Help Drafting Assignment

QualitativeDataWK3Reading.pdf

Home >Law homework help >Help Drafting Assignment

contents 157

Qualitative Data

15 Many of us in the social sciences come from arts rather than science backgrounds and find statistics difficult. Hence, we tend to look to qualitative data as the core of our research. Qualitative data is information that is represented usually as words, not numbers. If you have pages of text before you, recordings of interviews or notes from observation, for all practical purposes you have qualitative data. Nevertheless, words must be analysed as carefully as numbers. Naturalistic inquiry does not guarantee the meaning of your research any more than statistics guarantee its rigour.

In researching people’s subjective perceptions, we build up scientific knowledge about their personal knowledge by objectifying their perceptions systematically. However, the actual perceptions themselves do not suddenly become scientific by virtue of having been studied scientifically. Just because our research gives us a better idea of what some people think (and maybe, gives us an emotional attachment to these people and their ideas), it neither suddenly changes the nature of their thoughts nor makes them more important than other people’s. Different groups have ethical, social and political rights in consonance with their own culturally meaningful sets of social constructs. However, this does not necessarily mean that these groups have professionally or scientifically informed views, or that their constructs are valid beyond their own group.

Carefully done qualitative research is just as demanding as other research. However, it is easy to do badly, by being intellectually lazy and hiding the fact from yourself and others. The rules of the game are not as transparent as they are in quantitative research. The result can be a weak report that has neither technical nor intellectual rigour. Rather, research reports must demonstrate a careful systematic approach to data analysis.

This chapter will:

1. review the key principles about qualitative data established earlier in this book, including the key hierarchy for presenting data;

Co py ri

gh t © 2 01 0. S ag e Pu bl ic at io ns P vt . Lt d. A ll r ig ht s re se rv ed . Ma y no t be r ep ro du ce d in a ny f or m wi th ou t pe rm is si on f ro m th e pu bl is he r, e xc

ep t fa ir u se s pe rm it te d un de r

U. S. o

r ap pl ic ab le c op yr ig ht l aw .

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 1/22/2022 2:32 AM via PURDUE UNIVERSITY GLOBAL AN: 340339 ; Gerard Guthrie.; Basic Research Methods : An Entry to Social Science Research Account: ns019078

158 Basic ReseaRch Methods

2. show how to present observational and unstructured interview data in more detail; and

3. outline some basic techniques for manual and computer analysis of short open- ended questions.

15.1 Qualitative data principles

First, we need to revise some of the underlying principles that apply to qualitative data:

1. Words are data that express qualities and attributes (Chapter 14). 2. Words particularly come from available sources (Chapter 9), naturalistic

observation (Chapter 10), unstructured interviews (Chapter 11) and open-ended questions in questionnaires (Chapter 12).

3. Words are most often classified on the nominal measurement scale (Table 14.1).

Although much qualitative research does not analyse data formally using measurement concepts, understanding the principles will often help solve problems that arise during analysis. Does something not seem logical? You are possibly confused about the measurement scale or maybe, the underlying semantic differential is not a polar opposite. Does a paragraph seem jumbled? Are you not sure about where some material belongs? Chances are that you are mixing up more than one variable.

A hierarchy drawn from Bloom’s Taxonomy helps present data clearly. First, describe; then analyse; and later, draw conclusions or interpret.

1. Describe

(a) Write out the ‘facts’ of the situation observed or heard about in open-ended interviews. This should be clear descriptive reporting, free of adjectival colour.

(b) Do not present everything. Filter out those matters which are not relevant to the research problem and themes.

2. Classify

(a) Group the material to identify similarities and differences in the data. (b) Break paragraphs when they start getting complex so that there is one main

idea in each paragraph. (c) Use more headings if you have trouble categorising the literature ap-

propriately. Rather than making the headings more abstract, make them more concrete.

Co py ri

gh t © 2 01 0. S ag e Pu bl ic at io ns P vt . Lt d. A ll r ig ht s re se rv ed . Ma y no t be r ep ro du ce d in a ny f or m wi th ou t pe rm is si on f ro m th e pu bl is he r, e xc

ep t fa ir u se s pe rm it te d un de r

U. S. o

r ap pl ic ab le c op yr ig ht l aw .

qualitative data 159

3. Interpret

(a) Present your interpretation of the material separately. Pick out key features that identify patterns and keep your mind open to new ideas that arise from the data.

(b) You do not need to be highly conceptual here—that can come later in the concluding chapter with a wider analysis of the findings in light of the literature review.

