Grounded Theory (GT) Approach

It is essential in developing innovative educational practices to give ed-ucational stakeholders in the field an opportunity to be heard and to contribute to the design process. The grounded theory approach pro-vided a meaningful tool to start the pedagogical investigations and to move towards identifying the salient features of the PLE. Grounded theory is useful when trying to understand how people interpret their relations to the environment, play and learning activities and peers (see also Hyvönen, 2008). In applying grounded theory, my approach relied most closely on that described in the work of the sociologists Strauss

and Corbin (1990; 1998), in particular as regards the analytic procedures and techniques.

Grounded theory was proposed by Barney Glaser and Anselm Strauss (1967); they were the first scholars to articulate qualitative grounded theory strategies and advocate developing theories from re-search grounded in data rather than deducing testable hypotheses from existing theories. In this type of inquiry, the researcher generates an ab-stract analytical schema of a phenomenon, a theory that explains some action, interaction, or process (Creswell, 2007). By ‘grounded theory’

Strauss and Corbin (1990; 1998) mean a theory derived from data that are systematically gathered and analyzed through the research process.

The authors (1998, p. 12) describe it as follows:

In this method, data collection, analysis, and eventual theory stand in close relationship to one another. A researcher does not begin a project with a preconceived theory in mind (unless his or her purpose is to elaborate and extend existing theory). Rather, the researcher begins with an area of study and allows the theory to emerge from the data. Theory derived from data is more likely to resemble the “reality” than is theory derived by putting together a series of concepts based on experience or solely through specula-tion (how one thinks things ought to work). Grounded theories, because they are drawn from data, are likely to offer insight, en-hance understanding, and provide a meaningful guide to action.

With reference to the quotation above, my aim in Studies I and II was to derive a theory from data that would likely to account for children’s real-ity, that is, children’s views and ways of doing things. The GT approach is considered well suited for pedagogical and educational research, be-cause it has been created primarily for examining social activities and interactions (Studies I and II) and for describing individual experiences and meanings (Study III) (Martikainen & Haverinen, 2004).

According to Strauss and Corbin (1990; 1998), grounded theory analysis is composed of three major types of coding: a) open coding, b) axial coding, and c) selective coding. These phases have been included in Studies I, II and III. However, the authors emphasize that the boundar-ies between each type of coding are “artificial” and the different types do not necessarily take place in stages (1990, p. 58). Grounded theory coding can be considered flexible: if a researcher wishes, he or she can return to the data and make a fresh coding (Charmaz, 2006). This was a principal consideration in Study II.

In the open coding process, categories of information were formed and several properties, or subcategories, were identified. After that, the data were put back together and assembled in new ways by making con-nections between categories. In selective coding, the researcher often de-velops the theoretical model that best explains the subject under study.

The following table (Table 3) describes the objects of the studies and the conclusions, that is, the theoretical contributions of the studies in terms of the GT approach:

Table 3. The studies and the contributions based on the GT approach.

Grounded theory

studies Object of study Focus of analysis Theoretical contributions

Study III

One characteristic feature of the GT approach is that formal theory is defined in the study applying it (e.g. Martikainen & Haverinen, 2004).

As Table 3 shows, the studies have produced a richer understanding and richer models of:

1. emotion-rich play environments,

2. narrative thinking in creative collaboration, and

3. the Broadening and Engaging Learning Environment (BELE).

As noted above, in the GT approach the researcher analyzes and codes data in stages to reach the phase where he or she chooses a central cat-egory, such as the BELE, and posits theoretical accounts of the data.

However, rather than seeing theoretical accounts as static constructions, researchers using GT work more with models that will achieve the sta-tus of theories after they have been tested in subsequent and related studies (Glaser & Strauss, 1967).

According to Charmaz (2006), the researcher creates the codes by defining what he or she sees in the data. She points out that coding is an emergent process in which unexpected ideas emerge and language plays a crucial role in how and what to code. Having children as infor-mants has provided me as a researcher with a fascinating opportunity to deal with the rich and imaginative language of children and to use in the studies the vivid expressions they suggest. Additionally, creativity on the part of the researcher in data coding and analyses is a vital compo-nent of GT (Glaser, 1978; Strauss & Corbin, 1998). Coding can be seen

partly as work, but, according to Charmaz (2006), it is also play with data. The researcher plays with the ideas he or she gains from the data.

He or she becomes involved with the data and learns from it. Cod-ing offers a focused way of viewCod-ing data. Through codCod-ing the researcher makes discoveries and gains a deeper understanding of the empirical re-ality. In this respect, playfulness once again becomes a central factor:

Charmaz points out how theoretical playfulness allows the researcher to try out ideas and to see where they might lead (see Charmaz, 2006). In this study, the “playfulness” required by the theory seemed to fit this account: the categories are based on children’s responses, primarily their creativity, imagination and word-play (Studies I and III).

For analyzing the data, I used NVivo, qualitative coding and analyz-ing software designed in keepanalyz-ing with grounded theory, which proved to be an appropriate tool for managing the data in the first phases. Af-ter transcribing the research maAf-terial – discussions (study I and II) and handwritten texts (study III) – I imported the data into the software and segmented it in order to identify concepts and categories, as well as their properties and dimensions (see Strauss & Corbin, 1990). This first step produced the basic categories, such as schoolyard, nature, learning activities, fantasy and adventure. Among the questions I asked during this initial coding were the following: (cf. Glaser, 1978; Charmaz, 2006):

• What do the data reveal and suggest with regard to the PLE and learning activities?

• What theoretical category do these specific data indicate?

The openness of initial coding stimulated my thinking and allowed new ideas to emerge (cf. Charmaz, 2006; Strauss & Corbin, 1998). In the first studies¹⁸ (I and II), coding was carried out in our research team and

18. Study I: data analysis in collaboration with Pirkko Hyvönen (using NVivo

the data were coded separately, then compared and combined. In addi-tion to increasing validity, such researcher triangulaaddi-tion facilitates the emergence and elaboration of different theoretical terms and concepts before final categories and a theoretical model are defined that best ex-plain the focal phenomenon.