Psychological Contracts, Work Behaviour and Turnover Intention in the Hong Kong Telecommunications Industry

This paper reports a quantitative study of psychological contract between employees and employers in the telecommunications industry in Hong Kong, and the relationships between the psychological contract, work behaviour and turnover intention. Previous research in the telecommunications industry in Hong Kong has focused mainly on consumer protection and telecommunications market competit ion, and research has seldom been done on the human resources issues in this industry. This research investigates the informal relationship between employees and employers in the telecommunications industry in Hong Kong. A quantitative research methodology was used in this research by surveying employees. A total of 227 valid responses were received, and the data was analysed by Partial Least Squares Structural Equation Modelling (PLS-SEM). The results found that the psychological contract has a significant relat ionship with work behaviour and turnover intention. The psychological contract is a better means of exp laining turnover intention than work behaviour. Business managers can leverage this finding to develop better plans and approaches to deal with work behaviour and turnover intention issues. The outcomes also provide empirical evidence to validate the psychological contract inventory developed by Rousseau [1]. This research contributes to the development of human resource theory and practice.


Introduction
The deregulation of telecommun ications services in Hong Kong not only creates business opportunities for companies but also provides substantial career development opportunities for emp loyees. The business growth of teleco mmunications services has established a strong workforce in this industry. However, based on the statistical informat ion fro m the Office of Teleco mmunications Authority of Hong Kong (OFTA) [2], the total number of teleco mmunicat ions full-time staff has fluctuated between 15,862 and 26,042 in the past decade in a declining trend. Staff turnover has been a co mmon issue encountered by telecommun ications service providers in Hong Kong. Despite this, there has been little research on human resources issues in the industry, and previous research has focused main ly on consumer protection and telecommun ications market co mpetition. It is essential to investigate the relationship between the emp loyer and emp loyee and how it is affecting work behaviour and turnover intention.
Emp loyee turnover has been a challenge for many organisations, in particular for human resources professionals [3]. It has been a crit ical organisational phenomenon that has attracted considerable managerial and scholarly attention for many decades [4]. Porter and Steers [5] argued that turnover has potentially crit ical consequences for both the person and the organisation. This study aims to investigate the relationship between the emp loyee and employer in the teleco mmun ications industry with regards to the psychological contracts involved, and also the relat ionship between the psychological contract, work behaviour and turnover intention. Lam [6] proposed effective retention strategies to tackle the turnover of the mailing staff in one of the largest telecommun ications services operators after conducting an emp irical study. To understand the relationship between the emp loyee and emp loyers, many scholars have proposed studying the psychological contract. A psychological contract is defined as "An individual's beliefs regarding the terms and conditions of a reciprocal exchange agreement between that focal person and another party" [7]. Prev ious research on the psychological contract has largely been undertaken in Western countries, so the results might not apply to Hong Kong given the cultural differences. The objective of the study is to deepen managers ' and researchers' knowledge and understanding in psychological contracts so as to use them to formu late appropriate human resources strategies and plans.

Literature Review
The concept of the psychological contract can b e traced back to the 1930s, and significant academic interest in it has grown since the late 1990s [8]. The psychological contract is an important research tool in organisational studies for understanding how the emp loyee-organisation relationship is constituted and how it develops and changes, and for understanding the consequences of a breakdown in the relat ionship [9]. It is about the mutual expectations [10], the relationship between the rewards and the work performed [11], and the rights, priv ileges, and obligations between employee and employer [12]. This psychological relationship develops around the explicit process of co-operative exchanges that operate between an employee and an employer [13].

Psychological Contract
The psychological contract involves the employee receiving benefits such as pay, training and job security, in return for prov iding effort and contribution to the organisation [14]. Rousseau recognises the psychological contract as a particular type of pro missory contract and, because it involved beliefs, it existed "in the eye of the beholder" [15]. It is important to have a healthy psychological contract to ensure a harmonious relationship between the employee and employer, and once violated by either party they no longer have a co mmon set of goals [16]. Rousseau [1] mentioned that researchers had proposed four different typologies including relational, balanced, transactional and transitional to conceptualise the forms of promissory contracts.
