MASTER THESIS
The role of corporate
mobile communication on
employee engagement
A study amongst retail associates
Raymond Kulsen
Student number: 10901795 Supervisor: E. Peelen Master Thesis
University of Amsterdam, MSc Business Studies 4/7/2017
2
Abstract
This study measures the effect of corporate mobile communication on employee engagement. Employee engagement is a cognitive, emotional, and psychical construct that can be defined as; “a positive, fulfilling, work‐related state of mind that is characterized by vigor, dedication, and absorption” (Schaufeli et al., 2002)’.
The study is conducted via a pilot and final measure in a fashion retail setting. The results show that with corporate mobile communication, employee engagement can be influenced positively. Especially the level of dedication of an employee towards their job can be affected when using a mobile communication tool on a functional as well as an emotional level. On the other hand, when focusing too much on the emotional side of employees, in for instance a competitive manner, the corporate communication tool can have a negative effect on employee engagement. In conclusion, corporate mobile communication is an antecedent of employee engagement.
Next to this antecedent of employee engagement, the positive consequences of highly engaged employees are studied and as in other studies (Bakker & Demerouti, 2008; Cole et al., 2012; Christian et al., 2011; Macey & Schneider, 2008; Rich et al., 2010; Saks, 2006; Schaufeli et al., 2002), this positive effect is present on several performance factors. This indicates that highly engaged employees are less likely to quit, are more satisfied with their job, show a higher level of organizational citizenship behavior, and show a higher level of organizational commitment.
In short, the role of corporate mobile communication on employee engagement is positive and in return has a (in)direct positive effect on performance factors. This study, therefore, contributes to the few studies which investigate the role of internal communication on employee engagement and adds a contribution to the field of employee engagement consequences.
3
Acknowledgement
I would like to thank my supervisor prof. dr. Ed Peelen for all the guidance and feedback during the thesis process. His enthusiasm for the subject made my enthusiasm grow even more. Also, a special word of gratitude to Fashion Retailer x and Kega and especially to Mr. Kees Verkade for giving me the opportunity to research this particular case and for the support in the overall process of the thesis.
June 2017
Raymond Kulsen
Statement of Originality
This document is written by Student Raymond Kulsen 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.
4
Content
1. Introduction ... 6
1.1 Employee engagement in Retail ... 6
2. Literature review ... 8
2.1 Status of engagement worldwide ... 8
2.2 Employee engagement ... 9
2.3 Comparable with engagement ... 10
2.4 Dimensions of employee engagement ... 11
2.5 Internal communication as antecedent of employee engagement ... 14
2.6 New media as communication tool ... 15
2.7 Hypotheses ... 17
2.7.1 Corporate mobile communication tool ... 18
2.7.2 Specific hypotheses ... 20
2.7.3 Peer to peer communication ... 21
2.7.4 Antecedents of employee engagement ... 22
2.7.5 Net Promoter Score ... 23
2.7.6 Consequences of employee engagement ... 23
2.8 Conceptual framework ... 25
3. Research design ... 26
3.1 Data collection and design ... 26
3.2 The research procedure ... 26
4. Results baseline... 31
4.1 Reliability measures Pilot-measure... 31
4.2 Mean ... 32
4.3 Testing Normality ... 33
4.4 Current engagement levels ... 33
4.5 Net Promoter Score ... 34
5. Results total measure ... 36
5.1 Reliability measures ... 36
5.2 Testing Normality ... 37
5.3 Use of internal communication application ... 38
5
5.5 Structural Equation Modeling ... 47
5.6 Employee engagement dimensions on performance factors ... 57
6. Discussion & further research ... 62
6.1 Functional & emotional ... 64
6.2 Performance factors ... 65
7. Limitations... 67
Appendix ... 69
Attachment: ... 75
A1: invitation pilot survey (Dutch) ... 75
A2: invitation final survey (Dutch) ... 76
A3: questionnaire pilot (in Dutch)... 77
A4: questionnaire total measure (in Dutch) ... 81
B1: frequencies survey ... 92
B2: reliability performance measures ... 94
B3: usage levels InTouch application Fashion Retailer x... 95
B4: tables pilot and total measures ... 96
B4: output parametrical tests ... 97
C1: structural equation model ... 122
Stripped model: model fit; ... 122
Stripped model: statistics ... 122
C2: total structural equation model... 125
Total model: model fit ... 125
6
1. Introduction
Imagine entering a store to experience a store associate who is not performing their job with enthusiasm, dedication, energy, and interaction. Even though organizations experience increased economic demands (Ruck & Welch., 2012), this is actually the state of one-third or even two-thirds of employees according to reports of AON Hewitt (2016) and Gallup (Adkins, 2016). Meaning, that for every three employees, one or two of them is not performing their job with their “hands, head, and heart” (Rich et al., 2010), which in return has a negative effect on business outcomes such as employee turnover, job performance, organizational citizenship behavior (Saks, 2006) and even revenue (Harter et al., 2013). This state of employees can be summed up as actively disengaged employees.
Employee engagement can be described as “a positive, fulfilling, work‐related state of mind that is characterized by vigor, dedication, and absorption” (Schaufeli et al., 2002) and even though the positive effects of internal communications on employee engagement are known, only half of all employees receive appropriate internal communication (Kahn, 1992; Welch, 2011).
1.1 Employee engagement in Retail
Generally, engagement amongst store associates of retailers is in line with the numbers given by Gallup and AON Hewitt. Meaning that at least one third is not actively engaged while performing their job (Kega, 2016, van Sante, 2016). Store associates are the spokesperson of retailers towards the customer. Having up to two-thirds of spokespersons who are not engaged at their job or even actively unengaged is seen as a real threat in retail (Kega, 2016, Adkins, 2016).
Next to the low engagement levels, it is difficult for retailers to communicate directly towards the store associate. In an interview with the Deputy Director of Personnel & Organization at Sligro, these difficulties are described as many retailers struggle with. According to van Sante, it
7
is difficult for corporate headquarters to reach the store associates as they mostly do not have a corporate email address or another personal way to reach them (van Sante, 2016). This lack of ways to communicate directly makes it difficult for the corporation to engage employees through internal communication (van Sante, 2016, Kega, 2016).
Dutch Fashion Retailer x works with a mobile communication platform that does make it possible to communicate on a personal level with store associates (Kega, 2017). Although there is much research carried out on the dimensions and consequences of employee engagement, little is known about how internal communication through new media (e.g. mobile communication) can influence engagement of employees (Welch, 2012). As the communication platform of Fashion Retailer x makes it possible to directly communicate with store associates, it is possible to study the effects of internal communication via new media on employee engagement. This study will therefore focus on the effect of such a corporate mobile communication tool on employee engagement. The main research question that is answered in this report is:
What influence has corporate mobile communication on employee engagement and what is the effect of engaged employees on the business results?
