Among all the HR issues present in hospitality businesses, employee retention needs to be considered as the hardest issue to be managed (Enz, 2009). Retention issue is significant due to the financial cost of recruiting and retaining employees (Hinkin & Tracey, 2000). For instance, employee turnover could cause emotional costs as well as recruiting and training (Hinkin & Tracy, 2000). Emotional cost explains downward reduction of service quality and employee morale (Hinkin & Tracy, 2000). Hospitality business shows relatively higher costs in both of turnover and emotional issues because hospitality employees are more likely to be influenced by other colleagues (Hinkin & Tracy, 2000) and business productivity can be lowered (Dipietro & Milman, 2004). Thus, investigating the causes of the retention issues and the retention strategies becomes inevitable.

One of the primary factors comes from the candidates' unclear career goals. The more they have clear career goals, the more passionately they explore and acquire the requisite knowledges and/or techniques to accomplish them (Damon, 2008). The deficient knowledge and/or information create unmet expectations between their perception and real career experiences, and this process makes them experience frustration and change career often. To this end, universities and/or colleges which breed up those candidates need to help their students develop career exploration based on long-term career goals. This is one of their primary roles because not only the paradigm of university makes a shift to career focuses but also they are evaluated through the employment rate and the graduates' term of service since hired.

In other words, employment plays an important role for survival of universities and students perceive universities as a main step for their career development. Yoon (2010) also emphasized the university educators’ role to build up their students’ sound work values. Therefore, the universities need to motivate students to explore their own careers and help their career development. As a result, the urgent issues of hospitality businesses are interlinked with those of universities.

In general terms, career exploration has been found to influence on vocational interests (Savickas, 1990) and job satisfaction and attainment (e.g., Werbel, 2000). Moreover, career exploration contributes to the career adaptation of university students (Ahn & Cho, 2012), and that encourages their personal growth (Urakami 1996) and career identity (Robistscheck & Cook, 1999). To this end, the need to investigate the conditions of career exploration is raised.

Within the existing literature, one main stream of research has centered on how the individual motivational variables make an impact on exploratory activities (e.g., Blustein, 1988). In the previous literature, the studies about the relationship between individual motivational factors and the hospitality students' academic performance (e.g., Kim & Byun, 2011) and career exploration (Shim & Lee, 2012) have been developed as well. However, the theoretical assertions upon career factors and individual realms need to be reconfirmed because the vocational issues tend to be disregarded in counseling psychology (Keller, 2006).

In addition, there are few studies to search for the individual motivational factors towards career exploration focused on the specific majors regardless of the raised issues for current educational needs. At the same time, although in the previous literature, self-esteem and anxiety have been frequently examined as the motivational variables towards the college students’ academic and other decision-making behaviors (e.g., Kim, Kim, Cho, & 2002; Chapman, 2006), there are little studies to clarify their influence on career exploration behaviors in a model. As a result, self-esteem and contextual anxiety are likely to be the matters of concern as predictors of career exploration behaviors

Based on the study needs, this study chose self-esteem and contextual anxiety as the predictors of career exploration. Further, researching how anxiety related to career exploration and self-esteem take a part in a student's life can imply significance for university personnel making an effort to encourage students' personal growth (Parisi, 2011).

To this end, it is suggested that there should be renewed studies on the predictors of career exploration behaviors. Hence focused on university students having in hospitality-related majors, who are potential candidates in hospitality business, this study aims to clarify the following goals: 1) investigating whether self-esteem significantly influences contextual anxiety in relation to career exploration; 2) investigating whether the anxiety positively or negatively influences career exploration behavior; 3) presenting the hierarchical relationship among the variables in a structural equation model. Through this study, university personnel may have clear direction of how to conceptualize and treat students' career development.

Self-esteem, commonly developed at a young age, refers to how one perceives about himself or herself. It usually consists of cognitive, behavioral, affective and physical components. It is considered to be the evaluation on a self (Babbio, 2009). Based on the standard of excellence in every society or group, self-evaluation depends on whether he or she fulfills the framework of these particular standards (Rosenberg, 1989). Between the age of 15 and 18, individuals are experiencing a lot of changes and become more care of themselves. Because this is a period of life change, he or she becomes more concerned about his or her own image (Rosenberg, 1989). In a word, self-esteem is a perception about his or her capability and attractiveness from others point of views rather than reality. Noticeably, self-esteem does not reflect a person’s actual attractiveness, and it derives from his or her self-perception (Baumeister, Campbell, Kruenger, & Vohs, 2003). According to previous studies, people are willing to retain their self-esteem because they like to feel positive about themselves based on an intrinsic need.