"Relat ional contracts are based on mutual trust and loyalty; it is a long-term or open-ended employ ment relationship. The rewards are deriv ing fro m membership and participation in the organisation and are only loosely conditioned on the performance of emp loyees." [1, p. 4]. Rousseau [1] proposed two constructs to measure the relational contracts, they are stability and loyalty. Stability is a long-term emp loyment relationship in which the emp loyer co mmits to offering stable wages, and the emp loyee is obligated to do what is required to keep the job and remain with the firm. In loyalty contracts, the emp loyees demonstrate that they are good organisational 'citizens': they are obligated to support the firm, manifest loyalty and co mmit ment to the needs and interests of the organisation. In return, the emp loyer is co mmitted to supporting the wellbeing and interests of emp loyees and their families.
"Balanced contracts are dynamic and open-ended emp loyment arrangements, which are conditional on the economic success of the firm and emp loyee opportunities to develop career advantages. Emp loyee and employer make valuable contributions to each other with learning and development. The rewards to employees are performance-based, which can contribute to the firm's comparative advantages, particularly under the market pressure of changing demands", [1, p. 4]. There are three constructs proposed by Rousseau [1] to measure the balanced contracts, namely External Emp loyability, Internal Advancement and Dynamic Performance. External emp loyability measures the obligation to career development on the external labour market. The employee is obligated to develop their o wn skills wh ich are demanded in the market, and the employer makes a commit ment to enhance the employee's long-term emp loyability whether it is inside or outside the organisation. Internal advancement measures the obligation to career develop ment inside an organisation. The emp loyee has an obligation to develop their own skills which can be of value to their current employer, and the emp loyer has made a co mmit ment to create internal career advancement opportunities for the emp loyee. Dynamic performance measures the obligation to perfo rmance support. The emp loyee is obligated to take up and perform well new and mo re demanding goals, which constantly change, in order to help the firm become and remain competitive. The emp loyer is committed to providing continuous learning opportunities to help employees achieve their performance requirements.
"Transactional contracts are a short-term or limited time employ ment arrangement. They are primarily focused on the economic exchange that the employee's duties provide specific, narrow and only limited involvement in the organisation." [1, p. 4]. Rousseau [1] proposed two constructs to measure transactional contracts, namely Narro w and Short-term. Narrow measures the relationship between the emp loyee and emp loyer in a limited or fixed scope of duties. The employee is obligated and paid to perform only an agreed set of duties. The emp loyer has committed to offering the employee only limited involvement in the organisation with little or no training and emp loyee development opportunities. Short-term measures the relationship between the emp loyee and employer in a specific and limited time. The emp loyee basically has no obligation to remain with the firm and is committed to wo rking only for a specific and limited t ime. The emp loyer is not obligated to any future commitments.
"Transitional contracts are a cognitive state reflect ing the consequences of organisational change and transition that are at odds with previously established employment arrangements." [1, p. 4]. Rousseau [1] proposed three constructs to measure transitional contracts, namely Mistrust, Uncertainty and Erosion. Mistrust measures the lack of trust between the emp loyee and the employer. This is because the emp loyee believes the organisation sends unclear, inconsistent and mixed in formation about its intentions. The employer does not trust its workers either, and withholds important informat ion about the organisation. Uncertainty measures the relationship between the emp loyee and employer in an environment of the ambiguity of obligation and future commit ments. The emp loyee is not clear about the nature of h is or her o wn obligations to the firm and the emp loyer evaluates the extent that the employee is uncertain regarding the emp loyer's future co mmit ments to him or her. Erosion measures the relationship between the employee and emp loyer in a corrosive environment. The employee anticipates a declining of returns fro m his or her contributions to the firm co mpared to the past and the emp loyer has introduced changes that reduce employee rewards, e.g. wages and benefits, eroding the quality of work-life compared to previously.

Work Behaviour
The concept of organisational citizenship behaviour (OCB) was introduced in the organisational research literature in the early 1980s [17]. Smith, Organ and Near [18] defined the last of these aspects as "Organisational Citizenship Behaviour" (OCB), and Organ [19] argued that OCB should be regarded as the most important of the three aspects of behaviour as it can pro mote the efficient and effective functioning of an organisation. Th is is because the sense of responsibility felt by employees can be expected to extend well beyond the job remit. OCB does influence the organisation's performance, as the available empirical ev idence has indicated [20]. Kazt and Kahan [21] and Jex [22] argued that OCB is a form of productive behaviour that affects the organisation's effectiveness. Spector [23] also argued that OCB is an essential aspect of the emp loyee's behaviour that can contribute to the overall effect iveness of the organisation. In research undertaken by Mackenzie, Podsakoff and Fetter [24], it was indicated that there was no positive correlation found between individual's performance and his/her OCB score; however, other research undertaken by Podsakoff, Ahearne and MacKen zie [25] found that group performances are correlated to the extent of the OCB.
