21 Chapter III Research Methodology III.1 Introduction This chapter will cover the explanation regarding the research methodology chosen. It will also include the step-by-step description on how the research is being conducted. III.2 Research Methods This study examined the phenomenon by following the exploratory mixed method design. As both qualitative and quantitative data provided different sets of information, mixing both methods was seen to develop a stronger understanding toward the issues (Creswell & David Creswell, 2018; Gogo & Musonda, 2022). Exploratory mixed method design allowed the researcher to explore the phenomenon before validating, giving the research more room to discover the issues (Gogo & Musonda, 2022). Researchers could also obtain a better understanding of the cultural setting of the population, giving a stronger context on the study (Creswell & Plano Clark, 2018). Thus, given the insufficient past studies on this topic in the Indonesian setting, this study was able to better highlight the relevance and environmental influence on e-commerce waste issues in Indonesia. Following the exploratory sequence, the study consisted of three phases: (1) qualitative data collection, (2) intermediate phase, (3) quantitative data collection. The qualitative data collection intended to explore the phenomenon and emphasise the research purposes (Creswell & Plano Clark, 2018). The researcher performed semi-structured interviews as the qualitative data collection method. This part of the research was identified as exploratory research design, which allowed the researcher to further identify the issue and act as the input for the next phase (Maholtra, 2010). As the second step, the intermediate phase utilised the theme and quotes obtained from the qualitative data collection to be the base of instrument creation for the quantitative data collection (Creswell & David Creswell, 2018). It allowed the researcher to revise the pre-existing instruments to 22 better fit the population, or even created a new variable based on the qualitative results (Edmonds & Kennedy, 2017). The third phase, quantitative data collection, was performed by administering the instruments into a self-administered survey. This phase was identified as a conclusive and descriptive research design. As the continuation of the previous exploratory research, the survey had a clearer to goal to quantify the relationship between the variables previously discussed in semi-structured interview (Maholtra, 2010). The findings obtained in this phase was aimed to generalise the previous qualitative findings on a new population (Creswell & Plano Clark, 2018). We also performed another exploratory research in a form of netnography using social media data from Twitter. It provided the researcher with consumers’ honest opinion and experience that they posted themselves (An & Alarcón, 2021; Vidal et al., 2018). Performing exploratory research with secondary data, like data from social media, allowed the researcher to better identify the phenomenon and confirmed the findings from the primary data collections (Maholtra, 2010). Thus, it might complement the data from quantitative nnd qualitative research (An & Alarcón, 2021). III.3 Research Design This study was conducted in several phases. The steps of how the research being conducted is illustrated on Figure III.1. 23 Figure III.1 Research Design 24 The first phase consisted of problem identification and literature review to identify the gap from the existing literature. We began the data collection with the semi- structured interview, where the interview protocol was generated from the literature review regarding pro-environmental behaviour and waste management. Conducting the exploratory interview allowed the researcher to better understand the phenomenon in the context of Indonesian consumers. The coding result from the semi-structured interview was one of the inputs for the next phase, the intermediate phase or the measurement items creation. In this phase, we adjusted the pre-existing measurement items based on the interview results to generate questionnaire items that were fit to be tested on Indonesian consumers. The obtained items were tested using self-administered questionnaire, with the respondents coming from across Indonesia. The last step of the data collection was Netnography, where the researcher gained additional insight on the Indonesian consumers behavior toward e-commerce waste. On the last phase of the research, we combined the findings obtained from all three data collections to answer the research questions and provide managerial recommendations regarding e-commerce waste issue in Indonesia. III.4 Data Collection III.4.1 Settings This study targeted Indonesian consumers who had done a minimum of one transaction in e-commerce for the past six months. While some e-commerce also provided food delivery services, we excluded this type of service in the study and focused on the online shopping feature. As the study aimed to understand the consumers’ behaviour in dealing with e-commerce packaging waste, we did not limit the name or platforms used by the consumers. The study focused on both men and women consumers at the age of 18-40 years old. The early Gen-Z and Millennials were seen as the generations who utilise e-commerce the most in their daily lives. A study from Das et al., (2018) on McKinsey & Company’s digital report showed that consumers at the age of 20-39 made up around 68% of total e- retailing consumers. It signified the relevance of the said age range on the usage of e-commerce in Indonesia. 25 III.4.2 Semi Structured Interview An interview is a guided, and purposeful conversation conducted to obtain insight regarding a specific issue (Sekaran & Bougie, 2016). We performed the semi-structured interviews as exploratory research to gain a better understanding on the issues and confirm the findings from previous studies (Maholtra, 2010). Performing exploratory interviews with small number of samples allowed the researcher to get a more detailed response on the respondents’ point of view about the specific phenomenon (Creswell & Plano Clark, 2018). Semi-structured allowed the researcher to guide the interviewee by providing predetermined topics and questions, while still giving room for other questions or related topics to emerge in the conversation (Whiting, 2008). We followed the laddering technique where the interviewer started with one question at a time, then guided the interviewee to explore more about the topic by asking some follow up “why” questions. This technique helped the interviewer to unveil the interviewee’s motivation and resistances behind their action (Heinze et al., 2017). The interview utilised open- ended questions, which were helpful in getting unforeseen responses from the participants and further exploring their answers (Goonewardene & Persad, 2018). This type of interview is suitable for understanding before conducting a large-scale survey (Adams, 2015). Depending on the agreed circumstances, the interviews were conducted either through face-to-face meetings or virtual meetings in ZOOM. The interviews were conducted in February and March 2023. We selected the interviewees by following the purposive sampling method, as we specifically chose the interviewees who had specific experience in using e-commerce (Creswell & Plano Clark, 2018). The participants needed to fulfil these criterias: male or female, 18-40 years old, living in Indonesia, and have conducted at least one shopping transaction in their preferred e-commerce platforms.