• Keine Ergebnisse gefunden

SECTION IV: P ROCEDURE OF D ATA A NALYSIS

4.3 Classical Grounded Theory

researcher’s construction of meaning becomes an important part of the analysis. The limitation was that constructive grounded theory does not provide the empirical reality to emerge as it is.

For the constructive grounded theory, realities are a social construction of the mind and thus for constructive grounded theory multiple realities are constructed (Mills et al. 2006; Guba and Lincoln 1989). Realising the difference between constructing realities in constructive grounded theory and objective reality in Glaser’s classical grounded theory (Mills et al. 2006), the researcher switched to the classical grounded theory. Researcher did not choose Straussian grounded theory as Straussian grounded theory proceeds on theoretical grounds which creates preconceived bias (Evans 2013). To avoid any preconceived bias in order to discover a theory rooted in empirical data, the researcher chose Glaser’s classical grounded theory

4.3 Classical Grounded Theory

Choosing CGT over Straussian and constructive grounded theory allowed the researcher to grasp participants’ perspective and to study participants concerns (Glaser 1998, P 32). Thus, the methodology gives importance to participants’ perspectives, concerns and their reflections to gain in depth insight on the issues of concern (Holton 2008; Glaser 1998). The issue of concern for the current study was to gain an understanding of women’s education and violence to understand empowerment process from an NGO expert perspective. As suggested by Glaser, the pre-literature review is avoided for the main study field work, consequentially the coding procedure was started without any preconceived ideas.

There are two types of coding in classical grounded theory: substantive coding and theoretical coding. Some authors have divided the substantive coding process into two sub processes; open and selective coding (Evans 2013; Hernandez and Andrews, 2012; Walker and Myrick 2006). In substantive coding, the researcher works with data directly and it is the process of conceptualizing the empirical substance of the area under study: the data in which

the theory is grounded (Holton 2007). Holton further explained, indicators of category or concept are grounded in the incidents of the empirical data from which a grounded theory is generated. The aim of using classical grounded theory was not to verify facts but to generate plausible integrated set of conceptual hypotheses in a substantive area of concern (Glaser 1998). Grounded theory is a general methodology of analysis linked with data collection that uses a systematically applied set of methods to generate an inductive theory on a substantive area (Glaser 1992, P 16). As Glaser (1998) said “All is Data” signifies the flexibility of using all types of data. The researcher used, empirical data collected by BNI and EI, previous scientific literature, newspaper articles, video clips to critically analyse the empirical data. The data collected by various methods such as EIs and BNIs were used creatively to generate an inductive multivariate theory (Holton 2007; Glaser 1998).

The substantive theory that emerged from the systematic data analysis is the theory of victimization of educated women in India, an inductive theory. The systematic data analysis in classical grounded theory engaged a process of constant comparison, coding and theoretical sampling. The following sub section lays out details of each of the components of the CGT process starting from theoretical sampling for classical grounded theory. The section discusses the sampling process laid down by the CGT and pre-determined sampling size in the current study.

4.3.1 Sampling

CGT gave preference to data collection guided by theoretical sampling, the theoretical sampling, as defined by Glaser and Strauss (1967) is, “The process of data collection for generating theory whereby the analyst jointly collects, codes and analyses his data and decides what data to collect next and where to find them, in order to develop his theory as it emerges”

(Glaser and Strauss 1967, P 45). An emerging theory guides the sampling and prevents the researcher from collecting the same data again (Glaser 1998). Thus, the theoretical sampling

prevents repetitive or same data but collection of diversified data for the emergence of a holistic and all-inclusive theory.

As the sample size was predetermined for the current study, the theoretical sampling was still achieved in the current study. During the main study field work, due to lack of time, the researcher could only write observations and notes during the interview. After the interview sessions, the researcher listened to the recorded interview again which helped the researcher to keep a check on the same data and decide where to go, such as which caste and religion NGO experts to investigate next in order to collect diversified data. Thus, due to this process, the researcher could gather data from experts who are working on different forms of violence, following different approaches for empowerment and from different castes and religions of the society. During the main study field work, the researcher analysed the questions based on the themes and revised the questions to get more relevant data for the emergence of a grounded theory.

