Introduction

Presently, digital technology is advancing at an incredible rate. The term "Digital Transformation" has grown increasingly associated with information technology and the practical environment over the last few years (Berman, 2012). An essential component of digital transformation is to enhance the organization's performance continually, as the information technology sector is highly competitive. If an organization does not take a forward step and strengthen itself, it risks collapsing. A business must adapt and develop continually due to changing consumer behavior and technological advancements. Everyone may easily and quickly have access to knowledge and data from all around the world (Agarwal et al., 2010).

An essential component of digital transformation is to enhance the organization's performance continually, as the information technology sector is highly competitive. If an organization does not take a forward step and strengthen itself, it risks collapsing. A business must adapt and develop continually due to changing consumer behavior and technological advancements. Everyone may easily and quickly have access to knowledge and data from all around the world. Even though there are several administrative system in the real practical such as adhocracy, bureaucracy and flat-organization management, but for some organizations, such as government sector agencies or multinational corporations which are unlikely to be able to adopt themselves due to their size and administrative system. These organizations are typically managed by a bureaucratic administrative system that appears to be out of date and is characterized as a sluggish method of management.

The bureaucratic theory of organization and management was proposed by Max Webber (Donaldson, 2001). Bureaucracy is an administrative system that efficiently and effectively manages the responsibilities of several institutions (Rondinelli, 1980). Max Weber classified authority of power into three categories: traditional, charismatic, and rational-legal or bureaucratic and also said that the best form of authority is bureaucratic. Weber's Bureaucracy contains a range of bureaucratic characteristics (Evans & Rauch, 1999). (Figure 1)

In the past, talent management was one of the human resource management processes that were hardly ever prioritized because people were more concerned with other influences; as a result, businesses were constantly recruiting and losing skilled and experienced personnel; as a result, businesses needed to find innovative ways to engage and retain talented employees (Cappeli, 2008). Efficient talent management is about identifying, recruiting, developing, rewarding, compensating, engaging, and retaining high-value or high-performance workers who have a promising future or who are essential to the organization's success. It demonstrates the sensitive significance of not just attracting people, but also managing, developing, and retaining employees with the appropriate capabilities when evaluating the return on the organization's investment. Effective people management may help organizations achieve various strategic goals and objectives (Stahl et al., 2016). So the researchers decide to study about the mediation effect of talent management process on the effect of talent value proposition toward high-performance working system for bureaucratic organization in the digital transformation era to obtain the answer for the research question of the dissertation.

Literature Review

As Gubman (1998) described, the Talent Management Process (TMP) is a continual process of attracting high-quality employees, developing their abilities, and continuously encouraging them to improve their performance. The word "talent management" is complex to describe because of its ambiguous associations with descriptions and the numerous hypotheses advanced by researchers using the interchangeable terms "talent management," "talent strategy," "succession management," and "human resource planning." Schweyer (2004) defined personnel management processes as "...controlling the supply, demand, and movement of talent through the human capital engine." The talent management process may be split into four categories: identification and recruiting, training and development, reward and recognition, and retention, although not all of these components get equal weight (Eghbal et al., 2017).

For the previous study in the talent management process, there are several researchers which mention the elements of the talent management process in their research. The talent identification and recruitment was mentioned by Pandita & Ray (2018); Pasha & Ahmed (2017) while the training and development was mentioned by Dhanabhakyam & Kokilambal (2014); Yeswa & Ombui (2019); Mercer (2005). The talent reward and compensation was mentioned by Johennesse & Chou (2017) while the talent retention was studied by Rabbi et al., (2015); Johennesse & Chou (2017).

For the talent value proposition, Minchington (2006) describes the Employee Value Proposition (EVP) as a collection of affiliations and offerings made by a business in exchange for an employee's talents, capabilities, and experiences. Therefore, employee value propositions are sometimes referred to as talent value propositions. Additionally, DiVanna (2003) defined TVP as a collection of talents and experiences classified as assets and included on a company's balance sheet. Because it was informed by existing workers and the external target audience, the TVP is an employee-centered approach that is consistent with existing, integrated workforce planning methods. If a TVP is a significant driver of talent acquisition, engagement, and retention, it must be original, relevant, and engaging. The elements of talent value proposition which are employee branding, talent branding and organization branding were mentioned by Mishra & Subudhi (2017); Noutel, et al., (2020).