Self-esteem is influential to people’s overall life like personality, emotions, anxiety, etc. Kowalski & Western (2005) asserted that under a general level of self-esteem, "people have feelings about themselves along specific dimensions, such as morality, physical appearance, and competence" (p245). This helps us to understand those with high self-esteem versus those who don't. Although a person with high self-esteem doesn't need to be good at everything they do, those with high self-esteem are inclined to consider their strength more valuable than those with low self-esteem. They are aware of what they are weak at, and don’t put much weight on it.

Furthermore, high self-esteem is related to psychological well-being while low self-esteem is likely to associate with sadness and depression (Leary, 1999). High self-esteem has also been found to be positively correlated with academic performance because people with high self-esteem may have higher career goals than the ones with low self-esteem (Parisi, 2011). High self-esteem may lead people to have positive level of self confidence to deal with difficult tasks. They are more likely to consistently make an effort regardless of failure while those with low self-esteem may lose self-confidence at the moment of failure (Baumeister et al., 2003).

Finally, school age children’s self-esteem predicts their adulthood’s important performances and overall behaviors (Chung & Yuh, 2009). However, the college students’ self-esteem can be developed through training for communication competency, collective counseling, and wring therapy (Lee, Choi, & Hwang, 2009; Suh, 2009). In other words, self-esteem can be increased among college students if professional training is provided.

Anxiety is a “stress-induced emotion that figures in the group-think model” according to Chapman (2006, p1396). In addition, anxiety includes a distressed state in a fear of future insecurity (Chapman, 2006). Contextual anxiety refers to a kind of trait anxiety generated from stress related with a given activity in the context of career decision making and./or exploring behaviors (Stumpf, Colarelli, & Hartman, 1983). This kind of anxiety can influence the process of information seeking (Mittal & Ross, 1998) and decision-making (Baradell & Klein, 1993). In detail, contextual anxiety consists of exploratory and decisional stress (Stumpf et al., 1983). Exploratory stress means to which extent the exploratory behavior for a career stresses a person compared to the other life issues. Similarly, decisional stress means to which extent the decision-making process for a career stresses a person compared to the other life issues.

Researchers have verified a negative relationship between contextual anxiety and the students' self-esteem as follows. The existing studies asserted that self-esteem and stress were interrelated (e.g., Parish, 2011). According to Emil (2003), university students who have higher self-esteem showed lower stressful life events. Additionally, Abouserie (1994)'s study, undergraduate students with high self-esteem are less suppressed by stress than low self-esteem. Further, high self-esteem are unlikely to lower one’s sense of self in the face of stressors, thus looking for intrinsic coping (Guinn & Vincent, 2002). In addition, Park, Bae, & Jung (2002) presented the negatively significant relationship between self-esteem and job exploration. Withdrawn from the previous studies, the relationship between self-esteem and contextual anxiety has been hypothesized, as follows.

3.1. Career Exploration: theory and practice

The theory of career exploration has been extended since the early 1960s, when it was first considered as a concept of general exploratory behavior (Jordaan, 1963). Four different conceptions of career exploration were reviewed by Taveira (1997, 2001). He or she clearly illustrated the evolution of the construct during the several decades as follows.

The first concept, based on Krumboltz's learning theory of learning choice and counseling (Krumboltz, 1979), explains career exploration as an information-seeking behavior or as a career problem-solving behavior. A second concept was originated from career decision-making included the activities of identifying and evaluating options, and seeking information (e.g., Gelatt, 1962). A third concept was in relation with general career development theories considered exploration as a major life stage, that of adolescence (from age 14 to 24). It consisted of the career development behaviors such as collecting, experiencing and implementing an occupational options (Super, 1957). The final concept described exploration as a life-span process to train and develop careers (Taveira, 1997).

In the context of the last position, career exploration has been involving a complex psychological mechanism. For instance, career exploration continues to search for information related and examine self and environment on the basis of clear career goals. In other words, people need to be cognitively and emotionally motivated to interpret and recreate past and current experiences and utilize them for future career to exert the behaviors to exert career exploration. Finally, according to the existing views of the construct, selfand environmental exploration are convinced as two sub-dimensions of the career exploration.

Furthermore, career exploration is different from general exploration in the core goals of exploratory behaviors, and not in the process or condition of that activity. In other words, although someone merely seeks information or examine one’s own past or present experiences, it cannot be assured that he or she is in the process of career exploration. For example, career exploration needs to be conducted through clear goal set up.

