39 CHAPTER 3 METHODOLOGY 3.1 Research Design This study uses a quantitative approach to investigate the relationships between the variables discussed earlier. This chapter explains the methodology, outlining the process from identifying the problem to drawing conclusions and making recommendations, as visualized in Figure 3.1 below. Figure 3.1 Research Design CHAPTER III METHODOLOGY 40 This study created a research design, as illustrated in the figure above, to test the hypothesis from Chapter II. The design provides a detailed outline of the study, and all steps shown in the figure will be explained further in this chapter. 3.2 Research Approach In research, quantitative and qualitative methods are the main data collection approaches. This study focuses on the quantitative method, using surveys to collect numerical data, which is then analysed statistically to draw conclusions (Balnaves & Caputi, 2001). Additionally, before the main study, a pilot test is conducted to ensure the validity and reliability of the survey instruments. This step helps identify any issues or shortcomings in the questions and protocols. Once the pilot test confirms reliability and validity, data collection from a larger sample can proceed. 3.3 Problem Identification According to Holyoak (1984), people build mental blueprints based on their experiences to solve problems, including their goals, steps, information, and rules. This study identifies the research problem through literature reviews and gap analysis. Chapters 1 and 2 gather evidence on the insufficient exploration of graduates' soft skills (communication, teamwork, time management, problem-solving, and adaptability) versus those required by employers in the digital job market. Secondary data supports understanding which specific soft skills are most lacking among graduates. These findings establish the research question and objective 3.4 Literature Review To better understand the topic, the researcher conducts a literature review to establish the theoretical basis of the variables involved in this research. A literature review, or review article, is defined as a study that examines and integrates existing research by identifying, challenging, and enhancing the foundational elements of a theory through the analysis of prior work (Post et al., 2020, p. 352). This study developed a hypothesis and theoretical framework by examining past research, including academic journals, textbooks, and other credible sources on the topic. This review helps to understand the methodologies previously used, evaluate their effectiveness, and identify challenges faced in earlier studies. The literature review is crucial for defining the variables tested in this research. 41 • Variables A variable is something that can be measured and can change for the same object or person, or among different objects or persons at the same time (Sekaran & Bougie, 2016). It represents a characteristic or quantity of the phenomenon being studied. Statistically, a variable is defined as a quantity that can take on different possible values (Onen, 2016). This research discusses three main types of variables: 1. Dependent variables are the primary interest because they help us understand how other factors influence them. These variables are expected to change because of manipulating the independent variable(s) in the study. Essentially, they represent the presumed effect (Cramer & Howitt, 2004). In this research, the dependent variable is a student’s perceived work readiness (PWR). 2. Independent variables, on the other hand, remain stable and are unaffected by the variables being measured. They are the conditions that the researcher systematically manipulated to test an idea. These variables are considered the presumed cause (Cramer & Howitt, 2004). In this research, the independent variables are soft skills, including communication, problem-solving, time management, adaptability, and teamwork skills. These skills are expected to influence or affect the outcome on the dependent variable. 3.5 Data Collection In this chapter, the researcher will describe the data collection process, the types of data used, and the method for determining the sample size. These details will be outlined in the following sections. 3.5.1 Type of Data The researcher will gather primary data through online surveys targeting recent graduates from 2021-2024 from various public universities, including ITB, UI, UGM, UNPAD, and ITS. The aim is to thoroughly evaluate the effect of soft skills development on perceived work readiness among these graduates. Additionally, secondary data will be collected from online sources such as journals, websites, other publications, and books. This secondary data will provide accurate information to support the analysis and context of the research. 3.5.2 Population and Sample The population for this research consists of fresh graduates from various public universities in Indonesia, including ITB, UGM, UI, UNPAD, and ITS. However, the sample calculation will 42 focus on fresh graduates from Institut Teknologi Bandung (ITB) from 2021 to 2024, covering the April, July, and October graduation periods. These numbers will then be used to estimate the graduate numbers for the other four universities. According to Rector's Decree No. 924, ITB graduated 3,639 students in 2021, 3,693 in 2022, 4,252 in 2023, and 658 in April 2024, adding up to a total of 12,242 graduates from 2021 to 2024. Furthermore, the PDDikti website lists the average annual graduates from UI as 405. As per the ITS website, ITS had 559 graduates in 2021, 3,525 in 2022, 1,280 in 2023, and 1,355 in 2024, making a total of 6,719 graduates during these years. Additionally, the UGM website indicates that UGM had 917 graduates in 2021, 1,254 in 2022, 1,852 in 2023, and 1,345 in 2024, totaling 5,368 graduates from 2021 to 2024. Lastly, according to the UNPAD annual report, UNPAD had 2,986 graduates in 2021, 3,302 in 2022, and 1,363 in 2023, resulting in a total of 7,651 graduates over these three years. Roscoe (1975) suggests a sample size of 30-500 for most behavioral studies (Sekaran & Bougie, 2016). Therefore, the minimum number of respondents will be calculated using Slovin’s formula.