Perceived servant leadership and job satisfaction among culturally diverse restaurant employees

(1)

Perceived Servant Leadership

and Job Satisfaction among

Culturally Diverse Restaurant

Employees

Céline Fauvelle

Stockholm Business School

Master’s Degree Thesis 30HE credits Subject: Management

Program: Operations Management and Control Spring term 2017

(2)

Acknowledgements

I would primarily like to give a warm thank you to my academic advisor Alexander Chakhunashvili for steering me in the right direction, for the time he has committed to provide patient guidance and support for my statistical calculations and for the helpful, prompt and kind replies that I received whenever I contacted him. I would also like to thank my uncle Davis Hayden for his valuable help with my struggles regarding SPSS, for the hours he spent with me to come up with the most interesting statistical analyses and for sharing his knowledge with me. To Isa Abell and Alice Maurel, a huge thank you for challenging my ideas, providing guidance, for always being sensible and for all the support and motivation. Finally, I want to thank my father for always helping me see things clearly and my mother for all her encouragement and for having faith in me.

Thank you! Céline

(3)

Abstract

High employee turnover is a costly problem for the restaurant industry. Previous research has shown that job satisfaction is one factor that is negatively associated with employee turnover in different fields and that the leadership style servant leadership is positively related to job satisfaction. Servant leadership is a management style where the leader shares power, puts the needs of others first and helps employees develop and perform as highly as possible. The relationship between job satisfaction and servant leadership had not previously been examined among culturally diverse restaurant employees across organizations, and that was the purpose of this study.

The research question asked was: Is there a statistically significant correlation between the perception of servant leadership and job satisfaction for culturally diverse restaurant employees? To answer this question, 45 restaurant employees spread over 7 cultures completed a survey with a demographic questionnaire and short forms of the Servant Leadership Scale and the Minnesota Job Satisfaction Questionnaire. The data was analysed through the utility of SPSS and a correlation analysis revealed a statistically significant relationship between perceived servant leadership and job satisfaction among culturally diverse restaurant employees (r = +0.692; p < 0.01). The result suggested that the use of servant leadership in restaurant organizations may increase the job satisfaction among employees of different cultural backgrounds and thereby possibly reduce employee turnover.

The findings of the study contributed to the body of knowledge regarding the utility of servant leadership in the multicultural restaurant industry and should be of interest to research in the fields of restaurant management, servant leadership and cross-cultural leadership.

(4)

1. Introduction

The stereotypical view of restaurants shows a culturally diverse field with working conditions and corporate cultures characterized by strict hierarchy, low educational levels among the employees and a low focus on employee well-being. This is the picture we are shown in media from movie comedies like Ratatouille (2007) to books like Heat by Bill Buford (2006) or Kitchen Confidential by Anthony Bourdain (2001). These stories are indeed entertaining and portrait the extreme, but in my own experience as a chef, reality is often not that far from the fiction when it comes to the leadership styles in real restaurants.

What is known for the restaurant industry is that it is affected by a high employee turnover compared to other fields (Piong, 2016). Previous research has shown that one factor that may help reducing high employee turnover is the choice of leadership style (Waldman et al., 2012). Employee turnover has been shown to be negatively associated with job satisfaction (Mobley, 1977). Therefore, logically, a leadership style that increases job satisfaction may be worth applying in order to reduce the rate of employee turnover.

In order to be effective with respect to the multicultural nature of the restaurant industry, the leadership style of choice would take the differences in national culture into account. Previous studies have focused on disciplines that are culturally transferrable, meaning that they represent leadership traits that are appreciated and effective across cultures (Wilson, 2013). One of these leadership theories is called servant leadership. This is a theory characterized by leaders who share power and put the needs of others first in order to help employees develop and perform as highly as possible.

Several studies have previously shown a positive correlation between servant leadership and job satisfaction in different fields (Wilson, 2013), but very few studies – with small samples, have examined the restaurant industry. No studies that quantitatively examined the correlation between servant leadership principles and job satisfaction for culturally diverse employees in the restaurant industry have been found in the research preceding this study.

The motivation for this study was a desire to understand to what extent a relationship existed between perceived servant leadership and job satisfaction among restaurant employees from

(6)

different cultures. This was done by quantitatively examining the correlation between the perceived servant leadership and job satisfaction among culturally diverse restaurant employees through a survey in social media. The possible predictions based on previous research were that the results would show a statistically significant relationship between servant leadership and job satisfaction across the cultural groups among employees. However, with the particularities of the stereotypical, hierarchical restaurant leadership in mind, it was not sure that the servant leadership style, which represent the opposite features, would be as appreciated among restaurant employees as within other fields. That uncertainty made this study an interesting topic within leadership research as well as for the restaurant industry.

1.1 Purpose

This study aimed to contribute to existing research by observing the relationship between the perceived servant leadership and job satisfaction among culturally diverse restaurant employees. This is a combination of factors that constituted a gap in the research. Other organizational outcomes had previously been studied in relation to servant leadership within the restaurant industry, presenting varying results. Most recently, Piong (2016) found a weak but positive relationship between servant leadership and employee commitment in Florida restaurants. Liden, Wayne, Liao and Meuser (2014) found that the serving culture in a restaurant chain was positively related to employee performance. Carter and Baghurst (2013) found through a qualitative study that there was a positive relationship between servant leadership and employee engagement at a restaurant in Dallas. In the hospitality industry, Wilson (2013) studied servant leadership, job satisfaction and cultural diversity within one small hotel in the Mid-Atlantic and found a positive correlation between servant leadership and job satisfaction and did not find any statistically significant variety among the cultural groups studied.

These previous studies linking servant leadership to employee performance and commitment in the restaurant industry share the geographic limitation to the United States and only one of the studies was conducted while simultaneously studying the moderating effect of cultural diversity. The lack of previous studies observing the correlation between perceived servant leadership and job satisfaction in the restaurant industry through the lens of cultural diversity constituted a research gap that this study aimed to fill. Such knowledge could provide

(7)

restaurant managers with information that may help them manage their multicultural staff in a more appreciated way by adapting an appropriate leadership model and ultimately saving money by reducing employee turnover.

1.2 Research method

This quantitative study examined the relationship between the perceived servant leadership and job satisfaction among culturally diverse restaurant employees. Servant leadership was used as the independent value and job satisfaction as the dependent value. The cultural diversity among the participators was observed as a moderating factor in order to observe any cultural differences in the relationship between servant leadership and job satisfaction.

