Data analysis – using computers for presentation

MEMBERS

Lim Pau LinZiana binti Hj Mahmud

Hjh Normaslinah binti Hj Mohd Noor/AsriNur Adyani binti Hj Yusop

Nuwairani Azyyati binti Hj Sahidi Abdul Wafi Sia bin Abdullah Sia

DATA ANALYSIS – USING COMPUTERSGROUP X

CONTENTSTypes of softwareFunctionsSelection of softwareResearch articlesValidity and reliability issuesReferencesQuestions

GENERAL TYPES OF SOFTWARE

Word processors: designed for production and revision of text and helpful for transcribing, writing up or editing field notes, etc

Word retrievers: Specialise in finding all of the instances of words, phrases, and combinations of these you are interested in locating. Some have content-analytic capabilities

Text based managers: Organise text more systematically for search and retrieval. They search for and retrieve various combinations of words, phrases, coded segments, memos, or other materials

Code-and-retrieve programs: Software developed specifically by qualitative researchers – helps divide texts into segments or chunks, attach codes to the chunks, and find and display all instances of coded chunks (or combination of chunks)

Theory builders: Also researcher-developed. Usually include code-and-retrieve capabilities but also allow you to make connections between codes, to develop higher-order classifications and categories, to formulate prepositions or assertions. Organised around a system of rules or based on a formal logic.

Conceptual network builders: Helps you to build and test theory but you work with systematically-built graphic networks. You can see your variables as nodes, linked with other nodes by specific relationships. Networks are not just casually hand-drawn but are real ‘semantic networks’ that develop from your data and the relationships you see among them.

WHAT FUNCTIONS TO LOOK FOR?

Coding Memoing/annotation Data linking Search and retrieval Conceptual/theory development Data display Graphics editing Other things to consider: flexibility and

user friendliness

WHICH IS THE ‘BEST’ CAQDAS PACKAGE…?

This is perhaps the most frequently asked question we receive – however, it is impossible to answer! As such, they each have their own advantages and disadvantages.

Software packages do not provide you with a methodological or analytic framework.

The tools available may support certain tasks differently However, as the researcher you should remain in control of the

interpretive process and decide which of the available tools within a software can facilitate your approach to analysis most effectively.

Thinking about and using CAQDAS software should not necessarily be any different from other types of software package – just because a tool or function is available to you does not mean you will need or have to use it.

You therefore need to think clearly about what it is you are looking to the software to help you with. Do not choose a package simply because it seems the most sophisticated.

SOME GENERAL QUESTIONS TO ASK WHEN CHOOSING A CAQDAS PACKAGE

• What kind(s) and amount of data do you have, and how do you want to handle it?

• What is your theoretical approach to analysis and how well developed is it at the outset?

• Do you have a well defined methodology? • Do you want a simple to use software which will

mainly help you manage your thinking and thematic coding?

• Are you more concerned with the language, the terminology used in the data, the comparison and occurrence of words and phrases across cases or between different variables ?

• Do you want both thematic and quantitative content information from the data?

• Do you want a multiplicity of tools (not quite so simple) enabling many ways of handling and interrogating data?

• How much time do you have to ‘learn’ the software?

• How much analysis time has been built into the project?

• Are you working individually on the project or as part of a team?

• Is there a package – and peer support – already available at your institution or place of work?

LIST OF CAQDAS SOFTWARE Atlas.ti HyperRESEARCH MAXqda(&MAXdictio) N6 Nvivo Qualrus QDA Miner Code-A-Text Dedoose The Ethnograph HyperResearch Kwalitan Qualifiers

TAMS Analyser

Transana webQDA Weft QDA & many more

Specialised software is not free. Free demos are provided for 30-days for some. Price range from US$150 – US$1200 depending on academic, institutional or professional users. Some require you to email them for pricing.

RESEARCH ARTICLESTitle Purpose Data

Analysis

Research 1 Simplifying Qualitative Data Analysis Using General Purpose software tools

Show how MS Word can be used for coding, retrieval and other qualitative analysis functions

MS Word

Research 2 Using Nvivo to Analyse Qualitative Classroom Data on Constructivist Learning Environments

To find out what successful teachers who use CLE, especially authentic learning, do in the classroom and how students behave in such context.

NVivo

Research 3 Teacher leadership in (in action)

Qualrus

Title Purpose Data Analysis

Research 4

Understanding mothers’ experiences of infant daycare: a new approach using computer-assisted analysis of qualitative data

Small scale study on Australian mothers' experiences of infant day care

NUD*IST

Research 5

Learning, Beliefs, and Products: Students' Perspectives with Project-based Learning

To find out the student perspective in project based learningTo explore how learners created projects and how they chose to complete the learning tasks

QSR N6 (NUD*IST)

Rssearch 6

SIMPLIFYING QUALITATIVE DATA ANALYSIS USING GENERAL PURPOSE SOFTWARE TOOLS

Analyse text from key informant interviews, focus groups, document reviews and open ended survey questions.