The aim should be reporting that is clear and analytical. It should work from the data as recorded in observation and interviews to the analysis generated by the literature review, the fieldwork and interviews.

15.2 Presenting available and observation data

How do we present unstructured qualitative data gained from available sources and field observation? This can be confusing at first because now you have large chunks of text, usually written in a note pad. Once writing begins, it is very easy to start com- mentating, and quickly there is a problem. What is observation or description and what is commentary or interpretation? The answer to the problem remains, first, describe, then analyse and, later, draw conclusions or interpret.

Computers can be used to analyse this data, but for decades, qualitative researchers survived without them. Some manual techniques are quite usable, especially when there are large volumes of text. Computer processing will be faster than manual processing once the text is set up, but the trade-off is that learning to use specialist computer packages and setting up text analysis functions might take more time than you save in processing. For small amounts of data, you can transcribe from your notes into a word processor.

In writing up the material, you have two main options:

1. Narrate it as a chronological story, which is usually the most straightforward for both writer and reader.

2. Analyse it systematically, which will add insight and academic value.

Examples of chronological narration have been given already in Box 9.1 (based on available documentary data about school inspections classified under appropriate headings) and Box 11.5 (using observation and interview about highway crime). You can revisit those examples, so now we will focus on systematic analysis.

The basis for systematic analysis is issues or themes identified in the literature review or grounded in the data. You can either work straight from a note pad (which

Co py ri

gh t © 2 01 0. S ag e Pu bl ic at io ns P vt . Lt d. A ll r ig ht s re se rv ed . Ma y no t be r ep ro du ce d in a ny f or m wi th ou t pe rm is si on f ro m th e pu bl is he r, e xc

ep t fa ir u se s pe rm it te d un de r

U. S. o

r ap pl ic ab le c op yr ig ht l aw .

160 Basic ReseaRch Methods

I did with the 18 interviews in the highway study) or transcribe the material into a word processing document, which you should do as soon as possible after collecting it. Entering the whole text into the computer is time consuming, but the entry process leads to revision of the material and ignites your thought process about patterns and connections in the data. Then the following steps apply:

1. If a note pad is the source, strike out material in the pad as it is copied. If you have typed up the material, make a backup document and use another copy as a source for cutting and pasting. In both cases, keep coming back to the remaining material.

2. In a new document, insert headings using key themes from the literature review. Use them as destinations to copy different phrases, sentences or paragraphs from the notes or document.

3. If using more than one period of observation or more than one interview as source material, keep a reference to it with each item that is entered. When you analyse the data, you can comment on who said what and how their different perspectives might illuminate the topic under discussion.

4. If source material straddles more than one heading or does not fit under an existing heading, consider making up a new one.

5. If something still does not fit, perhaps it is of minor importance and can be omitted, but do not leap too quickly to this conclusion. The minor elements might actually be important and the problem is that you have not woken up to this.

To clarify the data, follow the describe/classify/interpret steps given in the previous sub-section. Box 15.1 shows an example, being a summary of two days observing a school inspection visit during which I followed the inspector and took detailed notes. The first part, taken from eight pages of the original report, is a straight chronological description of the inspector’s day and his interactions in the school. However, the description reports only his activities and filters out other aspects of the school that were secondary to this (for example, details of lessons observed). The second part is classification of the events to show themes, being key summary points cross-referenced back to parts of the observation. Broader academic and professional interpretation about the nature of inspections in a formalistic educa- tion system was saved for the concluding chapter.

15.3 Presenting open-ended interview data

The techniques for presenting long pieces of data from unstructured interviews are similar to unstructured observation. Once you have notes or text, you can describe, classify and interpret it using the five steps listed earlier. Co

py ri

gh t © 2 01 0. S ag e Pu bl ic at io ns P vt . Lt d. A ll r ig ht s re se rv ed . Ma y no t be r ep ro du ce d in a ny f or m wi th ou t pe rm is si on f ro m th e pu bl is he r, e xc

ep t fa ir u se s pe rm it te d un de r

U. S. o

r ap pl ic ab le c op yr ig ht l aw .

qualitative data 161

Box 15.1 Presenting oBservation data

A School Inspection Visit

An Advisory Visit

‘7.50am. The Inspector arrived at the school, met the Headmaster and introduced the researcher. Arrangements for the day were discussed and the timetable checked.