The sample size was predetermined, the researcher had to pre-plan the field work. In order to develop an all-inclusive and holistic theory, the researcher made notes of different fields of women’s empowerment and searched online, and contact was established. Initially only few contacts were established. During the main study field work, the researcher went through the recorded interview and decided which sample to look for next. Initially, the researcher looked for NGO experts working for child sexual abuse, sexual education and intimate partner violence and family problems. After going through the recorded interviews, the researcher felt a lack of data from the caste and class perspective. Thus, the researcher started to look for NGO experts working for marginalized castes and classes. Due to observations and notes made during interview sessions, the researcher analysed experts’

different approaches of women’s empowerment and their methods to empower women. To achieve these, the researcher collected data from NGO experts who practice different

empowerment methods. To get the contact of these participants, the researcher contacted pre-study NGO experts and got contact details from the participants, and some contact information was collected from the internet.

Post main study field work, the researcher analysed the data with CGT. The following sub section explains the procedure of classical grounded theory analysis. Constant comparison is first step towards forming a multivariate theory. Following section discusses the aim and purpose of the constant comparison in CGT.

4.3.2 Constant comparison

Constant comparison is a concept-indicator (Holton 2007), an underpinning of grounded theory to such an extent that all three streams of grounded theory namely CGT, Strauss and Corbin’s Grounded Theory and Charmazian Constructivist Grounded Theory were used for constant comparison in the data analysis process to generate a plausible theory (Evans 2013). The CGT laid strong focus on constant comparative method for concept-indicator initial categorization of incidents and then eventually for conceptualization (Glaser 1992). The data is compared at three levels- a) Incidents to incidents to identify the underlying meaning and pattern within a set of descriptive incidents for the emergence of concepts. Furthermore, b) concepts are compared to more incidents for the best fit data and for the emergence of properties of concepts for verification and densification of the concepts and lastly, c) concepts are compared with concepts to generate a range of concepts and integration to discover a theory (Holton 2007). In CGT, a unit of analysis is data and not participants; the researcher needs to see incidents and constantly compare the incidents to generate categories, properties of categories and integration of concepts to form integrated and parsimonious conceptual hypotheses (Glaser 1998).

For the current thesis, the researcher compared different aspects of the data such as different approaches for women’s empowerment followed by experts, status of women in

different communities, aim and purpose of education and its impact on women of different communities to gain a holistic picture on changing forms of subjugation in contemporary India.

In the following sections, the researcher discusses CGT coding illustrated by sample coding with different comparative approaches.

4.3.3 Open coding

Initially the researcher struggled not only with large amounts of data (820 pages) but also a lack of self-belief. As Holton stated the researcher also experienced lack belief in the grounded theory method (Holton 2009). The coding process for the researcher was not less than a roller coaster ride. In coding process, open coding provides a direction and answer to deal with these ambiguities emerged at the beginning of grounded theory analysis. In open coding a researcher fractures the data and relevant data is deduced to a conceptual level (Glaser 1978). Open coding is the first analytical step to yield initial categories, it is an initial stage of analysis before delimiting a theory. At this stage a researcher remains open to the data without any preconceived codes and lets the pattern emerge from the incidents which are compared (Glaser 1998). Open coding is “the initial step of theoretical analysis that pertains to the initial discovery of categories and their properties” (Glaser 1992, P 39).

Though Glaser gave a systematic method of constant comparison for emergence of concepts and categories at open coding stage, at this stage a researcher struggles with selectivity and relevance. Deciding which incidents should be coded and which should be omitted is the biggest challenge at this stage. To address the challenge, Glaser (1978) stated, the researcher codes the data every way possible, that is ‘running the data open’ (P 56). To code data every way possible, transcriptions are read line-by-line, but due to the overwhelming volume and repetitive data, the researcher followed paragraph to paragraph coding (Glaser 1992). In order to avoid pre-conception, the researcher asked questions while coding the incidents such as “what is this data about?”, “what does this incident indicate?” and the pattern

emerging from incidents are focused while open coding. The researcher coded the data every way possible, such as coding interpretations, different incidences into as many categories as possible (Glaser 1978, P 56). This allows the most workable ones and the core relevant data to emerge on their own (Glaser 1978, P 56).

