exploratory research which considered being around .50 to .60 (Nunnally, 1978). Factor analysis helps to make items and database more structured for next step analysis. Erroneous items which do not load with the same factor could be removed to improve reliability (Bryman, 2012). Based on the findings from factor analysis and reliability test, this study will proceed with four items from the personnel control construct from the second factor analysis. The other four items which are originally designed to load with one component are removed because they do not measure the same component as desired and cannot capture the implicit personnel control. After taking out those four items, the remaining items conducted an individual sampling adequacy test, also called MSA test. Results show the remaining constructs all have figures higher than .50 and can be processed to further research (Hair et al., 1995). Statistical results can be seen in Table 4. The scores of the four items on this dependent variable from the second loadings analysis are calculated and sum up into composite score.
Table 4: Factor analysis – Hiring in personnel control
First Loadings Analysis Second
Loadings Analysis
Items Component Component
1 2 3 1
Extensive hiring process .744 .826
Go through many steps to be hired .767 .854
Interview with several people .630 .730
Evaluation at hiring process .607 .602
27
Consistency in the type of professional .592
Same kind of education and training .662
Competence within job title .564
Variance explained (%) Cronbach’s alpha 66.25 .67 57.63 .75 3.3.4 Control Variables
Regression analysis needs to be conducted in order to test the hypotheses. Thus, it is important to consider variables that are not the subjects of this paper but might have impact on results of the analysis and control for them. In this section, several control variables will be set and taken into consideration in regression model. Two control variables are used in the following regression analysis.
3.3.4.1 Size
Organizational size is commonly used as control variable in management control system literature. King and Clarkson (2015) list size as control variable since size could possibly have a relationship with performance in MCS design. The paper from Chenhall (2003) concludes 6 contextual factors that can be studied by applying contingency theory, among one of the factors proposed is the company size. Therefore, size is considered as control variable given the possibility that it could have impact on personnel control discussed in this paper. This control variable is designed in the following two statements to analyze organizational size from business unit to company as a whole: How many people are employed by your entire company? (Q15)
How many people work in your organizational unit? (Q16)
The first item includes four choices which vary in numbers of employees within the organization as a whole (less than 100, 100 to 500, more than 500 but less than 5000, above 5000) and the categories are recorded from 1 to 4 scales with the increase in total employee numbers. The other item also
28
contains four categories vary from small amount to a large amount in organizational units. The answers are recorded from 1 to 4 in scales from less than 10 people in one unit to more than 100 people in unit. For sizes variables, the organizational size is referred to ‘ORG_SIZE’ and business unit size is referred to as ‘UNIT_SIZE’.
3.3.4.2 Firm structure and ownership
Results of survey sample of 120 managers in PSFs confirm the authors' prediction that organizational ownership type and MCS design are correlated and performance outcome is positively affected by the interplay between ownership and management control system design. Findings also conclude that organizational structure exerts impact on MCS design in PSFs (King and Clarkson, 2015). Based on their findings, a firm's structure and ownership type influence the hiring process since personnel control is part of MCS design. Thus, from the questionnaires, two items related to organizational structure and ownership are proposed, which can be seen in the following:
Which of the following best describes your job? (Q18)
Which of the following best describes the ownership type of your organization? (Q19)
For organizational structure question, respondents have to choose if the services they provide represent the primary service provided by their firms, and answers are recoded as 0 or 1 as dummy variable ‘STRUCT’. For another question related to ownership type, respondents have to choose from three types of ownership: partnership, owned by shareholders and investors, or non-profit organizations. Answers to this question are coded as two dummy variables to classify different types; they are expressed as ‘IN_ORG’ and ‘OUT_ORG’ to represent ownership types as partnership and owned by people from outside the organization own the firm, for instance, shareholders and investors.