As related to this study purpose, the study of career exploration process is considered as one of the main stream of research in the context of career exploration. The contextual and personal dimensions of career exploration are still questioned although they have been studied over the several decades under various perspectives. Under the social learning paradigm, the first studies indicate that individuals more reinforced to information and/or indirect experiences explore careers more frequently (e.g., Krumboltz & Schroeder, 1965).

Other studies suggest that an individual personality also makes a difference on career exploration behaviors (Lazarick, Fishbein, Loiello, & Howard, 1988), pinpointing the importance of external reinforcement in the qualitatively exploratory conditions.

In this context, mostly the studies regarding the motivational theory have suggested that personal factors predict to which extent individuals have career exploration activities. The personal dimension is explained by work role salience, the value of career domain, accomplishment of valued career goals, satisfaction from decision made and usefulness of career exploration (Bluetein & Flum, 1999; Taveira, 2001). In short, it can be said that career exploration is conducted through the value of the career development dimension and through the previous success with career results and the consequent balance between expectation and outcomes.

The studies including self-determination theory by Deci and Ryan (1985) are based on motivation theory as well, and thus suggest that career may be explored though self-determination to some extent (Blustein, 1988). This delivers that career exploration is processed by intrinsic motivation or internalization of extrinsic regulation and this explains the individual variables. Having delineated the elements involved in career exploration models, a motivational factors are revealed self-efficacy (e.g., Creed, Patton, & Prideaux, 2007; Park, 2011), contextual anxiety (Keller, 2006), self-determination, perceived competence, internal locus of control, and self-esteem move towards career exploration intentions and/or behaviors (e.g., Betz & Voyten 1997; Bluestein, 1988, 1989; Solberg, 1998; Taylor & Pompa, 1990).

According to Flum & Bluestein (2000), both individual and environmental variables motivate the exploratory behaviors. Individual factors include self-interest, abilities, other self-related factors while environmental variables include parents' level of education and relationship with important others such as parents, teachers, and friends. Savickas (2002) also indicated the both dimensions explained career exploration behaviors. However, this study focuses on the self-esteem and the contextual anxiety as individual variables. The following review of literature addresses how these factors related to career exploration.

3.2. The Relationship between Self-Esteem and Career Exploration

Students' psychological well-being such as self-esteem and life satisfaction influence their access (participation) to the experiences related to their future employment (Creed et al., 2007). Self-esteem is a perception on self through reflecting how similar a person’s ideal self and real self are. Therefore, self-esteem motivates an individual to make and retain specific goals for themselves (Leary, 1999). For example, adolescent females in a loss of self-esteem experience low academic achievement and loss of career goals (Sadker & Sadker ,1994). In addition, Cho & Ha (2010) asserted that self-esteem, so called psychological characteristic, had a significant and positive impact on career commitment, which explained participation and attachment to career development. Further, Shim & Lee (2011) also proved the positive impact of self-esteem on the hospitality students' career exploration behaviors. Based on the previous studies, the following hypotheses have been drawn.

3.3. The Relationship between Contextual Anxiety and Career Exploration

When concerning the relationship between contextual anxiety and career exploration, there are conflicting theories as to whether anxiety play a positive or negative role for career exploration..

For example, Greenhaus and Sklarew (1981) presented that the people who were less stressed showed active exploration and consequently resulted in positive outcomes. Nichols & Schwartz (1998)’s study also appears to emphasize that the lower level of emotional reactivity towards stress are beneficial when it engages with specific activities.

Further, Kowalski & Westen (2005) depicted that those who have more positive attitude towards a stressor have a resilience of mind than those with a negative attitude. In other words, less stressed individuals from a stressor have more control over their mind. This leads them to the quality of decision making is better than those who are more stressed (Chapman, 2006). Further, in Mittal & Ross (1998)’s study, those in the mood of anxiety tend to process information systematically. On the other hand, the people in a positive emotion are prone to search for information strategically. Those in a positive emotion are likely to be the efficient researchers.

On the other hand, Bluestein and Phillips (1988)’s study on the relationship does have different idea. They stated, "...some forms of contextual anxiety may provide a necessary incentive for individuals to engage in career exploration" (p213). It means that some anxiety about career decision making and exploration moves students towards exploratory behavior. In addition, Bartley (1998)'s study centered on university students suggested that the exploratory stress experienced by male students was likely to result in environmental exploration. Further, the study result added that the both of male and female students stressed by career exploration process are likely to engage in self-exploration. Keller (2006)'s study also indicated career-related stress makes students seek more career information and/or opportunities.