With the growing interest in OCB, the constructs were expanded and diversified [26]. OCB can be categorised using two broad dimensions [18]: altruis m and generalised compliance. In this research, these two dimensions are used to measure the OCB of the respondents. Altruism captures the direct and intentional behaviour in a specific face-to-face situation, such as orienting a new colleague, or helping a co lleague with a heavy workload [18]. Podsakoff et al. [20,p516] defined altruis m as helping behaviour, follo wing the work done by Organ [19], who defined altruis m as a voluntary action that helps people to solve work problems, such as teaching new staff to use equipment, help ing peers to clean up their work backlog, and searching for informat ion for someone. Altruis m is labelled as OCBI-behaviour and immed iately benefits specific individuals, and indirectly through contributions to the organisation (e.g. covering for colleagues who are absent from their work) [27]. In this sense, altruism involves doing something extra to help sup ervisors or peers such as taking up the work of an absent co -worker [23]. General co mp liance "is doing what needs to be done and following ru les, such as going to work on time and not wasting time" [23, p236]. Generalised compliance "pertains to a more impersonal form o f conscientiousness that does not provide immediate aid to any one specific person, but rather is indirectly helpful to others involved in the system" [18, p 657]. The behaviour (e.g., punctuality, not wasting time) seems to represent something akin to compliance with internalised norms that define what a "good employee ought to do". Williams and Anderson [27,p601] labelled general co mpliance as the OCB-O dimension, behaviour that benefits the organisation in general (e.g. p roviding prior notice of absence fro m work, or following informal and self-defined rules to maintain order).

Turnover
Turnover refers to "employees both entering and leaving organisations" [28]. It is often implied that emp loyees who leave an organisation do so on their own initiat ive rather than as a result of organisational decisions. It is "the cessation of membership in an o rganisation by an individual who received monetary compensation from the organisation" [29]. Turnover intention is related to the various informa l relationships between the employee and emp loyer. The informal relationships include job satisfaction [30], emp loyee expectations from the emp loyer for career development and working, opportunities for alternative emp loy ment, participation in job-related decisions, receiving job-related informat ion, friendships with peers, the organisation's commit ment, job nature [31], and working conditions [32]. All these informal relat ionships are highly relevant to the psychological contract model as proposed by Rouss eau [1]. The respondents were observed to have moderate occupational and organisational co mmit ment and low turnover intention [6].

Conceptual Framework
The conceptual framework consists of three main components, namely Psychological Contract, Work Behaviour and Tu rnover Intention (See Figure 1). The literature indicates that the psychological contract will lead to a positive or negative effect on work behaviour; and psychological contract commit ment and work behaviour will lead to a turnover decision. Putting these three factors together in this study would provide a comprehensive view of their relat ionships and a better understanding of the consequences of psychological contracts, and how they affect the employees' work behaviour and their intention to leave the co mpany. This research aims to answer the research questionswhat is the relationship between psychological contract, work behaviour and turnover intention?

Methods
There were three main concept measures in this study: the Psychological Contract (PC), Work Behaviour (W B) and Turnover Intention (TI). Psychological contracts are described by the instruments developed by Rousseau [33], which measure both the emp loyee and emp loyer obligations. The description of work behaviour is based on the two dimensions of Organisational Cit izenship Behaviour (OCB) proposed by Smith, Organ and Near [18,, namely Altru ism and Generalised Co mpliance. TI is measured using the instrument created by Mobley, Horner and Ho llingsworth [34]. Th is instrument uses three items: thinking of quitting, intention to search for other employment, and intention to quit the present job.
The Psychological Contract was measured by four constructs, namely Psychological Contract -Balanced (PCB), Psychological Contract -Relational (PCR), Psychological Contract -Transactional (PBTa), and Psychological Contract -Transitions (PCTi). Work Behaviour is re ferred to as Organisational Cit izenship Behaviour (OCB) within this research and was measured by two constructs namely Work Behaviour -Altru ism (WBA) and Work Behaviour -Generalised Co mpliance (WBGC). TI had only one construct. All variables were measured by using a five-point Likert-scale in which 1 was "Not at all", 2 was "Slightly", 3 was "So mewhat", 4 was "Moderately", and 5 was "To a great extent".
