Effects of tour guides’ self-efficacy levels and autonomy perceptions on their job crafting behaviours

The main purpose of the study is to determine the effects of tour guides’ self-efficacy levels and their autonomy perception levels on their job crafting behaviours. It is also aimed to determine their job crafting, self-efficacy, and autonomy perception levels. Besides, comparing guides’ job crafting behaviours in terms of their personal properties and demographic characteristics is also conducted within the framework of the study. Data were collected from 203 tour guides via survey method. Multiple linear regression analysis results showed that selfefficacy and autonomy perceptions of tour guides had an effect on their job crafting behaviours. Besides, job crafting, self-efficacy, and autonomy levels of tour guides were high. Lastly, there were significant differences in job crafting behaviours of tour guides in terms of their ages, license types, experiences, and active days at work in a year.


Introduction
Tour guides, frontline employees in the tourism industry (Lin et al., 2008;Tsaur & Lin, 2014) hold a mediatory position between the local community and tourists (Ap & Wong, 2001). As leading visitors with preferred language and having skills and ability to interpret the cultural and natural heritage of a specific destination (WFTGA, 2020), tour guides have a notable influence on tourists' satisfaction and impression related to the destination they visit and have the ability to transform a visit into a touristic experience (Reisinger & Steiner, 2006). They can enrich tourists' knowledge and experience. Having a great impact on tourists' knowledge and experience, their emotional and spiritual changes; tourist guides' qualifications and their professionalism are among critical issues. Such that, they must provide a positive destination image and maintain it (Pratiwi et al., 2019). To be successful, tour guides have to be fully prepared for tourists' different demands and expectations. It is widely known that tour guides frequently experience circumstances that they need to respond to tourists' unexpected requests. In their qualitative study, Wong & Wang (2009) have found that tour guides try their best to fulfil tourists' nonsense demands and expectations; they have no chance to show their real emotions (anger, irritation, etc.); all they can do is to act taking tourists expectations into consideration. In such cases, they should take initiative and have the opportunity to make changes in their work tasks. Otherwise, today online platforms provide tourists with splendid opportunities for sharing both their negative and positive travel experiences with others as soon as they complete their tours (Ažić & Bačić, 2020).
It is known that becoming a tour guide is not an easy process. Being a leader of a group of tourists with unique characteristics and having different needs and demands cannot be categorized as a standardized job with a concrete job design. Tour guides play a crucial role during tourism product transactions as the providers of tourism products and services (Luoh & Tsaur, 2014). They also contact tourists for a long time. This could be an opportunity or a challenge depending on tour guides' skills and knowledge. Accordingly, their autonomy perceptions and job crafting behaviours can reinforce the service provided and give them space for solving extreme problems. Many jobs and tasks are specially designed by the managers initially and assigned to the employees. Yet, there are differences as employees can make modifications in assigned tasks with their skills, knowledge, and expertise while performing their jobs (Lazazzara et al., 2020). That makes job crafting a well-known, yet interesting concept. It enables employees to moderate their work relations, change the meaning of work and work tasks assigned (Wrzesniewski & Dutton, 2001).
Job crafting, defined as making changes in social, cognitive, and interpersonal aspects of the jobs (Wrzesniewski & Dutton, 2001); is an important phenomenon for tour guides, highly interacting with tourists. Meged (2017) has expressed that guides frequently use job crafting as they have flexible work schedules, they do not regard themselves just as mediators between locals and tourists. Instead, they think that they are valuable sources who enlighten and strengthen tourist groups when they design their jobs' idealistic and social aspects according to their values and opinions. Job crafting is different from job design. Traditional job design is a top-down process that managers determine job characteristics for the employees (Oldham & Fried, 2016). In job crafting, the process is just the opposite; employees determine the necessary changes related to their jobs or tasks and conduct them without any management approval (Leana et al., 2009).
Many studies (Ghitulescu, 2007;Tims & Bakker, 2010;Petrou et al., 2012;Wrzesniewski et al., 2013;Berg et al., 2013;Slemp et al., 2015;Cheng et al., 2016) have focused on the outcomes of job crafting especially because of its positive reflection to the organizations. Tims et al. (2012) have conducted research to develop a scale to measure tourist guides' job crafting behaviours. Yen et al. (2018) have developed a scale for measuring tourist guides' job crafting in a similar way. Cheng et al. (2016) have found that 3 tourist guides' individual and collaborative job crafting behaviours have a positive impact on their satisfaction, performance, and organizational commitment. Yet, it is highly important to find out antecedents of the behaviour. However, there are few studies on the determinants of job crafting. Teng & Chen (2019) have aimed at investigating the determinants of job crafting behaviour and they have figured out there is a relationship between proactive personality and job crafting. Rudolph et al. (2017) have found that there is a strong relationship between proactive personality and job crafting. Additionally, Kim et al. (2018) have determined perceived organizational support, autonomy, and creative self-efficacy have a direct effect on job crafting. Therefore, revealing job crafting determinants has become an issue of considerable concern for tourist guides. Taking the aforementioned studies in literature into account, it is assumed that self-efficacy, one of the individual determinants, has an effect on job crafting. Besides, defined as employees taking initiatives in their works or jobs, autonomy may influence job crafting behaviour. Yet, no study has been reached focusing solely on examining the effects of these variables on job crafting behaviour of tour guides. Accordingly, the main aim of the study is to find out whether tourist guides' self-efficacy levels and their autonomy perception levels have an effect on their job crafting behaviours. In order to investigate these effects, tourist guides' self-efficacy levels, their autonomy perception levels and their job crafting levels are determined. Also, this study explores the relationship between tourist guides' job crafting levels and their personal properties and demographic characteristics. Evaluating all the aims of this research, it can be stated that the study will contribute literature by fulfilling the mentioned gaps.