Data for this quantitative study was collected through an online survey available for the respondents on social media. Restaurant workers that were currently employed at a restaurant were able to participate. The survey was completed in three parts. The first part was demographic with questions aimed to establish age group, gender, size of the company, time employed at the company and cultural identity. The rest of the survey consisted of two separate instruments that had previously been developed to measure servant leadership and job satisfaction. The second part of the survey used the Servant Leadership Scale developed by Liden et al. (2008) to measure the independent value: perceived servant leadership. In the third part of the study, the dependent variable: job satisfaction, was measured using the Minnesota Satisfaction questionnaire (Weiss, Dawis, England and Lofquist, 1967).

The following research question and hypotheses constituted the focus of this thesis:

RQ1: Is there a statistically significant correlation between the perception of servant leadership and job satisfaction for culturally diverse restaurant employees?

H1: There is a statistically significant correlation between the perception of servant leadership and job satisfaction for culturally diverse restaurant employees.

H0: There is not a statistically significant correlation between the perception of servant leadership and job satisfaction for culturally diverse restaurant employees.

(8)

2. Literature review

This chapter presents a review of the literature that was pertinent to the study. This review is divided into five main sections. The first part focuses on servant leadership where a background to the concept is given and the interpretation used for this study is explained. The second section focuses on job satisfaction as an organizational outcome and how it relates to a leadership model. The third section relates servant leadership to the hospitality industry and the fourth section discusses the cultural aspect and its relationship with the servant leadership theory. The fifth section explains the theoretical framework that constitutes the foundation of this study.

The literature presented in this review was mainly collected from the Stockholm University Library search engine and Google Scholar. While Google Scholar allowed an overview of articles and papers written within the concerned fields, the library search engine was mainly used to retrieve the full text material. Search terms used in literature searches included mainly the following, alone or in combinations: leadership, servant leadership, job satisfaction,

Greenleaf, Hofstede, Globe, Liden, restaurant leadership, universal leadership, multicultural leadership.

2.1 Servant leadership

The interest in leadership theories grew in the 21st century with a direction towards styles that focus on motivation and social responsibility in order to secure organizational success and profit (van Dierendonck & Nuijten 2010). One of the leadership styles considered to be meeting these objectives is called servant leadership. This concept was introduced by Robert K. Greenleaf in 1970 with the fundamental idea of servant leadership being that people want to serve others and do so by leading them. This, according to Greenleaf, can be contrasted to people who have a desire to lead rather than to serve (Greenleaf, 1970). The two principal qualities of a servant leader in Greenleaf’s definition are the ethical desire to serve people with the objective to improve their well-being, and the wish to provide these people with a vision, direction and purpose.

(9)

Extensive work on servant leadership has been done both by Greenleaf and by several other researchers who have continued after him. Today there is no generally agreed upon definition of servant leadership and several researchers have independently interpreted Greenleaf’s original work (van Dierendonck & Nuijten 2010). Some frameworks that have been frequently used in servant leadership studies are those of Laub (1999), Spears (1995), Liden et al. (2008) and van Dierendonck & Nuijten 2010). These disciples, although based on the same foundation, have created different ways of identifying and measuring the characteristics of servant leadership, all participating in determining the scope of servant leadership and how it differs from other leadership styles (van Dierendonck & Nuijten 2010). Although the servant leadership models vary slightly between Greenleaf’s disciples, the philosophy remains the same. This study used the instrument created by Liden et al. (2008) and their interpretation of Greenleaf’s philosophy will therefore be further discussed and detailed.

The decision to use the servant leadership instrument by Liden et al. (2008), the Servant Leadership Scale, was based on three reasons. Firstly, this scale was one of the instruments that was most closely related to Greenleaf’s original definition of servant leadership. Secondly, the Servant Leadership scale existed in a verified short version of 7 questions which was convenient as it helped keeping down the total number of questions in the survey as well as the time required to complete the survey. Thirdly, the author of the model, professor Robert Liden, provided a letter of consent according to which the Servant Leadership Scale could be used without any compensation.

Liden et al. (2008) explain that the foundations of servant leadership are personal integrity and serving others, including employees, communities and customers. A servant leader sets aside self-interest in order to focus on employees and organizations. Because of these unselfish criteria, servant leaders can play an important role in helping employees to reach their potential (Liden et al. 2000). This goes in line with Greenleaf’s original work (1977) stating that the focus on servant leadership is to develop employees to their fullest potential in areas including task effectiveness, community stewardship, self-motivation and future leadership capabilities. The servant leadership instrument developed by Liden et al. (2008) is based on the following nine servant leader characteristics:

(10)

2. Creating value for the community— a conscious, genuine concern for helping the community.

3. Conceptual skills— possessing the knowledge of the organization and tasks at hand so as to be in a position to effectively support and assist others, especially immediate followers.

4. Empowering— encouraging and facilitating others, especially immediate followers, in identifying and solving problems, as well as determining when and how to complete work tasks.

5. Helping subordinates grow and succeed— demonstrating genuine concern for others' career growth and development by providing support and mentoring.

6. Putting subordinates first - using actions and words to make it clear to others (especially immediate followers) that satisfying their work needs is a priority (Supervisors who practice this principle will often break from their own work to assist subordinates with problems they are facing with their assigned duties).

7. Behaving ethically— interacting openly, fairly, and honestly with others.

8. Relationships— the act of making a genuine effort to know, understand, and support others in the organization, with an emphasis on building long-term relationships with immediate followers.

9. Servanthood— a way of being marked by one's self-categorization and desire to be characterized by others as someone who serves others first, even when self-sacrifice is required.

Servant leadership is an adaptable theory that can be used and studied in various types of fields and organizations (van Dierendonck & Nuijten 2010) and a variety of positive organizational outcomes of servant leadership have been confirmed by previous studies. Job satisfaction is one variable which has shown to correlate positively with servant leadership (Wilson 2013). However, no study had previously examined the relationship between servant leadership and job satisfaction in the restaurant industry and in the light of cultural diversity, which was the subject of this study.