A High-Performance Working System (HPWS) is a physical or virtual workplace designed to maximize employee performance and achieve company objectives and add value. A high-performance workplace is achieved by constantly balancing investment in people, processes, the physical environment, and technology to improve employees' capacity to learn quantitatively, explore, innovate, collaborate, lead, and achieve efficiency and financial gain. The components of the high performance working system are maximized employee skill and commitment and flexibility which were studied by Shahriari & Allameh (2020); Gahan, et al., (2021). According to Mekhum & Jermsittiprasert (2019), the influence of supervisor support, coworker support, and meaningful work on individual work well-being has a positive effect on working flexibility and retention, and HPWS utilization is positively associated with the organization's sales growth and innovation (Messersmith & Guthrie, 2010) along with the relationship between HPWPs and service behavior (Oentoro, 2018).

According to idea obtained from the study of previous research, the research framework is shown in Figure 2. The talent management process includes four variables: talent identification and recruiting talent development and training, talent compensation and reward, and talent retention. Three variables are noticed in the talent value proposition: Organizational branding, employee branding, and talent branding. Two recognized factors characterize high-performing performance systems: maximized staff skills and flexibility. The talent value proposition has an indirect effect on high-performance working systems via talent management processes and a direct effect on high-performance working systems.

Research Methodology

The methodology section describes the research strategy, demographic, sample size, data collecting procedure, and statistical tool used to analyze the data (Iqbal & Hameed, 2020). The current study employed a quantitative research technique, also known as the deductive strategy, rather than a qualitative one. This study aims to analyze the mediation effect of the talent management process on the effect of talent value proposition toward a high-performance working system.

The population for this study included 1,725 information technology professionals who work in a bureaucratic government organization (Office of the Civil Service Commission, 2019). The study concentrated on information technology employees employed by government agencies since their job descriptions are comparable to the population. In addition, they are collaborating with a bureaucratic agency to represent accurate data. The rules of thumb indicate that a minimum required sample size of 200 ensures sufficient statistical power for data analysis (Hoe, 2008). The same N is proposed by several researchers as an adequate sample size for measuring up to 40 separate objects. A sample of 300 populations had been proposed for the data analysis (Tabachnick & Fidell, 2013). Comrey & Lee (1992); Comrey (1973) considered a factor analysis sample of 50 to be “very poor”, 100 to be “poor”, 200 to be “fair”, 300 to be “good”, 500 to be “very good” and a sample of 1000 to be “excellent”. Also along with Hair, et al., (2010) who stated that the minimum sample size for the structural equation modeling that have less than 7 latent variables are 300 samples. So this research’s sample size to be 360 information technology personnel, as the researcher add the extra 20 percent to the expected sample size for the buffer of error.

study’s research instrument was based on the quantitative method explanatory design. As a result, the questionnaire was used. After the questionnaire passed the Index of Item-Objective Congruence (IOC) pass criteria of 0.50 up for all questions (Tirakanan, 2007), it was tried out with 30 non-sample managers to check reliability by considering internal consistency based on Cronbach’s alpha coefficient. The data analysis of all background information of the sample was analyzed by frequency and percentage. Since all variables in the conceptual research framework were continuous variables, the study used Mean, S.D., Skewness, and Kurtosis to study variables' distribution characteristics. The hypotheses were tested by using Structural Equation Modeling (SEM). The interpretation of Mean in measuring TMP, TVP, and HPWS was considered from 5 levels of estimation based on Best & Kahn’s approach (2006). The appropriateness of the meta-correlation was investigated by considering the Kaiser-Meyer-Olkin (KMO) value, >0.5 (Hair et al., 2006), and the Bartlett Test of Sphericity (Hadi et al., 2016). They must have significant statistical significance (Sig.) 0.000, indicating that this set of variables is suitable for Confirmatory Factor Analysis (CFA). Multicollinearity by correlation coefficient (r) was used to find the linear relationship between the variables. The correlation value can be from negligible (±00.00-0.30) to Very high (±0.90-1.00). The correlation coefficients between variables in SEM not to exceed +0.80 were considered (Stevens, 2009). CFA was used to test the relationship between observed variables of 13 observed variables and four latent variables. The model fit measurement was based on the eight indices (chi-square: P> 0.05, relative chi-square<2, GFI, AGFI, TLI, and CFI> 0.95, RMR, and RMSEA<0.05) to test the consistency of the model based on hypothesis and empirical data. The researcher used these indices to validate the conformance of the model. If the calculated values do not meet the criteria, the model must be adjusted (Diamantopoulos & Siguaw, 2002).