‘8.40. A second teacher interview began...The Inspector went through the eight Inspection Report headings…then the Inspector went over a lesson observed the previous day…

‘10.30. The Inspector and the researcher walked down to the Social Science office to interview the Subject Master of Expressive Arts and Social Science. The Inspector first checked teachers’ timetables and whether there was an Expressive Arts syllabus, a problem raised by the third interview. He then checked how many subject meetings had been held…[and] checked the availability of Expressive Arts materials…He then turned to Social Science duties…

‘8.50 [the next morning]. A grade 8 Agriculture class lesson on picking coffee cherries was observed until 9.10 when note-taking from the blackboard began…

‘11.10 The staff meeting now started … The Inspector said that there would be ten points in his address…The first point concerned assessment…The Inspector’s next point covered the value of applying for in-service courses…For his fifth point, the Inspector discussed school maintenance…and praised the improvement in classroom displays…

Several features were apparent from this visit. ‘The first feature apparent was the volume of work covered. The first day spanned about

10 hours, the second 9½ hours… ‘A second and constant feature of the two days of the visit was an emphasis on quality, broadly

defined: the “tone” of the school, the need for thorough preparation and planning, the need to challenge and extend students…

‘A third feature was the close interrelationship between the three inspectorial roles [advice, evaluation and administration]. Within the space of a few minutes in an interview each role was frequently encompassed…Although it is easy in principle to draw a distinction between advising and evaluating, in practice it is somewhat more difficult…

‘The intermixing of roles parallels a fourth feature of the visit: the Inspector was the system-defined expert on everything…This was particularly evident in the meeting held with all the staff…

‘Finally, a seventh feature of the visit was the thoroughness of the investigations and the thoroughness of the cross-checking…’

Source: Adapted from Guthrie (1983b: 19–27).

Box 15.2 has reporting from an investigation of science teachers’ experience in the Philippines. It reports a trainee teacher’s narrative about the difficulty she had in relying only on a textbook that contained information contradicted by local knowledge. The article from which this story is taken then analysed the situation in more detail, bringing to bear further narrative about the teacher’s life experi- ences to show how they provided meaning for her teaching career and how that

Co py ri

gh t © 2 01 0. S ag e Pu bl ic at io ns P vt . Lt d. A ll r ig ht s re se rv ed . Ma y no t be r ep ro du ce d in a ny f or m wi th ou t pe rm is si on f ro m th e pu bl is he r, e xc

ep t fa ir u se s pe rm it te d un de r

U. S. o

r ap pl ic ab le c op yr ig ht l aw .

162 Basic ReseaRch Methods

fitted in with her social background. The researchers also commented on the role of English-language textbooks that might contain information conflicting with indigenous knowledge. These layers gave greater depth to the study.

15.4 Computer analysis of text

We now turn to an analysis of short pieces of data from open-ended questions and semi-structured interviews. The individual answers might be short but there are likely to be a large number of them so that manual coding is a burden. Computing is the path to follow.

Box 15.2 rePorting unstructured interviews

A Teacher’s Experience

‘I’m a student teacher of grade 1. As a teacher in science, there are times when unexpected situations will occur inside the classroom which create dilemmas. My dilemma is not really a big one, but when you look at it deeply, such a dilemma can create a serious situation that is hard to deal with in science teaching.

‘One day, I taught a lesson about places where plants grow. I first presented places where specific plants grow—in soil, water, air, wet and dry places. The plants I used as examples were taken from the science book that we were using in class. The pupils were confused about whether the water and wet places were the same? In real life situation there are plants that can grow in both places—in wet and aquatic places. On a test, I asked students to list two examples of plants that grow in soil, water, air, dry and wet places. One child wrote “Kangkong” plant under “wet” places; I marked it wrong because the book implied that Kangkong would be an “aquatic” plant since it grows in water.

‘The mother of one of our pupils came to the school. She had a correction for an answer her child had written that I had marked wrong. The mother protested saying that Kangkong can grow also in wet places. Because I followed what is written in the book, I marked it wrong. Besides, most of my pupils believed that the book is the source of knowledge. So, if I checked or accepted other answers that are not found in the book pupils will conclude that books can’t be trusted.

‘Actually, I believe that books are not the only source of knowledge. You can gain knowledge from other people, and also from real life situations. Answers that can be found in the book are also correct, but they are limited in the sense that other answers can also be found in other books.

‘My problem is: Am I going to stick to the book or consider other answers which are based on real life situation? What will happen if I stick or depend only on the book? What should I do to help my pupils understand our lesson?’

Source: Arellano et al. (2001: 214–15).