(11)

2.2 Job satisfaction

Research on job satisfaction has taken many different approaches since the 1920s (Wilson 2013). Some of the first works that are considered as the foundation of the domain are the Hawthorne experiments (Gillespie, 1993), but it was Edwin A. Locke who defined job satisfaction the way we know it today as “a pleasurable or positive emotional state resulting from the appraisal of one’s job and job experience” (Locke, 1974 p. 1300).

Job satisfaction is associated with several organizational outcomes, such as turnover and absenteeism (Akdol & Arikboga, 2015) which makes it an interesting variable to study. Factors that affect job satisfaction include working climate, organizational culture and leadership style. Hence, to some extent, examining job satisfaction can measure the effectiveness of a leadership model (Mueller et al, 2009). Multiple studies in various fields have shown that the characteristics of servant leadership, with a high focus on employees, are positively correlated with job satisfaction (e.g Drury, 2004; Van Dierendonck and Nuijten 2010; Wilson, 2013; Akdol & Arikboga, 2015; Piong 2016).

Several instruments have been developed to measure job satisfaction with examples like the Job Descriptive Index (Smith, Kendall and Hulin, 1969), the Job Diagnostic Survey (Hackman and Oldham, 1975) and the Minnesota Satisfaction Questionnaire (Weiss, Dawis, England and Lofquist, 1967). The short form (20 questions) of the Minnesota Satisfaction Questionnaire was the instrument used in this study. This questionnaire had previously been used to measure job satisfaction within the service industries. Previous findings that were pertinent to this study are described below.

Hancer & George (2003) studied job satisfaction within a chain in the American restaurant industry using the short form of the Minnesota Satisfaction Questionnaire. They found that 50% of the sample experienced low job satisfaction while 24% of the sample indicated a high level of job satisfaction. The authors pointed out the need to test the model for geographic differences among the respondents.

Ghiselli et al. (2001) studied the satisfaction levels among food service managers. They found that social services, security and moral values received the highest levels of satisfaction and that compensation, recognition and company practices received the lowest scores. Another

(12)

finding was that there were no significant differences among managerial categories, gender or ethnicity.

In a study relating the Minnesota Satisfaction Questionnaire to a different leadership model, the “Leader-Member Exchange” model, Mardanov et al. (2007) pointed out the supervisors’ importance for employee job satisfaction in the restaurant industry. They found that the supervisor alone determined more than 50% of the variance in the employee job satisfaction.

2.3 Servant leadership and organizational outcomes in the hospitality setting

As mentioned above, many studies have examined the relationship between servant leadership and job satisfaction. However, the research for this study found only one previous study in the hospitality setting. Wilson (2013) examined the relationship between servant leadership and job satisfaction within one small hotel in the Mid-Atlantic using Laub’s (1999) servant leadership instrument and found a positive correlation. The study included a culturally diverse aspect but no statistically significant differences in the outcome were found between the cultural groups.

Despite the lack of studies relating servant leadership to job satisfaction within the hospitality industry, some previous studies have addressed servant leadership strategies related to other organizational outcomes in this field.

Piong (2016) examined the relationship between servant leadership and employee commitment and perceived organizational support in a restaurant chain located in Florida. The findings indicated a weak correlation between servant leadership and both perceived organizational support and employee commitment.

Carter & Baghurst (2014) performed a qualitative study on a restaurant in Dallas with the objective to examine the influence of servant leadership on restaurant employee engagement. The findings showed that servant leadership influenced the employee engagement in a positive way, contributed to the employee loyalty and increased organizational commitment.

(13)

When studying an American restaurant chain, Liden et al. (2014) found that servant leadership culture was positively associated with serving culture. Several variables were taken into account and positive relationships were found between servant leadership and job performance, creativity and customer service while the relationship with turnover intentions was negative. Previous research indicates that servant leadership may have a positive impact on employee job satisfaction. However, a gap that still existed in the literature, and that this study aimed to fill, was to relate job satisfaction to servant leadership in the restaurant industry while taking cultural diversity into account.

2.4 Cross-cultural application of servant leadership

As it becomes increasingly easy to work across borders, the challenges related to multiculturalism need to be addressed, and leadership theories are getting difficult to assess without taking cross-cultural contingencies into account. The question has thus risen of whether universal leadership models (leadership models that are equally effective across national culture) could exist. Some researchers believe that it is impossible to determine a universal leadership theory due to the cultural differences between countries, and others have attempted to develop universal leadership theories based on cross-cultural research (Wilson, 2013). The research on cross-cultural research in field of management has been dominated by theories developed by Geert Hofstede and through the GLOBE project. These theories formed the base of the cultural dimension in this study.

Hofstede (1980) created a framework for identifying cultural differences and similarities across countries. This extensive work resulted in five cultural dimensions (power distance, individualism, masculinity, uncertainty avoidance and long-term versus short-term orientation). Countries are matched against this dimensions in order to create country-specific cultural profiles. Hofstede’s cultural dimensions, although criticised by some (Baskerville, 2003; Verbunt, 2011), have laid the foundation for cross-cultural research (Neuman & Nollen, 1996; Chow et al, 1999; Lair Sale, 2004).

Criticism to Hofstede’s cultural dimensions have two standpoints. Baskerville (2003) and Chow et al. (1999) presented three main criticisms to the validity of the theory. Firstly, they

(14)

argue that Hofstede’s studies use the words “culture” and “country” in mistakenly analogical ways. They also criticise the model for taking socio-economic factors into account and for not having been revalued since its development. Verbunt (2011) rejects the idea of cultural categorization for being too rigid. Through an intercultural standpoint Verbunt argues that cultures have been mixed since people began traveling and that all modern cultures are in constant change.

Despite the criticisms, Hofstede’s cultural dimensions have been frequently used for analysing the cross-cultural aspect of servant leadership (e.g. Winston & Ryan, 2008; Mittal & Dorfman, 2012; Wilson, 2013). Therefore, for lack of other cultural measures, and to facilitate the comparisons between this study and previous research, the work of Hofstede was used as the cultural foundation of this study.

The first of Hofstede’s dimensions is power distance. This refers to the amount of inequality or hierarchy that is normal for the population of a country (Hofstede, 1980). Effective servant leaders interact with employees in a participative way, with the aim that all employees feel free to express thoughts, ideas and feedback regardless of their societal or hierarchical position. In cultures with high power distance it is less likely to obtain the trust and feedback flows that are required for servant leadership (Hannay, 2008). This makes countries with a low power distance culture more naturally suitable for servant leadership (Davis et al., 1997).

