Proactive behavior at Philips : relationship with job performance and moderating effect of training

(1)

Master Thesis

Proactive Behavior at Philips: Relationship with Job

Performance and Moderating Effect of Training

Student: Maiya Assanova (6350925)

Academic Supervisor: Dr. Stefan T. Mol

Second Reader: Dr. Corine T. Boon

Company Supervisor: Felice Valente

HRM - OB Business Studies

Faculty of Economics and Business

(2)

2

Abstract

This thesis aims to further knowledge on proactive employee behaviors in relation to the job performance. In particular, behaviors of individual innovation, strategic scanning, networking, and helping others are explored. Also, the thesis investigates the moderating influence of the number of trainings received on the proactive behaviors listed above. The interest in the effect of training was triggered by recent research which demonstrated that proactive behaviors are not merely contingent on personality: they can also be nurtured via appropriate organizational interventions. This quantitative study was performed within an internationally well-known company Philips. In particular, two studies were conducted in order to test the assumed relationships. Results of Study 1 empirically support the positive effect of proactive behaviors on job performance. The longitudinal Study 2 showed that proactive behaviors exert a significant mediation effect on job performance at any value of trainings. The positive moderating effect of training on proactive behavior was also supported. However, it was shown that less proactive employees benefit from training to a greater extent than more proactive employees.

(3)

3

Introduction

This research aims to investigate a range of proactive behaviors and the influence of HR training and development interventions on them. The first assumption is that expression of proactive behaviors has a positive effect on job performance. The second assumption is that the number of trainings received by individual employees has a positive effect on their proactive behaviors. The main goal of this thesis is to find ways to promote the development of a more flexible, innovative, and proactive workforce in a large and dynamic organization. Nowadays, organizations compete in a rapidly changing global market arena. Both researchers and practitioners have realized that the main source of competitive advantage for successful firms is their human capital (Crant, 2000; Parker, 2000; Frese & Fay, 2001; Grant & Ashford, 2008). Increasing external pressure on firms leads them to expect a different type of contribution from their employees. Today, employees are asked to be flexible, self-directed, and open-minded (Crant, 2000; Friedman, 2005). Such individuals proactively use their abilities and show personal initiative without requiring close supervision (Sonnentag, 2003). Also, researchers suggest that there is a trend in firms to reduce organizational input in individual career path development (Crant, 2000; Unsworth & Parker, 2003, Belschak & Den Hartog, 2010). Thus, employees are now expected to be proactive in their career management efforts as well.

Proactive behavior has been shown to enhance individual and organizational outcomes, such as personal innovation, learning and development, mental health as well as profits, entrepreneurial behavior, and overall organizational success (Unsworth & Parker, 2003; Stobbeleir et al., 2010; Griffin et al., 2010; Thomas et al., 2010). In the beginning, many researchers were skeptical of the value that the proactivity construct can add to already existing research domains of socialization, feedback seeking, issue selling, innovation, and personal initiative (Hatch, 1993; Crant, 2000; Parker, 2000; Thomas et al., 2010). To them, the concept seemed too vague and too broad. However, as the construct began to take form, the researchers realized the importance of proactivity. Crant (2000) acknowledged proactivity as a “high-leverage concept rather than just another management fad” (p. 435).

No matter how important, the proactivity construct has not yet been addressed comprehensively. Its antecedents and specific behaviors need further investigation and

(5)

5 validation, just as the HR interventions that could stimulate proactive behavior on the part of employees.

Theory and Hypotheses

In research on employee work orientation, there has been a major shift from the mentality of passively fulfilling duties to an active behavior within broad work roles (Frese & Fay, 2001). The following subsections address the implications of this shift and create theoretical and empirical ground for this thesis.

Definition of Proactive Behavior

Researchers define proactive behavior as a particular form of motivated behavior at work (Bateman & Crant, 1993). It is a “self-initiated and future-oriented action that aims to change and improve the situation and oneself” (Parker et al., 2006, p. 636). Thus, any self-started activity of an employee that is focused on increasing individual and organizational effectiveness can be viewed as proactive behavior. The current investigation will adopt this general definition in order to include a broad range of behaviors associated with this construct.

There is much theoretical and empirical work on the construct of proactivity. Especially since 1990s, the scholarly input into this topic has flourished (Crant, 2000; Unsworth & Parker, 2003; Grant & Ashford, 2008; Thomas et al., 2010). According to the accumulated research, proactivity includes such behaviors as voicing issues and challenging the status quo (Dutton & Ashford, 1993; Frese & Fay, 2001; Parker & Collins, 2010), feedback seeking (Stobbeleir et al., 2010; Miller & Jablin, 1991; Ashford et al., 2003), expanding roles and job crafting (Grant et al., 2009), proactive problem-solving and self-initiated learning activities ( Parker et al., 1997; Frese & Fay, 2001), building networks (Ashford & Black, 1996), improving work routines, and helping others (Organ, 1988; Axtell & Parker, 2003; Griffin et al., 2010). From this list of behaviors it can be seen that proactivity includes a wide spectrum of organizationally beneficial activities that are collected under one umbrella construct.

Antecedents of Proactive Behavior

Based on the literature review, the major aspect of proactivity that interests both theorists and practitioners is its antecedents. In other words, researchers have become interested in finding out what causes proactive behavior of employees. According to Fay and Frese (2001),

(6)

6 proactivity has both dispositional and contextual antecedents. Dispositional influences involve stable individual differences such as proactive personality, knowledge, skills, and abilities, or KSAs (Fritz & Sonnentag, 2009). From this perspective, Bateman and Crant (1993) developed a proactive personality concept, defining it as a relatively stable tendency of an individual to actively influence his/her environment. Proactive persons possess such KSAs as abilities to anticipate issues, proactively prevent and resolve problems, and actively socialize within a new setting (Frese & Fay, 2010; Parker & Collins, 2010). Such individuals are not constrained by the situational factors and persevere in achieving high standards of performance (Major et al., 2006). On the other hand, contextual influences include the dynamic environmental context such as work design, and social and organizational factors (Fay & Frese, 2001). Examples of these factors include job autonomy, supervisory support, effective communication practices, work climate, socialization and training interventions (Unsworth & Parker, 2003). Thus, one way a firm can gain a proactive workforce is by recruiting and selecting employees with appropriate personalities and KSAs. The other way is to train and develop existing employees in the ways that trigger them to express proactive behaviors.