The thesis is set up as follows. First, the literature review is described with the state and construct of employee engagement, the role of internal communications on this engagement and especially the role of the mobile phone as a communicator, and lastly the consequences of engaged employees. In this literature review, the hypotheses will be described based upon the existing literature. Second, the research design is described with the conceptual framework and research method. Finally, the results, implications, and suggestions for further research will be described.
8
2. Literature review
2.1 Status of engagement worldwide
Worldwide engagement amongst employees has been growing. According to a report of AON Hewitt (2016), which is published yearly, 65% of all employees worldwide are moderately or highly engaged at their work. This is an increase of about 3% in comparison to 2015 and even 4% compared to 2014 (AON Hewitt, 2015, 2016). Engagement in Europe is lacking a bit behind with 60% of employees who are engaged in their work, compared to 58% in 2015. Europe has seen the biggest increase compared to the other continents in engagement in the last five years. The level of engagement went from 52% towards 60% in only five years. Zooming in on West-Europe, there is a difference in the percentage of engagement. Just more than half of the Western Europeans are engaged according to the report of AON Hewitt (2016). In other words, five out of ten Western European employees are passive or actively disengaged at their job.
According to a Gallup study in 2013, the state of employee engagement is worse compared to the numbers of AON Hewitt (Harter et al., 2013). The Gallup study shows that only 13% of employees worldwide are engaged at their jobs, which contradicts the numbers of AON Hewitt (2016). Gallup is a consulting agency who measures the state of engagement worldwide through their own employee engagement survey. They indicate that the topic of employee engagement will grow in importance as economic changes are rapid. Later Gallup surveys show that the percentage of engaged employees is growing but it is still fairly low with only one in three US employees engaged at their work (Adkins, 2016). Next to that, when comparing different groups the results show that Millennials, a group of people born between 1980 and 1996, are the age group that is least engaged at their job. “The millennial workforce is predominantly "checked out"― not putting energy or passion into their jobs. They are indifferent about work and show
9
up just to put in their hours” (Adkins, 2016). Even though the results of Gallup and AON Hewitt are not close, they do both indicate that there still is a large group of unengaged employees active in the workplace.
2.2 Employee engagement
The topic employee engagement is popular (Macey & Schneider, 2008). The last ten to fifteen years there has been a vast amount of research about employee engagement among organizations, scholars, and consultants (Saks & Gruman, 2014). Employee engagement is a popular term which describes the “simultaneous employment and expression of a person’s preferred self in task behaviors that promote connections to work and to others, personal presence, and active, full role performances” (Kahn, 1990). According to Kahn (1990), engaged individuals are engaged when they are committed in their work roles in a cognitive, emotional, as well as in a physical way. In comparison, disengaged individuals bring less commitment towards those three aspects. Engaged employees invest their “hands, head, and heart in their performance” (Rich et al., 2010). Christian et al. (2011) also describe the cognitive, emotional, and physical dimensions as the constructs for engagement.
In addition, when focusing on the psychological aspects, engagement is “a positive, fulfilling, work-related state of mind that is characterized by vigor, dedication, and absorption” (Schaufeli et al., 2002). A high engagement level leads to individuals that are “attentive, connected, integrated and focused” about their roles (Kahn, 1992).
Next to the psychological aspects of engagement in the workplace, Christian et al. (2011) mention that engagement is a state for employees to be in rather than a personality trait for employees to have. They, therefore, defined engagement as follows: “relatively enduring state of
10
mind referring to the simultaneous investment of personal energies in the experience of work” (Christian et al., 2011).
Since Kahn’s (1990) article about the engaged individual, quite a few articles in HR, psychology, and internal marketing practices have been published about this construct (Saks & Gruman, 2014). Even then, employee engagement was seen as one of the most important challenges for organizations (Wah, 1999). But although the first signs of research are more than thirty years old, there is still no consensus on the definition of employee engagement (Saks & Gruman, 2014). As a basis for this research employee engagement is defined as:” a positive, fulfilling, work‐related state of mind that is characterized by vigor, dedication, and absorption” (Schaufeli et al., 2002), as this definition is most mentioned in academic journals (Albrecht et al., 2015) and as the definition finds its origin in the cognitive, emotional, and physical aspects of engagement firstly described by Kahn (1990).
2.3 Comparable with engagement
Christian et al. (2011) have shown that engagement is not to be confused with a state of satisfaction, commitment or job involvement. Although the term satisfaction seems fairly similar to engagement there is a clear difference. Engaged employees are passionate and committed to helping the organization grow, whereas the construct of satisfaction is focused more on basic happiness, meeting expectations towards the organization (Macey & Schneider, 2008).
Engagement is a complete construct of the self and is shown in cognitive, emotional, and physical ways. Job satisfaction is much smaller (Rich et al., 2010). Satisfaction may still be relevant in measuring engagement. In fact, Christian et al. (2011) even found that engagement and satisfaction correlate (r 0,53). However, the construct alone is not enough to cover engagement (Macey & Schneider, 2008).
11
Commitment is different from engagement as commitment is about attachment to the company’s values and beliefs whereas engagement is more about the job at hand, the work to be done (Christian et al., 2011). Furthermore, commitment is more an emotional and cognitive state whereas as mentioned before engagement is also physical. Macey & Schneider (2008) again comment that commitment is a part of engagement but it is not the whole dimension.
Job involvement is cognitive and can be seen as how an employee identifies oneself with their job as well as during work as beyond work (Brown, 1996). Job involvement is not emotional or physical and represents a smaller aspect of an individual (Saks & Gruman, 2014). Therefore Job Involvement differs from engagement but can also be seen as part of employee engagement (Christian et al., 2011).
Christian et al. (2011) conclude that “engagement is unique although it shares conceptual space with job attitudes” (job involvement, satisfaction, and commitment). Next to Christian et al. (2011), Hallberg & Schaufeli (2006) also conclude that the constructs are empirically different. 2.4 Dimensions of employee engagement
Employee engagement cannot be captured by one dimension. As mentioned, Schaufeli et al. (2002) describe employee engagement in three dimensions: vigor, absorption, and dedication. 2.4.1 Vigor, dedication, absorption
When an employee shows high levels of vigor, the employee has energy and mental resilience in their working behavior. Vigor is a physical expression of engagement. Employees who experience a “sense of significance, enthusiasm, and challenge” (Bakker & Demerouti, 2008) have a strong dedication towards their job. Dedication is more an emotional expression of engagement. Lastly, being absorbed is more a cognitive expression of engagement and can be described as “fully concentrated and happily engrossed in one’s work, whereby time
12
passes quickly and one has difficulties with detaching oneself from work” (Bakker & Demerouti, 2008).
