Terms Used in Qualitative Data Analysis
One slide carries eleven definitions and is the densest definitional page in the week. The words the week leans on hardest are not among them, and this page marks the two sets as clearly as each other.
The eleven terms, quoted exactly
The slide is titled "Terms used in qualitative data analysis" and carries a two-column table with no header row and eleven rows. The count of eleven is derived here by enumeration; the source states no count, gives no numbering and names no source for any definition.
THE WEEK'S ONLY GLOSSARY, VERBATIM
| Term | Definition as printed | What it is good for |
|---|---|---|
| Interim analysis | "Cyclical process of collecting and analysing data during study" | A design decision: it says analysis begins before collection ends |
| Memoing | "Recording reflective notes through- out the process - can be used as additional data" | The only mechanism named for capturing your reasoning as you go |
| Data entry | "Text is transcribed into word processing documents" | Names the clerical step between recording and analysis |
| Segmenting | "Dividing the data into meaningful units" | The step that makes coding possible; absent from the study notes entirely |
| Coding | "Marking the segments of data with symbols or names" | The second of three definitions of coding in the week |
| Master list | "List of all codes used in study" | The artefact an examiner will ask to see |
| A priori codes | "Refer to codes developed before data is examined" | Half of the only coding distinction the slides make |
| Inductive codes | "Refer to codes developed whilst data is examined" | The other half - a distinction by timing, not by function |
| Diagramming | "Process of making a sketch, drawing, or outline to clarify the relationship between the parts of a whole" | A named way of working out structure |
| Network diagram | "diagram showing the direct links between categories, variables, or events over time" | Named; no example appears anywhere in the supplied material |
| Matrix | "Columns and intersecting lines showing relationships" | Named; no example appears anywhere in the supplied material |
Source: the week's slide deck, the terms slide. The third column is this page's, not the source's. The broken hyphenation in "through- out" is in the slide's stored text and is reproduced rather than tidied.
Six of the eleven are worth carrying into a real project as they stand. Segmenting, coding, master list, a priori codes, inductive codes and memoing together describe a workable minimum: divide the data into units, mark the units, keep one list of every mark you use, know which marks you brought with you and which the data gave you, and write down why as you go.
Interim analysis, and the model it contradicts
The consequence is real. If collection and analysis are cyclical, then what you find in your first six interviews can change the questions you ask in the next six - which changes your ethics wording, your instrument, your timeline and what you must disclose in your methods chapter. None of that is drawn out here, and no example of an iteration appears anywhere in the supplied material.
Four definitions that will not carry the weight put on them
Read these four with care
A definition with no rows
"Columns and intersecting lines showing relationships." A matrix has rows and columns; this names one dimension and replaces the other with intersecting lines. It is also the deck's only statement about matrices, and the only instruction anywhere in the supplied source to use one sits in an unlabelled figure in the study notes. No matrix appears anywhere in the material.
Adjacent to diagramming, unrelated to it
Diagramming is defined without any temporal element; a network diagram shows links "over time". Whether a network diagram is something diagramming produces, or a separate device, is not stated. The two rows sit next to each other and neither refers to the other. No network diagram appears in the supplied material.
One practice, two word forms, no link
The glossary defines memoing. The study notes' software passage lists "Memos" as a concept in a proprietary package and defines nothing. Same practice, two forms, in one week, with no cross-reference. Reproduce the broken hyphenation in "through- out" or normalise it, but say which you did.
Transcription under another name
"Text is transcribed into word processing documents." Transcription is treated at length in the data-collection week; neither week refers to the other, and this one-line definition adds nothing to what that week supplies. It also sets up the coding slide's requirement for transcribed text, which no other part of the week requires.
The terms the week uses and never defines
This is the half of the vocabulary that matters more, because these are the words the week's own instructions are built out of. Each row below is a term the material uses to tell you what to do, and does not define anywhere in its notes or its slides.
USED, LOAD-BEARING, AND UNDEFINED
| Term | Where the week relies on it | Definition supplied? |
|---|---|---|
| Theme | Six of the notes' eight sections, three slides, six of the ten bubbles in the notes' figure, and the endpoint of the stage diagram | Not defined in source |
| Topic · Idea · Concept · Term · Phrase · Keyword | Named together as the six other things a code can be based on, alongside themes | Not defined in source |
| Pattern · Trend · Consistency · Unusual event | The objects you are told to look for in step 1 of the four-step model and in the figure's bubbles | Not defined in source |
| Category · Sub-category | The unit of nearly every instruction in the week; the notes state what categories do, not what they are | Function only |
| Core category | The pivot of the definition of selective coding, and an item in the ordering exercise | Not defined in source |
| Grounded theory | Named as the basis of qualitative analysis in the notes' opening bullet | Not defined in source |
| Triangulation · Member checking · Audit trail | Used as the means of achieving credibility, dependability and confirmability | Not defined in source |
| Respondent validation | Named on the week's ethics slide; never linked to member checking, which is the same practice | Not defined in source |
| Theoretical understanding | The terminal state of the stage diagram - the last word of the week's process content | Not defined in source |
| Discrete variable · Linear cause and effect | The things a holistic perspective is said not to reduce to | Not defined in source |
| "Large amounts of data" | The material's only stated criterion for abandoning manual analysis for software | Never quantified |
Triangulation is defined in an earlier week of the subject and the two treatments are never linked. Everything else in this table is undefined across the whole supplied source.