This thesis focuses mostly on the contextual component of training and development practices in an organization. Such a choice is influenced by the view of a cognitive affective system theory. This theory suggests that situations differ in terms of how many opportunities they provide for expressing certain behaviors (Tett & Burnett, 2003). Thus, the dispositional component such as proactive personality is just a slice of a bigger pie of proactivity. However, in research, the proactivity construct has mostly been addressed from the dispositional perspective, more specifically, as a personality trait. Therefore, this thesis attempts to close the gap by focusing on the contextual influence of training and development interventions in order to predict and enhance proactive behaviors of employees.

Training and Development Focus

The training and development principle is central to the concepts of the learning organization, empowerment, and organizational citizenship behavior (OCB) (Spreitzer, 1995). Some scholars empirically showed that proactive behavior is a learning process. For instance, Argyris (1964) suggested that adults at work can progress from passive and reactive states to active and proactive ones; from dependent to independent; from persons with few abilities to ones with many abilities. From the practical angle, the urgency of emphasizing

(7)

7 training and learning is outlined by Bratton (1993) who stated that management needs to re-orient towards treating “employees as a valued asset rather than a variable cost, see training and development as an asset, and view empowerment and high trust employment relations as prerequisites to retain an effective and committed workforce” (p. 398). Therefore, by training employees to be proactive, a firm can empower them, develop their trust in the company, and create personal career success as well as organizationally valuable outcomes.

The interest in this thesis topic was triggered by the 2003 longitudinal study of Axtell and Parker called “Promoting Role Breadth Self-Efficacy through Involvement, Work Redesign and Training.” It showed that the amount of training received was positively correlated with proactive motivation. The following types of training were examined: management/leadership skills, team-building, communications, on-the-job training, technical/maintenance training, and health and safety training. In the same manner, Gist, Stevens, and Bavetta (1991), showed that training role-relevant behaviors and skills increases employee proactivity. Therefore, estimating the effect of the number of trainings on the proactive behavior appears to be a reasonable approach for furthering empirical research in this direction.

Levels of Operationalization

Proactivity can be operationalized at the individual, team, and organizational levels (Seibert et al., 1999; Unsworth & Parker, 2003; Grant & Ashford, 2008). This research is narrowed down to the individual-level proactivity due to two reasons. First of all, most of the current research has focused on proactivity from the individual perspective. Therefore, this thesis was able to utilize the abundant theoretical and empirical material on individual’s proactive behaviors. Secondly, researchers have addressed proactivity from the individual standpoint for a reason. Individuals are the fundamental building blocks of an organization. By understanding the motives of individual employees it becomes possible to explain team dynamics and organizational trends.

In general, there is much theoretical and empirical evidence on the topic of proactive behavior. However, the question of how organizations can actively promote proactivity has not been fully addressed (Axtell & Parker, 2003). Also, studies were conducted in isolation from one another and were testing different types of proactive behaviors (Grant & Ashford, 2008; Belschak & Den Hartog, 2010). Thus, there is a need to further investigate the expressions of proactivity empirically. However, change-oriented behaviors such as

(8)

8 proactivity are still considered novel and difficult to standardize or describe in terms of specific actions (Griffin et al., 2007).

In examining proactive behaviors, researchers mostly used dependent variables such as job satisfaction, organizational commitment, turnover intentions, and job strain (Parker, 2000). These outcome variables are typically used in research as a first point of reference in order to see whether a specific behavior has an impact on the organizationally important variables (Parker, 2000; Parker et al., 2006; Grant &Ashford, 2008). However, these measures include behaviors that are quite passive in nature, such as compliance with procedures, punctuality, and attendance. In this investigation, the focus will be on four proactive behaviors and their relationship with job performance as a behavioral outcome. The choice of this outcome variable is explained in the following subsection.

Job Performance

There is a number of ways to define job performance, since it is a multifaceted construct. The most commonly accepted definition is provided by Campbell and colleagues (1993) who describe job performance as an individual behavior that results in certain outcomes. However, according to them, performance is not limited to observable actions and behaviors, but also includes mental output, such as decisions and answers (Campbell et al., 1993). To be more specific, this investigation will adopt the definition of job performance provided by Motowidlo (2003). He interprets job performance as the total expected value to the organization of distinct behaviors that an individual expresses within a certain period of time (Motowidlo, 2003). This way, organizationally valuable behaviors are considered the main criteria of job performance.

The definition of Motowidlo (2003) and his associated taxonomy are in line with the performance appraisal method at Philips. At this company, job performance is also understood in terms of categories of behaviors. First of all, the objective results (called performance “what”) are evaluated within a range of three possible behaviors:

1. Partially meets 2. Solid results 3. Exceptional results

Secondly, the performance behavior (called performance “how”) is appraised by the following three levels:

(9)

9 1. Improvement required

2. Valued player 3. Role model

Further description of this performance management technique is given in the Method section and in Appendix 2.

The decision to use job performance as an outcome variable was influenced by the meta-analytic study conducted by Thomas and colleagues (2010). They used a total of 103 samples with 32,967 participants to reveal a significant correlation between overall job performance and proactive personality (ρ = .26), proactive personal initiative (ρ = .35), and taking charge (ρ = .46). Also, according to Seibert, Crant, and Kraimer (1999), “proactive individuals select and create situations that enhance the likelihood of high levels of performance (p.417).” Thus, proactive employees identify job-related opportunities, show initiative, take action, and persist until they get considerable results (Dutton & Ashford, 1993; Fuller & Marler, 2009; Seibert et al., 1999; Crant, 2000). Indeed, proactive employees show better results than non-proactive ones in both subjective and objective performance appraisals. This finding was supported for various aspects of individual job performance, such as performance on core tasks (Grant et al., 2009; Belschak & Den Hartog, 2010), sales volume (Crant, 1995), entrepreneurial success (Fay & Frese, 2001), and customer service performance (Rank et al., 2007). Moreover, proactivity positively affects individual career outcomes (Seibert et al., 1999), leadership (Bateman & Crant, 1993; Crant & Bateman, 2003; Deluga, 1998), organizational innovation (Crant, 1996; Parker, 1998), and organizational involvement and commitment (Thomas et al., 2010). Therefore, empirically testing the effect of proactive behavior on organizationally valued outcome, such as job performance, will enlarge the evidence on the proactivity construct. Also, it will benefit the company in understanding the role of proactive behaviors within its boundaries.

