Segmentation

Being a follow-up step to the creation of the coding frame, segmentation means dividing ʿmaterial into smaller units so that one unit fits exactly one subcategory within a dimension⁶⁰ʾ (Schreier 2012: 134). These smaller units are called units of coding. They are ʿthose parts of […] material that can be interpreted in a meaningful way with respect to […] categoriesʾ (Ibid.: 132). The proces of segmentation consists of three stages (Ibid.: 139):

1) marking the relevant parts of your material;

2) deciding on your criterion of segmentation;

57 The codes refer to the corresponding codes in Appendix 1.

58 In Appendix 1, each unit coded for Practice is accompanied by a paraphrase to better illustrate the reasons for assigning it to this category

59 No paraphrase is provided for [M_CP_1]. Since this unit is rather long, it was difficult to come up with a paraphrase that would sum up the main idea of the practical suggestion expressed in this unit. However, since the unit provides practical examples of projects that contribute to developing studentsʾ intercultural awareness, the unit is considered as being practical, which is why it was subsumed under Culture (practice).

60 This is in line with the requirement of mutual exclusiveness (see Schreier 2012: 75), which dictates that each segment can be assigned to only one subcategory within a dimension.

44

3) marking your units of coding.

Below, I will describe how the above stages relate to my research:

1) I first marked the relevant parts of my material. I did this by applying my research questions, as well as my coding frame, to the textual material, i.e. the four analyzed publications. If a particular stretch of text corresponded with the main category, and the subcategories, of my coding frame, I marked it as relevant. Only the ʿprimaryʾ⁶¹ textual material was used for the purpose of the research.

2) There are two ways of approaching segmentation, i.e. using a formal or thematic criterion (see Ibid.: 134-138). A formal criterion is useful if the material has ʿan inherent structureʾ (Ibid.: 136). If the material does not have an inherent structure, a thematic criterion is more appropriate. In such cases, it is important to keep the coding frame in mind when segmenting the material. The four analyzed publications are clearly structured in different ways. As such, a formal criterion alone was not appropriate. Nevertheless, there are some similarities among the analyzed publications. These lie in the way that the individual chapters are structured.

In all publications, chapters are divided into a number of shorter sections. In Scrivenerʾs Learning Teaching and Urʾs A Course in English Language Teaching, each chapter is divided into numbered subchapters, which are further divided into shorter sections introduced by headings⁶². In McDonough et al.ʾs Materials and Methods in ELT, chapters are also divided into numbered subchapters, but no shorter sections introduced by headings are present. As regards Watkinsʾs Learning to Teach English, chapters are also divided into shorter sections introduced by headings, but these are not numbered. On the whole, each publication contains sections introduced by headings. Although the functions of these headings differ in the individual

publications, and even within the publications, i.e. they may introduce numbered

61 The term primary material refers to all textual material minus certain publication specific aspects of the material; In Urʾs A Course in English Language Teaching, sections entitled ʿOverviewʾ, ʿTaskʾ and ʿReviewʾ were not used for the purposes of the analysis as they did not bring any new information; In Scrivenerʾs Learning Teaching, on the other hand, sections entitled ʿTaskʾ and ʿCommentaryʾ were used for the purpose of the analysis as they contained new information; In McDonough et al.ʾs Materials and Methods in ELT, sections providing practical tasks did not contain any new information, and as such, they were not used; The same can be said about Watkinsʾs Learning to Teach English, which contained sections entitled ʿTry it out!ʾ and ʿSummaryʾ.

62 Ur makes a frequent use of numbered bullet points followed by an introductory statement in the form of a heading. These bullet points, however, do not count as headings because they do not serve to introduce a new section (see e.g. [U_CT_1]).

45

subchapters, or merely shorter sections within subchapters, the resultant structure is somewhat similar.

As such, I decided to make use of this structure and consider it when creating the segmentation rules, thus using a formal criterion. At the same time, I also considered the content of the material, i.e. the thematic aspect, thus using a thematic criterion.

What follows is a list of the segmentation rules reflecting both approaches, i.e. formal and thematic.

a) The paragraph was chosen as the basic unit, i.e. an entire paragraph is marked as a unit of coding even if only part of it matches the selected category.

b) Since the analysis is qualitative, and the nature of the material makes it problematic to determine the ideal length of a unit of coding, the individual units vary in length.

c) The first rule, i.e. marking the entire paragraph as a unit of coding, does not apply if two different categories are identified in one paragraph. In such cases, the paragraph is included in two separate units, and the part of the paragraph matching the other category is underlined, and

enclosed in paratheses, as in [U_PR_7] and [U_RE_5].

d) If there is a longer stretch of text matching one category, and it runs over several paragraphs, the entire stretch of text is marked as one unit of coding, i.e. the unit of coding consists of more than one paragraph, as in [S_PR_1].

e) In order to limit the length of the individual units of coding,

segmenting respects the boundaries created by the division of the text into sections. In other words, if there is a section heading, the text following the heading is marked as a new unit of coding⁶³, as in [U_PR_7] and [U_PR_8].

f) Since a unit of coding does not need to be continuous (as long as it respects the above-mentioned boundaries), there are cases where part of

63 This boundary is not respected as far as U_CP_2 is concerned. Here, the two stretches of text (on p. 204 and pp. 207-208) are explicitly interconnected (thematically), therefore it would not make much sense to treat them as two separate units.

46

text had to be ellipted (for reasons of space). In such cases, the ellipted part is marked by the following symbol: ʿ….ʾ, as in [U_CP_2].

g) Every unit of coding is introduced by the heading of the section under which it is subsumed (to provide context, and make orientation easier).

Where the heading does not immediately precede the paragraph.

containing the unit of coding, the text in between the two is ellipted and marked by the ʿ….ʾ symbol, as in [W_CT_1].

3) I then had another look at the parts of material that I had marked as relevant and segmented them following the above segmentation rules.