Spence (1973), in his theory of Job-Market Signaling, assumes that hiring new employees is like investing in uncertainty. The attributes that applicants have, such as education and training, serve as signals for employers to assess an applicant's productive abilities (Maier, 2022). These credentials demonstrate relevant qualifications, thus helping to increase the employability of applicants in the labour market who do not have complete information about them (Brown et al., 2024). In this context, the employability of accounting students can be seen as their ability to use signals such as Big Data Analytics skills to demonstrate their job readiness.

Employability is a set of skills and achievements that make a person more likely to obtain a job and thrive in his or her career (Yorke, 2004). According to Hillage and Pollard (1998), employability includes the knowledge, skills, and attitudes that a person possesses and how they present those assets to employers (Wang et al., 2022). (Xu et al., 2022) divide employability into three main dimensions: human capital (knowledge and skills), individual attributes (adaptability, flexibility), and career development (career management and self-reflection).

Human capital in the context of accounting students includes the technical knowledge and skills needed for professional work (Jackson et al., 2022). With technological advancements, accounting students are also expected to have skills in Big Data Analytics to solve problems and make data-driven decisions (Suarta et al., 2023). Big Data Analytics allows students to gain a deeper understanding of complex financial data analysis, as well as improve their technical competence (Mahmudi, 2024).

(Xu et al., 2022) showed that hard skills and soft skills in Big Data Analytics contribute significantly to increasing student human capital. In addition, (Jackson et al., 2022) emphasized the importance of technological skills and critical thinking in improving the employability of accounting students. Proposed hypothesis:

H1: Hard skills Big Data Analytics have a positive effect on increasing the human capital of accounting students.

H2: Soft skills Big Data Analytics have a positive effect on increasing the human capital of accounting students.

Development

The career development of accounting students depends not only on technical skills but also on the ability to manage their careers independently (Zhu, 2023). Students who master Big Data Analytics can leverage data insights to make better career decisions, as well as develop sustainable leadership and career management skills (Jackson et al., 2022). Data-driven decision-making also allows students to be more competitive in the ever-changing job market (Ceschi et al., 2017).

(Xu et al., 2022) showed that Big Data Analytics has a positive influence on career development, especially in decision-making and effective communication. Research by (Jackson et al., 2022) adds that understanding future technology can improve accounting students' career opportunities. Proposed Hypothesis:

H5: Big Data Analytics hard skills have a positive effect on improving the career development of accounting students.

H6: Big Data Analytics soft skills have a positive effect on improving the career development of accounting students.

The population in this study is students of the Accounting Study Program who have an understanding of Big Data Analytics. Since the exact number of the population is unknown, the sampling technique uses the Lemeshow formula with a confidence level of 95% and a margin of error of 5% ((Adam, 2020). Based on these calculations, the minimum sample needed in this study is 286 respondents.

Research Instruments

The instrument is in the form of a questionnaire, consisting of several parts designed to measure independent variables, namely hard skills and soft skills in Big Data Analytics, as well as dependent variables, which include three dimensions of employability: human capital, individual attributes, and career development. Each item in the questionnaire is measured using a 4-point Likert scale: Strongly Agree (SS), Agree (S), Disagree (TS), and Strongly Disagree (STS). Neutral options were not included to avoid response bias from respondents who might choose a middle answer (Raaijmakers, 2000).

The respondents in this study are students of the Accounting Study Program who already have an understanding of Big Data Analytics. The selection of respondents was carried out using the purposive sampling technique, where respondents were selected based on the criteria that they had been exposed to the concept of Big Data Analytics through relevant courses or training.

Data collection was carried out through a survey distributed online using Google Forms. This method was chosen because of its efficiency in reaching respondents in a relatively short time and its ability to allow extensive data collection without geographical constraints. The survey was disseminated through student social networks and relations to ensure that the questionnaire was accepted by respondents under the research criteria.

The data collected was analyzed using descriptive and inferential statistical methods. Descriptive analysis was used to describe the respondents' profiles and research variables. This study uses the Structural Equation Modeling (SEM) method with SMART-PLS software (Borkowski, 2022) to test the hypothesis. The SEM approach was chosen because of its ability to evaluate the relationships between variables simultaneously, including the measurement of latent variables and causal relationships between hypothesized variables (Limberg et al., 2021). The model used in this study is based on a strong theory and validation from previous research.

The results of the descriptive analysis show that the mean value of each variable is in the category of "Agree." It shows that respondents feel that they have good hard skills and soft skills and consider that the quality of their human capital, individual attributes, and career development is also good.

The initial measurement model is shown through the following path diagram generated from SmartPLS before eliminating some invalid indicators.

After eliminating several indicators with loading factor values below 0.60, the measurement model is improved, as seen in Figure 2 below.

The eliminated indicators are HS2, HS4, and IA2, so all remaining indicators have a loading factor value above 0.60. Thus, all indicators can be declared valid for use in further analysis.

