Data collection methods

My research aim was to describe and explain EAL’s pedagogy. For this, I needed profound insight into the project’s practice and conceptual considerations.

The questions that I had in mind to approach my aim were: What are

professionals doing when they educate hard-to-reach youngsters on-site?

On what resources do the professionals rely? In order to find information to elucidate my questions regarding the praxis and conceptual superstructure, I used three sources for data collection: field trips, documents, and interviews.

Table 2 presents the data collection methods.

Table 2. Overview of data collection

Field trips Project documents Problem-centered interviews

Data collection 2009–2015 2015 2016–2018

Quantity 5 12 in total, of which

10 for deeper analysis

2 supervisor interviews 14 key worker interviews Data management Field protocols Pdf files stored

and assessed for conceptual characteristics

Interviews recorded and fully transcribed

Table 2 informs about the type and amount of data collected and denotes how it was managed for further assessment. Since the field trips were part of my preparation for the study, they are not addressed here in detail. I took field notes during these visits to record the impressions, feelings, and emerging questions that were the precursors of my interest in the area under study. The next paragraphs concentrate on my two main methods of data-gathering: document analysis, and problem-centered interviews with professionals on the project.

Documents

In addition to my previous visits on-site, another source for my research was written material about the EAL project. These documents were found during my initial information-gathering at the beginning of my doctoral studies. Besides informative documents such as scripts for speeches and presentation, PowerPoint slides, and professional articles, they also included

conceptual material written by the founders of EAL to inform third parties about the project or guide the project’s practice. I accessed this material, which mainly documented conceptual developments, with the help of the project’s secretary, who manages the EAL paperwork. As a result, I found 12 documents which I stored on a memory stick, and printed out. After an initial review, the written material about EAL was set aside until I analyzed some interviews and developed initial categories grounded in the data.

At that stage I revisited the documents, applying open coding to see what concepts would emerge in order to compare them with the preliminary concepts derived from the first interviews. After a systematic evaluation of the retrieved documents to assess their type, content, purpose, and relevance for the current study (Bowen, 2009; Prior, 2008), I considered 10 exploitable documents relevant and retained them for open coding. For a compilation of the documents, organized by document type and outlining their role and impact for practice on-site, see Table A1 in the Appendix. Table A1 also lists the year each document was written, as a reference point for the historical perspective that infused my thick description of the EAL project. Although documents can have limitations concerning retrievability, insufficient content, or biased selectivity (Bowen, 2009, p. 31), researchers can ensure their quality by following a systematic procedure for finding, reviewing, and evaluating relevant documents, especially when they are to be used as a complementary data source.

Interviews

The major part of the data was harvested from interviews with two different kinds of professionals involved on the project: firstly, key workers,¹³ i.e. the staff working on-site with the adolescents; secondly, their supervisors, who held superordinate positions on the project and fulfilled administrative, monitoring, and supporting tasks from Germany. Witzel’s problem-centered interview (2000, Witzel & Reiter 2012) was used to gather information from the participants. This semi-structured interview approach focuses on

13 I have chosen the term “key worker” to designate those working on-site with the adolescents. It refers to their role as the adolescents’ care persons, attachment figures, and direct contact persons. It is a translation of the German term “Bezugserzieher.”

social problems explained by participants in the field. Drawing on previous knowledge about the field, the researcher develops questions to keep in mind, but also allows for spontaneous questions during the interview. This restraint creates a dialogical relationship between researcher and interviewee, with an emphasis on the process of developing and understanding the topic as a joint enterprise between participant and researcher (Helfferich, 2009, pp. 37, 49). It is especially suitable if the researcher is closely related to the field, because it allows the researcher to encounter the participant as a knowledgeable observer. I chose this kind of interview for several reasons related to my research interests and my attempt to ensure the scientific quality of the project. As I explain more deeply in Section 2.4, I had a close occupational relationship to the EAL as a field of investigation. This entailed some professional encounters with staff and participants, and enabled me to extend previous knowledge about the project even before my investigation started. The problem-centered interview approach takes the unavoidable prior knowledge of the researcher into account, and allows it to be openly shared in the interview situation. This stance connects to the notions of theoretical sensitivity (Glaser, 1978) and sensitizing concepts (Bryant, 2017, p.

79; Kruse, 2015, p. 99) in the grounded theory approach.¹⁴ Helfferich (2009, p.

105) adds that a guiding structure also ensures the comparability of different interviews, and this is also a core technique in grounded theory, which uses constant comparisons to develop concepts from the data. Therefore, my choice of the problem-centered interview approach supported my study in three respects. Firstly, problem-centered research focused on the views and experiences of professionals on the project. Secondly, it facilitated transparency about my position as researcher in the field. Thirdly, it utilized my prior knowledge to improve my investigation.

Moreover, grounded theory’s requirement for constant comparisons of the data to develop concepts was supported by the use of a flexible

14 The notion of sensitizing concepts is one of the central points of discussion in grounded theory. It was a central issue in the disagreement between Glaser and Strauss, which ended in a split between the two founders of grounded theory. While Glaser (1978, 2008) retained a conservative attitude toward the inclusion of previous knowledge in the research process, Strauss favored the transparent use of prior knowledge (1987, Corbin & Strauss 2015).

content structure. However, there is one crucial limitation in semi-structured interviews: It is difficult to be sure that the participant’s issues and problems are covered. While an unstructured interview provides the most abundant source of data and facilitates the free flow of information, a semi-structured interview takes the lead to a certain extent, and runs the risk of limiting participants’ expressions. The researcher needs to be aware of this risk and maintain a free flow for the interviewee (Corbin & Strauss, 2015, p.

39). In accordance with Witzel’s (2000, Witzel & Reiter 2012) approach, my interviews were informed by a set of questions developed from my previous occupational encounters and the knowledge gained by my scoping review of extant literature in the field. Furthermore, I took account of information gained from two supervisor interviews conducted as preliminaries. Finally, further orientation information was collected during my prior informal meetings with former key workers on the project.