Conceptual Framework
The Relationship Between Job Crafting and Self-Efficacy As wellbeing has started to get attention from organizations, job design theories focus on employees' abilities to change their jobs' characteristics autonomously. This is called job crafting (Letona-Ibanez et al. 2019). As a concept, job crafting is related to the changes in jobs' relational and task boundaries either physically or cognitively (Wrzesniewski & Dutton, 2001). Employees making changes in number or types of work activities refer to the task boundaries; while adjusting their perceptions on their jobs is related to cognitive boundaries. Making changes in relational boundaries mean making choices with whom an employee interacts during work (Berg et al., 2008;Cheng et al., 2016). Altering in job design and relations at the workplace, an employee can contribute to the meaning and identity of work (Yen, 2018). According to Lyons (2008), employees make the aforementioned changes without managerial sanction. Yet, these changes can be approved or detested by the managers.
Taking job crafting definitions into consideration, it is underlined that employees conduct job crafting process (Slemp & Vella-Brodrick, 2013) and demonstrate proactive behaviours in those changes (Tims & Bakker, 2010;Niessen et al., 2016). In their study, Leana et al. (2009) argue that employees show proactive behaviours while changing work boundaries and shaping their works. In such circumstances, they demonstrate their abilities and perform better. Accordingly, Berg et al. (2008) state that job crafting is effective on employees' performance and satisfaction. In case it is enacted properly, job crafting can improve employees' lives at work and have a positive outcome for the workplace. Previous studies show that job crafting has a direct effect on job satisfaction (Ingusci et al., 2016;Cheng & Yi, 2018), job performance Tims et al., 2015), motivation (De Beer et al., 2016) and organizational commitment (Iqbal, 2016).
Self-efficacy, one of the factors in social cognitive theory, is defined as a person's perception of the ability to exhibit specific behaviour, meet challenges, overcome specific circumstances or accomplish a task (Bandura, 1977). A person's beliefs play an important role in one's approach to goals, tasks, and challenges. Expectations regarding the outcomes of a specific task or behaviour that needs to be done and their judgments about their abilities motivate individuals to perform that task or behaviour (Wood & Bandura, 1989). Self-efficacy is not a passive personal characteristic; instead, it is a mixture of variables forming self-esteem. Furthermore, ineffective behaviours can be observed taking individuals with low self-efficacy into consideration (Ustuner et al., 2009). Griffin et al. (2007) express that individuals' selfefficacy levels determine their self-confidence, perceptions, and capacities to perform a specific behaviour. Thus, individuals with high self-efficacy have a tendency to exhibit proactive behaviours and perform better. Self-efficacy is one of the determinants of proactive behaviour. Individuals, relying on their abilities, believe that their activities will be successful (Morrison & Phelps, 1999).
In literature, there are studies focusing on investigating the relationship between self-efficacy and job crafting. Tims et al. (2014) have found that individuals in need to control the work environment; in other words, individuals who have high self-efficacy perceptions, and those who feel responsible for making changes, tend to demonstrate more job crafting behaviour. Researchers have also added that the relationship between daily work performance and self-efficacy can be explained by job crafting behaviour. Kanten's (2014) study on hotel employees has revealed that employees' self-efficacy perceptions have a positive effect on job crafting. In their quasi-experimental study, Van den Heuvel et al. (2015) have expressed that employees' self-efficacy levels increase in job crafting interventions. However, there is no strong relationship between self-efficacy and job crafting behaviour. Mäkikangas et al. (2017) have tried to reveal the relationships between self-efficacy and job crafting; study findings show that self-efficacy perception and team job crafting are strongly correlated. Taking previous studies into account, it is proposed that: H1: Self-efficacy has a significant effect on job crafting behaviours of tour guides.