The questionnaire was carefully designed based on the suggestions of Ticehurst and Veal [35] and Frazer and Lawley [36] who reco mmend that jargon, leading questions and amb iguities be avoided. The survey was carried out using a self-ad ministrated questionnaire with a target sample size of 200. The data was prepared in three interactive steps: data entry, data editing and data coding. The target respondents were emp loyees in the telecommun ications industry in Hong Kong. Bayesian approach is applied to determine the appropriate sample size with N 0 = t 2 x s 2 / d 2 . Where t = 1.96 for 0.05 alpha level, s = 0.833 for 5-point scale and 6 standard deviations, d = 0.15 for 5-point scale and 0.03 marg in of error. Therefore the minimu m samp ling size is 118. The targeted sampling size for this research is 200.

Findings
There were a total of 235 questionnaires returned, and 227 were valid. The respondents were mostly aged 40-to 49-years-old (35.3%); female (60%); married (56.5%); holding a Bachelor's degree (42.3%); working at officer/senior officer level (34.7%); working in sales and market ing areas (30.2%); having 15 to 20 years working experience; and 1 to 5 years working in their current organisation. The respondents' organisations were largely providing mobile services (44.2%); had more than 15 years' industry experience (32.4%) the size of the co mpanies comprised 1000 to 1999 staff (22.6%); and more than half were private companies (52.7%).
The validity and reliab ility of each construct was assessed using exp loratory factor analysis (EFA) and Cronbach's Alpha Coefficients respectively. Normality was checked to determine if any transformation was required prior to performing mu ltivariate analysis. Principal co mponents analysis (PCA) and Pearson Correlations were used in EFA. The assessment was done based on the suggestions by Hair et al. [37] that the item-to-item correlat ions in PCA should exceed 0.3, the item-to-total correlations in PCA should exceed 0.5, and there should only be one eigenvalue greater than 1 and the loading factors are all > 0.5. In the exp loratory study, values of >0.6 were considered acceptable for the Cronbach's Alpha coefficient. In the main study, Cronbach's Alphas >0.7 were considered acceptable, >0.8 were considered good and >0.9 were considered to be of excellent reliability [37 p118].
Hair et al. [38] suggested evaluating the collinearity issues of the structural model before proceeding with any analysis. The high collinearity can lead to incorrect estimation of the weights and their statistical significance. To assess the level of co llinearity, Hair et al. [38] suggested calculating the tolerance (TOL), which represents the amount of variance of the formative indicator not exp lained by the other indicators in the same model. A related measure of collinearity is the variance inflat ion factor (VIF), wh ich is defined as the reciprocal of the TOR. The VIF value of 5 represents 80% o f its variance, wh ich is accounted for by the remaining formative indicators associated with the same construct. When the VIF is higher than 5, the suggestion is to consider removing one of the corresponding indicators if it still sufficiently captures the content of the construct from a theoretical perspective [38]. In this research, the VIF of the inner model was checked, and all values are below the threshold 5; therefore, the collinearity among the predictor's constructs is not a critical issue, and we can continue the examination of the results.
Gefen, Straub and Boudreau [39] argued that the model of fit for a PLS model be established with the acceptability of high R 2 values, with significant path coefficients. Hair et al. [38] suggested calculating the standard bootstrap error, which means computing the emp irical t values and p values for the assessment. The coefficient is statistically significant at a certain error probability when it is larger than the critical value. In this research, 5 % significant level is assumed, the crit ical is 1.96 for a two-tailed test. In this model, the R 2 for Turnover Intention (TI) is 0.53, Work Behaviour -A ltruis m (W BA) is 0.62, and Work Behaviour -Generalised Co mp liance (W BGC) is 0. 23. The t values and p values are reported in Table 1 and the paths with p>0.05 are found significant. In addit ion to assessing the R2 values, Hair et al. [38] suggested reporting the effect size f 2 of the endogenous constructs. The effect size f 2 is to measure the impact of the change in the R 2 when a specified exogenous construct is omitted fro m the model. Cohen [40] mentioned that the f 2 with a value of 0.02, 0. 15 and 0.35 represent small, moderate, and substantial effects. In this research, the f 2 values were calculated by SmartPLS. In the research, the effect size f 2 for the significant paths is between small to mediu m; whereas PCB to WBA has the large f 2 values (0.32), wh ich is between mediu m to large effect, and the remaining paths with small (0.02< f 2 ) to medium effect (0.02< f 2 <0.15).  The results from the Smart PLS analysis for the hypothesis test is as shown in Table 1 and Figure 2. The sub-hypothesis H1.1 tested the relationship between the Psychological Contract Balanced (PCB) and Turnover Intention (TI) with a path coefficient 0.25 (t=2.5, p=0.01), even the path was statistical significant, this hypothesis was not supported as the path coefficient is a positive number, wh ich implies a positive correlation instead of a negative correlation.