Four Proactive Behaviors

The four behaviors that will be explored in this thesis are individual innovation, strategic scanning, networking, and helping others. These behaviors were retrieved from the article “Taking Stock: Integrating and Differentiating Multiple Proactive Behaviors” by Parker & Collins (2010). In this article, the authors classify multiple types of proactive behaviors and empirically show that they are distinct from one another with a help of factor analyses. This

(10)

10 way, the thesis made use of the labels and categories of proactive behaviors that were outlined in the article.

The choice of these specific four proactive behaviors was also determined by the Philips’ “leadership 360° assessment questionnaire.” This assessment tool was introduced at the company in 2006 in order to allow individual employees to receive feedback on their leadership skills. These skills are outlined in the following list of competencies that are evaluated by managers, colleagues, direct reports and others:

- Drive for results

- Create innovative strategies - Inspire commitment

- Champion people's growth - Pursue market insight - Leverage capabilities

Even though this tool is said to measure leadership, there is a substantial overlap with the proactive behaviors as indicated in the article by Parker and Collins (2010). The link between four proactive behaviors that are used in this investigation and the leadership 360° framework is provided in the following subsections.

Individual Innovation

This proactive behavior involves creation and implementation of ideas (Scott & Bruce, 1994). The innovative component of proactivity includes the process of identifying opportunities, generating new ideas or methods, and implementing them. Nowadays, this behavior can be illustrated by such actions as searching out new techniques, technologies, and product ideas (Parker & Collins, 2010). In research, the construct of innovation is closely linked to proactivity due to various innovative actions associated with being a proactive employee (Unsworth & Parker, 2003; Ohly et al., 2006; Parker et al., 2006; Garcia-Morales et al., 2007). The associative link with the leadership 360° framework is via competency to “Create Innovative Strategies.” According to Philips (2011), it entails developing and implementing novel solutions and encouraging innovation and disciplined risk-taking. Leadership 360° tool provides illustrative behaviors for this competency, such as:

• Proposes new, unconventional approaches; • Develops new business opportunities;

(11)

11 Strategic Scanning

This is an ability to keep a broad focus beyond the borders of the organization. This behavior is characterized by proactively surveying the organization’s environment to identify opportunities for improvement. The main purpose of strategic scanning is to ensure a good fit between the organization and its external environment, such as finding ways for the firm to use emerging markets or actively investigating the environment for future threats and opportunities (Kickul & Gundry, 2002; Parker & Collins, 2010). Strategic scanning is also a proactive behavior because it is self-initiated and focuses on overall improvement of the organization and on one’s own success (Howell & Sheab, 2001). The link with the Philips’ leadership 360° is via the competency to “Pursue Market Insight.” In the company, this competency is explained as “showing a deep understanding of the global business environment and end-user insight” (Philips, 2011). This competency can be illustrated by the following behaviors provided by the leadership 360° tool:

• Keeps well informed about market: customers, products, competitors and suppliers; • Identifies new commercial opportunities with customers;

• Responds flexibly to shifting market conditions and related customer needs. Networking

Networking behavior can be defined as expressing initiatives to build and sustain interpersonal networks in which to seek information, advice, or help (Claes & Ruiz-Quintanilla, 1998). The ability to socialize, network, and build relationships has become increasingly important, since no one considers a career to be a static concept anymore (Miller & Jablin, 1991). Therefore, researchers have proposed that proactivity may positively affect social networking. For instance, Thompson (2005) stated that “proactive people are likely to seek ways to construct a social environment conducive to their own success on the job (p. 1012).” At Philips, this proactive behavior is explained by the competency to “Leverage Capabilities.” In the company’s terms, it is defined as using human resources across internal and external boundaries to maximize value for Philips (Philips, 2011). Illustrative behaviors outlined by Philips for this competency include:

• Actively strengthens existing relationships;

• Builds an effective network both within and outside the organization; • Promotes teamwork between people from diverse backgrounds.

(12)

12 Helping Others

Networking would be impossible if there was no opportunity for reciprocal help. Integration into the workplace and successful performance involves some level of team engagement. According to Thompson (2005), helping behaviors involve such actions as peer motivation, help with ad hoc problem resolution, support with the satisfaction of learning needs, and mentoring and coaching informally. Therefore, helping other colleagues proactively is also an essential part of the proactivity construct: it enables people to increase chances of other proactive behaviors, such as issue selling, feedback seeking, and role expansion (Van Dyne & LePine, 1998). This proactive behavior is associated with two Leadership 360° competencies “Inspire Commitment” and “Champion People’s Growth.” They are defined as mobilizing people around the Philips’ strategy, developing an inclusive and high-energy environment, and committing personally to the performance and development of others (Philips, 2011). The illustrative behaviors for these competencies include:

• Coaches others to deliver top performance; • Gains support and commitment from others;

• Encourages the expression of uncommon / unconventional views; • Ensures others are kept well informed.

Based on the above-stated descriptions of proactive behaviors and associated Philips’ leadership 360° competencies, the following figure summarizes the interrelations between them:

Figure 1: Proactive Behaviors in the Philips’ Leadership 360° Framework

Individual Innovation

Strategic Scanning

Networking

Inspire Commitment Champion People’s Growth

Leverage Capabilities Pursue Market Insight Create Innovative Strategies

(13)

13 Further explanation of the leadership 360° tool and its relation to the four proactive behaviors will be given in the Method section and in Appendix 1.

Hypotheses

The theoretical and empirical evidence mentioned above lead to the following goals of this thesis:

1. Explore the relationship between the four proactive behaviors and job performance; and

2. See how the number of training and development interventions may moderate the proactive behavior over time and produce intra-personal change.

Therefore, this thesis attempts to answer two research questions: Is there a positive relationship between proactive behaviors and job performance? Can HR managers enhance proactive behavior of their employees via increasing training and development interventions?

These research questions trigger the exploration of the associated hypotheses. As stated above, the theoretical and research evidence showed that proactivity is positively correlated with individual job performance. Therefore, the thesis first assumes:

Hypothesis 1

There is a positive relationship between proactive behaviors (individual innovation, strategic scanning, networking, and helping others) and individual job performance.

Then, going beyond the correlation analysis, the moderated mediation relationship is explored. The researcher assumes that proactive behaviors over time mediate results on job performance. Also, based on the cognitive affective system theory, there should be a positive relationship between training and development interventions and proactive behavior. Thus, it is assumed:

Hypothesis 2

Proactive behaviors measured at Time 2 serve as a mediator of the relationship between proactive behaviors measured at Time 1 and job performance at Time 2. In turn, the number of training and development interventions received by individual employees between Time 1 and Time 2 moderates the positive relationship between proactive behaviors at Time 1 and proactive behaviors at Time 2.