The Relationship Between Job Crafting and Autonomy
Autonomy or job autonomy is a critical work characteristic, highly effective on employees' satisfaction, commitment, and motivation. Also, there is a strong relationship between autonomy and higher work performance. Autonomy has a direct impact on reduction in role ambiguity and turnover (Spector, 1986). Turner & Lawrence (1965) define autonomy as an opportunity for employees to decide how to conduct work procedures. Hackman & Oldham (1975) consider autonomy as a part of job characteristics and define it as the degree of freedom, independence, or discretion that the job provides for employees while planning their work or determining the procedures in their jobs. According to Leach et al. (2003) job autonomy is related to workplace practices that increase employees' authority on decision making. These practices directly focus on increasing their control over works. Employees' control over their jobs increases if they are empowered with access to information, resources, and development opportunities (Lin et al., 2013). Autonomy provides an opportunity for them to choose the work project and decide how to complete a task (Morgeson et al., 2005); that's why it is an important concept especially for organizational success.
It is thought that there is a relationship between job crafting and autonomy. Tims & Bakker (2010) underline that workplace properties enhance employees' opportunities to make changes in their work designs and providing autonomy is also an important issue for employees' health. They can also evaluate how to handle critical problems with the help of autonomy. As supported by Bakker et al. (2004), restriction on autonomy can increase employees' work stress. Slemp et al. (2015) have found that there is a relationship between job crafting and employees' well-being. According to the researchers, perceived autonomy support enables employees to participate in more job crafting behaviour. However, Petrou and Bakker (2016) have found that even in jobs with low autonomy, employees can make changes regarding the demands and resources of the job. In other words, they demonstrate job crafting. Taking all these studies into account, it is proposed that:

H2: Autonomy has a significant effect on job crafting behaviours of tour guides.
Lastly, job crafting is a proactive behaviour that can also be related to personal properties and demographic characteristics of tour guides. Rudolph et al. (2017) have stated that determining descriptive relationships between job crafting and demographic characteristics can contribute to job crafting theory even if they do not give a critical insight into the behaviour. The researchers have also underlined that in order to understand the nature of the behaviour and compare the findings of other forms of proactive behaviour, examining demographic variables becomes a necessity. In their study, Petrou & Bakker (2016) have found that men have been more active in exhibiting job crafting than their female counterparts. Cheng et al. (2016) have stated that there have been relationships between demographic variables and job crafting behaviours of tour guides. They have found out that age has a significant correlation with job crafting behaviours of tour guides. Additionally, organizational tenure has a strong relationship with the job crafting behaviours of tour guides. Carrillo et al. (2020), have focused solely on gender differences in the tour guiding profession and expressed that there are differences in performing the profession. That can also be evaluated as a shred of indirect evidence related to the behaviours exhibited while performing the job. Taking all these studies into account, the following hypothesis is proposed: H3: There are significant differences in job crafting behaviours of tour guides in terms of their personal properties and demographic characteristics.