According to So et al., (2014), the construct of engagement consists of five dimensions. When all dimensions are high they lead towards highly engaged employees. These dimensions are enthusiasm, attention, absorption, interaction, and identification. (So, 2014). Three of these dimensions find their origin in the definition of engagement by Schaufeli et al. (2002). All dimensions are described below.
2.4.2 Enthusiasm
Harter et al. (2003) found enthusiasm to be an indication of engagement. This enthusiasm can be described as “strong excitement or zeal about the focus of engagement” (Vivek, 2009). In other words, people are passionate and visibly excited about something. When enthusiasm is high people tend to take risks, overcome obstacles and take ownership of their own actions (Glassman & McAfee, 1990; Vivek, 2009). Enthusiasm can be seen as a combination of the constructs vigor (physical) and dedication (emotional) as mentioned in the definition used in this research by Schaufeli et al. (2002). Vigor and dedication as well have a focus on strong excitement and enthusiasm as an indicator (Bakker & Demerouti, 2008). Therefore enthusiasm can be seen as a dimension of engagement (So et al., 2014).
2.4.3 Attention
“Attention refers to cognitive availability and the amount of time one spends thinking about a role” (Rothbard, 2001), making attention comparable with absorption as a positive mental or cognitive state in which the employee has the mental resilience to facilitate a high state of attention (Schaufeli et al., 2002). In a highly engaged state, people tend to focus on the job at hand, given it a great deal of attention (So et al., 2014). In the case of highly engaged employees,
13
this would indicate that they have a great focus on work and spend a significant amount of time thinking about it. An employee who is engaged with the company or the brand has a higher need for information related to the company (So et al., 2014). The employee is feeling attentive, integrated and has a high deal of focus when engagement is high (Kahn, 1992). Concluding that, attention is another dimension of engagement.
2.4.4 Absorption
Absorption is seen as a dimension of engagement by several researchers (So et al., 2014) and according to Rothbard (2001), is related to the construct attention, as they are both motivational aspects of engagement. When absorbed, employees have a high level and effortless state of concentration. They lose self-consciousness and track of time and enjoy themselves intrinsically (So et al., 2014). In other words, employees who arehighly absorbed in their job enjoy what they are doing, are in a state of concentration and are not focused on the end of their shift. Absorption is for that reason a dimension of employee engagement.
2.4.5 Interaction
Interaction is an employee’s participation with the company and their customers (So et al., 2014). According to Vivek (2009), interaction in customer engagement can be characterized by consumers who exchange ideas, feelings, and thoughts about a brand. This characterization can easily be translated towards the employee engagement construct. In which an employee also exchanges ideas, feelings, and thoughts towards the company and also the customer. Employees scoring high on interaction question the status quo and want a dialogue about their role (Macey & Schneider, 2008). In that respect, interaction is a dimension of engagement.
2.4.6 Identification
The last dimension is constructed from social identity theory in which individuals identify themselves and others around them in different social classes (So et al., 2014). According to
14
Bakker & Demerouti (2008), a strong identification with a job is a clear indicator of work engagement. Macey & Schneider (2008) underline this and also indicate that identification is an important aspect of employee engagement. Furthermore, Brown (1996) wrote that highly engaged employees have a strong identification with the job at work and beyond work. Therefore identification is added as the fifth dimension of engagement.
The study of So et al (2014) uses five constructs of engagement compared to the three constructs Schaufeli et al (2002) uses. Looking closely the study of So et al (2014) includes two new constructs as only interaction and identification do not have a strong overlap with vigor, dedication, and absorption. As the study of So et al (2014) introduces new aspects to the engagement literature, a combination of the dimensions of engagement of Schaufeli et al (2002) and So et al (2014) is used throughout this study.
2.5 Internal communication as antecedent of employee engagement
Kahn (1992) recognized communication to be a factor which contributes to employee engagement. Internal communication towards employees can even be seen as a critical antecedent of employee engagement. High-quality internal communication improves employee engagement, indicating that employees want high-quality communication to perform their tasks with high engagement. Next to that, unclear communication is seen as an important reason for disengagement amongst employees (Welch, 2011).
Despite these indications that engagement can be increased by effective internal communications, this construct has received little attention in research. Of the vast amount of research on employee engagement, in comparison, only a few contributions are made from the
15
perspective of internal communication. Hence, internal communications research is needed (Welch, 2011).
There are many types of internal communication, from office chit chat to communication on a corporate level (Welch, 2011). Welch & Jackson (2007) describe a framework of four types of internal communication which includes: “line management, team peer, project peer, and internal corporate communication.” The possible effects of internal corporate communication and peer to peer communication are described further.
2.5.1 Internal corporate communication
Internal corporate communication has the potential to successfully transfer the goals and values of an organization and to immerse employees in these goals (Welch, 2011). It can even lead to higher engaged employees (Welch, 2011). From a practitioner’s perspective, corporate communication can “reinforce the commitments to employees and customers” (Gallup Consulting, 2010). Performed badly, internal communication also has the power to weaken the organizational relationship between employee and organization (Welch, 2011).
When communicating to and with employees, the focus of internal communication should be about personal and job-related subjects. Next to that, employees want more information about decisions made and room to communicate back to and evaluate management (Goldhaber et al., 1978). However, only half of all employees are informed by management about what is going on in their organization (Welch, 2011).
2.6 New media as communication tool
As mentioned, internal communication can be a factor of employee engagement. For this to happen, the communication efforts need to be appropriate and used in formats that reach the employees (Welch, 2012). Apparently, as only half of all employees are reached by their
16
organization (Welch, 2011), the current communication types are not sufficiently appropriate and evidently, do not reach all employees. A possible way to improve this is through the use of new technologies like electronic media. Electronic media, as for example a mobile phone, can provide employees without access to other media, a chance to receive internal corporate communications directly and not through management (White et al., 2010). Technology has had a major impact on the way business is performed and how companies communicate with their employees. Web 2.0 has given way for several new communication tactics (Constantinides et al., 2008). However, the effect of mobile phone communication towards employees has received very little attention in research. It does seem that electronic media, such as the use of a mobile phone to communicate with employees, is gaining in popularity (Welch, 2012). De Bussy et al. (2003) found that the use of new (electronic) media in corporate communication has a positive effect on the effectiveness of the communication. Part of this is because the use of electronic media in communications facilitates two-way communication between the organization and the employee (Bussy et al., 2003).