(14)

14 With relation to the above-mentioned hypotheses, the following conceptual model depicts the proposed relationships:

Figure 2: Conceptual Model

This conceptual framework depicts both hypotheses. On the one hand, hypothesis 1 addresses the direct relationship between the four proactive behaviors and job performance. On the other hand, hypothesis 2 focuses on intra-personal changes in proactive behavior over time due to training.

Method

To test these two hypotheses, two studies were conducted. Study 1 focused on testing hypothesis 1, while Study 2 addressed the hypothesis 2. The following subsections will explain how the data were collected and analyzed as well as how different variables were operationalized and interrelated.

Context

This thesis was a result of an internship within corporate headquarters of the Royal Philips Electronics (Philips) in Amsterdam for three months. Philips is a large and dynamic firm that provides innovative solutions in healthcare, lifestyle, and lighting technologies. The company’s people-centered focus creates benefit to both customers and personnel. Since Phillips considers its human capital an essential source for success, the company attempts to stay up-to-date with high performance initiatives, various HR interventions and best practices. Proactive Behaviors Time 1 - Innovation - Strategic Scanning - Networking - Helping Others Outcomes Time 2 Job Performance Contextual Factors Number of Training&Development Interventions Proactive Behaviors Time 2 - Innovation - Strategic Scanning - Networking - Helping Others

(15)

15 A special interest was triggered by the Talent Management (TM) program at Philips. This program was initiated at Philips because of the need to develop future leaders internally. The company’s vision 2015 focuses on accelerated growth and strengthening of leadership and proactive innovation (Philips, 2011). The desire of Philips to achieve and sustain solid business goals helped produce a step-by-step process to find and nurture individuals who exert exceptional performance and organizationally valuable behaviors. These leaders would proactively resolve issues posed by the rapidly changing conditions within and outside the firm (Philips, 2011). The TM team called them “High Potentials”, because they have shown a lot of potential to become future business leaders in the first five or more years of their career at Philips.

There are approximately two thousand of Philips’ High Potentials (Philips, 2011). The selection (calibration) of employees into the Talent pool of High Potentials is based on their annual performance appraisals and discussions of the management team. This group of exceptional employees receives particular attention in the form of development centers, learning curricula, internal vacancies, and coaching. Therefore, the training and development interventions apply especially to this group of employees.

The corporate grades (CGs) for High Potentials vary from 50 to 80. In total, there are eight CGs or organizational levels through which an employee can progress in the organization over time. The GG 50 is the starting grade for many inexperienced employees when they join Philips. In the course of years, they may advance in their positions and move into CGs 60, 70, and 80. After GC 90, there are three executive levels.

The TM team is responsible for all activities that help accelerate and develop this group of people. Besides identifying High Potentials, the company also offers programs to their high-performing employees in the Early Potentials, Top Potentials and Executive Potentials categories. Through increased levels of communication and coordination the TM team offers special HR practices to employees in the Talent pool. These interventions are focused on promoting job autonomy, engagement, proactivity, employee confidence in their skills, and advancement of their individual careers. The goal for this internship was to fully understand the process of developing High Potential employees, their leadership, proactivity, functional competencies, and the quality of communication.

(16)

16

Participants

Various data were collected on two sets of participants, each of them containing 100 employees. The first set consisted of Philips employees in the Talent program called “High Potentials”. The second set was retrieved randomly from the employee database of professionals who are not involved in any Talent program. The decision to limit the samples to 200 people was made by the researcher due to the complicated and laborious process of retrieving the appropriate data from the web tools and databases. For instance, finding 100 employees who have gone through the 360° questionnaire took investigation of around 5000 employees in the database and manually typing their email addresses in the leadership 360° web tool. This web tool takes considerable time to load individual reports on people’s results, and some results are missing. Also, since the researchers’ access to the Leadership 360° web tool was limited, she had to rely on the willingness to help and time availability of the tool supervisor.

The participants chosen for this investigation have CGs ranging from 50 to 80 in both of the studies. The major reason to limit the research samples to these corporate grades was because High Potentials are located within this CG range. The choice to limit investigation to only High Potentials was made by the Philips’ TM team during the initial discussion of the possible sample for this thesis. Also, the High Potentials constitute the majority of the so-called “Talent community” within Philips. Therefore, investigation of this group may benefit the largest sum of people within the community. Moreover, High Potentials share homogenous formal and informal training and development interventions, which became helpful for using the number of training as a variable. In general, employees in higher grades are treated differently: they receive more informal training and coaching, rather than standardized programs; they occupy roles that differ a lot in their orientation from roles in CGs 50-80; and they are treated more confidentially by Philips. Therefore, it was more feasible for the purposes of this investigation to limit the sample to the High Potentials. For Study 1, testing the first hypothesis, both sets of 100 people were merged and used simultaneously. The reason to merge the High Potentials with the set of Philips’ professionals was to avoid the restriction of range problem. In other words, by including the professionals, the researcher prevented from having a biased sample of over-achievers (High Potentials). Such restriction usually has no or limited statistical significance and prevents the researcher from seeing the whole picture. Therefore, merging the two sets was supposed to increase the

(17)

17 variability of results on the leadership 360° questionnaire and produce a more reliable outcome.

Here is the descriptive demographic information on the sample of 200 people: 78% of the sample was males. The average (mean) age is 37.8 (SD = 6.38); average tenure in the company is 7.97 (SD = 5.24) years. The geographical origin and nationality composition: Dutch – 39%; other citizens of the EU (e.g., French, Belgian, Italian, German, Czech) – 10%; Asian (e.g., Chinese, Thai, Singaporean, Japanese) – 11%; North-American – 8%; South-American – 5%; Unspecified – 27% of the sample. The corporate grade constitution is as follows: 5% of the sample was from CG 50, 20% - from CG 60, 52% - from CG 70, and 22% - from CG 80. The educational background is: Bachelor and Associates Degrees – 8%; Masters Degree – 49%; Doctorate degree – 7%; Unspecified – 36%.

For Study 2, testing the hypothesis 2, only the data on 100 High Potentials was used. The first reason for that was that there were more available data on this group than on regular employees. For example, it is very hard to locate professionals who had leadership 360° evaluations in two points in time. Also, performance evaluations for many professionals were recorded in the system only at one point in time. On the contrary, information on High Potentials is rather complete, because of the special attention they receive by the TM team. This made it possible to investigate intra-personal changes in proactive behavior over time. Time 1 for completing the questionnaire in majority of the cases was either year 2007 or 2008, and the Time 2 was in most of the cases either 2009 or 2010. To be more specific, the mean difference between the two points in time was 1.96 years (SD = 0.55).