Research Results

The testing for the measurement model for this study considered the normal distribution of values from the Skewness values of -3 to +3 and Kurtosis values of 3. It also found that all factor loadings of the first-order CFA were between 0.3-1.0, and Cronbach’s Alfa coefficients were greater than 0.7, as shown in Table 1. These values were acceptable, as proposed by Rangsoongnern (2011). Besides, the model fit measurement was passed in eight indices, as suggested by Diamantopoulos & Siguaw (2002). These indicated that the measurement models were acceptable. Moreover, the measurement models of TMP, TVP and, HPWS were considered in terms of reliability, convergent validity, and discriminant validity with the criteria of CR>0.70; Convergent validity: AVE>0.50; Discriminant validity: AVE>MSV. CR=composite reliability; AVE=average variance extracted; MSV=maximum shared variance; ASV=average shared variance (Hair et al., 2010) as shown in Table 2. After that, the Goodness of fit by the second-order CFA depicted passing values based on indices suggested by Diamantopoulos & Siguaw (2002), as shown in Table 3.

Table 1 Testing Results of the Measurement Model
	Items’ no.		SD	Interpret	1st order loading	α	Remarks
Talent Value Proposition
Organization Branding	6	4.23	0.59	High	(0.65-0.79)	0.90	Acceptable
Employee Branding	6	4.01	0.63	High	(0.65-0.86)	0.94	Acceptable
Talent Branding	6	4.15	0.59	High	(0.60-0.70)	0.96	Acceptable
Talent Management Process
Talent Identification and Recruitment	4	4.31	0.63	High	(0.67-0.80)	0.90	Acceptable
Talent Training and Development	5	4.24	0.64	High	(0.67-0.75)	0.89	Acceptable
Talent Compensation and Rewarding	4	4.45	0.54	High	(0.74-0.81)	0.94	Acceptable
Talent Retention	6	4.33	0.59	High	(0.70-0.77)	0.96	Acceptable
High-Performance Working System
Maximized Employee Skill	3	4.13	0.70	High	(0.76-0.84)	0.87	Acceptable
Commitment and Flexibility	3	4.22	0.64	High	(0.75-0.84)	0.89	Acceptable

Table 2 Reliability, Convergent and Discriminant Validity
	CR	AVE	MSV	ASV
Talent Value Proposition (TVP)	0.837	0.520	0.491	0.476
Talent Management Process (TMP)	0.801	0.507	0.438	0.426
High Working Performance System (HPWS)	0.843	0.525	0.476	0.464

Table 3 The Goodness of Fit By Second Order CFA
Index	P-value	𝑥2/df	CFI	GFI	AGFI	TLI	RMR	RMSEA	Remarks
	>0.05	<2	≥0.90	≥0.90	≥0.90	≥.90	<0.05	<0.05
TVP	.70	1.64	0.99	0.95	0.92	0.98	0.17	0.042	pass
TMP	0.55	1.25	0.99	0.96	0.93	0.99	0.16	0.026	pass
HPWS	0.09	1.81	1.00	0.99	0.97	0.99	0.10	0.047	pass

The testing result of the causal relationship model is shown in Table 4. The effects of TVP on TMP depicted R-square of 0.763. The effect of TMP on HPWS displayed R-square of 0.723. As shown in Table 5, hypotheses 1-3 were supported with the statistical significance at a P-value level <0.001 – P-value<0.05. Such table showed C.R. (Critical Ratio) or T-Value of testing results of all hypotheses. Testing results of hypothesis 1 described the T-value of 19.470. Testing results of hypothesis 2 described the T-value of 4.600. Testing results of hypothesis 3 described the T-value of 3.678. These T-values were meet the acceptable criterion of C.R. or T-value>1.96, as suggested (Henseler et al., 2009).

Table 4 Total Effects, Direct Effects, and Indirect Effects
DV	TMP				HPWS
IV	TE	DE	IE	S.E.	TE	DE	IE	S.E.
Talent Value Proposition (TVP)	0.873***	0.873***	-	0.046	0.813***	0.362***	0.451***	0.141
Talent Management Process (TMP)	-	-	-	-	0.516***	0.516***	-	0.155
R2		0.763				0.723

Table 5 Hypothesis Testing Results
Hypothesis	Path	(β)	T-Value	Results
H1	Talent Value Proposition → Talent Management Process	0.873***	19.470	Supported
H2	Talent Management Process → High-Performance Working System	0.516***	4.600	Supported
H3	Talent Value Proposition → High-Performance Working System	0.362***	3.678	Supported

Path analysis by structural equation model was used to test the three hypotheses comprising the proposed model of TMP and TVP's effects on HPWS, the effect of TVP on TMP, and the effect of TMP on HPWS. The model fit analysis results were acceptable (Chi-square= 18.505; degree of freedom=18; P=0.423; relative chi-square=1.028; GFI=0.989; AGFI=0.974; TLI=1.000; CFI =1.000; RMR=0.005; RMSEA=0.009). The results are presented in Figure 3

The second model is the structural model assessment used to analyze the hypothesis made in the literature part of this study. Results revealed that elderly personality traits like (positive mindset, good self-esteem, and social relationships have a significant impact on employees’ decision to choose employment of entrepreneurship by examining the t values. T values are greater than 1.96, which means the hypothesis is accepted. The intensity of relationships that whether the relationship is positive or negative analyzed by the value of beta. The positive value showed positive relations. All values of this structural model assessment depicted in Figure 3 and Table 3.