For labelling each incident in vivo codes are taken from respondents’ interview and subject to revision through constant comparative method to form newer conceptual codes from the empirical data (Holton 2007). In open coding the data is compared until a core category emerges. At this level the researcher verifies and saturates the categories, irrelevant categories are omitted, thus the delimiting theory procedure starts with an open coding (Holton 2009).

Open coding sample is discussed in the following table for related category ‘approaching’. The table illustrates the procedure followed while coding the data with comparative approaches.

Questionnaire Interview Data Questions

supervising. She had certain assumptions regarding husband which caused problems between them.

regarding her husband.

Negative thinking cause of action.

Monitoring and

supervising as action.

Solving problem.

Working to solve

intimate partner violence.

Table 2: Example of coding for open coding

The sample open coding in table 2 illustrates, initial coding step, the researcher coded every data possible. While coding the data, questions such as ‘what is this data about?’ which revealed not only the pattern of empowerment method followed but also the causes of violence.

Thus, asking questions at each level of coding assisted the researcher in finding relevant data and emergence of patterns of action and belief system. The core-relevance emerged at selective

coding. The following section explains the procedure of selective coding and emergence of core relevance.

4.3.4 Selective coding

Selective coding followed open coding. At the open coding level, the researcher provides the direction to find relevant data whereas in selective coding process the researcher’s focus is on emerging core variable and delimiting coding to variables which are relevant and sufficiently used in parsimonious theory. In this process, the irrelevant codes are not lost but are of secondary concern to the main concern (Glaser 1978). The delimiting data in selective coding continued until the researcher elaborated and integrated the core category, its properties, and its theoretical connection with other relevant categories (Holton 2007). The table 3 below explains the sample of selective coding.

Questionnaire asked while doing selective coding.

Interview Data Open Coding Selective coding

supervision. She had

Table 3: Example of coding for selective coding for the concept ‘reaching roots’

Table 3 explains the strategy followed while doing selective coding for the CGT. At open coding, the data is fractured to get relevant codes, at the second step the coded relevant data is further fractured to form concepts, properties and related categories. In the table 3, the main concern for the data mentioned above is related category ‘approaching’ category and its property ‘reaching roots’. While coding the data, the main concern for the substantive grounded

theory ‘Assumption as cause’ has emerged. The main concern emerged at every level of coding as cause for experts’ intervention. The procedure of the emergence of core category is discussed in the sub-section of ‘memo’ to explain how writing memos had assisted the researcher in determining the core category ‘way of living’. The next step of coding is theoretical coding which assisted in the integration of various categories and emergence of the substantive grounded theory.

4.3.5 Theoretical coding

Theoretical coding is the last level of the classical grounded theory analysis and one of the most problematic areas for the researcher (Hernandez 2009). The analysis process started with open coding where the researcher starts with an open mind and emergence of a core category. At the second level the researcher focussed on the core category, related categories and relevant properties to form substantive codes whereas at the last stage the researcher identified and used appropriate theoretical codes to achieve an integrated theoretical framework for the overall grounded theory (Holton 2010). As rightly put by Glaser (1978) theoretical codes established a relationship between different patterns and weaved the fractured story back together again (P 72). Theoretical codes provided the researcher clarity of the relationship of substantive codes, integrative scope, broad pictures and a new perspective (Glaser 1978). As theoretical codes are flexible in nature and several theoretical codes may fit the data, it provides the researcher enough space for creative yet logical and grounded integration (Glaser 1978). During the open coding and selective coding, the researcher already has begun delimiting the theory and establishing an interconnection between core categories, substantive codes, and properties through writing memos. Sorting of these memos is an essential step to discover theoretical outlines (Glaser 1978). Theoretical sorting is based on theoretical codes, the theoretical codes are not imposed but they are fit according to similarities, connections and underlying uniformities, the decision about the precise location of a particular

memo is based on the researcher’s logic, creativity (Holton 2010) and understanding of the data.