The demographic composition is very similar to the 1st sample of 200 people: 77% of the sample was males. The average age of participants is 36.1 (SD = 4.07); the average tenure is 7.65 years (SD = 3.24). The geographical and educational backgrounds are very comparable to the first sample. However, the corporate grade distribution is rather different: 3% people are from CG 50, 2% people - from CG 60, 67% - from CG70, 28% - from CG 80.

Measures

In this section the variables used for this investigation are explained. In both studies, the researcher controlled for gender, age, tenure, corporate grade, and MD category (whether in the Talent program or not).

(18)

18 Proactive Behavior

As mentioned earlier, for testing the four proactive behaviors at two points in time the leadership 360° tool was used. The 360° feedback is a widely recognized HR method of collecting opinionated information from a wide range of people who are in direct contact with a given employee (Church & Bracken, 1997; Ghorpade, 2000). It allows the measurement of competencies of individuals in a more objective manner than self-ratings or subjective appraisals by a single supervisor.

In Philips, every job function from grade 50 and up to the Board of Management has a set of defined competencies – a combination of KSAs and behaviors – in order to ensure the best output for the company (Philips, 2011). This set of competencies was developed in 2006-2007 in order to promote leadership and proactive behaviors among employees. The following pie diagram provides the gist of this approach:

(19)

19 This framework contains six leadership competencies that can be seen as segments of the pie chart. In turn, each competency has six to nine behaviors on which each participant is evaluated (see Appendix 1 for more detailed information on exact behaviors). The questionnaire is filled out by a number of required respondents (self, 1 manager, and 2 colleagues) and optional ones (2 direct reports, 1 other manager, and 2 others). On average, there are three to five raters per rater category (Philips, 2011). After rating a person against various behavioral statements on the 0-100% scale, the results are automatically aggregated per competency.

Based on its setup, this 360° questionnaire to some extent addresses the issue that specific external raters may draw on a general impression across all behaviors, the so-called “halo” effect (Lance, LaPointe, & Fisicaro, 1994). Also, the drawback of self-ratings - the desire to put forward socially desirable answers - is addressed by requiring other people to provide ostensibly more objective feedback. The feedback is expected to be more objective because the questionnaire is used not for assessment purposes, but for the personal growth and development of the person that is being reviewed (the focus). Also, even though the focus chooses his/her evaluators, the feedback is kept anonymous for the categories that involve more than one rater (e.g., colleagues). However, for the manager role, there is a minimum requirement of 1 person. Therefore, it cannot be kept anonymous in case only one manager rates. In any case, the results of the feedback are highly confidential: they are available only to a given employee, his/her HR administrator, and the supervisors of the tool. This fact may trigger the focus to be more honest in self-rating.

In general, this framework explains the new mindset and style of leadership Philips expects – energetic, focused and empowered (Philips, 2011). Leadership competencies are aligned with the performance management program at Philips, since they aim to achieve leadership mentality and reflect on not only what has been achieved (results), but also how it has been achieved (behavior).

In sum, leadership behavior is a prominent variable for organizations that try to foster proactive behavior (Belschak & Den Hartog, 2010). In line with this notion, the work by Strauss et al. (2009) showed that employees who score high on transformational leadership behavior also have high ratings on proactive behavior. Therefore, the use of the leadership 360° questionnaire for this investigation was considered appropriate and even essential.

(20)

20 Training and Development Interventions

For this investigation, the number of training and development interventions was assumed to have a moderating effect on the proactive behavior overtime. The following types of training were examined: management/leadership skills, team-building, communications, on-the-job training, developmental programs, technical/maintenance training, and health and safety training. Unfortunately, it was not possible to estimate the number of trainings in which participants engaged on a voluntary and self-initiated basis. Therefore, the gross number of trainings was used as an estimate.

Performance

Performance at Philips is measured via the so-called “People Performance Management” (PPM) program. It is a global process for performance evaluations within Philips. The target group is all employees of grade 50 and above. Approximately 42,000 Philips employees worldwide participate in the electronic PPM tool (Philips, 2011). Therefore, it was considered suitable for collecting performance data on employees.

The process of the PPM is a dialogue between the employee and his/her direct manager about the employee’s performance and development. The performance part focuses on “what you do in your job (results)” and “how well you do your job (behavior).” The development part entertains the question “How do you develop yourself?”

As discussed earlier, this process identifies three levels of performance per each of the two criteria of “what” and “how”:

• What: partially meets, solid results, exceptional results; • How: improvement required, valuead player, and rolde model.

Appendix 2 contains definitions and further descriptions of the specific performance levels outlined in the diagram.

Procedure

Philips collects descriptive data on employees (such as age, gender, and number of trainings) as well as their “what” and “how” results of the PPM in the Talent@Philips database. The personal reports of results in leadership 360° questionnaire are stored on a web tool that is specifically designed for this purpose. After recording percentage results of 360° feedback for six competencies at two points in time for 100 High Potentials and at one point in time for

(21)

21 100 professionals, a special filter in Talent@Philips database was created to retrieve necessary descriptive data on the collected pool of participants as well as their scores on performance. Then, the data was made anonymous and was compiled in a single SPSS dataset.

The control variables in the dataset included age, gender (coded as “1” for females, and as “0” for males), employee tenure in organization, corporate grade (ranging from 50 to 80), MD category (those in the Talent program were coded as “1”, those not - as “0”), and academic degree (High School diploma coded as “1:, Bachelor – as “2”, Associates – as “3”, Masters – as “4”, and Doctorate as “5”). However, since many participants among professionals did not have any academic degree stated in their report, this variable was dropped. The numeric representation of the performance scores for both “what” and “how” was coded in SPSS as “1” (partially meets/improvement required), “2” (solid results/valued player), and “3” (exceptional results/role model) accordingly.

To test hypothesis 1, a correlation matrix was produced in the SPSS that included data on both control variables and the results of individual scores on six competencies of the leadership 360° questionnaire at Time 1 and the Performance value at Time 1. The value of Performance at Time 1 was used because it was in each case measured after the questionnaire. Then, the data was further investigated with the regression analysis.