Hofstede’s second dimension is named individualism. The level of individualism/collectivism is defined as the degree to which people from a country act as individuals or as part of a group (Hofstede, 1980). In servant leadership, success is considered a joint venture between employees and superiors. Focus on teamwork is central as well as responsibility and loyalty to the organization. Because of the importance of teamwork for servant leadership, cultures with low to moderate individualism are good fits (Hannay, 2008).

Masculinity is the third of Hofstede’s dimensions and is opposed to femininity in that masculine

cultures value assertiveness, performance and competition and feminine cultures value quality of life, care and service. (Hofstede, 1980). Although servant leadership does not neglect the importance of performance and results (Liden et al. 2008), feminine values are central in achieving positive outcomes in servant leadership (Hannay, 2008).

(15)

The fourth of Hofstede’s cultural dimensions is uncertainty avoidance which is defined as the level to which cultures prefer structured or unstructured situations (Hofstede 1980). In servant leadership the employees are pushed to take responsibility and rules and work assignments are less strictly set by the leader. Although this does not imply that the organisation is unstructured, cultures with a higher tolerance for uncertainty may be more comfortable under an organization that applies servant leadership (Hannay, 2008).

Hofstede’s fifth dimension was identified in his later work (Hofstede, 1993) and concerns the

long-term or short-term orientation where cultures with long-term orientation value saving and

persistence while short-term oriented cultures are oriented towards the past and present with respect for traditions and social obligations (Hofstede, 1993). One of the main objectives in servant leadership is to help employees develop in order to maximize their potential (Van Dierendonck and Nuijten, 2010), and this implies a good fit with long-term oriented cultures. Following the cultural profile above, it would seem possible that servant leadership may be a model which is more appreciated in some cultures than in others. Nevertheless, when attempting to establish leadership models that are universally applicable, studies have found servant leadership to meet the requirements.

Rooted in Hofstede’s (1980) cultural work, the Global Leadership and Organizational Behaviour Effectiveness (GLOBE) project had the objective to determine universal characteristics of organizational leadership (House et al., 2002). The GLOBE project identified six leadership value-dimensions that were attributed to effective leadership across cultures (charismatic, team-oriented, participative, autonomous, humane-oriented and self-protective). These dimensions created a general framework to assess leadership models and to adapt them to a culturally diverse setting. If these six global leadership values are universally valid – and servant leadership include these values, servant leadership could be a leadership model that would be equally appreciated across cultures (Wilson, 2013).

The disagreement about the universality of servant leadership motivates the cultural variable applied in this study. Previous studies have shown varying results regarding the universality of servant leadership, and the creators of servant leadership models have noted the need for more research concerning the influence of national culture on the effectiveness of the leadership method (Van Dierendonck, 2011; Rodréguez-Carvajal et al., 2014; Liden et al., 2008).

(16)

2.5 Theoretical framework

This study was conducted to obtain an indication of whether a positive relationship existed between perceived servant leadership principles and job satisfaction among culturally diverse employees in the restaurant industry. The theoretical framework for this study consisted thus of the three components servant leadership, job satisfaction and cultural diversity.

2.5.1 Servant leadership

The theory behind servant leadership was first outlined by Greenleaf (1970). It was based on the idea that leadership should be seen from the perspective of the followers. The servant leader builds new leaders by serving others including employees, customers and communities (Liden et al. 2008), and it is an approach to leadership that focuses on developing employees to their fullest potential in areas like task effectiveness, community stewardship, self-motivation and future leadership capabilities (Greenleaf, 1977).

Several frameworks have been developed by Greenleaf and his disciples. Some that have been frequently used in servant leadership studies are those of Laub (1999), Spears (1995), Liden et al. (2008) and van Dierendonck & Nuijten 2010). These disciples, although based on the same foundation, have created different ways of identifying and measuring the characteristics of servant leadership, all participating in determining the scope of servant leadership and how it differs from other leadership styles (van Dierendonck & Nuijten 2010). Liden et al. (2008) created one of the instruments to measure servant leadership that are the closest to Greenleaf’s original definitions. Therefore, their instrument, the Servant Leadership Scale (SLS), was used in this study. The SLS is originally a 28 item questionnaire – with a validated short version of 7 items that was used for this study (Liden et al. 2015).

2.5.2 Job Satisfaction

Job satisfaction is associated with several organizational outcomes – such as turnover and absenteeism (Akdol & Arikboga, 2015) which makes it an interesting variable to study. Several factors affect job satisfaction including working climate, organizational culture and leadership

(17)

style. Hence, to some extent, examining job satisfaction can measure the effectiveness of a leadership model (Mueller et al, 2009), and several studies in various fields have shown that the characteristics of servant leadership, with a high focus on employees, are positively correlated with job satisfaction (e.g Drury, 2004; Wilson, 2013; Van Dierendonck and Nuijten 2010; Akdol & Arikboga, 2015; Piong 2016).

The instrument of choice for this study was the Minnesota Satisfaction questionnaire (Weiss, Dawis, England and Lofquist, 1967). This instrument has been used several times in the hospitality setting (e.g. Mardanov et al. (2007); Hancer & George (2003) & Ghiselli et al. (2001)). The short version of 20 questions was used in this study because of the wish to not make the final questionnaire too time consuming for the respondents.

2.5.3 Cultural diversity

The multicultural component of this study was examined as part of a set of demographic questions in the beginning of the survey to collect descriptive cultural data. The question concerning culture included 9 answer choices based on the nine cultural clusters determined by House et al. (2004) according to Hofstede’s (1980) cultural dimensions. These were: Latin Europe, Northern Europe, Germanic Europe, Anglo (cultures with English as the native language), Latin America, Middle-East, Sub-Saharan Africa, Southern Africa and Confucian Asia.

The three variables perceived servant leadership job satisfaction and cultural diversity were studied as shown below to examine the relationship between them.

(18)

Perceived servant leadership was used as the independent variable, measured by the mean score of the 7 questions from the Servant Leadership Scale. The dependent variable, job satisfaction, was measured by the mean score of the 20 questions from the Minnesota Satisfaction questionnaire. The relationship between these two variables was investigated for a culturally diverse sample, adding the question regarding cultural identity as a moderating variable to the study.