This electronic media could be a good way to reach employees in a useful and acceptable format. As the density of smartphone users in Europe is likely to grow up to 70 or 80% in the next few years (The Mobile Economy, 2015 p13) this electronic media type might be an effective way for internal communication and thus a way to directly reach more employees and increase engagement.
In interviews with retail experts, it seems that the lack of communication is especially high amongst retailers with a great distance between headquarters and physical stores which are spread around the country (van Sante, 2016; Verkade, 2016). In comparison with office based businesses, retailers have less access to the low-level employees as they mostly do not have
17
access to direct communication as a computer or a company e-mail address. According to Verkade (2016), CEO of retail agency Kega, most of the communication towards the store associates goes via a store manager with little to no means to interact directly with this level of employees. In other words, the place where policies are made and the people who carry out policies hardly interact with one another. As mentioned, internal communication has an effect on the level of engagement (Kahn, 1992; Welch, 2011), but for this effect to take place, communication with the employees must exist and be appropriate (Welch, 2012). As for retailers with store associates without or low-level access to direct sources of internal communication, new media communication via a mobile can be an appropriate way to communicate to and with employees and thus increase the engagement of these store associates.
2.7 Hypotheses
The literature about employee engagement, internal communications, and the information about communication with operating employees leads towards the first hypothesis:
1) Internal communication via mobile phone has a significant positive effect on employee engagement.
To measure the employee engagement level a combination of the dimensions of So et al. (2014) and Schaufeli et al. (2002) is used to test five dimension of engagement as well as the overall level of employee engagement. Each of these dimensions is tested separately to be able to discover which dimension has the biggest impact on employee engagement via mobile corporate communication. The effect of communication via a mobile phone is compared to the five separate dimensions of employee engagement leading towards the following hypotheses:
18
1a) Internal communication via a mobile phone has a significant positive effect on the level of dedication, vigor, and absorption off employees.
Although interaction is not mentioned by Schaufeli et al (2002), research performed later on does imply that interaction is a dimension of employee engagement (Macey & Schneider, 2008). The platform of a mobile phone could make it easier for employees to interact with headquarters as well as with co-workers (Welch, 2011; Anitha, 2014). The several modules of the mobile communication tool of Fashion Retailer x are an example of how interaction is made easier by such an instrument. Therefore the following hypothesis is constructed:
1b) Internal communication via mobile phone has a significant positive effect on the interaction level of employees.
Next to general interaction between the corporation and employee, the dimension identification is also mentioned as a construct of engagement in papers about this subject appearing after the Schaufeli et al (2002) paper (Bakker & Demerouti, 2008.; Macey & Schneider, 2008). Furthermore, it is interesting to see whether internal communication via the mobile phone enables employees to create a higher level of identification as this tool provides the possibility to engage with work beyond work (So et al., 2014). The hypothesis for identification is:
1c) Employees have a higher level of identification with the company when receiving internal communication via the mobile phone.
2.7.1 Corporate mobile communication tool
To measure these above-mentioned effects, a sample is needed of employees who have experience with a mobile internal communication tool. Fashion Retailer x, a fashion retailer based in The Netherlands, uses such a tool in the form of a mobile application (Kega, 2017). This
19
mobile application, called InTouch, is focused on providing internal communications via the mobile phone of individual employees. Fashion Retailer x is able to communicate via different modules in the InTouch app. As this corporate mobile communication is used to specify several hypotheses the components of this communication tool are described below:
News: a news timeline comparable with Facebook where employees can like and react on news items. Latest news about the brand, company, and employees is posted here by different departments of Fashion Retailer x headquarters.
Your Biggest Fan: comparable to the news timeline described above but even more interactive as employees themselves can post news items in the form of text and pictures, creating a way to interact with other employees and headquarters.
Chat: a group and one on one chat function, comparable with WhatsApp which enables employees to directly interact with anyone (including the CEO) of the company.
To-do: the To-do function enables employees to very functionally see their tasks at hand. Next to that, the module shows the birthdays of all employees.
Poll: the Poll creates the possibility for the company to interact with multiple employees and ask their opinion about work related issues as for instance marketing plans or HR topics.
Figures: a menu item where on a daily basis the sales numbers of the previous day are posted. These are the numbers of an employee’s own store and the means of comparable stores.
My Suggestion: a menu item where employees can directly suggest an idea towards the appropriate department, making it possible to exchange ideas, feelings, and thoughts about the brand.
20
Videos: here videos about the brand are placed. This can be commercials and social video’s originally designed for the consumer but also video’s with news and (training) instructions. Frequently asked questions (FAQ): a menu item in which the frequently asked questions of several departments are visible. This menu item also offers the possibility to interact directly with the regarding department if a question is not answered.
My HR: a portal towards the HR web page of the company. This menu item makes it easier for employees to directly access information about their paychecks and other HR related items.
Training videos: a menu item comparable to ‘video’, where employees can reach all posted training videos and keep up to date with training.
2.7.2 Specific hypotheses
As mentioned, the effect of internal communication via the mobile phone on engagement will be measured via five dimensions of employee engagement (vigor, dedication, absorption, identification, and interaction) (Schaufeli et al., 2002; So et al., 2014). Next to that, as the modules who are described above differ, different impacts on the several dimensions of employee engagement are expected. Therefore specific hypotheses are defined. For instance, the use of the module Figures could have a bigger impact on the dimension dedication, as employees who often use the module Figures are more likely to have a “sense of significance, enthusiasm, and challenge” (Bakker & Demerouti, 2008), which fits with highly dedicated employees.
1d) High usage of the menu item Figures has a higher positive effect on the dimension dedication compared to the other dimensions of employee engagement.
Another and logical example can be made with the dimension interaction. Several modules (Chat, My Suggestion, News, YBF, and Poll) provide employees the opportunity to interact with
21
the company and colleagues. Especially the possibility to react on and ‘like’, messages makes it easy to interact with the company. It is likely that making use of these possibilities will have a stronger effect on the dimension interaction instead of other dimensions of employee engagement. This leads to the hypothesis:
1e) High usage of the react and ‘like’ possibility has a higher positive effect on the dimension interaction compared to the other dimensions of employee engagement.
Being absorbed in work means that, “one has difficulties with detaching oneself from work” (Bakker & Demerouti, 2008). The mobile app from Fashion Retailer x enables employees to be physically absorbed with work outside of work, leading to the last specific engagement hypothesis:
1f) High usage of the mobile communication tool in general has a higher positive effect on the dimension absorption compared to the other dimensions of employee engagement.