Anxiety associated with career exploration drives a person towards career exploration behaviors through the activation of common defence mechanism, however, defective decision-making could occur (Chapman, 2006). On the other hand, positive emotion associated with a stressor such as career exploration may result in strategic exploration according to Mittal and Ross (1998)’s idea. To this end, the significant influence of contextual anxiety on the college students’ career exploration behaviors can be assumed. However, the direction of the relationship could not be decided based on the literature. As a result, the hypothesis between contextual anxiety and career exploration have been suggested as follows.

Based on the propositions developed from the literature review, the conceptual model to explain the relations among the hospitality students' self-esteem, exploratory stress, decisional stress, environmental exploration, and self-exploration is illustrated in Figure 1.

1.1. Self-Esteem

Self-esteem defined as a perception about individuals’ capability and attractiveness reflected by others was evaluated through Rosenberg Self-esteem scale (RSE:Rosenberg, 1989). The ten items were used to measure global evaluation of self worth. Participants were asked to rate to which extent they agree with each statement on a five-point scale from "strongly agree" to "strongly disagree".

1.2. Contextual Anxiety

Contextual anxiety refers to a kind of trait anxiety from stress regarding a mandatory activity within the context of decision making or career exploration in this study. Contextual anxiety was measured using the Exploratory Stress (ES) scale and the Decisional Stress (DS) scales of the Career Exploration Survey (CES; Stumpf, Colarelli, & Hartmen, 1983). The ES scale include three items and the DS scale have five items. The Participants compared career-related stress with other stresses related to important life issues. Answers were put on a 5-point Likert scale from “insignificant compared to other issues with which I have had to contend” to "one of the most stressful issue with which I have had to contend." Based on Stumpf et al. (1983)’s scale, modified items by Keller (2006) was used since he or she adjusted wording so as to apply to university sample.

1.3. Career Exploration

Career exploration is a long term process to train and upgrade careers through overall life and include exploration on self and environment. The dimensions of Environmental Exploration (EE) and Self Exploration(SE) scales were measured employing Exploration Survey (CES; Stumpf et al., 1983). Participants rated how much they were enrolled in career exploration activities over the previous three months. Environmental exploration consists of six items and self exploration consist of five items. Items were measured based on a five-point Likert scale raging from “little” to “a tremendous amount.”

The people targeted for this study are university students majoring in hospitality related management. First of all, draft survey questionnaires had been drawn after translating the questionnaire in English to Korean. In addition, hospitality students of K university in Seoul and D university in Daegu were identified to conduct preliminary survey to presume the validity of the structural model. In preliminary survey, the researcher mailed draft questionnaires to a total of 50 students. Moreover, the respondents were recommended to pinpoint any inadequate and/or unclear items. This preliminary survey has been executed from September 15th to 30th in 2012.

To conduct a main survey, a convenience sample of universities which have department of hospitality in Korea were surveyed; K1, K2 and S universities in Seoul, D1, D2, Y and G universities in Daegu, and D and P universities in Pusan. Further, a mail survey was employed to carry out the survey after each department professor's approval had been given. The main survey was conducted from October 15th to November 25th in 2012. A total of 315 responses were received out of the 400 cases. Cases with missing value were subsequently dropped from the analysis and 277 faithful cases have been analyzed.

The collected data was analyzed employing the SPSS and AMOS 20 software program. Descriptive statistics, multi-variate analysis of variance, and structural equation modeling (SEM) were utilized. Frequency analysis, reliability analysis after using Cronbach's alpha, and confirmatory factor analysis were operated. Further, the correlation analysis was conducted to verify the reciprocal relationship among the variables. To verify the hypotheses and model of the study, confirmatory factor analysis was conducted to examine conformity of the causal relationship among each factor and covariance structure analysis was used to investigate a path coefficient.

As presented in

, 39.2 % of respondents were males, and 60.4% of them were females. It reflects that mainly females are majored in hospitality management. All the respondents are in twenties and average age is 20.78. Moreover, freshmen consist of 37 people (13.3%); sophomores consist of 71 people (25.5%); juniors consist of 131 students (47.5); seniors consist of 38 people (13.7%). Finally, their current majors were as follows; tourism management (18.7%), hotel management (32.1%), hotel and tourism management (19.1%), food service management (7.9%), convention management (3.2%) and hospitality management (18.7%).