3.3.4.3 Education level
Abernethy et al. (2004) found that the education level of individuals has a positive relationship with the level of trust. They used education level to examine if the degree of trust has impact on the management control system and found education level is linked the degree of trust and a higher level of trust between employees and organization, the more effective MCS is. Another reason for using this construct as control variable is based on the characteristics of PSFs. PSFs enjoy high human
29
capital intensity and professional workforce, which are linked to employees’ degree of education. Education level represents the knowledge one has and it is the building stone of professions. The education received by employees and potential new hires can possibly effect hiring processes and practices. The knowledge professionals own provide them better employment, better skills and higher level of job security. For professionals who are highly educated, mostly they do not want to be subject to formal control mechanisms (Goodale et al., 2008). Thus, PSFs might apply more informal controls such as personnel control and cultural control compared with other types of firms. This indicates that education level is an important factor for PSFs and should be added into the analysis as a control variable. One item is taken from the questionnaire asking respondents the highest education level they have achieved. The item is measured based on scales, where 1 represents ‘Bachelor degree’, 2 for ‘Master degree’, and 3 for ‘PhD and other professional doctorate degree’.
3.4 Hypotheses Testing Models
In literature review part, two hypotheses are proposed and the figure for hypotheses testing is shown. In this section, models for testing hypotheses will be described. To test them, multiple linear regression approach is applied. Model 1a consist of independent variable uncertainty, dependent variable hiring process while model 1b add various control variables; together two models are used to test the first hypothesis. Model 2a and 2b add the moderator reputation mentioned in last section, that is, independent variable, dependent variable, moderating variable in model 2a and control variables are added in model 2b to test the second hypothesis.
Hypothesis 1 states that firms face higher environmental uncertainty conditions use more extensive personnel controls. Environmental uncertainty is the independent variable and the composite scores are measured on ordinal scale. Personnel control here refers to as hiring process and the composite scores are also treated as an ordinal scale. Model 1a first does a regression analysis without all control variables to see the possible relationship between the environmental uncertainty and the hiring process. Model 1b adds all the control variables - size, firm structure, firm ownership and education level - into the formula to test the relationship between environmental uncertainty and hiring process in a more complex situation. The models are proposed as followings:
30
Model 1b:
Hypothesis 2 proposes that when firm reputation is added as a moderator, this will positively impact the link between environmental uncertainty and personnel control. That is, together with a high firm reputation, environmental uncertainty will lead to a higher degree of personnel control. The model is similar to model 1, while this time reputation is a moderator should also be considered in the formula. To further investigate moderating effect of firm reputation, the reputation variable is split into two groups: high reputation (REP) and less reputable firms. The aggregate mean split (4.21) is used as a cut-off since the items on reputation are in continuous scale. A higher figure than the mean is classified as more reputable firms and coded as ‘1’ in REP, ‘0’ if the standard is not met. Model 2a is designed to first test the moderating effect of firm reputation on environmental uncertainty and hiring process without all control variables, and model 2b refines model 2a by introducing all control variables size, firm structure, firm ownership and education level into regression model to test if the moderating effect still exists. This gives the regression models in the followings:
Model 2a:
Model 2b:
For the two models above, in the regression model stands for regression coefficients and stands for residual deviation when results are interpreted in the next section.
31
4. Results
4.1 Descriptive Statistics
Descriptive statistics of the survey can be seen in Table 5. A total sample size of 414 is observed and table reports the range of the corresponding variables, minimum and maximum scales, the mean and median and standard deviation. For different items there are different scores. For example, for organization size item, there are four choices thus the scores are set from 1 to 4, and the range is from 1 to 4. On average, respondents consider the environment they are in quite uncertain (3.53). The relatively high median score is consistent with the literature review that PSF is subject to uncertain environment due to the nature of PSF. As for company reputation, most of our respondents think highly of the organizations they are working for, thus we can see a higher mean and higher median compared with other variables (4.21 and 4.25 respectively). Additionally, the standard deviation for this reputation variable is low too; in general, respondents rate their companies high and they perceive the firms in good reputation. From the mean (2.94) and median (3.00) it can be perceived that a large percentage of respondents we surveyed work for medium to large companies. As for highest education received, the statistics indicate that a large amount of our respondents have obtained Bachelor degree and on average a Master degree (mean=1.74, median=2.00).