The sub-hypothesis H1.2 tested the relationship between Psychological Contract Relat ional (PCR) and Turnover Intention (TI) with a path coeffic ient -. 35 (t=3.72, p=0.00), this result supports our hypothesis of a negative relationship between the Psychological Contract Relational (PCR) and Turnover Intention (TI).
The sub-hypothesis H1.3 tested the relationship between Psychological Contract Transactional (PCTa) and Turnover Intention (TI) with a path coefficient -0.36 (t=4.39, p=0.00). Since the questions for PCTa were asked in a reverse direct ion and the items were reverse transformed during the data preparation as discussed, the negative path coefficient is representing a positive correlation. This result supports our hypothesis of a positive relationship between the Psychological Contract Transactional (PCTa) and Turnover Intention (TI).
The sub-hypothesis H1.4 tested the relationship between Psychological Contract Transitional (PCTi) and Turnover Intention (TI) with a path coefficient -0.2 (t=3.78, p=0.01). Since the questions for PCTi were asked in a reverse direction and the items were reverse transformed during the data preparation, the negative path coefficient is representing a positive correlation. This supports our hypothesis of a positive relationship between Psychological Contract Transitional (PCTi) and Turnover Intention (TI).
The sub-hypothesis H2.1 tested the relationship between the Psychological Contract Balanced (PCB) and Work Behaviour -A ltruis m (WBA) with a path coefficient of 0. 55 (t=8.99, p=0.00). This result supports our hypothesis of a positive relationship between Psychological Contract -Balanced (PCB) and Work Behaviour -Altruism (WBA).
The sub-hypothesis H2.2 tested the relationship between the Psychological Contract -Balanced (PCB) and Work Behaviour-Generalised Co mp liance (W BGC) with a path coefficient of 0.28 (t=2.62, p=0.04). This result supports our hypothesis of a positive relationship between Psychological Contract -Balanced (PCB) and Work Behaviour -Generalised Co mpliance (WBGC) as p<0.05, which is statistically significant.
The sub-hypothesis H2.3 tested the relationship between Psychological Contract -Relational (PCR) and Work Behaviour -A ltruis m (WBA) with a path coefficient of 0.11 (t=1.61, p=0.11). This result does not support our hypothesis of a positive relat ionship between the Psychological Contract -Relational (PCR) and Work Behaviour -Altru ism (WBA ) as p>0.05, wh ich is statistically not significant.
The sub-hypothesis H2.4 tested the relationship between Psychological Contract -Relational (PCR) and Work Behaviour -Generalised Co mpliance (W BGC) with a path coefficient of -0.09 (t =0.89, p=0.37). This result does not support our hypothesis of a positive relationship between the Psychological Contract -Relational (PCR) and Work Behaviour -Generalised Co mpliance (WBGC) as p>0.05, which is statistically not significant.
The sub-hypothesis H2.5 tested the relationship between Psychological Contract -Transactional (PCTa) and Work Behaviour -A ltruis m (WBA) with a path coefficient of 0.14 (t=2.02, p=0.1). Since the questions for PCTa were asked in a reverse d irection and the items were reverse transformed during the data preparation, the positive path coefficient is representing a negative correlation. This result supports our hypothesis of a negative relationship between the Psychological Contract -Transactional (PCTa) and Work Behaviour -Altruism (WBA).
The sub-hypothesis H2.6 tested the relationship between Psychological Contract -Transactional (PCTa) and Work Behaviour -Generalised Co mpliance (W BGC) with a path coefficient of 0.14 (t=1.46, p=0.14). This result does not support our hypothesis of a negative relationship between the Psychological Contract -Transactional (PCTa) and Work Behaviour -Generalised Co mp liance (W BGC) as the p>0.05, which is statistically not significant.