Research Design
In this study, quantitative research methods were adopted. To test the constructed hypotheses, correlational design and causal-comparative research design were used. As one of the quantitative research methods, the correlational research design is especially used to look at relationships between two or more variables that occurred naturally (Field, 2009). Accordingly, correlational design was adopted in order to test the effects of self-efficacy and autonomy perceptions of tour guides on their job crafting behaviours at the first stage of the research. Secondly, causal-comparative design was preferred as job crafting behaviours of tour guides were compared in terms of their personal properties and demographic characteristics. Causal comparative design is used to establish cause-effect relationships between variables and it is highly recommended to make group comparisons (Charles, 1998;Gay & Airasian, 2000).

Data Collection Procedure and Participants' Characteristics
With a detailed investigation of studies in literature, the theoretical framework of the study was constructed. According to TUREB (Tour Guides Association) (2020) statistics, there are 8053 registered tourist guides active in the field. Accordingly, the population is formed by 8053 tourist guides in Turkey. Researchers reached each of the chambers located in different regions and cities in Turkey via telephone or electronic mail. As the data collection procedure was conducted under extreme conditions (restrictions implemented due to COVID 19 pandemic) between April-September, 2020; the convenience sampling method was adopted in the study. Convenience sampling is easy to conduct, affordable and makes it easier to reach the data (McMillan & Schumacher, 2010). All chambers were asked to share research questionnaires with their members on online platforms. During the six-month data collection period, 203 usable questionnaires were collected. The demographic profile of the participants is presented in Table 1. With a detailed investigation of Table 1, out of 203 participants, 120 (59%) of them were male and 83 (41%) were female. Taking their marital status into consideration, 116 (57%) of them were single and 87 (43%) of them married. Out of 203 respondents, 87 (43%) tour guides were aged between 29-40 years old. 61 (30%) of them were aged between 20-28. So, it can be stated that the participants were relatively young. 86 (42%) out of 203 had an undergraduate degree and 63 (31%) tour guides had an associate degree. That means, education level was high among participants. Furthermore, 64 (31%) of them had experience between 2-7 years. 60 (30%) of them had 8-14 years of experience. Lastly, 65 (32%) of the participants were active in the field between 0-50 days in a year. 56 (28%) of them worked actively between 151 days or more. 136 (67%) of them held their positions as full-time jobs.

Measuring Instruments
In order to collect data, a self-administered questionnaire consisting of two sections was formed. The first section of the questionnaire covering 10 questions focusing on determining personal properties and demographic characteristics of the participants with close-ended questions. In the second section, three scales were used. One of them was Job Crafting Scale (a 15 item scale) developed by Slemp & Vella Brodrick (2013). Validity and reliability analysis of the scale was conducted by Kerse (2017) and its Cronbach Alpha value was found as 0.91. In our study, this value was found as 0.94. The second scale was Self-Efficacy Scale (a 12 item scale) developed by Bosscher and Smit (1998). Its validity and reliability analysis was conducted by Tarakcı (2009) and its Cronbach Alpha value was found 0.69. In our study, the Cronbach Alpha value of the scale was determined as 0.81. Lastly, Job Autonomy Scale was used in the study. The scale was developed by Breaugh (1985) and its short form was formed by Dude (2012). The short form of Job Autonomy Scale (a 4 item scale) was used in other studies (Pekdemir et al., 2014;Bora, 2019). In Pekdemir et al.'s (2014) study, its Cronbach Alpha value was found as 0.65 and Bora (2019) found this value as 0.86. In our study, the Cronbach Alpha value of the scale was determined as 0.80. In order to check its validity and reliability, two language experts controlled the four item scale and translated the scale back to English. Additionally, researchers consulted four tour guides to control its comprehensibility. All items were rated along a five-point Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree) in order to provide consistency for the participants. All scales used in the study were developed by other researchers in the literature. Validity and reliability analyses were also conducted by the researchers who developed them. Accordingly, the scales shown in Table 2 were used without any intervention in the study.