(19)

3. Methodology

This chapter explains the empirical setting with the purpose and contribution of this study, the research design, methods for data collection, the validity of the study and finally ethical considerations important to the study.

3.1 Empirical setting

The aim of this study was to obtain an indication of whether servant leadership may be an effective leadership choice in the multicultural restaurant industry. Servant leadership is a popular leadership style that has shown to increase various positive outcomes across industries by focusing on leaders that share power, put the needs of others first and help employees develop and perform as highly as possible through participative and encouraging behavior. The researcher had noticed from several years as a chef, that leadership in the restaurant industry was often strictly hierarchical and authoritative, properties that express the opposite of those advocated in servant leadership. The apparent discordance between traditional restaurant leadership and servant leadership makes the restaurant industry particularly interesting for a study relating servant leadership with job satisfaction.

The problem addressed in this thesis was that it was not known to what extent a correlation exists between the perception of servant leadership and job satisfaction among culturally diverse restaurant employees. The lack of previous studies observing the correlation between perceived servant leadership and job satisfaction in the restaurant industry through the lens of on cultural diversity constituted a research gap that this study aimed to fill. Such knowledge could contribute to the research within the fields of leadership, cross-cultural research and restaurant management. This could practically be of use to restaurant managers in that it would provide them with information that may help them manage their multicultural staff in a more appreciated way by adopting their leadership model and ultimately to save money for the company by reducing employee turnover.

(20)

3.2 Research design

A quantitative, non-experimental descriptive approach was selected for this study with the aim to examine the correlational relationship between servant leadership and job satisfaction among culturally diverse restaurant employees. A quantitative approach was appropriate because the study focused on numerical relationships between the variables (Bryman & Bell, 2005). A quantitative design focuses on observing phenomena in a sample of a population that is descriptive of the whole (Creswell, 2009). For this study, a sample was taken to represent culturally diverse restaurant employees in a variety of organizations.

The research question for this study was concerned with what, if any, correlation exists between the two variables “servant leadership” and “job satisfaction”. For this type of study, a survey design is appropriate as it permits to study a sample of restaurant employees (Bryman & Bell, 2005) and to make inferences that can generally be applied to the whole industry with regard to behaviour (Creswell, 2009).

The following research question and hypotheses constituted the focus of this thesis:

In order to answer the question above, a questionnaire was made available for restaurant employees through an online survey. The survey consisted of three sets of questions combining the short version of Liden et al.’s (2009) Servant Leadership Scale to measure perceived servant leadership and Weiss et al.’s (1967) Minnesota Satisfaction Questionnaire to measure job satisfaction. The third set of questions (placed first in the survey) was demographical to measure the cultural variable. The survey was expected to take 5 to 10 minutes to complete.

(21)

3.3 Philosophy of social science

The understanding that individuals have of reality is based on culture and meanings are drawn from this cultural world (McCracken, 1990). However, standardized procedures are developed within societies and organizations and individuals are forced to adapt to these. Organizations represent social orders and individuals in the organization are pressured to follow the requirements and behavioural guidelines. The power of the organization affects and constructs limitations to its members (Bryman & Bell, 2005). The same can be argued about culture. Cultures as well as subcultures constitute collections of common values and guide the individuals in the sense that we internalize the beliefs and values associated with them. This ontological perspective of reality, referred to as objectivism, was adopted as the foundation of this study.

From an epistemological perspective, this study adopted a positivistic standpoint, advocating the use of scientific methods while studying the aspects of the social reality. Quantitative methods provide numeric data through a positivistic lens with the aim to test theories. (Bryman & Bell, 2005). The positivistic perspective implies that the study was deductive by generating hypotheses that could be tested in order to confirm theory and inductive in that knowledge was produced through the collection of facts in order to reach conclusions. Essential for the positivistic philosophy is also that the research was objectively conducted (Bryman & Bell, 2005).

3.4 Data generation method

This quantitative, non-experimental descriptive study was conducted through an online survey that was made available to restaurant employees through social media in international networks for the restaurant industry, thus providing the chance to collect responses from respondents with a wide range of cultural identities.

The survey was designed to take no longer than 10 minutes to complete and it was possible for the respondents to complete the survey at the time and place of their convenience. The link to the survey was available for two weeks and a reminder was sent halfway through the period. The survey application used to create the survey was Survey&Report, a tool available for users

(22)

at Stockholm University to create surveys and collect data. Before releasing the survey, it was tested on 5 people to make sure that the questions were clear and that the time limit of 10 minutes was not exceeded.

3.4.1 Sample

This study was conducted by distributing an online survey to restaurant employees by the means of social media channels available to the researcher. This sample was chosen due to the accessibility and because a random sample of culturally diverse restaurant employees was not judged to be reasonable to collect due to the time and scope limits of this study.

The sample selection was a nonprobabilistic, purposeful sample based on the availability of contacts through social media. This choice provided a way to collect a large number of responses from participators across the world. Generalization of the results is difficult through this sample strategy as it is not possible to determine the population for which the results are representative. However, the results could provide a springboard for further research and create links to other existing results (Bryman & Bell, 2005).

3.4.2 Research instrument

Data for this study was collected through an online survey with a set of questions available for respondents during time period of two weeks. The two variables servant leadership and job satisfaction were then studied with the objective to find correlational patterns between them. A third, moderating variable of diverse culture as it relates to employee perception of servant leadership and job satisfaction was studied through a demographic question about the participants’ cultural identity.

The questionnaire used for this study consisted of three parts with a total of 33 questions (see appendix 1). The first part was demographic with 6 questions aiming to establish age group, gender, size of the company, time employed at the company and cultural identity of the respondent. The second part of the survey used the short version of the Servant Leadership Scale developed by Robert Liden in 2008 with 7 likert-style questions to measure the independent value: perceived servant leadership. In the third part of the questionnaire, the dependent variable: job satisfaction, was measured using the short form (20 likert-style

(23)

questions) of the Minnesota Satisfaction questionnaire (Weiss, Dawis, England and Lofquist, 1967). The use of the Servant Leadership Scale was permitted by the author, Professor Robert Liden (appendix 1) and the Minnesota Satisfaction Questionnaire is open for research purposes without the need for written consent (University of Minnesota, 2017).