The sub-hypothesis H2.7 tested the relationship between Psychological Contract -Transitional (PCTi) and Work Behaviour -Altru ism (W BA) with path coefficient of 0.05 (t=0.90, p=0.37). Since the questions for PCTi were asked in a reverse direct ion and the items were reverse transformed during the data preparation, the positive path coefficient is representing a negative correlation. Nevertheless, this result does not support our hypothesis of a negative relationship between Psychological Contract -Transitional (PCTi) and Work Behaviour -Altru ism (W BA) as the p>0.05, which is statistically not significant.
The sub-hypothesis H2.8 tested the relationship between Psychological Contract -Transitional (PCTi) and Work Behaviour -Generalised Co mpliance (W BGC) with a path coefficient of 0.22 (t =2.97, p=0.00). Since the ques tions for PCTi were asked in a reverse direction and the items were reverse transformed during the data preparation, the positive path coefficient is representing a negative correlation. This result supports our hypothesis of a negative relationship between Psychological Contract -Transitional (PCTi) and Work Behaviour Generalised Compliance (WBGC).
Sub-hypotheses H3.1 tested the relationship between the Work Behaviours Altruism (WBA) and Turnover Intention (TI). The result indicates that the path coefficient was 0.18 (t=2.39, p=0.02). The hypothesis was rejected as the path coefficient was a positive value even it is statistically significant (p<0.05).
Sub-hypothesis H3.2 tested the relationship between the Work Behaviours Generalised Co mp liance (WBGC) and Turnover Intention (TI). Th is result rejects the hypothesis of a negative relat ionship between these two variab les as the path coefficient was -0.06 (t=1.03, p=0.3), which is statistically not significant.

Discussions
The psychological contracts (in terms of balanced and transactional) positively influence the emp loyee's work behaviours in terms of altruis m. It also positively influence the employee's work behaviours in terms of generalised comp liance. There is no significant relationship between work behaviours and turnover intention found; and the relationships between psychological contacts and turnover intention are mainly a direct effect, only the employee's work behaviour in terms of alt ruis m has a partial co mpetit ive med iation effect between balanced psychological and turnover intention. Psychological contract, co mpared with wo rk behaviour, is a better predictor of turnover intention. Three out of four types of psychological contracts having a significant relationship with turnover intention.
A negative correlation was found between the balanced psychological contact and the turnover intention, and a negative correlation was found between the relational psychological contract and the turnover intention [41,42,43,44,45,46,47] and that the binding of the transitional and transactional psychological contract can have a positive correlation with the turnover intention [43,48,49]. Such findings were also supported by the research undertaken by Larwood,Wright,Desrochers and Dahir [50], Freese and Schalk [51], Tekleab and Tay lor [52] and Meyer & Allen [53].
This research echoed the findings of the prev ious research, such as Kotter [54], Robinson [55], Shore and Barksdale [56], Turnley and Feld man [57], Lester and Kickul [58], Sutton and Griffin [59], and Lemire and Rouillard [60], related to the relationship between the psychological contract and turnover intention. Turn ley and Feld man [57] demonstrated from their research that any negative factors perceived by the employees with regards to non-fulfilment of the psychological contract, lead the individual to have a greater intention to quit in order to escape the perceived work environ ment.
[50] also confirmed fro m their research that non-fulfilment of the psychological contract was associated with a greater expressed intention by employees to quit their current jobs. This finding is also supported by others scholars, e.g. Freese and Schalk [51] and Tekleab and Taylor [52]. Rousseau [1] classified the relat ional psychological contract as a long-term and open-ended employ ment relationship, the transactional psychological contract as a short-term or limited time emp loy ment relat ionship, and the transitional psychological contract, which influences the previously established employ ment arrangement, as a cognitive state reflecting the consequence of organisational change. These three types of psychological contract have significant influence on the employee's turnover intention. This implies that the organisation has to carefully manage the informal relationship between the emp loyee by improving the communication to remove mistrust and uncertainty, it will definitely help with respect to assisting staff retention.
In this research, work behaviour was measured by Organisational Citizenship Behaviour [18] in wh ich there are two constructs, namely Work Behaviour -Altru ism (WBA) and Work Behaviour -Generalised Co mpliance (WBGC). For the psychological contract, the same constructs were used in H1 as proposed by Rousseau [1]. There are a total of eight sub-hypotheses, and these were divided into two groups. The first group (H2.1, H2.3, H2.5 & H2.7) tests the relationship between the four types of psychological contract with work behaviour in terms of altruis m wh ile the second group (H2.2, H2.4, H2.6 & H2.8) tests the relationship between the four types of psychological contract with work behaviour in terms of generalised compliance.