Co py ri

gh t © 2 01 0. S ag e Pu bl ic at io ns P vt . Lt d. A ll r ig ht s re se rv ed . Ma y no t be r ep ro du ce d in a ny f or m wi th ou t pe rm is si on f ro m th e pu bl is he r, e xc

ep t fa ir u se s pe rm it te d un de r

U. S. o

r ap pl ic ab le c op yr ig ht l aw .

qualitative data 163

Your choices are to use whichever package is available on your computer system, buy a programme and obtain manuals and textbooks that teach how to use it, or use a spreadsheet. Several software packages are available for text analysis with large volumes of data, for example, ATLAS.ti, HyperRESEARCH and NVivo. Computer packages can allow audio, video and photo analysis as well, which are beyond the scope of this book. In general, the more powerful and flexible a package, the more time required to learn how to programme it. This can divert time from the primary purpose, which is the research report.

Instead, you can use word processing and spreadsheet packages with which you are already familiar. They are not as good as the specialist packages for advanced analysis, but they can still go a long way. Despite being primarily for quantitative analysis, Excel and other spreadsheets have basic functions that allow text analysis too. This section gives you some procedures for searching individual questions using Excel and Word 2003. Despite being elementary, they were sufficient for the text analysis in the 16 crime surveys and they will suffice for many other projects.

First, enter each item of text into a spreadsheet cell. The shorter the text units, the easier this is. With a structured or semi-structured questionnaire, column head- ings should be the question numbers that contain open-ended responses. Row names should be the questionnaire code numbers. If there are 10 open-ended questions and 50 interviews, there will be 500 cells, but ‘no answers’ on questionnaires will mean that many cells do not contain text. Those with text will usually contain only a few sentences of comment from each interviewee, which is quite manageable with this process.

The next steps involve using the count functions to tally numbers, word search functions to list cells that contain key words and pasting the table text from the spreadsheet to the word processor document. Table 15.1 gives procedures for pasting answers to individual questions direct into the word processor. There you can further cut and paste the items to give the data a logical flow.

Table 15.2 uses a commentary to illustrate further the procedures. This example comes from the 2006 crime surveys in Bougainville, reporting on people’s comments on reasons for changes in crime levels that were found in responses to an open-ended question. The result was a mixture of descriptive quantitative data (the number and frequency of responses) and qualitative comments in respondents’ own words.

You can also use the Excel ‘Edit’ > ‘Find All’ function to search the whole spreadsheet for a key word that might have come up in answers to different questions. However, the copy function for the search results is extremely cumbersome, so you have to do a manual selection of individual quotes from the search results in Excel and copy them one by one into the categories in the document.

Co py ri

gh t © 2 01 0. S ag e Pu bl ic at io ns P vt . Lt d. A ll r ig ht s re se rv ed . Ma y no t be r ep ro du ce d in a ny f or m wi th ou t pe rm is si on f ro m th e pu bl is he r, e xc

ep t fa ir u se s pe rm it te d un de r

U. S. o

r ap pl ic ab le c op yr ig ht l aw .

164 Basic ReseaRch Methods

15.5 Summary

Qualitative data is information that is usually represented as words, not numbers. Both must be analysed carefully.

Qualitative data principles

1. Words are text data that express qualities and attributes. 2. This type of data comes particularly from available sources, naturalistic

observation, unstructured interviews and open-ended questions. 3. Words are usually classified on the nominal measurement scale. 4. To help present text clearly, follow a hierarchy drawn from Bloom’s Taxonomy:

describe first, then classify and later interpret.

Presenting available and observation data

1. Some manual techniques are quite practical, especially for small volumes of text. 2. In writing up the material, the two main options are narration as a chronological

story or systematic analysis.

Table 15.1 Text analysis guidance using spreadsheets

Task Operations

Enter text in spreadsheet Columns = questions, rows = questionnaire IDs, cells = responses.

Add number of responses COUNTA(…) underneath first column of data > ‘Copy’ formula to other columns.

Copy responses to Word and format

Highlight all data cells in a column > ‘Copy’ > ‘Paste’ cells in word processor > set ‘Paste Options’ drop-down box to ‘Keep Text Only’ > format to document quotation style, e.g., bullet points, indented, italics.

Classify text Inspect text > cut and paste like items to adjacent lines > count number of items in each group and calculate percentage of COUNTA total > cut and paste groups in order from largest to smallest > the final classification is ‘Other’ for items that do not fit previous categories.

Comment on classifications

Write brief commentary describing each classification, including number and percentage of total cells (from COUNTA).