There are several advantages in using a questionnaire as data collection instrument. The possibility to collect data from geographically dispersed participants through an online service provides economical as well as timesaving advantages. Costly and time consuming travels for interviews can be avoided for the researcher – and the participants can answer the questions at their own convenience. Another advantage is the reduced risk of human factors affecting the answers and altering the results.

Using existing survey questions has two additional advantages (Bryman & Bell, 2005). The first advantage is that the questions have already been tested in research so their reliability and validity have been assessed. The second advantage of using existing questions is the possibility to compare the results of the questionnaire with previous studies. The two sets of validated questions in the questionnaire for this study were preceded by a set of demographic questions to measure the cultural variable. This type of questionnaire design makes it possible for the researcher to create a survey that is sensitive to the context of the research and relevant for the research question of the study. The choice of demographic questions was based on previous studies on servant leadership (Wilson, 2013, Liden et al. 2014).

3.5 Data analysis method

The first step in the data analysis was to check that the registered responses were completed. If more than 15% of the questions were not answered by a participant, that participant would be eliminated from the data analysis (Hair et al. 2014). To be able to analyse the data, the results of the questionnaire were imported from Survey&Report to Microsoft Excel. The data was then analysed in the software SPSS, version 24.

The internal reliability of the survey instruments was calculated to measure the relatedness of the questions from the two instruments Servant Leadership Scale and Minnesota Satisfaction Questionnaire were. This was done using Cronbach’s Alpha. A values of Cronbach’s Alpha

(24)

between 0.7-0.95 is considered adequate in exploratory research (Hair et al., 2014).

In order to determine the relationship between the independent variable of participants’ perceived servant leadership and the dependant variable of job satisfaction, Pearson’s correlation procedure was used. This procedure determines the correlation between two variables with a coefficient between 0 (no relationship) and 1 (perfect relationship) (Bryman Bell, 2005). The 0.05 probability level (p < 0,05) indicated statistical significance.

A linear regression analysis was run in order to observe if any of the demographic questions were able to predict the outcome of job satisfaction in the study. This analysis would rule out the importance of the demographic variables for the outcome of the data analysis.

3.6 Threats to validity

The validity of a study concerns the assessment of the conclusions drawn in the study and whether one or several indicators used to measure a concept really measures that specific concept (Bryman & Bell, 2005).

Selection bias may have occurred due to the non-randomness of the sample. It is possible that there was a difference in attributes between those restaurant employees who selected to participate in the study and those who did not. This possibility was attempted to be counteracted by sending a reminder as well as by presenting the study in the invitation note using non-scientific words. An additional threat to validity was the risk of receiving biased responses due either to social desirability to provide certain responses or participants’ mood or alertness while completing the survey. The researcher attempted to minimize this risk by giving written instructions to answer the questions truthfully and accurately.

Additional threats to the validity of the study include threats based on instrument reliability. In order to prevent this, the researcher used survey questions that have already be proven in previous research to effectively measure the variables servant leadership and job satisfaction. The survey instrument Survey&Report was provided by Stockholm University and was therefore judged by the researcher to be the most reliable choice available.

(25)

This study was limited to a quantitative non-experimental descriptive approach and therefore to the use of a single instrument to collect data. The study’s population consisted of a single-sample size of self-selected restaurant employees. This limitation restricted the generalizability in that all participants had made an active choice to participate which may have affected the results in that employees sharing certain attributes, for example a deeper interest in leadership, may have been over represented.

The choice of using job satisfaction as the choice of organizational outcome to measure constitutes limitations to the findings of the study. Other organizational factors could have been examined to demonstrate effective leadership. The use of Liden et al.’s (2008) model to measure servant leadership as opposed to other models for examining the same leadership style also restricted the generalizability of the study.

3.7 Ethical considerations

The Swedish Research Council has established four basic ethical principles of research ethics (Forskningsetiska Rådet, 1990) that have been respected in the process of this study. 1. The researcher has informed the participators of the purpose of the study. The purpose of the study was also briefly explained in a note before the link leading the participators to the survey. 2. The participators had the right to choose whether or not they wanted to take part in the survey. They could leave the survey at any time if they wished to not complete it. 3. All personal information was given confidentiality. The recorded information did not make it possible to identify the respondents and the collected data was only available to the researcher. 4. The data collected through the survey was only used for the purpose of this study and was destroyed after the completion of the study.

(26)

4. Findings and analysis

The purpose of this study was to investigate whether a relationship could be found between perceived servant leadership and job satisfaction among culturally diverse restaurant employees. The independent variable for the study was the restaurant employees’ perception of their immediate supervisor’s servant leadership practices as measured with the Servant Leadership Scale (short version) by Liden et al. (2008). The dependent variable was the employees’ job satisfaction as measured by the Minnesota Satisfaction Questionnaire developed by Weiss et al. (1967). The employees’ cultural designation was studied in order to observe whether the correlation was moderated by national cultural clusters as defined by House et al. (2004).

This chapter contains the details of the data collection and the results of the data analysis. Descriptive data findings are presented in the first part to support the analysis of the collected data. The data analysis is explained and illustrated in the second part. This part contains the Pearson’s correlation analysis that was used to answer the research question. Separate correlation analyses were also run for each cultural subgroup in order to observe cultural differences. A regression analysis was run to observe the statistical influence of servant leadership and the demographical variables on job satisfaction. The final section of this chapter consists of a summary of the findings and presents the answer to the research question.

4.1. Descriptive data

A total of 45 restaurant employees chose to take part in the study. The respondents filled out a survey online and the responses were collected through the survey instrument Survey&Report. The data was exported to the statistical program SPSS (version 24) for analysis. To prepare for the data analysis the data was checked for straightlining and completion. Straightlining, when a participant has given the same response systematically, indicates that the participant may not have answered to the best of their ability. Therefore, straightlining constitutes a reason to delete that participant’s responses (Hair et al. 2014). If 15% or more of the questions had not been answered by one same respondent, that respondent was to be withdrawn from the analysis due to lack of completion. No straightlining or lack of completion was found among the registered responses and all collected data was thereby considered suitable for further analysis.

(27)

4.1.1 Demography

The statistical program SPSS was used to acquire descriptive data concerning the demographic information about the respondents. This data defined six attributes: gender, age, cultural designation, professional role, time employed at the company and total number of employees in the company. While all of these demographic variables were taken into account in the analysis, the most important factor was the question concerning cultural designation as it represented the cultural moderating variable in the study. The other demographic factors were collected in order to observe their effect on the findings in a regression analysis. The details of the responses regarding these questions are found in appendix 2.