For the relationship with Work Behaviour -Altru ism (WBA), it is found that H2.1 (balanced psychological contract) and H2.5 (transactional psychological contract) were supported, while H2.3 (relational psychological contract) and H2.7 (transitional psychological contract) were rejected. For H2.1, the path coefficient between Psychological Contract -Balanced (PCB) and Work Behaviour -Altru ism (WBA ) was 0.55 (t=8.99, p =0.00) and for H2.5 the path coefficient between the Psychological Contract -Transactional (PCTa) and Work Behaviour -Altru ism (WBA) was 0.14 (t =2.02, p=0.04). This imp lies a h igher obligation fro m emp loyee and emp loyer to improve training, internal career advancement and external emp loyability opportunities would lead to the employee being more willing to offer help in the organisation to other colleagues, e.g., providing orientation to a new colleague and helping a colleague with a heavy workload [18]. The short-term relationship with limited o r fixed scope of duties or a specific or limited time would lead the emp loyee to be less willing to offer help in the organisation. The results of H2.3 and H2.7 indicated that there was no significant relationship existing and the path coefficient was also comparatively low. However, it is reco mmended that the organisation should also take notice of the potential impacts on work behaviours fro m these two types of informal relationship with the employee. This is because such findings may be specific to the telecommun ication industry in Hong Kong, so the interpretation of the results should only be applied in this context unless we can demonstrate generalised behaviours in future research. Based on the definition, the relational psychological contract represents a long-term and open-ended relationship[1], which is often in the form of job security. Indeed, Staufenbiel and Konig [61] argued that job insecurity will have a negative impact on work performance, so the employer must make an effort to build a good relationship with the employee by removing mistrust and uncertainty to create a shared understanding working environment [43].
For the second group of hypotheses (H2.2, H2.4, H2.6 & H2.8) relating to the relationship between the four types of psychological contract and work behaviour in terms of generalised compliance, the results indicated that H2.2 and H2.8 were supported and H2. 4 and H2.6 were rejected. For H2.2, the path coefficient between the Psycholo gical Contract -Balanced (PCB) and Work Behaviour -Generalised Co mpliance (WBGC) was 0.28 (t=2.62, p=0.01), and for H2.8, the path coefficient between the Psychological Contract -Transitional (PCTi) and Work Behaviour -Generalised Co mpliance (W BGC) was 0.22 (t=2.97, p=0.00). The path coefficients were not particularly high, but there was a significant relat ionship. The definit ion of work behaviour in terms of generalised compliance refers to "a more impersonal form of conscientiousness that does not provide immediate aid to any one specific person, but rather is indirectly helpful to others involved in the system" (17, p657). It focuses more on individual-level behaviours, e.g. punctuality, not wasting time and self-defined rules to maintain order as stated by Williams and Anderson (17,p601). The results of the hypotheses test imply that the higher obligation by both employee and employer to work on train ing and career develop ment (balanced psychological contract) together with efforts to remove mistrust and uncertainty (transitional psychological contract) will lead to the emp loyee having better work behaviours, which can also help the organisation to build a good corporate culture. It is noticed that H2. 4 and H2.6 were not supported as the paths are not statistically significant. The path coefficient between the Psychological Contract -Relational (PCR) and Work Behaviour -Generalised Co mpliance (WBGC) was 0.11 (t =1.61, p=0.11) and the path coefficient between the Psychological Contract -Transactional (PCTa) and Work Behaviour -Generalised Co mpliance (WBGC) was 0.14 (t=1.46, p=0.14). However, it is reco mmended that the organisation not underestimate the potential impacts of such relations even though no significant relations were supported by the collected data. As is discussed in previous paragraphs, the transactional psychological contract does influence the OCB in term of altruis m. Such easy-to-exist agreement does influence the emp loyee when it co mes to offering help to other emp loyees. The four types of psychological contact and the two aspects of OCB are interdependent and form a complex web o f exchanges between employees and emp loyer [62, p 124]. The psychological contract is a dynamic relationship and, given that mu ltiple psychological contracts exist, the organisation should manage them as a meta-contract across the firm [62].