Hypothesis 2 was tested using a relatively novel and parsimonious technique called moderated mediation (Preacher et al., 2007; Lippke et al., 2009). This term was introduced by James and Brett (1984) who suggested that some mediation models require addition of a moderator to the relationship. This moderator influences the strength of the indirect effect between the two variables. The reason to use this method in this study is the desire to establish specific relationship between variables, since correlation is not sufficient for claiming that two variables are significantly related (Preacher & Hayes, 2008). The moderated mediation analysis might permit to see by what means certain results in 360° questionnaire may be associated with a certain job performance.

As stated in the Theory and Hypothesis section, proactive behavior measured at Time 2 was proposed to mediate the relationship between proactive behavior at Time 1 and job performance. It is expected that the proactive behavior will have a conditional indirect effect on job performance. The number of trainings is expected to moderate the relationship between proactive behaviors at Time 1 and Time 2.

(22)

22 For testing hypothesis 2, individual items of the Leadership 360° at both Time 1 and Time 2 were aggregated into single numbers. They were also tested for moderated mediation independently in order to reveal peculiarities per item. As to the performance scores on “what” and “how”, they were averaged to produce single scores at two points in time.

Results

This section provides the outcomes of the statistical analyses that were conducted to test the two hypotheses. The subsections for Study 1 and for Study 2 will explain results accordingly.

Study 1

For the Study 1, the descriptive statistics and correlations of control variables, individual items of the leadership 360° questionnaire, and the job performance are given in Table 2.

(23)

Table 1: Correlations, Means and SDs of Variables in Study 1 (Cronbach’s alphas on diagonal; N = 200)

M SD 1 2 3 4 5 6 7 8 9 10 11 12

1. Job performance 2.31 0.36 (.69)

2. Drive for Results 79.85 6.12 .47** (.89)

3. Create Innovative Strategies 76.25 6.12 .31** .74** (.88)

4. Inspire Commitment 79.04 5.09 .42** .68** .61** (.89)

5. Champion People's Growth 75.34 6.33 .40** .74** .71** .72** (.88)

6. Pursue Market Insight 76.10 7.41 .24** .53** .80** .53** .57** (.91)

7. Leverage Capabilities 79.44 6.10 .33** .52** .54** .75** .63** .46** (.90) Control variables 8. Age 37.86 6.39 -.26** -.07 .03 -.17** -.13* .05 .00 9. Gender {0 = male;1 = female} .25 .75 .01 .05 .05 .06 .02 .02 .06 .00 10. Tenure 7.97 5.25 -.08 -.11 -.04 -.11 -.16* -.01 .07 .50** -.09 11. Corporate Grade 69.15 8.00 .16* .22** .14 .01 .14* .04 .02 .20** -.02 .17** 12. MD Category

{0 = not;1 = yes, in Talent pool}

.50 .51 .20** .17** .06 .10 .13* -.03 .01 -.27** -.04 -.07 .36**

* Correlation is significant at the 0.05 level (one-tailed). ** Correlation is significant at the 0.01 level (one-tailed).

(24)

The overall Cronbach’s α for six items = .91. However, the reliability of the performance measure is smaller than of any of the six proactive behavior competencies (α = .69). This is mostly due to the fact that the performance measure included only 2 items (“what” and “how”). In general, from this table it can be seen that deletion of any of the items does not produce a significant change in the overall Cronbach’s alpha. This is also a good indication of the reliability of each item.

As to correlation magnitudes, it can be seen from the table above that all items of the questionnaire had significant correlations with the job performance variable. The highest correlation for questionnaire items turned out to be for the “drive for results” item (r = .47), and the lowest - for the “pursue market insight” item (r = .24). Also, there was a significant relationship between the variables themselves. This might be due to the possible “halo” effect that raters expressed when giving feedback. As to the control variables, they do not correlate with job performance significantly. However, there is a negative impact of age on the performance (r = .26), and it seems that that those in the Talent pool have higher performance than those who are not (r = .20). However, these results are preliminary.

The following Table 2 provides results of the multiple regression analysis of all variables, including the control variables:

Table 2: Multiple Regression Analysis for Study 1 (N = 200)

Variable B SE B β t Sig. (p) Step 1 (Constant) 2.31 .24 9.53 .00 Age -0.02 .01 -.30 -3.64 .00 Gender 0.01 .03 .03 0.38 .71 Tenure 0.00 .01 .04 0.54 .59 Corporate Grade 0.01 .00 .20 2.57 .01 MD Category 0.04 .06 .06 0.70 .48 Step 2 (Constant) 0.04 .42 0.10 .92 Age -0.02 .00 -.27 -3.50 .00 Gender 0.00 .03 .00 0.02 .99 Tenure 0.01 .01 .07 0.99 .32 Corporate Grade 0.01 .00 .13 1.78 .08

(25)

25

MD Category 0.02 .05 .03 0.39 .70 Drive for Results 0.02 .01 .39 3.60 .00 Create Innovative Strategies -0.01 .01 -.18 -1.36 .18 Inspire Commitment 0.01 .01 .10 0.83 .41 Champion People's Growth 0.00 .01 .01 0.07 .95 Pursue Market Insight 0.00 .01 .09 0.81 .42 Leverage Capabilities 0.01 .01 .10 1.06 .29

Notes: R2 = .12 for Step 1 (ps < .001), ∆R2= .20 for Step 2 (ps < .001).

As the Table 3 shows, the item “drive for results” has the highest significance in terms of the dependent variables (β = .39, p = .00). Other items contribute considerably less to the relationship with the job performance. As to the control variables, the negative effect of age is notable on the job performance (β = -.27, p = .00). In general, since the change in R2 = .20 after introducing the proactive behavior measures of the questionnaire, the proactive behaviors do contribute to the job performance ratings.

Study 2

The descriptive statistics for the sample of 100 High Potentials did not vary much from the larger sample of 200 that included professionals. However, the average results for items of the Leadership 360° questionnaire and for the job performance are slightly higher than when merged with the pool of professionals. However, the difference is almost negligible.

The moderated mediation hypothesis was tested by the ModMed macro designed for SPSS (Model 2). This Model 2 assumes the interaction between training and proactive behavior before proactive behavior at Time 2 influences job performance at Time 2 (Hayes, 2011). This macro calculates the Sobel test for the conditional indirect effect. It also quotes percentile-based, bias-corrected regions of significance (Johnson-Neyman technique), and accelerated bootstrap confidence intervals (Preacher et al., 2007). Since the sampling distribution of the indirect effect is rarely normal, the bootstrap method (5000 samples) was applied to the sample.