2.7.3 Peer to peer communication
In a study of the antecedents of employee engagement, Anitha (2014) found that amongst other factors, team and co-worker interaction has a big impact on employee engagement. This type of interaction even explains 36% of the construct, showing the importance of a positive relationship between co-workers (Anitha, 2014). Kahn (1990) also wrote that worthwhile interactions with co-workers have a positive effect on engagement. But with regard to the current theme of the mobile internal communication platform as a communicator, little is known about team and co-worker interaction via the mobile phone (instead of directly in the workplace) and its direct or indirect effect. Next to the general interaction between the corporation and employee, it is interesting to see whether the mobile phone enables a higher level of engagement via team and
22
co-worker communication. In previous research, team and co-worker communication is proven to have a direct link to employee engagement (Anitha, 2014). Since in this research the effect of peer to peer communication is measured through internal communication via the mobile (Chat module in the Fashion Retailer x application), team and co-worker communication is seen as an indirect effect on employee engagement. As the internal communication tool via the mobile makes it easier for team and co-workers to interact outside of the workplace, it is possible to measure this effect. Consequently, the following hypothesis is setup:
H2: Team & co-worker communication through a corporate mobile tool has a positive effect on employee engagement.
2.7.4 Antecedents of employee engagement
It makes sense that the engagement and disengagement of employees are influenced by the internal communication of the firm towards the employees (Bakker et al., 2011). However, there are several other aspects that are considered to be an antecedent of engagement. For instance, job characteristics, type of leadership and personal characteristics (Anitha, 2014; Christian et al., 2011; Gruman & Saks; 2011; Macey & Schneider, 2008; Saks, 2006). As the focus of this study is to measure the effect of internal communication via the mobile phone on employee engagement and as several other antecedents are already well known and studied, these antecedents will not be a part of this research because of practical limitations (time frame research, agenda Fashion Retailer x and focus on internal communication). On the other hand, peer-to-peer communication, which is already known as an antecedent for employee engagement, will be measured as the current study provides the means to test if the effect is also present under such conditions.
23 2.7.5 Net Promoter Score
Next to the five dimensions of employee engagement, this study also measures the Net Promoter Score (NPS) amongst employees. The NPS measures customers’ intention to recommend a service or product to friends and colleagues resulting in a single positive or negative number which can be easily interpreted (Reichheld, 2003). As the measure is useful for several stakeholders of a company, the measure is used to get a NPS score of the employees of Fashion Retailer x. Claims are made that the NPS has a direct link with the level of engaged employees (AON Hewit, 2015) and thus it is interesting to test whether such a link exists.
2.7.6 Consequences of employee engagement
Besides the antecedents, the consequences of employee engagement are also explored. Several studies about engagement suggest that engagement has an influence on organizations bottom-line results (Saks & Gruman, 2014; Macey & Schneider, 2008). Additionally, having engaged employees is positively related towards organizational citizenship behavior, job performance, attitude, and wellness outcomes (Bakker, 2008; Cole et al., 2012; Christian et al., 2011; Rich et al., 2010; Saks, 2006). Furthermore, engagement can lead to lower employee turnover (Saks, 2006; Schaufeli et al., 2002). Other consequences of high employee engagement can be found in business outcomes. According to Harter et al. (2002), engagement amongst employees has a positive effect on customer satisfaction, profitability, productivity, turnover, and safety in the company. Even greater return on assets, higher employee performance, and shareholder value have been found to be positively impacted by employee engagement (Anitha, 2014; Macey & Schneider, 2008). The difference between highly engaged companies and highly disengaged companies are significant according to Gallup (Harter et al., 2013). The top-quartile and bottom-quartile of their study show big differences in measurable business outcomes: 10% higher customer ratings, 22% higher profitability, lower safety incidents (48%) and so on (Harter et al.,
24
2013). Maybe even more notable, Macey & Schneider (2008) found that having highly engaged employees can be a form of competitive advantage over your competitors. The factor people is very difficult to imitate and highly engaged employees are therefore highly valuable (Saks, 2006).
All this leads to the next hypothesis:
3) High (vs. low) engaged employees have a higher positive effect on corporate performance factors, such as: employee turnover, organizational citizenship behavior, job satisfaction, and organizational commitment.
25
2.8 Conceptual framework
In summary of the literature review, employee engagement is an individual’s cognitive, affective, and behavioral state towards an organization. The construct of engagement consist of five dimensions; vigor, dedication, absorption, interaction, and identification, which can be positively affected by appropriate communication (Schaufeli et al., 2002; So et al., 2014; Welch, 2011). When these dimensions score high, employees are engaged in their work and this, for example, results in, desired organizational outcomes as, job satisfaction, organizational commitment, lower employee turnover, and desired behavior towards the end-user (Schaufeli et al., 2002; Saks, 2006).
Based on the hypotheses formulated in the literature review, the conceptual framework is set-up:
26
3. Research design
In the following paragraph, the research design of this study is described.
3.1 Data collection and design
In order to test the formulated hypotheses, a quantitative research is conducted. The research is performed amongst store associates of fashion retailer Fashion Retailer x. Fashion Retailer x is a fashion retailer with a focus on jeans and denim. This retailer is chosen because they have been working with a corporate mobile communication tool called the InTouch App since early 2017 (Kega, 2017). The application is available for every employee with a smartphone. This research consists of a survey based research and database research and is thus a mixed method research type. The survey-based research is a cross-sectional study with two separate measures three months apart. The database-based research can be seen as longitudinal as data is acquired over a period of three months via Google Analytics (2017). For the survey, data is gathered twice from the same population. The aim of the study is explanatory as the goal is to explain the effect of corporate mobile communication on employee engagement. As the data collected for this research is collected especially for this project, the research will consist of primary data.
3.2 The research procedure
As the goal of the study is to observe what naturally happens with employee engagement levels when corporations communicate with store staff via a mobile phone, the design of the study is correlational. The study is survey-based and is complemented with database data collected from the usage of the mobile employee application of Fashion Retailer x. During the months February till April application usage is obtained in order to cross check the usage data collected from the survey. By doing this, the validity of the study is improved as conclusions are made from two sources.
27
In January via a survey, a baseline measure is performed to measure the engagement levels of Fashion Retailer x employees before the introduction of the mobile communication tool. In May another survey is set up to, again, test engagement levels and to be able to test the hypothesis and see if there are any significant differences between the baseline and second measure. Next to that, the differences between non-users, low users, and high users are tested through several analytical tests performed on IBM’s SPSS and IBM’s AMOS. By using SPSS and AMOS, tests are performed to examine if the usage levels of the InTouch App leads to significantly different engagement levels.
