Scholarly and Legal Terms

Common terms

• Abstract – A concise summary of what a paper or article is about, briefly touching on key themes/concepts, arguments, and conclusions. See how to section of links under 'Abstracts.' See the 8-sentence abstract formula.
• Advanced editing – Holistic consistency work across a whole manuscript: whether methodology stated in one chapter matches what later chapters report, whether terminology stays stable, whether the abstract still describes the finished argument, and whether the voice stays scholarly throughout. Universities and handbooks usually call this content editing or *substantive editing* — same tier, different name. Distinct from basic (sentence-level) editing and from coaching. See Editor, Advanced Editor, or Coach? and What Help Do You Need?.
• Automated cadence – The rhythm that gives machine-written prose away to a human reader — as distinct from the statistical patterns a detector measures. In academic writing the tell is evenness. Sentences of similar length and weight, paragraph after paragraph, with none of the variation that human writing under pressure produces. Nothing is wrong with any single sentence. The uniformity is the signal. Be careful searching for the term. In content marketing, AI cadence means almost the opposite — clipped, staccato, one-line paragraphs that read like a sales pitch. Both descriptions are accurate for their own register, because a model reproduces the prose it was trained on. Academic AI output is flat; marketing AI output is punchy. A cadence objection is not a detector score. No number is involved — a reader simply notices. It is answered by prose that carries your reasoning, not by argument.
• Bibliography – A complete and formal alphabetized list of all sources and footnotes used in a document, consistently formatted in a major reference style. University programs will indicate which reference styles they require. See link, How to Write a Bibliography in Links section at bottom of page.
• Citation – the source of arguments or information you are “citing” (referring to) when asserting a fact, making a point, or giving a quotation or paraphrase.
• Conceptual Framework – A structure you build yourself, drawing on several theories, concepts and prior findings, to fit a study that no single existing theory covers. Distinct from a theoretical framework, which adopts an established theory whole. The difference matters practically: with a theoretical framework you must choose well and stay inside it; with a conceptual one you must justify every connection you made, because you made them. See which one you need.