The Sobel test results of the moderated mediation analysis show that mediation exists on the path from 360° results at Time 1 to Time 2 (β = .31, p<.01) and from 360° results at Time 2 and the job performance (β = .37, p<.01). These results show that the hypothesis was

(26)

26 supported and there is a strong support for the mediation relationship. The support for the mediation was found across values of the moderator.

These findings were still valid after addition of the control variables (gender, age, tenure, educational background, and corporate grade). The control variables did not have any significant effect. However, due to missing data on some of these covariates, the second sample was much smaller (N = 86). Therefore, conclusions were drawn on the analysis that excludes covariates. Moreover, executing the ModMed analysis on the individual items of the leadership 360° questionnaire did not produce any significant changes to the existing mediation.

However, this ModMed analysis showed that the interaction term between training and proactive behavior is not significant. At the same time, the conditional indirect effect at the specific values of the moderator was significant. The Johnson-Neyman significance regions showed that the trainings have substantial indirect effect before they reach the approximate value of 11 trainings. The effect decreases as the number of training increases, but after 11 trainings, the significance ceases to exist at all. This finding was also supported with the bootstrap confidence intervals analysis shown in the table below:

Table 3: Bootstrap 95% Confidence Intervals for Conditional Indirect Effect of Training (Bootstrap sample size = 5000)

Lower Upper Percentile

-.0016 .0368 Bias corrected

-.0001 .0378 Bias corrected and Accelerated

.0006 .0387

This controversial interaction between training and proactive behavior led to further investigation with the help of the ModProbe script designed for SPSS (Hayes, 2011). This script generates moderator values (M, +/- SD) and estimates the effects of a focal predictor at these points. The output was plotted using an associated Excel tool.

The test results of the moderator probing analysis show that the effect of training is higher on employees with a low proactivity score at Time 1 than those with a high initial proactivity score. This becomes clear from the following Figure 4:

(27)

27 This figure shows not just that training an employee positively affects his proactive behavior, but more specifically it shows that a less proactive employee benefits more from training than a more proactive employee. However, interestingly enough, this figure also indicates that the proactive behavior of employees who score high at Time 1 actually seems to decrease at Time 2, regardless of the number of trainings. Yet, this effect requires further investigation by researchers in order to make any conclusions. Finally, the lack of significance starting from value 11 of training was found due to the very small number of participants who have gone to more than 11 training between Time 1 and Time 2. Therefore, the calculation of proper significance level became impossible.

In general, mediation effect of proactive behavior on job performance was supported at any value of the training and development intervention. The moderating effect of training was also supported, showing that employees do benefit from training. Moreover, less proactive employees benefit from the training more than already proactive employees. The discussion of these results is provided in the following section.

Discussion

Combined, the results of these two studies show that there is positive relationship between the four proactive behaviors and job performance. The results answer the research questions posed for this investigation. Also, the conceptual framework was shown to be valid by the

60

65

70

75

80

85

90 Low Proactivity T1

High Proactivity T1

P

ro

ac

ti

v

e

B

eh

av

io

r

T

2 Low Training

High Training

(28)

28 results of testing Hypothesis 2. The further two subsections will address theoretical and practical implications of these findings.

Theoretical Contributions

This investigation responds to calls for an empirical investigation of the proactive behavior construct. Few researchers have considered how to promote increases in employees’ proactive behavior. However, this direction of research is especially important during this day and age. At the same time, proactive behavior cannot be simply prescribed in organizations. For instance, managers cannot simply instruct their employees to be more proactive. Therefore, there is a need to further investigate situational antecedents of proactive behavior, such as organization’s training and development interventions.

Proactive behaviors have been addressed in the literature from differing angles and with various purposes. This study adds to the existing research by showing a positive relationship between proactive behaviors and job performance as well as the significant effect of training on the proactive behaviors. The correlations, regression analysis, and moderated mediation analysis showed support for the both hypotheses that this thesis posed. Analysis of the relationships between proactive behaviors and key individual factors (age, gender, tenure, educational background and corporate grade) revealed low and statistically insignificant correlations.

This study contributes to the existing literature on proactive behavior by attempting to establish a relationship between the four proactive behaviors (both in distinct and aggregated form) and job performance. While these results may not generalize to all proactive behaviors, Grant and Ashford (2008) argued that by exploring one proactive behavior one can use that information to infer general characteristics that influence all proactive behaviors. Also, Parker and Collins (2010) investigated different categories of proactive behaviors and showed that these behaviors were distinct from one another. This research substantiated their framework by focusing on the content of the four proactive behaviors that were outlined in their categories.

These proactive behaviors explain unique variance in employee job performance, which is in line with Porath and Bateman (2006) who also found that proactive behavior acted as a predictor of individual job performance. However, the findings of this study refine these ideas, suggesting that there is a potential for a moderating effect of training.

(29)

29 In research, there is also a greater differentiation among performance constructs and definitions. Much research has focused on the performance of core tasks or other measures that are stable and passive in nature (Cliff, 2006). In recent years, researchers and practitioners have argued for expanded performance constructs to capture all aspects of behavior that have value for the organization. The current study responds to this argument by using the Philips performance measures of “how” and “what”, which go beyond the immediate task performance.

The moderation effect of training was supported. This is in line with the arguments offered by Axtell and Parker (2003). They showed a positive correlation between the number of trainings and proactive motivation. However, the findings also show that proactive employees benefit from the number of trainings not as much as less proactive ones do. Also, as the number of trainings increased substantially, there were too few participants who participated in them. Thus, it became impossible to properly measure the significance of the proactive behavior - training interaction. In any case, the presence of the effect of training on the proactive behavior results supports the cognitive affective system theory, which suggests that the situational factors affect the proactive behavior just as proactive personality does. Finally, the findings of this thesis highlight that proactivity researchers should focus more on the situational antecedents of the proactive behaviors. The present study attempts to stimulate greater investigation in that direction.

Practical Implications

Philips’ new market-centric focus is well in line with the proactive behaviors in many ways. The people-oriented, sensible and simple solutions are there to fuel the company’s growth. Employees are expected to challenge the status quo, experiment with new ways of doing things, learn to deal positively with resistance, take every opportunity to help others, and be flexible and adaptive when required (Philips, 2011). The organizational value “Develop People” is there to keep employees engaged and inspired to constantly work on developing their leadership and skills. Therefore, the practical relevance is mainly to HR managers at Philips who want to enhance employee’s confidence to fulfill a number of proactive tasks. The following suggestions and commentary are outlined below to provide the practical advice to the management team:

First of all, some individuals who have particular KSAs and personality traits are more likely to be proactive. Thus, managers who are in need of a proactive workforce should recruit

(30)

30 appropriate individuals into the organization. This may imply some adaptation to the selection process, such as extending criteria beyond technical skills to include those personal factors that have been linked to proactivity. This is in line with the Philips’ policy to be “attracting and retaining the best people” (Philips, 2011).