Of the total number of respondents, 21 were female and 23 were male. One participant had chosen the alternative “other”. The three biggest age groups were 20-25, 31-35 and 26-30 with 15, 12 and 9 respondents respectively (33.3%, 26.7% and 20.0%). The groups 36-40 and 46-50 were represented by 3 (6.7%) respondents each, and the groups 51-55 and 56-60 were represented by one respondents 82.2%) each.

Of the 45 respondents, the kitchen workforce was the most well-represented group with 19 (42.2%) respondents. This was followed by the category kitchen management (9 respondents, 20%), top management (8 respondents, 17.8%), service workforce (5 respondents, 11.1%) and service management (4 respondents, 8.9%).

The data concerning the total number of employees employed at the company showed mixed results with both small and large firms represented. 10-30 employees, was the most well represented company size with 13 respondents (28,9%) followed by the smallest and largest groups, <10 employees and >100 employees with 11 respondents (24,4%) respectively. Medium sized firms of 31-50 and 51-100 employees were represented by 4 and 6 respondents respectively (8,9% and 13,3%).

The responses to the question regarding the time employed at the current employer reflected the high employee turnover rate for the restaurant industry. 34 respondents (75,6%) had been working for their current employer 3 years or less. The segment 4-6 years was represented by 9 respondents (20%), and 2 respondents (4,4%) had worked for their current employer 7 years of more.

(28)

For the question regarding cultural designation, 10 culture clusters were available for the participants to choose from and it was only possible to select one option in order to keep the analysis as simple as possible. The cultural clusters used in the study were those identified by the GLOBE project (House et el. 2004). The clusters available were: Latin Europe, Northern Europe, Germanic Europe, Anglo, Latin America, Middle-East, Sub-Saharan Africa, Southern Africa, Confucian Asia. 7 of these clusters were represented among the respondents of the study (table 1.). This demographic question was central for the analysis as its outcome corresponded to the cultural variable in the theoretical framework used to answer the research question.

Table 1. Cultural subgroups

4.1.2 Instruments

The responses for each of the questions in the two instruments measuring the perceived servant leadership and job satisfaction (the Servant Leadership Scale and the Minnesota Satisfaction Questionnaire) were coded numerically from 1 (strongly disagree) to 7 (strongly agree) for the questions in the Servant Leadership Scale, and from 1 (very dissatisfied) to 5 (very satisfied) for the questions in the Minnesota Satisfaction Questionnaire.

The total mean scores and standard deviations for each of the two instruments were calculated for every respondent. The average mean score for the 45 respondents was then calculated for

Cultural subgroups

Frequency Percent Valid Percent Cumulative Percent

Valid Latin Europe 14 31.1 31.1 31.1

Northern Europe 14 31.1 31.1 62.2

Germanic Europe 2 4.4 4.4 66.7

Anglo (Cultures with English as the native language) 5 11.1 11.1 77.8 Latin America 6 13.3 13.3 91.1 Middle-East 1 2.2 2.2 93.3 Confucian Asia 3 6.7 6.7 100.0 Total 45 100.0 100.0

(29)

the 7 items of the Servant Leadership Scale to 4.6142 out of 7.0, and the average standard deviation was 1.15120. For the 20 items of the Minnesota Job Satisfaction questionnaire, the average mean score was 3.7333 out of 5.0and the standard deviation was 0.57928 (tables 2 & 3). These means were used in the Pearson’s correlation analysis that answered the research question.

Table 2 & 3. Means and standard deviations for servant leadership and job satisfaction

4.2 Data analysis

The analysis of the collected data provided the answer to the research question: Is there a statistically significant correlation between the perception of servant leadership and job satisfaction for culturally diverse restaurant employees? The analysis was initiated by a reliability examination of the instruments using the Chronbach’s Alpha method. The analysis then proceeded with the calculation of Pearson’s correlation coefficient to answer the research question. In order to examine the interactions of all variables on job satisfaction, a regression analysis was performed with job satisfaction as the dependent variable.

4.2.1 Reliability

To begin the data analysis, the internal reliability of the survey instruments was examined. This was done by calculating the Cronbach’s Alpha. Internal reliability, or consistency, refers to how related a set of items are as a group. When testing the reliability of questions in a scale of Likert-type, as the ones used for the Servant Leadership scale and the Minnesota Satisfaction Questionnaire, Chronbach’s Alpha is the most commonly used method (Laerd, 2013). SPSS

Mean scores

Job Satisfaction

N Valid 45 Missing 0 Mean 3.7333 Median 3.9000 Std. Deviation .57928 Range 2.35 Minimum 2.50 Maximum 4.85

Mean scores Servant

Leadership

N Valid 45 Missing 0 Mean 4.6127 Median 4.7143 Std. Deviation 1.15120 Range 4.29 Minimum 2.00 Maximum 6.29

(30)

was used to calculate the Chronbach’s Alpha for the subscales of both the Servant Leadership Scale and the Minnesota Satisfaction Questionnaire. The results indicated a reliability of 0.800 for the Servant Leadership Scale, and of 0.904 for the Minnesota Satisfaction Questionnaire (tables 4&5). If alpha is too low, some questions might need to be discarded whether a too high score may suggest that some items are redundant as they are testing the same question. Acceptable values of alpha range from 0.7 to 0.95 (Travakol & Dennick, 2011). Both scales were considered to be satisfyingly reliable.

Tables 4 & 5. Reliability tests

4.2.2 Pearson’s correlation

To examine the relationship between servant leadership and job satisfaction, and thereby to answer the research question, a two-tailed Pearson’s correlation analysis was conducted. Pearson’s correlation procedure determines the relationship between two variables with a coefficient between 0 (no relationship) and 1 (perfect relationship) (Bryman Bell, 2005). The 0.05 probability level (p < 0.05) indicated statistical significance. This analysis was done using the mean scores of servant leadership and job satisfaction.