Delete extraneous text ‘Delete’ repetitive items. Retain number of comments in each classification in proportion to percentage of total.

Source: Author.

Co py ri

gh t © 2 01 0. S ag e Pu bl ic at io ns P vt . Lt d. A ll r ig ht s re se rv ed . Ma y no t be r ep ro du ce d in a ny f or m wi th ou t pe rm is si on f ro m th e pu bl is he r, e xc

ep t fa ir u se s pe rm it te d un de r

U. S. o

r ap pl ic ab le c op yr ig ht l aw .

qualitative data 165

Table 15.2 Annotated text analysis

Text Commentary

‘Open-ended responses to Q.2.2 expanded the reasons for the changes in crime levels that were believed to have occurred…In Buka, 36 comments were made. The largest number (47% or 17) was on reasons for peacefulness, for example:

• As soon as a problem arises, the community holds meetings to solve it.

• People are behaving. • This is a mission area and people respect it. • The new Task Force is doing its duties.

‘Two of this group thought alcohol was under better control, for example:

• No drinking in public is helping a lot.

‘Of the comments on problems, 17% (6) related to alcohol, for example:

• Drunkards are getting worse.

‘Five comments (14%) related to youth, for example:

• A lot of youths don’t have a job. • Youths are still causing troubles.

‘Two commented on guns still being in the community.’

1. The open-ended responses were typed from the questionnaires into the spreadsheet, taking similar time as any other package.

2. Q2.1 asked, ‘Do you think the level of crime in your area has changed since the last survey 12 months ago?’. Q.2.2 was a followup, ‘Why’.

3. Searching Q.2.2 following the process in Table 15.1 found 36 comments using the COUNTA total.

4. The responses column was copied to the text document and formatted. Blank lines were deleted. The remaining responses were categorised manually into two small groups (reasons for peacefulness, including alcohol-related) and problems (including alcohol, youth and guns).

5. The groups of comments were put in order from the largest with 17 to the smallest with 2. Repetitive comments were deleted.

7. Linking text was written at the start of each group.

8. Later, in the conclusions chapter, themes were taken up about community responses to crime and the perceived influences of alcohol, unemployment and guns.

Source: Adapted from Guthrie et al. (2007: 22).

Presenting open-ended interview data

The steps for presenting data from unstructured interviews are similar to observation.

Computer analysis of text

Large volumes of short text from open-ended questions and semi-structured interviews usually require computing. Techniques are provided for systematic analysis using spreadsheets.C op yr

ig ht ©

2 01 0. S ag e Pu bl ic at io ns P vt . Lt d. A ll r ig ht s re se rv ed . Ma y no t be r ep ro du ce d in a ny f or m wi th ou t pe rm is si on f ro m th e pu bl is he r, e xc

ep t fa ir u se s pe rm it te d un de r

U. S. o

r ap pl ic ab le c op yr ig ht l aw .

166 Basic ReseaRch Methods

Analysis of qualitative data is deceptive. Just because the data are words does not mean that they would fall easily into place. Text from naturalistic observation and interviews does not automatically provide meaningful results. Research reports must demonstrate a careful and systematic approach to analysis so that the report has both technical and intellectual rigour. Do this well, however, and you should end up with an interesting, meaningful and accessible report that is a pleasure to read.

15.6 Annotated references

Babbie, E. (2007). The Practice of Social Research, 11th edition. Belmont: Wadsworth. This sociology text has chapters on both qualitative and quantitative data analysis.

Best, J. and J. Kahn. (2005). Research in Education, 10th edition. Needham Heights: Allyn & Bacon. There is plenty of material in this book on qualitative research and data analysis.

Scheyvens, R. and D. Storey (eds). (2003). Development Fieldwork: A Practical Guide. London: Sage. A comprehensive collection on fieldwork in developing countries containing chapters

on both quantitative and qualitative research.

Vulliamy, G., K. Lewin and D. Stephens. (1990). Doing Educational Research in Developing Countries: Qualitative Strategies. London: Falmer. This book is heavily based on practical research experience using qualitative

approaches in Malaysia, Papua New Guinea, Sri Lanka and Nigeria, including extensive discussion of data analysis.

Co py ri

gh t © 2 01 0. S ag e Pu bl ic at io ns P vt . Lt d. A ll r ig ht s re se rv ed . Ma y no t be r ep ro du ce d in a ny f or m wi th ou t pe rm is si on f ro m th e pu bl is he r, e xc

ep t fa ir u se s pe rm it te d un de r

U. S. o

r ap pl ic ab le c op yr ig ht l aw .