] Result of the demographic analysis of the respondents

   2. Analysis of Validity and Reliability

, GFI (0.888) and AGFI (0.862) indicate unfulfilled indices, however, RMR (0.050), NFI (0.900) CFI (0.963), TLI (0.963), and RMSEA (0.043) indicate the reasonable fits of the data. The relatively small sample sizes limit the possibility of reaching the 0.9 cutoff value for the two fit indices and they are not dependable as "a stand alone index " (Hooper, Joseph & Mullen, 2008, p54). Further, a strict adherence to suggested cutoff values can result in the improper rejection of an acceptable model (Marsh, Hau, & Wen, 2004). Therefore, the relationship among the latent variables can be presumed with the reasonable fits of the data.

] Confirmatory factor analysis and reliability analysis

Further,

presents standard estimates for a measurement model. As illustrated, factor loading of all measures were moderate (ranging from 0.615 to 0.892). The factor loadings showed that relevant measurement items performed moderately well in reflecting the designated underlying construct.

2.2. Results of Validity and Reliability

As the survey items are adapted from different streams of studies, it is important to ensure construct reliability and validity. Cronbach's coefficient alpha was calculated to determine reliability of the measurement. As indicated in

, Cronbach's a of each construct in measurement model is ranged from 0.772 to 0.925, significantly a scale with high level of reliability; this value is adequate at Cronbach's a >0.60 (Lee, 2009). If construct reliability reaches above 0.7, convergent validity or internal consistency is secured 2007). Also, convergent validity is procured as long as AVE reaches above 0.5 (Kim, 2007). In terms of construct reliability, the values of five constructs are ranged from 0.788 to 0.937. At the same time, as illustrated in , factor loading of each variable is above 0.615, showing a moderate to high construct validity. Further, each average variance extracted (AVE) reaches between 0.500 to 0.636.

] Measurement model

Discriminant validity was established using the procedures outlined by Fornell and Larcker (1981).

shows the correlations between the latent variables and the average variance extracted (AVE) of each construct. Fornell and Larcker (1981) prescribe that the squared correlation between constructs must be less than the AVE of each underlying construct in order for the constructs to have discriminant validity. As suggested in , the squared correlations between two constructs are ranged lower than each AVE. These outcomes established discriminant validity. As a result, these values represent all five constructs and it is significant to analyze the relationship between those constructs.

   3. Test of Hypotheses Testing

illustrated the strength of the relationships among the constructs, showing path coefficients and overall goodness of model fit indices. Overall, the model was acceptable fit; (x2: df =329)= 470.952 (p = 0.000), GFI = 0.892, AGFI = 0.867, RMR = 0.057, NFI = 0.912, TLI = 0.965, CFI = 0.970, RMSEA=0.039.

] Parameter estimates in structural model

As shown in

, the postulated hypotheses were examined through investigating the path coefficients of the constructs in the final model. Firstly, the results indicate that exploratory stress (β = - 0.250, p<0.001) and decisional stress (β = - 0.332, p<0.001) were significantly predicted by hospitality majored student's self-esteem. In other words, the students' self-esteem drives them to less stress about exploration and decisional activities. Specifically, the negative impact of students' self-esteem on decisional stress is higher than the impact on exploratory stress. Therefore, hypotheses 1-1 and 1-2 were supported. Overall, the hypothesis 1 for the relationship between self-esteem and contextual anxiety was supported.

Secondly, results suggest that hospitality majored students' self-esteem significantly influences environmental exploration (β = 0.309, p<0.001) and self exploration (β = 0.568, p<0.001). That is, the students who have higher self-esteem tend to look for career opportunities and reflect themselves to fit into specific careers. Therefore, the hypotheses 2-1 and 2-2 were supported. That is, the hypothetical impact of self-esteem on career exploration was fulfilled.

Finally, results indicate that hospitality majored students' exploratory stress does not have any significant impact on their environmental exploration while it significantly influences their self exploration behavior (β = -0.141, p<0.05). At the same time, decisional stress is a negative predictor of environmental exploration (β = -0.202, p<0.01), while it is not a significant predictor of self exploration. Therefore, hypotheses 3-1 and 3-4 were rejected while the hypotheses 3-2 and 3-3 were supported. As a result, the contextual anxiety has a partially negative impact on career . This means, the students with positive emotion about career exploration have more research on career opportunities and self examination to some extent. In the hypotheses setting, the positive or negative direction was not decided due to the possible outcomes of both directions based on the existing literature. This study result supports the negative relationship between the contextual anxiety and career exploration.

V. Discussion

   1. Findings and Implications

   2. Limitations and Future Research