Hypotheses Data Collection Tools
H1: Self-efficacy has a significant effect on job crafting behaviours of tour guides. Self-Efficacy Scale Job Crafting Scale H2: Autonomy has a significant effect on job crafting behaviours of tour guides. Job Autonomy Scale Job Crafting Scale H3: There are significant differences in job crafting behaviours of tour guides in terms of their personal properties and demographic characteristics.
Job Crafting Scale

Data Analysis
In this study, firstly descriptive analysis (means and standard deviations) was conducted in order to reveal levels of participants' job crafting, self-efficacy and job autonomy. When analysing descriptive results, the arithmetic mean of each scale's all items was calculated over five as 5 point Likert type scales were used. The calculated arithmetic means were also used in all statistical analyses in the study. Secondly, correlations among variables were determined with Pearson correlation analysis as the data showed normal distribution. It was found that there were correlations among variables. No correlation coefficient exceeded 0.90 (Field, 2009) in the study. Accordingly, it can be stated that obtained correlation coefficients were between acceptable levels to conduct regression analysis. Thirdly, a model was constructed in line with the hypotheses within the framework of the study. As the data showed normal distribution, multiple linear regression analysis, one of the most suitable analysis methods to test a model, was used. Furthermore, the adequacy of the sample size was controlled. According to Green (1991) and Field (2009) sample size for regression analysis is calculated with the given equation: N=50+8k (k refers to independent variables or predictors). So with 2 predictors (self-efficacy and autonomy), it is necessary to have a sample size of 50+16=66. Besides, another equation can be used for calculating sample size as N=104+k where k refers to individual predictors. Here, 104+2= 106 is an adequate sample size. According to both equations, our sample size is adequate for multiple linear regression analysis.
Lastly, job crafting behaviours of tour guides were aimed to be compared in terms of their personal properties and demographic characteristics. Accordingly, independent sample t-test and ANOVA test were used to compare job crafting behaviours of tour guides in terms of their personal properties and demographic characteristics as the data showed normal distribution. Independent sample t-test was used in order to reveal differences in terms of gender, marital status and employment type as these variables had two groups. One-way ANOVA was used to reveal differences in terms of age, experience, license type, working type, number of active days in a year as these variables had more than two groups.

Descriptive Findings
Firstly, arithmetic means of participants' responses for each items in all scales are presented in detail in Table 3. Taking tour guides' job crafting levels into account, it can be expressed that their job crafting levels (M=3.91) are high. Obtained data is shown in Table 4. Taking the self-efficacy levels of tour guides into account, it can be stated that their self-efficacy levels (M=3.70) are high. Obtained data is presented in Table 5.
When self-efficacy levels of tour guides were compared in terms of their gender, it is clear that the selfefficacy levels of male participants (M=3.71; SD=0.744) were higher than their female counterparts. Besides, married participants (M=3.85; SD=0.620) had higher levels of self-efficacy than single counterparts. Looking into their ages in detail, it is clear that participants around 51 years old or more (M=4.07; SD=0.563), had the highest level of self-efficacy. Furthermore, participants with 15 years or more experience (M=3.93; SD=0.570) had the highest level of self-efficacy among all groups. Taking autonomy levels of tour guides into account, it can be stated that their autonomy levels (M=3.75) are high. Obtained data is presented in Table 6.