• Content editing – See advanced editing.
• Construct – The abstract thing you set out to study — self-esteem, job satisfaction, organizational commitment. A construct may be simple, reflecting one idea, or multi-dimensional, drawing several related concepts together. Constructs used in academia must be capable of precise definition. Tip: 'construct' and 'variable' are often used interchangeably. They are not the same thing. A construct is the abstract thing; a variable is the measurable stand-in you use to get at it — and the gap between the two is where measurement problems live.
• Content Analysis (qualitative) – methodology for identifying patterns in recorded communication. (e.g., identifying and creating codes for qual analysis etc.) See How To links section at bottom of page to find guides to content analysis.
• Discourse Analysis (qualitative) – The analysis of language in use: what language is doing, and what it is doing in a particular context. It asks what is this language being used for, rather than what does it mean. Formally it covers any semiotic event, but in doctoral practice it is nearly always language.
• Dissertation – The document a candidate submits in support of a degree, presenting their research and findings. Length varies enormously by institution and field — commonly 140 to 150 pages at online universities, and 300 or more at large research universities. There is no useful average. Find your own program's figure; see chapter two for how.
• Chair – The head of a doctoral candidate's thesis supervisory committee. In the UK, Australia and much of the Commonwealth the same role is called the supervisor, or principal supervisor where there is more than one. The vocabulary differs; the responsibility does not.
• Mentor – An individual who provides academic assistance and guidance to a dissertation or doctoral student but is not their Chair.
• “Comps” – see Comprehensive Doctoral Exam.
• Comprehensive Doctoral Exams – Exams taken by a PhD candidate prior to starting their thesis proposal and writing of their dissertation.
• Dissertation proposal (aka 'proposal') – a formal document outlining the proposed thesis's purpose, research topic, research questions, theoretical framework, and methodology.
• Grounded Theory (qualitative) – Is an analytic approach based on inductive reasoning. In grounded theory you collect data relevant to your topic and only propose your theories and hypotheses after gathering your data and by reasoning from your findings. Doctoral candidates can find this tricky.
• Literature Review (often just “Lit. Review”) – A comprehensive review of all existing scholarly publications directly relevant to your specific topic.
• Empirical – Based on observation or experiment rather than on theory or reasoning alone. Your own collected data is empirical the day you collect it; so is data reported in published studies. Peer review is where empirical work gets published, not where it comes from.
• Google Scholar – A version of the google search engine that searches scholarly publications, and (mostly US) legal cases, in response to searched queries.
• Grok – To understand something so completely that you are no longer standing outside it looking in — all of it, how the parts relate, what follows from what, where the edges are. Not an approximate hold. The word was invented by Robert A. Heinlein, and English has no native equivalent, which may be why the idea is so rarely discussed. Grokking your own study is what clarity actually requires, and it is why teaching something improves your grasp of it: you cannot teach what you have not grokked.
• Holistic – Something concerning the whole or overall understanding of a given topic or thing, as opposed to a specific part. Closely related to grok.
• Interpretive Phenomenological Analysis (IPA) (qualitative) – A qualitative approach examining how individuals make sense of significant lived experiences. IPA is doubly interpretive: the participant makes sense of their experience, and you make sense of the participant making sense of it. It requires small, homogeneous samples and intense, line-by-line engagement with each transcript. Tip: IPA is frequently chosen by candidates who underestimate it. It demands a real grasp of phenomenological philosophy and a great deal of time. If you are considering it as a first-time qualitative researcher, talk to someone experienced before committing at the proposal stage.
• Linear – Progressing from one thing to the next sequentially, as if in a straight line. In writing this translates to strong, logical narrative skills.
• Meta-analysis – A statistical analysis that combines the results of previous studies. These are often to be found when multiple studies address the same questions.
• Mixed Methods – utilization of two or more types of quantitative and qualitative research in one piece of research.
• Peer review – The concept and process of having scholarly work available to open review and critique from other academics or professionals in the given field.
• Periodical – A publication, such as a magazine or newspaper, published at regular intervals.
• Phenomenology – The philosophical study of subjective, first-person experiences. Tip: Be careful with phenomenology unless you are very confident of your grasp of this school of philosophy. Used as an analytical tool it has a habit of confusing first-time users who receive little outside support. That is to say, doctoral candidates. It is best to avoid as an analytical tool until the postdoctoral level.
• Phenomenological analysis – see 'Interpretive Phenomenological Analysis'.
• Plagiarism – the unacknowledged (no citation or footnote) use of someone else's work which then passes as your own. At once theft of intellectual property and intellectually and academically dishonest.
• Professional Journal – An academic peer-reviewed journal reporting scientific research. Alternatively for professional practitioners in various fields (engineering, law, medicine inter alia) a publication providing professional opinions, findings, research, and scholarship for practitioners.
• Proposal – see 'dissertation proposal'.
• Qualitative Analysis – Analysis of why humans behave the way they do. Thus, generally, analysis that does not rely on numeric value but on descriptive data. Data such as human motivations, feelings, opinions, beliefs etc. Generally, this research delves into people's lived experiences to discover data that has a lot of contexts. Data is usually gathered by observation, interviews (especially in the instance of doctoral theses), and focus groups.
• Codes – codes allow you to categorize data. It is critical that any coding scheme you use is uniform across your data set. A code applies to one category of data (even if the data is a concept). Remember also, your codes are foundational to finding/creating your themes.
• In vivo coding – this type of coding (and there are many) is often found in doctoral theses as it allows the coding of terms and expressions interviewee's themselves use.
• Substantive editing – See advanced editing.
• Themes – Patterns of meaning that run through your data. Human brains are relational engines and seek patterns, which is why you will usually find your themes while doing your coding, or while reviewing it, rather than afterwards. Once you have them you perform thematic analysis — working out what the patterns mean and how they answer your research questions.