Discussion and Conclusion

The findings indicate that Thailand's bureaucratic structure talents have a high degree of TVP, TMP, and HPWS. This might be because each talent today demands a high-performance work environment that will result in a superior experience for them. The findings on TVP's direct positive effects on HPWS and the direct positive effect of TMP on HPWS and TVP on TMP are consistent with previous studies as stated in the literature review. When the direct impact size of TVP on HPWS is evaluated, it demonstrates the structural equation model's smallest effect sizes. This may be due to a well-managed and balanced talent management strategy. The talent value proposition of the bureaucratic organization will have a marginally favorable influence on employee commitment and flexibility, which is lesser than the talent management process's effect on the high-performance working system.

Also with the mediation effect of the talent management process on the effect of talent value proposition toward high-performance working system showed that talent management process has a mediation effect between talent value proposition and high performance working systems which can be implied that the executive of the organization should execute both talent management process and talent value proposition to increase the high-performance working system for the organization in the digital transformation era.

For the organization management aspect, the bureaucratic organization will be benefit from applying the talent management process and talent value proposition in the organization as it will create the high performance working system or high performance working atmosphere which is needed by the information technology personnel who can move the organization forward through the digital transformation era. The working trend is a dynamic feature that can vary over time in Thailand and on a global scale. As the trend of working is a dynamic element that can vary over time, further research should focus on the organization's other variables and the trend of working in the current scenario. The research should involve crisis management in the case of an unforeseen occurrence.

Acknowledgement

Pimploi Tirastittam is the Ph.D. Candidate in Information Technology Management Division, Faculty of Engineering, Mahidol University. Dr.Sotarat Thammaboosadee, who is the corresponding author, is an Assistant Professor in Information Technology Management Division, Faculty of Engineering, Mahidol University. The corresponding e-mail is sotarat.tha@mahidol.ac.th. Dr.Rojjalak Chuckpaiwong is also a lecturer in Information Technology Management Division, Faculty of Engineering, Mahidol University.