Findings on Modelling Factors Affecting Tour Guides' Job Crafting Behaviours
In order to find out correlations between variables, descriptive and correlation analyses were conducted. Addressed means, standard deviations, and correlations among variables are shown in Table 7. Multicollinearity among independent variables acts as a rule that needs to be checked before further analysis. Field (2009) states that correlation coefficients between .80 -.90 is high. In other words, the results shown in Table 6 did not restrain further analyses. As the data had normal distribution, Pearson correlation analysis was used in order to determine correlation values. According to Pearson correlation analysis results, there were positive relationships among variables. Between job crafting and autonomy, the correlation coefficient was .617, p<0.01. Also, there was a relationship between job crafting and selfefficacy (r=.753, p<0.01). Lastly, the correlation coefficient value was determined (r=.620, p<0.01) between autonomy and self-efficacy. That means positive relationships were found between these variables. However, this was not a sign of a strong relationship. Also, Variance Inflation Factor (VIF) values were controlled and values between 1.16 and 1.97 are acceptable limits for regression analysis.
Hypothesis 1 suggests that self-efficacy has a significant effect on job crafting behaviours of tour guides. We also assume in Hypothesis 2 that autonomy has a significant effect on job crafting behaviours of tour guides. In order to test the constructed hypotheses and determine the effects of self-efficacy and autonomy perceptions of tour guides on their job crafting behaviours, multiple linear regression analysis was conducted. The analysis was used in order to find out the coefficients of the linear equation covering two factors (autonomy and self-efficacy) predicting job crafting behaviours of tour guides. Results are shown in Table 8. coefficients into consideration, a unit of increase in autonomy leads 23% increase in job crafting behaviour. Besides, a unit of increase in self-efficacy leads 82% increase in job crafting behaviour. In line with the results obtained, it can be expressed that Hypothesis 1 and 2 are supported.

Findings Regarding Tour Guides' Job Crafting Levels in terms of Their Personal Properties and Demographic Characteristics
In order to determine the differences in job crafting behaviours of tour guides in terms of their personal properties and demographic characteristics, t-test and ANOVA tests were conducted. According to ttest results, there was no difference between job crafting levels of tour guides according to their gender, marital status, and employment type. Results are shown in Table 9. Taking t-test results into consideration, it can be stated that there was no difference in job crafting levels of female and male participants. Also, it can be inferred that married and single participants exhibit similar job crafting behaviours. Lastly, no difference was found in terms of their employment type. In other words, they exhibit similar job crafting behaviours either working part-time or full time. Besides, Analysis of Variance (ANOVA) Tukey tests were conducted in order to assess job crafting levels of tour guides in terms of their ages, license types, experience, working types, and the number of active days in a year. ANOVA Tukey test results are shown in Table 10.
According to Table 10, there was a statistically significant difference in job crafting levels of tour guides in terms of their ages (p<.05). Results showed that there were significant differences between job crafting levels of participants around 20-28 years old and 29-40 and 51 or more. Participants between 51 years old or more (M=4.20) exhibited the highest level of job crafting among all groups. Besides, there were significant differences among groups in terms of their license types. There were differences in job crafting behaviours of participants who obtained their license via an associate degree, undergraduate degree, master degree, and Ministry of Culture and Tourism Certificate Program. Among all groups, participants with a postgraduate degree (M=4.70) had the highest level of job crafting. Looking into test results, there were statistically significant differences in job crafting behaviours of tour guides in terms of their experiences. Among all groups, participants with 15 years or more experience (M=4.13) showed the highest level of job crafting behaviour. Besides, statistically significant differences in tour guides' job crafting behaviours in terms of the number of active days in a year were determined. Among all groups, tour guides working 151 days or more in a year in the field (M=4.36) had the highest level of job crafting. However, there were no differences in job crafting levels of tour guides in terms of their working type. Tour guides working full time, part-time or seasonal exhibited similar levels of job crafting.

