3.9.2 Interview guides
In conducting the interviews, the researcher made use of interview guides. Questions were prepared in advance taking heed of Patton, (2002)’s idea that an interview guide lists the questions or issues that are to be explored in the course of an interview. The use of an interview guide has advantages as it ensures careful use of interview time, makes interviewing systematic and comprehensive by deciding in advance the issues to be explored while maintaining the interactions focused. Cohen et al. (2005) also points that interview guide enable data collection to be systematic.

The interview guides were designed in such a way that they addressed the key issues which needed to be answered as far as an understanding of the recycling industry was concerned. Issues such as companies motives and extent of involvement in recycling, policies and legislation for recycling, recycling products, recycling networks, benefit chains, value addition processes, awareness of recycling, challenges of the industry were covered in the guide.
Data gathered answered the main objectives of the study. Appendix F gives details of the general structure of the interview guides with some omissions or additions in the different guides depending on the areas of focus.
3.9.3 Direct observation
Observation as a research process offers the researcher the opportunity to ‘look directly at what is taking place in situ rather than relying on second-hand accounts'(Cohen, 2005,p. 415). Data is gathered from naturally occurring social situations.Rugg ;Petre (2007, p. 110) remark that ‘one strong point of observation is that it shows you something, without the filtering effect of language’.Observations, according to Morrison (1993 p. 80 as cited by Cohen, 2005, p.397) enable the researcher to gather data on physical, human, interactional and the programme settings. The researcher carried out direct observations during company visits. These were done in order to confirm data collected during interviews and documentation. Through observations, the researcher collected data on activities and processes carried out by the different recycling companies. The researcher was able to observe recovery, preprocessing and manufacturing activities that took place at companies A, B, C, D, E,K,F and N. Manufacturing of plastic packaging and plastic pipes at companies B and D was captured during observations, confirming what was said during interviews.

While the method is credited for providing first hand information, several limitations have been recorded. Kothari (2004) identified some of the following shortcomings:
1. It is an expensive method.
2. The information provided by this method is very limited.
3. Sometimes unforeseen factors may interfere with the observational task.