Four senses of one word in a single week
"INDUCTIVE" ACROSS THE WEEK
| Where | What it is applied to | What it means there |
|---|---|---|
| Study notes, section 2 | "An inductive process" | Developing theories from the data you have gathered |
| Slide 4 | "Inductive analysis" | Engagement in the detail of the data to discover patterns, themes and inter-relationships |
| The glossary slide | "Inductive codes" | A matter of timing - codes developed while the data is examined |
| The worked example | "Inductive Categories" | Categories derived from the responses themselves |
Four uses, one week, no cross-reference between any two. If you use the word, name the sense and cite where you took it from.
Two coding vocabularies that share no term
The glossary's a priori and inductive codes are the slides' whole account of types of coding. The study notes give five names of their own - descriptive, analytic or theoretical, factual, axial and selective - and not one of the seven appears on both sides. The notes divide coding by what the code does; the slide divides it by when the code is made. Nothing in the week says whether the two schemes are alternatives, orthogonal or compatible, and no criterion for choosing among the seven appears anywhere. All seven are set out on Coding Qualitative Data in Practice.
Using a glossary that is not complete
Before your methods chapter goes out
- Every analytical term you use appears once in a definitions list, with a source.
- "Theme" is defined, and used only in that sense throughout.
- You have said whether your codes were a priori, inductive, or both, and how the mix changed over the study.
- Your master list exists as a document and can be produced on request.
- If you claim your collection and analysis were cyclical, you have described the cycles rather than asserting them.
One piece of context to finish with. This week states two learning objectives, the first of which - "Evaluate presentation styles for qualitative data" - is not delivered anywhere in it. Across the six weeks of the subject now on record, fourteen objectives are stated, none is fully delivered, four are partially delivered and ten are not delivered at all. The reconciled audit is on The Six-Week Objective Audit, and the broader terminology of the subject is collected on Research Terminology: A Working Glossary.
What to carry forward
- Eleven terms, one slide, no header, no numbering and no attribution. Six of the eleven are usable as they stand: segmenting, coding, master list, a priori codes, inductive codes and memoing.
- Interim analysis is the one entry with a design consequence, and it contradicts the two linear process models beside it. Both readings are the source's.
- Matrix, network diagram and master list are named and never shown. You cannot evaluate a display you have not seen.
- The words the week uses most - theme, pattern, category, core category - are the words it does not define. Take those from a cited text.
- Say which sense of "inductive" you mean. The week uses four and reconciles none of them.
Frequently asked questions
What is a theme, according to this material?
It does not say. "Theme" is the most-used analytical word in the week - it appears in six of the study notes' eight sections, on three slides and in six of the ten bubbles of the notes' figure - and no definition of it appears anywhere in the supplied source. If your findings are organised by theme, take the definition from a methods text and cite that text.
What is the difference between a priori and inductive codes?
Timing, and only timing. A priori codes are "developed before data is examined" and inductive codes "developed whilst data is examined". The distinction says nothing about what the codes do or how they are constructed, and the study notes' five coding names divide the ground a different way entirely, with no term in common.
Does interim analysis mean I should start analysing before I finish collecting?
That is what the definition describes - a cyclical process of collecting and analysing during the study. The material states it in ten words and never develops it, and the two process models on the slides immediately before it are strictly linear. If you intend to work in cycles, describe the cycles explicitly in your methods chapter rather than relying on this line to carry the claim.
What should a master list contain?
The material says only that it is a "List of all codes used in study" - no format, no fields, no worked example. In practice the useful minimum is the code name, a one-line definition, the date it entered the list, and one anchoring quotation, because that is what lets you show an examiner where a code came from. The source does not prescribe any of that.
Why are there no matrices or network diagrams anywhere in the material?
There is no explanation given. Both are named in the glossary, a matrix is also named in the study notes' figure as something to use, and neither appears in any form anywhere in the supplied notes, slides or worked example. The absence is worth recording rather than working around, because the week's own first objective asks you to evaluate presentation styles.
Is member checking the same thing as respondent validation?
The material never says. Member checking appears once in the study notes, in three words after an "e.g.", as the means of examining dependability. Respondent validation appears once on the week's ethics slide. They are ordinarily the same practice, both are named once, and neither is defined - so if you use it, define it from a cited source.
References and source attribution
- O'Leary, Z 2010, Doing your research project, Sage Publications - the single reference listed on the week's own closing slide.
- O'Leary, Z. 2017, The Essential Guide to Doing Your Research Project, 3rd ed., Sage Publications, London - the later edition cited elsewhere in the supplied source.
- Bryman, A. 2016, Social Research Methods, 5th ed., Oxford University Press, Oxford - cited elsewhere in the supplied source; a place to obtain definitions of theme, pattern and triangulation that this week uses and does not supply.
- Quinlan, C. 2011, Business Research Methods, 1st ed., Cengage Publishing.
- The supplied teaching source: the qualitative analysis week's slide deck, the terms slide, read against the same week's study notes. The glossary table carries no header row, no numbering and no source attribution of any kind.
Suggested questions for Ask KEVOS
- Build me a definitions table for my methods chapter covering every analytical term I use, with sources.
- Which terms in this week can I cite the teaching material for, and which need an outside text?
- How should I define "theme" for a study of project handover interviews, and what should I cite?
- Draft the paragraph describing my code development, distinguishing a priori from inductive codes.
- What would a master list look like for a study with about forty codes?