Discussion
In this study, tour guides' job crafting, self-efficacy, and autonomy perception levels have been examined. Also, the effects of tour guides' self-efficacy levels and their autonomy perception levels on their job crafting behaviours have been determined. Lastly, comparing guides' job crafting behaviours in terms of their personal properties and demographic characteristics has been conducted within the framework of this study.
In line with the aims of the research, job crafting, autonomy, and self-efficacy levels of tour guides were determined at first within the framework of the study. According to the obtained results, the job crafting levels of tour guides were high. This is in consistent with studies in the literature. Chen et al. (2014) conducted a research on hotel employees and examined the relationship between job crafting and work engagement. They found that hotel employees' job crafting levels were high. In another study, Cheng & Yi (2018) stated that hotel employees' job crafting levels were high. Taking these results into consideration, it can be stated that employees in the tourism sector exhibit high levels of job crafting. This can result from jobs or tasks in tourism sector. As service quality and customer satisfaction act as determiners of continuity of a business; altering cognitive, interpersonal, or task-oriented aspects of a job; in other words, job crafting is evaluated as an option by employees to contribute to organizations' success arising from differentiating itself from other businesses. Considering the self-efficacy of tour guides, it can be stated that their self-efficacy levels were high. This result is in consistent with studies in the literature. In their study, Idrus et al. (2015) determined the self-efficacy levels of tour guides as 4.31. Guan & Huan (2019) also examined the self-efficacy levels of tour guides and they found out the self-efficacy levels of tour guides were 5.04. This can result from a personal characteristic. Self-efficacy affects internal motivation by accelerating sense of control, critical for proactive behaviour such as finding solutions to problems or making recommendations (Parker & Collins, 2010) which can be evaluated as a necessity while performing tour guiding profession. Taking tour guides' autonomy levels into consideration, it was found that their autonomy perception levels were high. According to Cheng et al. (2016), tour guides need autonomy in order to perform their jobs. Tsaur et al. (2011) have clearly stated that tour guides should have higher levels of autonomy in order to make arrangements in spontaneous situations. Researchers also add that tour guides exploit autonomy and other sources to find new ways to provide services to tourists. Yen et al. (2018) have also made a similar statement and expressed that tour guides have the ability to convert ideas into actions in order to fulfil tourist needs which is impossible with a concrete work design without autonomy. Our study findings can be evaluated as empirical support to these statements.
In this study, it was also determined that autonomy and self-efficacy had a significant effect on job crafting behaviours of tour guides. This is in line with Rudolph et al.'s (2017) study. In their study, Rudolph et al. (2017) created a model using meta-analysis technique and in the model, self-efficacy was determined as one of the individual differences resulted in job crafting and autonomy was categorized as one of the job characteristics. Besides, demographic characteristics such as age, gender, work hours, education background were also categorized as demographics affecting job crafting behaviours of the employees. The obtained results showed that job autonomy and self-efficacy had effects on job crafting behaviours of employees. As a profession, tour guiding needs space in order to improve complex problem-solving skills and the ability to make instant decisions. Also, job crafting makes it possible to alter various aspects of a job or a task that can give the necessary space to tour guides. So, it is not a surprising result that autonomy leads job crafting. Vanbelle et al. (2017) also determined that autonomy positively predicted job crafting behaviour. Furthermore, in their study, Wrzesniewski & Dutton (2001) have stated that autonomy encourages employees to realize opportunities and make changes in their jobs. Tims et al. (2014) determined that self-efficacy was one of the antecedents of job crafting behaviour. Moreover, Wang et al. (2017) have implied that autonomy is contextual, self-efficacy is a personal antecedent of job crafting behaviour. Our study findings support these statements and research results empirically.
Lastly, it was aimed to find out differences in job crafting levels of tour guides in terms of their personal properties and demographic characteristics in the study. Statistically significant differences in variables were determined. According to ANOVA Tukey test results, there was a significant difference in job crafting levels of tour guides in terms of their ages. Participants between 51 years old or more exhibited the highest level of job crafting among all groups. According to Wong & Tetrick (2017), job crafting is a valuable source especially for older workers at work, enabling workers altering physical, cognitive, and task-oriented aspects of their jobs. Researchers state that as workers get old, they may not keep up with the changes in the nature of work and job demands. Besides, they may strain in distinguishing opportunities at work. So as a bottom-up process, job crafting helps older workers realigning and improving their demands-abilities. Our study results support these statements empirically. Comparing license types, it was found that there were significant differences in job crafting behaviours of tour guides. Among all groups, participants obtained their license via a postgraduate degree had the highest level of job crafting. According to ANOVA Tukey test results, there were significant differences in job crafting behaviours of tour guides in terms of their experiences. Among all groups, tour guides with 15 years or more experience exhibited the highest level of job crafting. Rudolph et al. (2017) found a negative relationship between experience and job crafting. However, in their study, they did not examine the differences in the target group's job crafting behaviour in terms of experience. Our study findings give insight into the differences in terms of experience. As it was found that experienced tour guides (15 years or more) had the highest level of job crafting; this can be explained with guides' ages. Furthermore, it is known that there is a positive relationship between experience and job knowledge (Schmidt et al., 1986). Experience might provide necessary information about workflow and processes and accordingly, experienced employees might have the ability to craft accurate aspects of their jobs without creating negative side effects (Niessen et al., 2016). However, less experienced individuals without adequate job knowledge might alter tasks that can harm their jobs. Thus, they may be reluctant to exhibit high levels of job crafting behaviour. In the study, statistically significant differences in tour guides' job crafting behaviours in terms of number of active days in a year were determined. Among all groups, tour guides working 151 days or more in a year in the field had the highest level of job crafting. As the active days in a year increase, the necessity of crafting might increase.