Table 6 presents the correlation matrix of independent variable, dependent variable and control variables. From the results of the correlation matrix, it can be seen that there is no correlation larger than .70 for any two variables. Therefore, the possibility of multicollinearity is believed to be low (Tabachnick and Fidell, 2001). The variable construct HIRING measured four items on regards of hiring procedures, UNCERT stands for environmental uncertainty, REP stands for moderator reputation, and moderating effect is measured based on the interaction variable UNCERT*REP. From the correlation we can see that environmental uncertainty is significantly positively correlated with dependent variable hiring procedures, in other words, personnel control (r =.104, p .05). The moderator reputation is significantly positively correlated with both dependent variable hiring process (r =.152, p .01) and independent variable environmental uncertainty (r =.130, p .01). This is consistent with the hypothesis that reputation has an influence on the relationship between hiring process and environmental uncertainty. The moderating effect of uncertainty and high reputation has
32
a significant positive correlation with hiring procedures and environmental uncertainty Control variables are listed under moderating effect UNCERT*REP. Both the organization size (ORG_SIZE) and unit size (UNIT_SIZE) are positively correlated with hiring procedures at a significant level r =.207, p .01 r =.102, p .01 . Both variables are also significantly positively correlated with moderator reputation and moderating variable. However, organization size and unit size do not show significant correlation with the independent variable environmental uncertainty. For organizational ownership type it can be seen that firms owned by individuals inside the organization (IN_ORG) is significantly positively correlated with the independent variable environmental uncertainty . The control variable level of education (EDU), on the other hand, shows a positive relationship with the hiring process and the result is significant . This is consistent with the findings that education level has a positive effect on MCS design by Abernethy (2004).
33 Table 5: Descriptive Statistics
N= 414 Min Max Mean Median S.D. Range
Dependent variable Hiring procedures Independent variable 1 5 3.27 3.25 .816 1 – 5 Environmental Uncertainty Moderator 1 5 3.53 3.67 .841 1 – 5 Reputation Control variables 1 5 4.21 4.25 .667 1 – 5 Organization size 1 4 2.94 3.00 1.087 1 – 4 Unit size 1 4 2.58 2.00 1.065 1 – 4 Structure 1 2 1.45 1.00 .498 1 – 2 Ownership 1 3 1.60 1.00 .702 1 – 3 Education level 1 3 1.74 2.00 .699 1 – 3
34 Table 6: Correlation Matrix
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (1) HIRING (2)UNCERT .104* (3)REP .152** .130** (4)UNCERT*REP .155** .314** .657** (5)ORG_SIZE .207** -.028 .181** .175** (6)UNIT_SIZE .102* .070 .132** .157** .561** (7)STRUCT -.014 .042 -.046 -.029 -.095 .099* (8)IN_ORG -.029 .211** .052 .080 -.249** -.115* .233** (9)OUT_ORG .043 -.020 .042 .027 .196** .103* -.308** -.671** (10)EDU .114* -.121* -.048 -.053 .131** .107* .214** -.051 -.154** *
. Correlation is significant at the 0.05 level (2-tailed). **
35 4.2 Main Findings
4.2.1 Hypothesis Testing H1
To test hypothesis 1, regression analysis is conducted to first test model 1a and then model 1b. Independent variable environmental uncertainty and dependent variable hiring process are analyzed in the regression model 1a: . Table 7 shows the regression results. The F value is 4.474 at a significant level (p < .05), meaning that this model is suitable for analysis. The value of R2 is .107 and adjusted R2 is .083, meaning that model 1a explains for 8.3% of the variation with the rest explained by other factors. From the regression model 1a environmental uncertainty and hiring process are significantly positively correlated ( =.134, t = 2.115, p .05). The result from model 1a supports the previous stated prediction of environmental uncertainty and hiring process. The results of this model indicate that without any control variable, there is a positive correlation between independent variable environmental uncertainty and hiring process.
Table 7: Regression results of model 1a
Model Unstandardized Coefficients Standardized
Coefficients
t Sig.
Beta Std. Error Beta
1a (Constant) 11.614 .691 16.808 .000 UNCERT .134 .063 .104 2.115 .035 R2 .107 Adjusted R2 .083 F-statistic 4.474* *
. Correlation is significant at the 0.05 level (2-tailed).
Another regression analysis is conducted using model 1b to further test the relationship : HIRING= 0 1UNCERT 2ORG_SIZE 3UNIT_SIZE 4STRUCT 5IN_ORG 6OUT_ORG