Reliability

Servant Leadership Scale

Cronbach's Alpha Cronbach's Alpha Based on Standardized Items N of Items .800 .803 7

Reliability

Minnesota Satisfaction

Questionnaire

Cronbach's Alpha Cronbach's Alpha Based on Standardized Items N of Items .904 .903 20

(31)

Pearson's Correlation

Servant leadership and job Satisfaction

Servant Leadership

Job Satisfaction

Leadership Pearson Correlation 1 .692**

Sig. (2-tailed) .000

N 45 45

Job Satisf. Pearson Correlation .692** 1

Sig. (2-tailed) .000

N 45 45

**. Correlation is significant at the 0.01 level (2-tailed).

Table 6. Pearson’s Correlation

The result of Pearson’s correlation test showed that the mean score of the servant leadership questions correlated with the mean score of the job satisfaction questions to +0.692 (table 6). A positive coefficient indicates a relationship between the variables and the actual correlation is significant at the 0,01 level (p < 0.01). The positive correlation that was found comprised all cultural groups and answered thereby the research question by confirming the positive hypothesis H1: There is a statistically significant correlation between the perception of servant leadership and job satisfaction for culturally diverse restaurant employees. The null hypothesis was thus rejected. A scatterplot was generated to get a visual view of the result (scatterplot 1).

(32)

In order to observe any statistical differences between the cultural groups, Pearson’s Correlation tests were run with the mean results of the servant leadership questions and the job satisfaction questions for each of the cultural subgroups. Due to the small sample size, meaningful analyses could not be done for all cultural subgroups. Four of the cultural subgroups had five or more respondents and were taken into consideration for a separate analysis. These subgroups were: Latin Europe, Northern Europe, Anglo and Latin America. The results of these correlation test are shown below (table 8).

Pearson's Correlation

Cultural subgroups

Servant Leadership Job Satisfaction Servant Leadership Job Satisfaction Latin Europe 1 .845** Northern Europe 1 .908** Anglo 1 .551 Latin America 1 .002 Latin Europe .845** 1 Northern Europe .908** 1 Anglo .551 1 Latin America .002 1

**. Correlation is significant at the 0.01 level (2-tailed).

Table 8. Correlations cultural subgroups

When the four cultural subgroups which had five or more responses were run in separate correlation analyses, it was found that the two largest subgroups, Latin Europe and Northern Europe, showed statistically significant relationships between servant leadership and job satisfaction (0.845 and 0.908 respectively). For the other two groups that qualified for separate analyses, Anglo and Latin America, the correlation was lower with 0.558 for Anglo culture and no correlation at all for Latin America (0.002). Due to the small sample sizes for these groups, the results of Pearson’s correlation test must be observed with caution.

In order to get an overview of the separate results between the cultural subgroups, scatterplots were drawn for all groups. These scatterplots allow one to spot trends although no statistically significant correlation was found for some of the groups (scatterplot 2).

(33)

Figure 2. Cultural subgroups

The scatterplots above show the linear relationships between the mean scores of servant leadership and job satisfaction for all cultural groups separated. The trends indicate a positive relationship for all groups except Latin America where no correlation was found and the Middle- East which is only represented by one participant.

4.2. Regression analysis

In order to observe if any of the demographic questions were able to predict the outcome of job satisfaction in the study, a linear regression analysis was run. The mean score for the servant leadership questions was used as continuous predictor and the demographic data was used as categorical predictors. The results of the regression analysis are presented in the table below (table 9).

(34)

Table 9. Regression analysis

The regression analysis shows the statistically significant relationship between servant leadership and job satisfaction. No other variable was statistically significant, meaning that the demographic information controlled for in the survey, including the cultural subgroups, did not have any major influence on job satisfaction.

4.3 Interpretation of results

The aim of this study was to investigate if a relationship could be established between perceived servant leadership and job satisfaction among culturally diverse restaurant employees. The research question for this study was the following:

This question would be answered by confirming either of the two hypothesis:

Regression Analysis

Source _DF _{Adj SS} _{Adj MS} F-Value P-Value

Regression 27 11.5923 .42935 2.3 .039

Servant Lead. mean 1 3.6285 3.62852 19.44 .000

Gender 2 .4579 .22897 1.23 .318 Culture 6 .6474 .10790 .58 .743 Age 7 .5136 .07337 .39 .893 Time at company 3 .2074 .06913 .37 .775 No of employees 4 .3085 .32713 1.75 .185 Role 4 .1904 .04760 .26 .903 Error 17 3.1727 .18663 Total 44 14.7650

(35)

The research question was aimed to be answered by analysing the data generated through a 33 item questionnaire made available to restaurant employees through an online survey. The questionnaire was available for two weeks during which 45 respondents submitted completed questionnaires. The data collected from the questionnaires was exported to the statistical software SPSS for analysis.

The research question was answered through a Pearson’s correlation test where the relationship between the two overall means of the 7 questions concerning servant leadership and the 20 questions testing job satisfaction was investigated. The Pearson’s correlation test showed a positive relationship between the two variables at r = +0.692 and a statistical significance of p < 0.01. Based on the result of the Pearson’s correlation test, it was concluded that there was a statistically significant correlation between servant leadership and job satisfaction among culturally diverse restaurant employees. The null hypothesis was therefore rejected.

After the correlation was established it was investigated if any statistical variance existed between the correlations for each individual cultural subgroup. Due to the small sample size it was not possible to fully analyse the cultural differences and only four of the cultural groups had enough respondents (5 or more) to observe any trends. The results of the culture specific correlation analyses showed that the groups Northern Europe and Latin Europe strongly supported the answer to the research question. The group Anglo showed a weaker correlation. The trend for the group Latin America was not in line with the others. This group did not show any correlation between servant leadership and job satisfaction. This finding indicates that servant leadership may have a weaker relation to job satisfaction in some cultures than in others. However, this analysis should be handled with caution as the small sample size does not allow any generalisation of the results.

The relationship between job satisfaction and all other variables was further investigated through a linear regression analysis. Through the regression it was concluded firstly that servant leadership was the only variable that showed a statistically significant relationship with job satisfaction with a p-value of 0.000, indicating a perfect relationship. The regression analysis also ruled out the influence of any of the demographic variables. This implied that

(36)

there may be support for the theory of servant leadership being a universally applicable leadership style.

In summary, the statistical analysis of the data collected in this study showed that a servant leadership style is positively associated with job satisfaction and that this relationship is independent of national culture. Some non-significant variance was found in the correlational analysis of one of the cultural groups studied, but the regression analysis showed that neither culture, nor any other demographical variable studied, predicted job satisfaction on a statistically significant level.