Conclusion, Implications, and Future Recommendations
In this study, the effect of self-efficacy levels and autonomy perception levels of tour guides on their job crafting behaviours were determined. According to the results obtained, self-efficacy and autonomy had an effect on job crafting of tour guides. Besides, results showed that tour guides' autonomy, self-efficacy, and job crafting levels were high. Furthermore, there were differences in job crafting behaviours of tour guides in terms of their personal properties and demographic characteristics. Tour guides, 51 years old or more, exhibited the highest level of job crafting. Also, there was a significant difference in their job crafting behaviours in terms of their license type. Among all groups, tour guides who got their license via post-graduate education showed the highest level of job crafting behaviour. Taking their experience into consideration, among all groups, tour guides with 15 years or more experience exhibited the highest level of job crafting behaviour. Lastly, there was a significant difference in tour guides' job crafting behaviours in terms of the number of active days in a year. Tour guides performing their professions 151 days or more in a year exhibited the highest level of job crafting.
Although there are many studies on job crafting, studies specially designed on determiners of job crafting behaviours of tour guides have rarely been conducted. In literature, job crafting behaviours of tour guides have been measured in order to reveal the effect of the behaviour on positive outcomes. Since the theoretical framework of job crafting behaviour has been drawn by the studies, testing and supporting both antecedents and outcomes of the behaviour with empirical evidence may strengthen the literature. Also in this study, only self-efficacy and autonomy have been chosen as variables affecting job crafting behaviour. In order to contribute to the job crafting theory, different variables such as selfimage and self-reliance can be added into the model and it can be re-tested.
It is found that the job crafting levels of tour guides are high. Also, self-efficacy and autonomy perception levels of tour guides have a significant effect on their job crafting behaviours. That can be an important finding for travel agencies. With the responsibility of a group of individuals with different needs and demands, tour guides should exhibit job crafting behaviours often. In order to engage their jobs, fulfil the need to have control over their works or have a better self-image, tour guides should craft physical, cognitive or relational aspects of their jobs. So, it is highly recommended that tour or travel agencies should avoid implementing strict job designs, full of limitations. Also, as autonomy enables employees to take more responsibilities in their jobs, providing autonomy is an important work characteristic. It is also a critical phenomenon for job crafting behaviour. Therefore, in special professions such as tour guiding, responding to tourists' changing demands, and increasing customer satisfaction, providing job autonomy encourages them to exhibit more job crafting behaviours.
This study also has some limitations that may provide insight for future research. In our study, a model with two independent variables has been tested; future studies may consider including more variables into the constructed model. Besides, in this study, tour guides' job crafting, self-efficacy, and autonomy perception levels are determined as high. Researchers can frame their future studies with a different sample selected from other tourism employees and accordingly, can compare their findings with the data obtained in this study. Also, our study findings are limited to the data obtained from 203 Turkish tour guides. Therefore, future studies can conduct similar research with a larger number of participants from different countries. Lastly, quantitative research methods are adopted in this study. Researchers can design future studies using qualitative research methods in order to get deeper information on tour guides' job crafting behaviours.