Secondly, managers often tend to overemphasize selection as a strategy for obtaining the appropriate individuals (Gruman et al., 2006). However, the KSAs and even motivations that trigger proactive behavior can be enhanced via training. The training and development interventions are in line with one of the four key values of Philips, “Develop people.” This thesis provides support of the effect of training on the proactive behavior as well as the indirect effect of proactive behavior on job performance. By acknowledging the value of training on organizationally valuable outcomes, such as proactive employees, Philips can further enhance its talent pool and increase chances for better organizational performance. In any case, neither recruitment nor training would be able to compensate for the organizational environment that does not allow, encourage, and support proactive behaviors (Brown & Starkey, 1994; Gruman et al., 2006). Therefore, special attention needs to be given to the work context. Important aspects of the work context include immediate tasks, work design, and the wider organizational culture, structure and processes (Stamper & Van Dyne, 2001). Changing the work design to create autonomous and challenging jobs will encourage proactivity (Parker et al., 1997). On the other hand, facilitating employee self-management, combined with lateral integration devices (e.g. teamwork, liaison managers) will promote further collaboration (Hatch, 1993; Gruman et al., 2006). For example, social events, job sharing initiatives, and knowledge data systems can enhance communication within and across teams, and allow mutual respect and openness to new ideas (Brown & Starkey, 1994; Van Dyne & Graham, 1994; Fischer et al., 2005). These interventions correspond with another of the four Philips’ values, “Depend on each other.” Other instances on the Philips’ internal website that go along with the above include the following behaviors:

• Trusting and empowering each other to contribute our best;

• Teaming up and allocating resources to the most promising opportunities; • Creating strong and diverse teams by being open to others and their ideas.

In addition, Philips proposed a policy of self-directed careers, which means that employees are expected to invest in their careers themselves and rely less on their supervisors, mentors, or HR managers. This is where proactive behaviors might also come into play. According to

(31)

31 Seibert and colleagues (1999), proactivity is important for individual career success, because it entails a large range of behaviors (such as networking and strategic scanning) which influence individual career success favorably. Therefore, by developing proactive employees, Philips may produce higher employee career success rates without directly investing in individual careers.

In sum, in order to promote proactive behavior, an inclusive way is needed, which will adopt several approaches simultaneously. According to Unsworth and Parker (2003), this approach may use changes in organizational interventions (e.g. recruitment, selection, and training practices), individual work design (e.g. job autonomy and complexity), and social and organizational characteristics (e.g. team work, supervisory support, effective communication, and work climate).

Locke (1991) argued that individual performance depends on two things: self-efficacy and goals/intentions. In other words, an employee’s decision to engage in proactive behavior will be based on his/her self-confidence and goals that are prioritized. Examples of goals can be preventing a problem, fitting within an organizational culture, creating desirable impressions. People will consider potential costs and benefits of proactive behavior for their image, job performance, career growth, and other outcomes (De Stobbeleir et al., 2010). Therefore, creating a shared vision for individual goals of employees will align them better with the culture of Philips.

In line with potential concerns offered by Locke, one should also be aware of the costs of proactive behavior, the so-called “dark side” of proactive behavior (Ashford &Northcraft, 1992; Belschak et al., 2010; Bolino et al., 2010). Some recent research (Van Dyne & Ellis, 2004; Bergeron, 2007; Grant, 2008) suggests that proactive behaviors can also cause a number of overlooked and unexpected consequences for employees and organizations. First of all, when management of the firm expects proactive behavior, it may contribute to stress, role overload, and work-family conflict among employees (Bergeron, 2007). Also, it can cause friction between proactive and less proactive workers (Grant, 2008; Bolino et al., 2010).

Secondly, researchers have pointed out that proactive employees are involved in rule-breaking activities (Belschak et al., 2010). Such behaviors could occur when an employee proactively seeks out opportunities to “benefit” from the system and involves in certain

(32)

32 counterproductive work behaviors (Bergeron, 2007, De Stobbeleir et al., 2010). For example, the employee may take initiative to cheat on the employer and steal company supplies.

Another potential problem is that some jobs provide more opportunity for proactive behaviors than other jobs. The inability of the organization to recognize and mitigate such differences may result in employees perceiving proactivity rewards as unfair (Lim, 1996; Aspinwall & Taylor, 1997; Erdogan & Bauer, 2005). Thus, promoting proactivity should be in balance with interventions such as development and organizational socialization efforts. Philips should not over-depend on the premise “employee, manage thyself” (Bolino, 2010, p. 325), where employees are left alone and are expected to socialize and develop themselves independently. Proactivity should not be perceived as a substitute to organizational socialization efforts and other practices, because these two factors reciprocally stimulate one another and, therefore, should go hand-in-hand (Gruman et al., 2006).

Finally, proactivity may threaten the heterogeneity of the workforce. Overtime, when the organization highlights the importance of proactive behaviors, it tends to attract people who possess those skills and make the firm homogenous. In turn, homogeneity has shown to cause some problems, such as lack of innovative capability and narrow organizational focus (Bolino, 2010). Also, a homogeneous workforce comes with overdependence on the capabilities that reside within proactive individuals (Bolino, 2010). This makes the employer vulnerable to turnover intentions of these proactive employees. However, more research is needed in the comparison of short-term versus long-term costs and benefits of proactive behaviors.

Limitations and Future Research

This study bears its own limitations. First of all, it used a relatively small sample of 200 employees (and 100 for Hypothesis 2), which prevents the research from being generalizable. Also, most of the participants of this study occupy managerial positions and are involved in cognitive tasks. This fact also hinders generalization of findings across other employees who are involved in functional tasks or manual work.

Another issue in the current study is that there was no differentiation between volitional training and compulsory training. This was because employees partially discuss what training is required or wanted during appraisals with their Talent managers, mentors, and supervisors. Therefore, it was not possible to see which training is initiated by an employee and which is

Proactive behavior at Philips : relationship with job performance and moderating effect of training

Master Thesis