References

1. Agarwal, R., Gao, G., Desroches, C.M., & Jha, A.K. (2010). The Digital Transformation of Healthcare: Current Status and the Road Ahead, Information Systems Research, 21(4), 796-809.
2. Berman, S.J. (2012). Digital Transformation: Opportunities to Create New Business Models, Strategy & Leadership, 40(2), 16-24.
3. Best J.W., & Kahn, J.V., (2006). Research in Education, (10th Edition). UK, Pearson Education.
4. Cappeli, P. (2008). Talent management for the twenty-first century. Harvard Business Review, 86(3), 74-81.
5. Comrey, A. (1973). A first course in factor analysis. New York, NY: Academic Press.
6. Comrey, A., & Lee, H. (1992). A first course in factor analysis. Hillsdale, NJ, Erlbaum.
7. Dhanabhakyam, M., & Kokilambal, K. (2014). A study on existing talent management practice and its benefits across industries. International Journal of Research in Business Management, 2(7), 23-36.
8. Diamantopoulos, A., & Siguaw, J.A. (2002). Introducing LISREL: A guide for the uninitiated. Journal of the Royal Statistical Society Series D (The Statistician), 51(3), 412-413.
9. DiVanna, J.A. (2003). The relativity of value. In Synconomy, London, Palgrave Macmillan.
10. Donaldson, L. (2001). The contingency theory of organizations. In Organization Design. Information and Organization Design Series, Boston, Springer.
11. Eghbal, F., Hoveida, R., Sisdat, S.S., Samavatyan, H., & Yarmohammadian, M.H. (2017). The effect of talent management process on the research performance of faculty members with the mediating role of organizational justice. Foresight and STI Governance, 11(2), 83-91.
12. Evans J.E., & Rauch P.B. (1999) Bureaucracy and growth: A cross-national analysis of the effects of 'Weberian' state structures on, American Sociological Review, 64, 748-65.
13. Gahan, P., Theilacker, M., Adamovic, M., Choi, D.J., Harley, B., Healy, J., & Olsen, J.E. (2021). Between fit and flexibility? The benefits of high‐performance work practices and leadership capability for innovation outcomes. Human Resource Management Journal.
14. Gubman E.L., (1998). The talent solution- aligning strategy and people to achieve extraordinary results. New York, McGraw-Hill.
15. Hadi, N.U., Abdullah, N., & Sentosa, I. (2016). An easy approach to exploratory factor analysis: Marketing perspective. Journal of Educational and Social Research, 6(1), 215-223.
16. Hair.Jr. J.F., Black W.C., Babin B.J., Anderson R.E., & Tatham R.L. (2006). Multivariant data analysis. Upper Saddle River, NJ Pearson Education.
17. Henseler, J., Ringle, C.M., & Sinkovics, R.R. (2009) The use of partial least squares path modeling in international marketing new challenges to international marketing. Advances in International Marketing, 20, 277-319.
18. Hoe, S.L. (2008). Issues and procedures in adopting structural equation modeling technique. Journal of Applied Quantitative Methods, 3, 76-83.
19. Iqbal, J., & Hameed, W. (2020). Open innovation challenges and coopetition-based open-innovation empirical evidence from Malaysia. In Pablos, P., Zhang, X., & Chui, K. (editions.). Innovative Management and Business Practices in Asia (144-166). Pennsylvania: IGI Global.
20. Johennesse, L.A.C., & Chou, T.K. (2017). Employee perceptions of talent management effectiveness on retention. Global Business & Management Research, 9(3), 46-58.
21. Mekhum, W., & Jermsittiprasert K. (2019). Effect of supervisor support, co-worker support and meaningful work on work engagement of pharmaceutical firms' employees in Thailand, Systematic Reviews in Pharmacy, 10(2), 176-187.
22. Mercer, H. (2005). Talent and analytics: New approaches, higher ROI. Journal of Business Strategy, 32(6), 4-13.
23. Messersmith, J.G., & Guthrie, J.P., (2010). High performance work systems in emergent organizations: Implications for Firm Performance. Human Resource Management, 49(2), 241-264.
24. Minchington, B. (2006). Your employer brand. Australia, Hyde Park Press.
25. Mishra, Y., & Subudhi, R. (2018). Role of employer branding and organization attractiveness influencing citizenship behaviour: A review. Amity Journal of Management, 5(2), 24-40.
26. Noutel, R., Araujo, P., Fernandes, R., & Arriscado, P. (2020). An employer branding strategy: Developing a talent attraction model through action research. Economic and Social Development: Book of Proceedings, 282-291.
27. Oentoro, W., (2018). High performance work practices and service performance: The influence of employee engagement in call center context. PSAKU International Journal of Interdisciplinary Research, 7(1).
28. Pandita, D., & Ray, S. (2018). Talent management and employee engagement – a meta-analysis of their impact on talent retention. Industrial and Commercial Training, 50(4). 185-199.
29. Pasha, S.M., & Ahmed, B. (2017). A study on talent management strategies in it industry. International Research Journal of Engineering and Technology, 4(4), 1384-1387.
30. Rabbi, F., Ahad, N., Kousar, T., & Ali, T. (2015). Talent management as a source of competitive advantage. Asian Economic and Social Society, 5(9), 208-214.
31. Rangsoongnern K. (2011). Factor Analysis with SPSS and AMOS. Bangkok, Se-Education Publishing.
32. Rondinelli, D.A. (1982).The dilemma of development administration: Uncertainty and complexity in control-oriented bureaucracies, World Politics, 35(2), 43–72.
33. Schweyer, A. (2004). Talent management systems: Best practices in technology solutions for recruitment, retention and workforce planning. Hoboken, NJ, Wiley,
34. Shahriari, M., & Allameh, S.H. (2020). Organizational culture and organizational learning: Does high performance work systems mediate? Journal of Workplace Learning, 32(8), 583-597.
35. Stevens J., (2009). Applied Multivariate Statistics for The Social Sciences. New York, Routledge/Taylor & Francis Group.
36. Tabachnick, B.G., & Fidell, L.S. (2013). Using multivariate statistics. Boston, MA: Pearson.
37. Tirakanan S. (2007). Research methodology in social sciences: Practical guideline. Bangkok, Chulalongkorn University Printing.
38. Yeswa, S.L., & Ombui, K. (2019). Influence of talent management strategies on employee retention in hotel industry in Kenya. International Academic Journal of Human Resource and Business Administration, 3(7), 83-103.