The second goal of this research is to analyze the relationship between the engagement level of the employee and the effect on their personal performance as an employee. These consequences are measured through survey and database data and are used to test the impact of the different employee engagement levels.
3.2.1 Survey
For the pilot measure, a random sample is selected as 100 employees were invited to join the study. For the overall measure, no sampling technique is used in this research as the entire population received an e-mail with the request to participate in the study. Every employee of Fashion Retailer x had the chance to participate in this study.
The population consists of employees of Fashion Retailer x. Fashion Retailer x has nearly 100 owned stores in the Netherlands and around 685 employees. Previous research amongst store associates from Fashion Retailer x indicates that a 50% till 60% response rate is likely (Vermeer, 2016). The survey is available in Dutch. Some parts of the survey are translated to Dutch by peers and are cross-checked by peers.
28
The measures used in this survey can be split into four subjects; employee engagement, consequences of engagement, internal communication, and demographics (including usage of mobile communication).See attachment A for the invitation and questionnaire of the survey.
Employee engagement consists of 5 dimensions that are measured on a 7-point Likert scale. The original Likert-scale starts at 0 (NEVER) through 6 (ALWAYS). The original Likert scale is adjusted to 1 (Strongly disagree) through 7 (Strongly agree) so that the several Likert scales used in this study match. Next to that, the engagement dimensions identification and interaction cannot be measured via the original Likert scale from Schaufeli & Bakker (2003), so one overall Likert Scale is chosen. For this measure, the questions of the Utrecht Work Engagement Scale (UWES) (Schaufeli & Bakker, 2003) are adopted to measure the dimensions vigor, absorption, and dedication (Cronbach's α = .93). This scale is used as the authors are highly cited on the topic of employee engagement. An example of the UWES scale is the question: ‘At my job, I feel strong and vigorous’ which is one of the three statements about the dimension vigor. As Fashion Retailer x is a Dutch fashion retailer, the Dutch translation of the questionnaire, provided by Schaufeli & Bakker (2003), is used. Because of practical limitations, the question; ‘I get carried away when I’m working’ is adjusted from the original translation to Dutch as the translation of this question is not approved by Fashion Retailer x.
The dimensions interaction (Cronbach's α = .94) and identification (Cronbach's α = .86) are measured via the scale developed by So et al. (2014). As the scale originally measures customer engagement, it is adapted to a scale for employee engagement. An example of the So et al. scale is the question: ‘When someone criticizes this brand, it feels like a personal insult’, which is one of the four statements about identification.
29
So et al. (2014) use a different Likert-scale compared to the UWES-scale (1 = strongly disagree and 7 = strongly agree). As mentioned, one consistent Likert scale is chosen to increase the readability of the survey. As the questions of the dimensions interaction and identification cannot be measured on the 7-point Likert scale from Schaufeli & Bakker (2003), the Likert-scale of So et al. (2014) is used for this research.
As mentioned, the NPS is measured for the employees of Fashion Retailer x. ‘How likely is it that you would recommend Fashion Retailer x as an employee to a friend or colleague?’ is the question that is used to measure the NPS. Normally the NPS is only considered to measure a customer’s intention to recommend, but for this study, the NPS will be used to measure the extent to which employees would recommend their company as a place to work towards their friends and family.
3.2.1.1Questions baseline measure
A baseline measure is conducted amongst a sample of employees of Fashion Retailer x. For this baseline, only the employee engagement questions are asked in order to obtain a baseline result of the current engagement level at Fashion Retailer x. Furthermore, the NPS is also measured to identify if significant differences appear after experiencing corporate mobile communication.
3.2.1.2 Antecedents and consequences
In the study about antecedents and consequences of employee engagement, Saks (2006) adopted several scales to measure these concepts. For this research the consequences scale is adopted to measure the following consequences; job satisfaction, organizational commitment, intent to quit, organizational citizenship behavior and perceived supervisor support. As mentioned, Saks (2006) adopted several scales to measure these consequences (Cronbach’s α > .75). An example of this scale is: ‘I frequently think of quitting my job.’
30
The internal communication questions consist of questions about the usage level of the InTouch application and are only asked to employees who use the application. These questions are self-made and will be reviewed by peers. They are asked in order to statistically test whether the different items have a significant effect on the several dimensions of employee engagement. An example of a usage question is: “Indicate how many times you used the following menu items in the last two weeks”.
The demographics asked via the survey are adopted from the research platform Qualtrics. A question about the job-level of the employee is added in order to be able to analyze between different job levels. Furthermore, the usage level of the InTouch app in general and the several modules of the app are identified in order to statistically test whether the different items have a significant effect on the several dimensions of employee engagement. Questions about the module Chat will be used to be able to test the hypothesis about the effect of team and co-worker communication via a corporate mobile communication tool.
Because of practical limitations, the original scale of identification and all the scales of the consequences of employee engagement are adjusted so that the questionnaire takes less time and the completion ratio will be higher. The entire questionnaire had a test run amongst co-workers and the HR-manager of Fashion Retailer x, who work with the mobile communication application and thus are adequate testers.
3.2.2 Database
The sample for the database study consists of the population of users of the InTouch application as every usage of the application is measured. The variables that are measured over a period of three months are the usage levels of the InTouch Application and the usage of the different modules of the application. The database consists of one source, Google Analytics.
31
4. Results baseline
For the pilot measure, the goal is to acquire a baseline of the engagement level of employees at Fashion Retailer x. This pilot is conducted before Fashion Retailer x started an internal communications program via the mobile phone. A link is sent to 100 employees personal email accounts with the request to fill out the anonymous survey (see attachment A1 & A3). From the 100 invitations, 66 employees completely filled out the survey. As Fashion Retailer x has about 650 store-based employees, the sample consist of more than 10% of the population and is a decent representation. The sample age ranges between 15 and 43, and 77% of the respondents are women. The largest group of respondents has the function, Store Employee (87.9%), which is also the biggest group in the total population of Fashion Retailer x employees. The rest of the sample consists of Assistant Store Managers (4.5%) and Store Managers (7.6%).
4.1 Reliability measures Pilot-measure
All measured variables for employee engagement in the baseline measure were tested for their reliability in order to make sure that the measurements were consistent with one another. Below the reliability of the five dimensions of engagement are described. See Table 1 for an overview.
Identification has a high Cronbach’s Alpha (α .780) indicating that the measurement is reliable. Next to that, deleting any item of this measure does not show a significant increase in reliability. Furthermore, the corrected item- total correlation shows that all items have a good correlation (r > .480).