Used word functions such as table, table sort, insert file, find/replace, and insert comment

PROCESS FOR USING WORD FOR CODING AND RETRIEVAL OF QUALITATIVE DATA

Step 1

•Formatting interview data into tables

Step 2

•Develop a theme codebook

Step 3

•Add columns and codes to capture face sheet data

Step 4

•Coding text rows with one or with multiple theme codes

Step 5

•Sorting data tables and finding patterns

Step 6

and 7

•Code validation/correction and merging of data tables

STEP 1: FORMATTING INTERVIEW DATA INTO TABLES

STEP 2: DEVELOP A THEME CODEBOOK

STEP 3: ADD COLUMNS AND CODES TO CAPTURE FACE-SHEET DATA

STEP 4: CODING TEXT ROWS WITH ONE OR WITH MULTIPLE THEME CODES

STEP 5: SORTING DATA TABLES AND FINDING PATTERNS

STEP 6 AND 7:CODE VALIDATION/ CORRECTION AND MERGING OF DATA TABLES

Code validation within one interview data table Once sorted by theme code and sequence number,

analyse all text segment for each code and decide whether all segments are instances of a particular category or if corrections are needed.

Merging data tables Data tables from all interviews or subsets of them

appropriate to one’s study can be merged (e.g. particular gender)

Once merging has occurred, code validation actoss transcripts can be done. To do this, sort by theme code, face-sheet codes such as participant ID and sequence number

Once merged and sorted, analyse all text segments for each code and decide whether the text segments are all instances of a particular category.

EXAMPLE OF USING COLOUR AND INSERT COMMENT

RESEARCH USING NVIVO

Purpose of the study: Sheds light on the ‘nuts and bolts’ of the qualitative

data analysis. Address a more open approach to reporting and help

researchers better understand how NVivo is used in an actual classroom study.

Research Questions Why are the ‘family characteristics’ of constructivist

learning environments in which authentic materials and activities are using regularly?

What are the students’ responses, in terms of their social (interaction with other students and interaction with the teacher), cognitive, emotional/affective, and other learning behaviours – to these kinds of activities and materials?

Data collected through: Classroom Observations Formal and informal interviews (with both

students and teachers) Field Notes Students’ Products Artifacts.

Why Nvivo? Structural design of the software (easy). Nature of the research study. Ease of the searching for relationships. Time saved. Rigor.

USING NVIVO IN THE ANALYSIS OF DIFFERENT DATA SOURCES Interviews and Field Notes:

The transcribed interview data and filed notes were transferred into electronic formats in the early stages of the study. They were only converted from a word format (.doc.extension) into rich text file format (.rtf extension) in order to process them as NVivo document files and use the NVivo’s rich text and visual coding features. After completing these conversions all the interview files as well as field note files were transferred into the NVivo Document Browser.

 Observation Videotapes recorded in the real classrooms for the

observation purpose were transformed from visual and verbal expressions to written text after encoding the transcripts.

Students’ work Most of the student works were kept in student portfolio

folders and files. The written part of the student works was entered as text files using document browser of NVivo, and ready for coding and further analysis.

FINDINGS(WITH REGARDS TO USING THE NVIVO FOR THE STUDY)

NVivo helps tremendously from conceptualization and coding of the data to an entire research report saving time and energy of researchers.

Drawback – need time to explore the software and your computer can crash or you can forget to save your work.

Expensive

DOCUMENT BROWSER

NODE BROWSER

MAIN AND SUB CODES IN NODE BROWSER

EXAMPLE OF CODING

USING CAQDAS :VALIDITY

To enhance validity• Can assist the management of larger

samples• Given that a reliable and stable code is

applied, they offer facilities to retrieve all information about a certain topic.

This increases the trustworthiness of qualitative findings considerably because these facilities can ensure that the hypotheses developed are really grounded in the data and not based on a single and highly untypical incidents.

ETHICAL ISSUES

Similar to how data is collected and analysed in

other instances: Anonymity of participants Confidentiality

Make sure that the data is safe, password protected. If storing data online, needs ethics clearance.

Ensure that the data is rich in description but not skewed into the direction where the researcher wants it to be.

REFERENCES AND SOURCES

Fielding, N.G. and Lee, R.M (1998). Computer Analysis and Qualitative Research. London: SAGE Publications.

Grant, M.M. (2011). Learning, Beliefs, and Products: Students' Perspectives with Project-based Learning, Interdisciplinary Journal of Problem-based Learning:Vol. 5(2), Article 6.

Kelle, U (1995). Computer-aided Qualitative Data Analysis: Theory, Methods and Practice. London: SAGE Publications.

La Pelle, N. (2004). Simplifying qualitative data analysis using general purpose software tools. Field Methods, Vol 16 (1), 85-108

Miles, M.B. & Huberman, A.M. (1994). Qualitative Data Analysis. California: SAGE Publications.

Ozkan, B.C. (2004). Using Nvivo to analyse qualitative classroom data on constructivist learning environments. The Qualitative Report, Vol 9 (4), 589-603.

QUESTIONS

FILES

GROUP X

POST YOUR QUESTIONS

FOLDER

YAHOO GROUPS

YAHOO GROUPS

Create text file

Create text file

Please have your questions in by 9pm, Saturday 31st March