Out of 23 families of the theoretical coding (Hernandez 2009; Glaser 2005), the researcher integrated the core category, sub core category and related categories with the best suitable theoretical codes such as cause, strategy, intervening condition and basic social process. The finding chapter of the current thesis explains the integrated substantive grounded theory. Sample memos explaining the development of theoretical integration is explained in the appendix of this chapter.

4.3.6 Memos

Memos are the theorizing write-up of ideas about the codes and their relationships;

memos are the way to store ideas of an analyst (Glaser 1978). Memo writing is a constant process, which begins right after the first interview. However, due to time constraints the field worker could write only field notes and not memos. These field notes consist of different ideas that came up during the field work. Such as the researcher used to scribble some important words or sentences from the incidents and used to note down conclusion of the interview. Later these field notes were transferred into memo banks after concluding the field work. The purpose of the memo writing is to capture and preserve the conceptual ideas at every stage of coding (Glaser 1998). The memo bank, consisting of the field notes to theoretical coding not only helps the researcher with an in- depth theoretical understanding of the data but also helps the researcher to grow personally (Glaser 1998). Glaser gave no formal rules for writing memos, it is unique to the researcher and consists of a flow of ideas with total freedom (Glaser 1998). The free flow of ideas is not bound by any specific patterns or grammar rules, but the only purpose is to capture free flows of ideas while coding the data to prevent loss of creative ideas (Glaser 1998). Memo writing assists the researcher to determine the best suitable

theoretical codes which later help the researcher to integrate into a substantive grounded theory (Hernandez 2009; Glaser 2003).

As mentioned earlier, the researcher started drafting memos after completion of the field work, while the field notes were transferred into memo banks at the time of sorting. The initial memos were written into the margins of the data transcription and later written down on index cards. These memos were later referred to in order to generate a connection with different categories and properties to generate the theory. Memo writing provided the researcher an in depth understanding and the researcher’s ability to see the conceptual scope of the data (Glaser 1978). The sample memo for the emergence of the core category is discussed in the memo (CC-11).

Memo- (CC 11)

After confusion of many days I finally presented my theory with a core cateogry ‘intervention’ to

the colleagues in the summer seminar 2016. The core category emerged after discussion with my colleague and the supervisor was “way of living” which is pattern of living as mentioned before. Glaser mentioned peculiarites of core category as the Central to the integration. To find out whether way of living is central at intergration, it is necessary to find out how “way of living” is main concern of the emerged grounded theory. Following questions asked helped me to find out the main concern of the theory

Questions answers Do women get empowered due to expert intervention External factors impacts experts efforts What do external factors impact what? Impact way of living

Consequneces of impact on way of living new way of living

Why new way of living is important It is the foundation for the future way of living The question and answer above established “way of Living” as core category for the theory

Outline 1 Memo for the core-category

The memos were written by the researcher to solve confusion regarding the main concerns of the grounded theory. The confusion was regarding ‘intervention’ as main concern or “way of living”. Thus, to address this confusion the researcher asked various question to

determine the central theme. The sample memo above discussed how “way of living” is central to experts’ intervention and how external factors impact “way of living” and consequences due to intervention and external factors overall. Thus, the memo writing assisted the researcher to get conceptual clarity and integration of various categories to form the theory. The appendix attached to this chapter discusses some more sample memos for the grounded theory. Memo writing is logically interrelated and sequences in memo sorting. The following sub section

determine the central theme. The sample memo above discussed how “way of living” is central to experts’ intervention and how external factors impact “way of living” and consequences due to intervention and external factors overall. Thus, the memo writing assisted the researcher to get conceptual clarity and integration of various categories to form the theory. The appendix attached to this chapter discusses some more sample memos for the grounded theory. Memo writing is logically interrelated and sequences in memo sorting. The following sub section