Studying the measure interaction shows that the reliability is also high (α .750) and the item correlation is high enough (r >.380). The scale reliability could be improved by excluding measure IT2 (α .869), but as the scale only consists of three items, and the original Cronbach’s Alpha is high enough (α >.700), the choice is made not to exclude IT2.
32
Looking at the reliability of the engagement dimension, vigor, it shows that the reliability is very high (α.823), the item correlation is also good (r >.655), and the measure will not improve by excluding one of the three items.
The Cronbach’s Alpha of the dimension dedication is also very high (α.832) and again, item-correlation is high (r >.60), and no improvement will be made by excluding one of the scale items.
The item-reliability for the dimension absorption is not very high but also not extremely low (α.641). The item-correlation is good with correlations above .3. No improvement can be made by excluding one of the three items used to measure the absorption levels of employees. Overall, the reliability of the items is high, only absorption is a bit to low. Overall, four out of five items have a sound reliability. The fifth item is less reliable and hence, this will be taken into consideration when interpreting the results.
4.2 Mean
To be able to further measure the several items, the mean values for all the dimensions of employee engagement are computed into total scores per dimension. Testing the reliability of those total scores combined shows that the Cronbach Alpha stays high with several combinations of engagement measures (α >.754). Looking further at all the engagement dimensions combined, the item correlation is high (r >.55), and the measure, with a current Cronbach Alpha of .885, can not be significantly improved by excluding one of the dimensions.
A correlation matrix (see Table 1) shows that the correlations differs from average correlation (r = 0.36) to high correlations (r = 0.76). All the correlations are significant (ρ <0.01).
33 Table 1: Correlations and reliability engagement dimensions (n = 66)
Variables Mean SD 1 2 3 4 5 1. Absorption_Total 5.67 0,85 (.64) 2. Vigor_Total 5.71 0.85 .67** (.82) 3. Dedication_Total 5.84 0.91 .74** .76** (.83) 4. Identification_Total 5.29 1.10 .71** .68** .62** (.78) 5. Interaction_Total 5.34 0.93 .36** .55** .42** .61** (.75)
** Correlation is significant at the 0.01 level (2-tailed)
4.3 Testing Normality
According to the Kolmogorov-Smirnov test the engagement scales, measured in this pilot study, are not normally distributed as the test is significant (ρ < 0.05). The low p-value suggests that there is enough proof that the engagement scales are not normally distributed. Looking at the Skewness and Kurtosis scores, it is also visible that the scales are not normally distributed. As well as the Skewness scores as the Kurtosis scores are not close to zero. They all exceed the rule of thumb between -1 and 1 (Muthen & Kaplan, 1985).
4.4 Current engagement levels
Schaufelli & Bakker (2003) developed a norm to categorize the scores on the UWES. This norm is used and adjusted to be able to categorize the current measure. The scores are increased by one so that they are in line with the current measuring scale. As for the dimensions identification and interaction, no such norm is available. Therefore, only the norm of the total score of the UWES is used to categorize the five used dimensions of engagement. See Attachment D1 for an overview of the changes.
34 Table 2: Scores Engagement in percentages (5 dimensions) (n = 66)
Engagement measures Five
dimensions UWES Very low 1.5% 0% Low 3.0% 3,0% Average 47.0% 25.8% High 43.9% 62.1% Very High 4.5% 9.1% Total 100% 100%
As can be seen in Table 2, the baseline employee engagement levels are quite high when the scores are compared to the numbers given by Gallup (Harter et al., 2013) and comparable with the numbers of AON Hewit (2015, 2016). About 48% of the sample is highly engaged at their work. 52% of the sample is not engaged, but only a very small percentage (4.5 %) is actively disengaged at their job.
Comparing these numbers with the original UWES engagement scale (see Table 2 and 3) it shows that the engagement levels are numerically higher when only measuring the original three dimensions.
Table 3: Mean scores Engagement measures (n = 66)
Means engagement
measures M SD n Skewness Kurtosis
Five dimensions 5.55 0.79 66 -1.46 3.07
UWES 5.74 0.79 66 -1.42 2.68
4.5 Net Promoter Score
In the baseline measure, the Net Promoter Score (NPS) is asked. Looking at the NPS scores, the Promoters are the bigger group compared to the Detractors, resulting in a positive NPS. Table 4
35
shows the distribution of the NPS and it can be seen that the majority of the respondents are Passive employees, about 58%. 32% is a Promoter, and the rest are Detractors, resulting in a positive NPS of 21.2.
Table 4: Net Promoter Score (n = 66)
NPS Frequency Percentage Cumulative
Detractor 7 10.6 10.6
Passive 38 57.6 68.2
Promoter 21 31.8 100
Total 66 100
Comparing the NPS with the total employee engagement score there is a significant (ρ <0.01) correlation of .603, meaning that there is a positive relationship between ranking the NPS and an employee’s state of engagement. A regression analysis shows that in this sample, 52.2% of the variance of employee engagement is explained by the NPS (ρ <0.01). When the NPS of an employee goes up by one, employee engagement is increased by .327.
Table 5: Regression analysis, Net Promoter Score (n = 66)
Model R R2 R2 Adj. B SE β T
1 .728 .530** .522
Net Promoter Score .327 .039 .728 8.491
Note. Statistical significance: ** p <.01
Exploring more possible relations with employee engagement shows that there is also a medium positive correlation between the engagement scale and type of function of the employee (r = .345 ρ < 0.01). In other words, employees with a higher position are more likely to be highly engaged.
36
5. Results total measure
Three months after introducing the internal communication tool at Fashion Retailer x, a second survey is held among the employees of Fashion Retailer x. Via a personal e-mail, 685 employees were requested to fill out the anonymous survey. Next to the e-mail, a link to the survey was posted on the company’s intranet. This was done to make sure that every employee could fill out the survey. From the 685 e-mails, 219 employees took part of the study. Furthermore, 17 employees participated via the link posted on Fashion Retailer x’s intranet. See attachment A2 for an example of the invitation. From the 236 respondents, 83% is female and the average age is 29 with a range between 16 and 63. The largest group is Store Employee (49%), followed by Store Managers (21%) and Assistant Store Managers (11%). The last 19% of the respondents consists of employees from the headquarters of Fashion Retailer x. The overall spread of the respondents is different from the baseline as can be seen in Table 6. In relation to the pilot, in the total measure, the percentage of non-managerial store staff is lower. Further one in the research the differences in engagement levels between the managerial and non-managerial layer are discussed.