The results of the convergence validity test through outer loading analysis show that some indicators with values below 0.60 have been eliminated. After elimination, all remaining indicators showed an outer loading value above 0.60, which means they are valid for further analysis.

In addition, the validity of discrimination is tested through cross-loading, where each indicator has a higher correlation value with the variable it measures compared to other variables. It shows that each variable has good discriminatory validity.

For reliability testing, the results of composite reliability show that all constructs have a reliability value above 0.80, which indicates that each construct in this study is reliable and reliable for further analysis.

The R-Square test was conducted to assess the predictive strength of the model used. The test results showed that the variables of hard skills and soft skills were able to explain career development by 40%, human capital by 50%, and individual attributes by 40%. It means that there are still other variables that affect the individual's career, human capital, and attributes that were not studied in this study.

The Q-Square test shows that the predictive values for all dependent variables are more than 0, which indicates that the model has good and relevant predictions. Furthermore, the research hypothesis was tested using a bootstrapping approach. The test results (see Table 4) show that hard skills have a significant influence on human capital and career development but not individual attributes. Meanwhile, soft skills showed a significant influence on the three dependent variables, namely human capital, individual attributes, and career development.

The results of the hypothesis test show that hard skills and soft skills in big data analytics have a significant influence on human capital and career development. However, hard skills did not show a significant influence on individual attributes, indicating that technical skills were not always related to individual attributes. Soft skills, on the other hand, have a significant influence on all three variables, underscoring the importance of interpersonal skills in improving the employability of accounting students.

The Effect of Hard Skills Big Data Analytics on Human Capital

Based on the results of the hypothesis test, hard skills in big data analytics have a positive and significant influence on increasing the human capital of accounting students. The better the hard skills of big data analytics, the better the quality of human capital. Skills in analyzing and using big data platforms help students gain diverse knowledge, not only limited to accounting but also in understanding analytics technology and data-driven decision-making. By understanding big data analytics, students can develop critical thinking skills, which are very useful in solving complex accounting problems (Ahmed et al., 2017) (Jackson et al., 2022).

These results support the job-market signalling theory, which explains that hard skills can be a signal for employers to assess the quality of students' human capital (Spence, 1973). This study supports the findings of (Xu et al., 2022), which shows that hard skills have a significant effect on human capital.

In addition to hard skills, hypothesis tests show that soft skills in big data analytics also have a positive and significant influence on increasing the human capital of accounting students. Soft skills, such as communication, collaboration, and critical thinking skills, are essential in the world of analytics. The ability of students to interpret and communicate data clearly through big data analytics helps them understand information in depth and interpret it into more applicable knowledge ((Tsiligiris & Bowyer, 2021); (Xu et al., 2022))

This increase in human capital through soft skills supports the theory of job-market signalling, which states that soft skills can be used as a signal for employers to assess students' ability and potential to adapt and develop in a dynamic work environment. These results are also in line with the findings of (Xu et al., 2022), who found that soft skills have a significant effect on the human capital of accounting students.

The results of hypothesis testing show that although hard skills in big data analytics have a positive influence on the individual attributes of accounting students, the influence is not significant. It means that even though students have good hard skills, this does not necessarily affect aspects of their personalities, such as adaptability, flexibility, or creativity (Clarke, 2017)

Hard skills tend to be related to technical and practical abilities in completing tasks but are not directly related to the individual's personality. Therefore, these results do not support the job-market signalling theory in the context of individual attributes because hard skills cannot be a strong indicator to measure a person's personality qualities. This finding is in line with the study of (Xu et al., 2022), which also found that hard skills do not have a significant influence on the individual attributes of accounting students.

On the contrary, the results of the hypothesis test show that soft skills in big data analytics have a positive and significant influence on the improvement of individual attributes of accounting students. The ability to collaborate, communicate, and think creatively in a big data analytics environment encourages the development of a more adaptive and innovative personality (Xu et al., 2022).

Big data analytics also help students improve communication skills and emotional intelligence, which is important to avoid miscommunication and strengthen teamwork (Tsiligiris & Bowyer, 2021). These results support the job-market signalling theory, where soft skills are considered a strong signal to demonstrate the quality of individual student attributes to employers. The results of this study are also consistent with the findings of (Xu et al., 2022), which found that soft skills have a significant effect on the individual attributes of accounting students.

Hypothesis testing also shows that hard skills in big data analytics have a positive and significant influence on improving the career development of accounting students. Students who have hard skills in analyzing data and utilizing big data technology can understand relevant trends and strategies for their career development (Jackson et al., 2022).

Hard skills help students better plan career steps and provide a solid foundation in professional decision-making (Xu et al., 2022). This result is in line with the theory of job-market signalling, where students can use hard skills to show their readiness to pursue a career, thereby increasing their chances of being hired. This study supports the findings of (Xu et al., 2022), which stated that hard skills have a significant effect on the career development of accounting students.