3.9.4 Observation checklist
Observations were recorded through the use of an observation checklist. For systematic data collection, the use of checklists is encouraged as noted by Rugg ; Petre (2007). A list of activities to be observed is kept in diaries by the researcher for answers for a specific observation to be recorded. Appendix G shows the checklist used during the research.
3.9.5. Document search
Apart from primary sources of data collection, the researcher also made use of documents from companies and archived on the internet. Documents comprise of written material and other documents from the cases under investigation (Patton, 2002). Document search was important because it gave the researcher a general background and operation issues on the subject that was being studied. The researcher collected official documents in hard and soft copy such as the waste management policies, regulations and reports on waste reduction measures and waste audit plus other related documents from Company (O) while company (M) documents were a report and pamphlet about the recycling industry. Document search gave the researcher an insight into the activities taking place within the organization, for example, company (M) documents helped the researcher to verify what was happening in the recycling industry in Namibia as some of the issues raised like challenges of the industry were well documented. Unlike respondents, who are aware of being studied, documents have the advantage of “unobtrusive and non-reactive measures” (Hoyle et al., 2002, p. 361). However, not all companies provided this information. Therefore some information was obtained from the internet such as general policies and legislation governing the industry. As with other data collection methods, documents have limitations. They may be incomplete and in some instances inaccurate.
3.9.6 Piloting and pre-testing
Pre-tests and pilot studies are different types of mini studies carried out as part of the process of planning and preparing for a study. A pilot study is a “small scale replica and a rehearsal of the main study”(Sarantakos, 1993, p. 277). The purpose of the exercise is to check the effectiveness of the research instruments e.g. checking respondents comprehension of questions or checking on cases of ambiguity. (Bless ;Higson-Bless, 1995; Powell, 1997; Sarantakos, 1993; Yin, 2003). Sarantakos 1993, p. 277) states that with ‘case studies piloting can establish availability of respondents, accessibility of the research environment and effectiveness of the data collection technique, whether it will collect too much or too little information’. Following this argument, the researcher decided to carry out a pilot study. Piloting was carried out in March 2015 with one of the largest recycling companies in the country. It came out during the interview that questions had to be short as long questions ended up with part of them not addressed. Some adjustments were later made to the interview guide. Through piloting, it was possible to clarify as well as identify other issues pertinent to the study, which were then included in the inquiry. Piloting established that the time required to complete the interview guides was too long and that some of the questions were unintentionally repetitive and ambiguous. The interview guides were then adjusted accordingly.
3.10 Data Collection Procedure
This section describes the process the researcher went through to collect data from the different companies e.g. from seeking authority to conduct the study, arrangement for interviews, up to the collection of the data.
3.10.1 Seeking permission from the institutions and individuals
For research conducted on an institution, approval for conducting the research should be obtained from the institution (Bell 1999). At the beginning of the research, the researcher physically visited the City of Windhoek municipal officials since the researcher had no idea of who to talk to. After getting contact numbers, the researcher phoned all companies that were identified in order to seek permission to conduct the research in their organizations. This was a lengthy and frustrating process as very few responded in an amicable way. Having explained the purpose, the researcher then sent a letter to all institutions seeking permission to conduct the research in their companies as per request. Some of the companies did not respond despite endless efforts by the researcher to get feedback. In some instances, the researcher had to wait for a period of three weeks to get response. Of the twenty companies identified, 15 later gave the researcher permission to conduct the research. Data collection started in May 2015.
After getting permission to conduct the research in the organizations, the next step was to arrange interview appointments, another lengthy process. Setting up interviews with the research participants was not an easy task with some of the participants. They would cancel appointments or request to be interviewed at short notice or simply to dodge the interview after giving promises of the date and time of when the interview will take place.
3.10.2 The interview process
The researcher conducted all interviews personally. Before the interview commenced, consent letter had to be signed for agreement to be interviewed and audio tapped. All participants agreed to be interviewed and audio tapped. However, in all instances, the researcher was only able to use the voice recorder on two participants as the voice recorder malfunctioned in some instances and the researcher had no back- up plan except to take down notes as the interview progressed. Information sought from the companies was to do with motives, extent of involvement in recycling activities, policies and legislation governing their operations, value addition, linkages and benefits of the industry as well as constraints they faced in their operations and how they managed with them.Where given permission, photographs were taken during observations. The researcher reviewed notes at the end of each day for any insight on issues relevant to pursue in subsequent interviews. This is what Patton (2002, p. 383) refers to as the “emergent nature of qualitative research”.
3.10.3 Research ethics
Issues of informed concern, confidentiality, and integrity during the research are important as recommended by some authors such as Patton (2009). Ethics considers the good and bad or right and wrong with moral duty and obligations.
Observations of ethics were of great significance to the study. The researcher firstly debriefed the participants before carrying out the interviews. This was done by explaining the whole purpose and process of the study before highlighting the importance of the research. In this research, the participants were given assurance of confidentiality, and by so doing they were assured of no disclosure of information such as names of companies or respondents as such information obtained would be considered personal and private and was going to be used for academic purposes only.
Creswell (2003) advises on the importance of maintaining privacy and confidentiality during a research. The researcher used codes for companies in order to protect their identities and no participants were coerced to take part. During and after the study, the identity of the respondents will remain confidential. Research information from interviews and discussions were coded and kept in a safe place with confidentiality.
3.11 Data Analysis
Data analysis and presentation follows data collection (Cohen, et al. p.458).It involves making sense of data in terms of the participants’ definitions of the situation, noting patterns, themes, categories and regularities (Cohen, et al. p.458). Data analysis is important in order to generate findings from the raw data. It is the process of organizing the mass of collected data (Nichodemus , 2010).The process is not left until to the end as Jacobsen et al. (2006) pointed out. In fact data collection and analysis takes place simultaneously with qualitative research. The study applied content analysis procedure. Content analysis refers to the systematic set of procedures for the rigorous analysis, examination and verification of the contents of written data according to (Flick 1998: 192; Mayring 2004: 266 as cited by Cohen, 2005, p. 475. Desired information from a text is extracted during this procedure. Content was extracted from the interview transcripts, documents and observation notes. As the first step, transcripts of interviews were analyzed through reading them. Information from the transcripts was then broken down into categories resulting in the emergence of different themes. According to Trace (2001), when analyzing the data, themes should be allowed to emerge rather than attempting to impose preconceived set of themes on the data. Using the deductive approach, the researcher thematically analyzed the transcript data first and then analyzed these themes in the light of the research questions.

Data was analyzed manually although there were computer software packages that could be used to analyze qualitative data such as Atlas/it and Hyper Qual. Mayring (2000) & Rourke et al., (2001) reported that these soft wares have proved their worth. Hoyle et al. (2002; p. 399) argued that computerized content analysis can analyze large amounts of data very quickly but cannot handle “verbal subtleties such as sarcasm”. The researcher considered the data collected as not being large enough to warrant the use of software.
3.12 Data Presentation
Data presentation differs depending on the research methods used. Quantitative research method relies on the use of statistical reports to present data. On the other hand, qualitative research relies on the use of narrative reports, with contextual descriptions, direct quotations from research participants, graphs and tables as noted by Hancock(1998).Qualitative presentations were employed in the presentation of research findings of this study. This was in the form of descriptive narrative, illustrative quotes, tables and some few diagrams and graphs.
3.13 Validity and reliability
Cresswell (2009) emphasizes that reliability and validity are of great importance when doing a research. This idea is further supported by Patton (1990) who points that that validity and reliability are two factors which must be of great concern to the researcher in qualitative studies while designing a study, analyzing results and judging the quality of the study. According to Esposito (2002) validity refers to whether the researcher actually measured what he/she wanted to measure. Reliability on the other hand means that responses to the questionnaire were consistent. Qualitative validity means that the researcher checks for the accuracy of the findings by employing certain procedures, while qualitative reliability indicates that the researcher’s approach is consistent across different researchers and different projects according to (Gibbs, 2007 as cited by Creswell(2009).
To ensure validity and reliability of the study findings, Creswell, (2007), Hoyle et al. (2002) and Patton (2002) argue for triangulation as a strategy for improving validity and reliability of research. For this study, observation, document search and interviews were used to ensure validity of the findings.The reliability of interviews for this research was observed through pilot testing. This was done to ensure that no information was missed from the respondents. Questions that were not clear to the respondents were noted and rectified.