Table 6: type of position pilot (n=66) vs. total measure (n = 236) in percentages
Position Pilot Total
Store Employee 87.9% 49%
Assistant Store Manager 4.5% 11%
Store Manager 7.6% 21%
Headquarters 0 19%
Total 100 100
5.1 Reliability measures
Next to the engagement scales, the several consequences of employee engagement are tested for their reliability. Consistent with the pilot measure, all engagement scales have a high Cronbach
37
Alpha (α > .700) and have a strong positive correlation (r >.53 ρ <0.01) between each other as can be seen in Table 7. Where in the pilot measure, the Cronbach Alpha of the dimension absorption was relatively low (α .64), in the final measure the reliability increased to a more acceptable value (α. 74). Taking a further look at the scales of the measured consequences of employee engagement (see Attachment B2), it can be seen that all scales have a high Cronbach Alpha (α > .75), can not be significantly improved by removing items from the scales, and are moderately to highly correlated with each other (r > .45 ρ < 0.01). Indicating that the scales are reliable.
Table 7: Reliability engagement scales final measure (n = 236)
Variables Mean SD 1 2 3 4 5 1. Absorption_Total 5.53 0,98 (.74) 2. Vigor_Total 5.62 1.06 .73** (.88) 3. Dedication_Total 5.81 1.06 .76** .83** (.90) 4. Identification_Total 5.44 1.13 .67** .64** .68** (.77) 5. Interaction_Total 5.30 01.06 .61** .54** .68** .59** (.80)
** Correlation is significant at the 0.01 level (2-tailed)
5.2 Testing Normality
The data in the pilot measure is not normally distributed. The data in the final measure is more close to a normal distribution in comparison to the pilot. The Kolmogorov-Smirnov test indicates that the engagement scales, as well as the measured consequences, are not normally distributed (ρ < 0.05). According to Field (2009), the reliability of tests as the Kolmogorov-Smirnov is not high as in large samples and high Skewness and Kurtosis values are likely even when the distribution is close to normal. Furthermore, Tabachnick & Fidell (2001) point out that with large
38
samples (n > 200) risks of non-normality are reduced. Field (2009) also indicates that large samples reduce the level of Skewness and Kurtosis and thus normality levels due to the Central Limit Theorem. This is supported by Norman (2010) who also indicates that the Central Limit Theorem ensures approximately normal distributions. Next to that, Schmider et al. (2010) indicate that Skewness levels below 2.0 and Kurtosis levels below 9.0 are sufficiently normal to perform non-parametric tests. As all Skewness and Kurtosis levels meet these standards and the sample is large (n = 236), no further action is warranted to improve the normal distribution. In other words, the data is fit to perform parametric tests.
Graph 1: usage levels most used items in percentages (n = 236)
5.3 Use of internal communication application
The study shows that after three months of introducing the internal communication application at Fashion Retailer x, 78% of the employees use the application and 22% does not. Of that 78%, about half (45%) use the app more than four times a week. Comparing the usage levels with the
27,17 18,92 13,11 9,24 5,95 16,85 10,81 6,56 5,98 6,49 30,43 21,62 15,30 15,76 25,41 21,20 16,76 15,85 14,67 21,62 4,35 31,89 49,18 54,35 40,54 0 10 20 30 40 50 60 70 80 90 100
News Figures To do Chat YBF
Never Once a week 2-3 times a week 4-6 times a week Daily
39
Google Analytics database shows that the number of application users is higher in the survey compared to the actual percentage of users. According to Analytics, the last ten weeks the application had around 300 unique users per week. This is less than 50% of all employees. In short, the data shows that the number of non-users is not comparable with the actual usage levels (Google Analytics, 2017).
Graph 1 shows the usage levels of the five most used menu items available in the application. The remaining twelve menu items are used much less by the respondents. At least 75% of the users responded that they do not use or only use those particular menu items once a week. Compared to the statistics from Google Analytics (2017), the most used items match. For further information on the usage of all menu items, see Attachment B3.
The main reason for not using the app is that respondents do not know of the existence of the mobile communications tool used by Fashion Retailer x. About half of the non-users (47%) are not familiar with the employee application. Next to that, about 14% of non-users do not want work and privacy to get intertwined with each other. 11.76% of non-users think that the application does not have added value to them and about 12% does not have a phone that supports the application.
5.4 Engagement levels final measure
The employee engagement levels of the final measure are displayed in Table 8. As can be seen, only 8.5% of employees are disengaged at their jobs. More than a third is not engaged or disengaged and more than half of all employees are indeed engaged at their job. A comparison between the pilot, total measure, and total measure filtered on users of the internal communications tool is shown in Table 8 and Graph 2.
40 Table 8: Employee engagement levels pilot (n = 66) vs. total measure (n = 236)
ENG_CAT Pilot Total
measure Very low 1,5% 1,3% Low 3% 7,2% Average 47% 38,6% High 43,9% 43,2% Very High 4,5% 9,7% Total 100% 100%
To test the differences between the pilot measure and the final measure an independent samples t-test is performed in which the assumption of homogeneity of variances is tested and satisfied (F= .323, ρ > 0.05). To be able to perform the t-test, not the categorical groups of engagement as shown in Table 8 but the individual scores of employees are used (the means of these scores are shown in Table 9). Comparing the engagement levels between the pilot measure (n= 66) and the final measure (n= 236), the independent samples t-test shows that there are no significant differences between the groups (ρ > 0.05). Indicating that the engagement levels are not statistically different between the pilot and final measure. The engagement level of the pilot measure was indeed almost the same (M= 5.55, SD= 0.79) as compared to the final measure (M= 5.54, SD= 0.90), thus significant differences were not expected. Although the means are very similar and no significant differences are found, the percentage of very highly engaged employees doubled in the period between the pilot and final measure.
41 Table 9: means scores employee engagement pilot and total measure
ENG_Tot M SD n Skewness Kurtosis
Pilot 5.55 0.79 66 -1.46 3.07
Total; users and
non-users 5.54 0.90 236 -1.33 2.14 Total; only users 5.65 0.84 185 -1.26 2.22 Total; non users 5.14 1.02 51 -1.39 1.27
5.4.1 Engagement level users mobile communications tool
Comparing the engagement levels between the pilot measure and the final measure filtered on employees who actually use the internal communication app (n = 185, M = 5.65, SD = 0.84),
again no significant differences are found (ρ > 0.05) via the independent samples t-test in which the assumption of homogeneity of variances is tested and satisfied (F = 0.323, ρ > 0.05). Even though there are no significant differences, it is interesting to indicate that the percentage of very highly engaged employees gets even higher when filtered on users of the application. Compared with the pilot measure, the percentage of very highly engaged employees tripled in three months as can be seen in graph 2.