This research echoed the findings of previous research [63,64,65,66] on the relationship between the psychological contract and work behaviour given that significant relationships were found. These studies found that negative emotions are a likely consequence of an undesirable psychological contract. Undesirable psychological contracts also impact on workplace effectiveness [66, p669], while emp loyees' perceptions of politics and procedural justice influence their psychological contracts, work attitudes and contextual performance in terms of Organisational Cit izenship Behaviour [65, p213]. Emp loyees who are h ighly committed show less destructive behaviour and attempt more constructive behaviour [63, p596]. Harrison, Neu man and Roth found that work-related attitudes are related to the employee's contributions to his or her work role.
There were t wo sub-hypotheses proposed to test the relationship between work behaviour in terms of altru ism (H3.1) and work behaviour in terms of generalised compliance (H3.2) with turnover intention. The results fro m the data analysis rejected both hypotheses. For H3.1, the path coefficient between the Work Behaviours -Altruism (W BA) and the Turnover Intention (TI) were 0.22 (t=2.6, p=0.01), even though statistically significant, is positive, which imp lied a positive correlation instead of a negative correlation. For H3.2, the path coefficient between the Work Behaviours -Altruis m (BWA) and Turnover Intention (TI) were -0.09 (t=1.47, p=0.14) was not statistically significant, therefore, the hypothesis was rejected.
This research did not echo the findings in previous research by Chen and Wei [67,p8] and Khalid et al. [68,p112] which found that a significant relationship between work behaviour and turnover intention existed and that work behaviour and turnover intention were negatively correlated. Th is finding seems to contradict our understanding of previous literature; therefore, an additional analysis was conducted to evaluate the relationship between work behaviours with turnover intention alone. The result indicated that both Work Behaviour -Altruis m (WBA) and Work Behaviours -Generalised Co mp liance (W BGC) have a significant negative correlation with Turnover Intention (TI). These results aligned with our understanding of the previous literature, despite the fact it was noticed that the R 2 is just 0.15, wh ich means that only 15% of turnover intention can be explained by work behaviour. The p revious research undertaken by Castro, Armario and Ru ize [69], Chen and Wei [67], Mossholder, Settoon & Henagan [70], Khalid et al. [68], and Chen,Hui and Sego [71] demonstrated a significant relat ionship with a low to moderate effect. Therefore, the emp loyer must pay special attention to managing the emp loyees' informal work behaviour. The individual will increase their tendency to withdraw fro m his/her job when h is/her OCB at work decreases. OCB was high when emp loyees have no turnover intention.

Conclusions
The study found that the psychological contract between emp loyers and employees has a significant impact on work behaviour and turnover intention. It was also found that the psychological contract is a better means of exp laining turnover intention than work behaviour. The findings of this research provide a better understanding of the psychological contract between emp loyees and employers in the teleco mmun ications industry in Hong Kong. Business managers can leverage this finding to develop better plans and approaches to deal with work behaviour and turnover intention issues. The outcomes also provide empirical evidence to validate the psychological contract inventory developed by Rousseau [1] for the teleco mmunications industry in Hong Kong. This research fills a knowledge gap in this area and also provides a solid foundation for future research.

Implications
To better manage the psychological contract, employers should understand individual differences and perceptions about employee commit ment versus employer commit ment in order to align the indiv idual goals with the organisation goals. Hu man resources policy and practice should be able to address individual needs, to support their needs in order fo r them to perform well in their current roles, provide internal pro motion opportunities, provide necessary training opportunities to enable them to take up new roles in the organisation, and provide necessary training for employees to acquire the necessary skills and knowledge, which are externally recognised, e.g. professional bodies and other employers. Emp loyers should improve commun ication with employees such that they can have a better understanding of the strategic direction, goals and objectives of the company. This has the potential to improve mistrust and uncertainty. They should make use of the psychological contract inventory (e.g. the instrument used in this research, as developed by Rousseau [1] as a management tool for psychological contract measurement. Finally, they should establish psychological contract management, which includes four steps: Measure, Identify, Act and Check. The Measure relates to surveying the perception of the psychological contract between employee and employer. Identify relates to finding out the areas that can improve the psychological contract. Act relates to implement ing an action plan to improve the psychological contract, wh ile Check relates to reviewing if the imp lemented action can achieve the desired results or not. This is a continual process to help the organisation to find out the strengths and gaps in the psychological contract between the employee and emp loyer, to find out the core areas of imp rovement, and to keep track of the actions taken. The output of the psychological contract management process should fit into the overall Human Resources policy and planning.