• Thematic Analysis – There are a number of common approaches to thematic analysis. They vary by their underlying philosophies and conceptual assumptions. It is a very good idea to carefully read books and papers that discuss these approaches if you think that thematic analysis is in your future. Once you choose an approach, stick with it. Reading material about similar approaches can lead to real confusion during the analysis process.
• Qualitative – Research that collects and analyzes non-numeric data — interviews, observations, documents — looking for meaning and pattern rather than measurement. The strength is depth, and the ability to find what you were not looking for. The constraint is that findings describe the people you studied and are not usually claimed to extend beyond them. See also generalizability.
• Quantitative – Research that collects and analyzes numeric data — counts, scores, measurements — using statistical methods. The strength is that findings can be tested for significance and, where the sampling supports it, generalized. The constraint is that you can only measure what you decided in advance to measure.
• Quantitative Analysis – the application of mathematics and statistical algorithms, models, and research to understand measurable or quantifiable properties of phenomena or things.
• Scholarly – Pertaining to scholars. Scholar is a term that is synonymous with academic in modern times. Thus, we get “scholarly journals” – magazines in which academics publish research. “Scholarly interests” – the interests someone has in fields that are of scholarly interest, and so forth.
• Secondary source – a cited source of information, data, or scholarship that is not the primary (i.e., original) source. As a secondary source it will be given less scholarly weight and be less authoritative than a primary source.
• Semiotic – Concerning signs and how meaning is made — words, images, gestures, objects, anything that stands for something else. The study of it is semiotics. You will meet the word most often in discourse analysis and in qualitative work drawing on linguistics or media studies.
• Statistics – the organized interpretation and presentation of collected data.
• Stodge – Our word for the hedged, evenly balanced, endlessly qualified prose that AI produces by default — and that some candidates learn to produce by imitation. It is the most probable text there is, which is exactly why a detector finds it least suspicious and a reader finds it unreadable. Full treatment on our clarity page.
• Thematic Analysis (qualitative) – A method for identifying, analyzing, and reporting patterns of meaning across a data set. There are several distinct schools of thematic analysis, differing in their underlying philosophy and in how much interpretive latitude the researcher takes. Choose one approach, read its foundational sources carefully, and stick to it. Tip: reading across multiple schools mid-analysis is a reliable way to become thoroughly confused. Pick your approach before you start coding, not during.
• Theoretical Framework – The specific theory or set of related theories your study rests on, and which shapes what you look for, how you interpret what you find, and how you connect your results to prior scholarship. It is not a literature review, and it is not a general statement of your field's assumptions. A well-chosen framework makes your methodology almost obvious; a poorly chosen one makes analysis painful or impossible. Distinct from a conceptual framework, which you build from several theories and concepts when no single theory covers the study. Tip: most universities want one framework, not several. See our methodology at proposal stage page — this is the single most consequential choice you make before writing.
• Theoretical Construct – See construct.
• Thesis, plural: 'theses' – a paper presenting its author's research in support of the author's candidacy for a graduate degree or qualification. In the US and Canada often, theses are written in support of masters degrees and the thesis is called a 'dissertation' if it is for a doctoral candidacy. In the UK it is usually the reverse. In any case, 'thesis' and 'dissertation' are often used rather interchangeably and are essentially the same thing.
• Trade journal or magazine – A publication aimed at practitioners in a particular industry rather than at scholars. Trade publications report industry developments, practice, and opinion; they are not peer-reviewed and carry far less scholarly weight than academic journals. They may occasionally be acceptable as grey literature in fast-moving fields — verify with your Chair first, and use sparingly.
• Variable – The measurable stand-in for an abstract construct: a score on a named scale, a count, a category. It is what you actually record. Often used interchangeably with 'construct,' and they are not the same. Your construct is what you want to study; your variable is what you can measure. Whether the second genuinely captures the first is the question your committee will ask.
• Dependent – Dependent variables depend on the actions of independent variables. If an independent variable changes its value many or all of its dependent variables will or are likely to change value. To state this differently, you check if a dependent variable changes in value based on changes in the value of an independent variable.
• Independent – Think of the independent variable as the cause of something. Its value is manipulated in your study to see what happens, that is measured by seeing if there are changes in the dependent variables. In quantitative studies that use statistical analysis, often estimates are created of the extent to which changes in the independent variable predicts/explains changes in the values of one or more dependent variables.
• Moderating – A moderating variable changes the strength or direction of the relationship between an independent and a dependent variable.
  An example. You study how many hours a week of training produce how much new muscle mass in bodybuilders. Age is a moderating variable — the same hours produce less gain in an older bodybuilder than a younger one. The relationship still holds; age changes how strongly.
  Not to be confused with a mediating variable, which explains how the effect travels rather than how much of it arrives.
• Mediating – Sometimes changes in an independent variable affect the dependent variable indirectly via a third variable that is the mediating variable. Not all real relationships are direct relationships, and the mediating variable captures that.
• Control – Control variables are variables whose values are maintained as constant during an experiment or study. For example, if you are studying germination of bean seeds. Daylight might vary but if you keep the bean seedling in a greenhouse you can keep the temperature constant throughout your experiment.

• Writing cycle coach – Our term for the tier of help above advanced editing. Most people call it a dissertation coach or an academic coach, and it is the same work. We say writing cycle because what gets coached is the whole cycle — topic, outlining, planning, milestones, alignment, and the Chair relationship — not the prose. An editor works on what you wrote. A coach works on how you are working. Not the same as a sherpa, which is who you are assigned rather than what you buy.

Research Methods & Design

• Alignment – The single most important structural concept in doctoral work, and the one candidates most often cannot define when asked. Alignment means every part of your study pulls in the same direction: your problem statement, purpose, research questions, theoretical framework, methodology, instruments, analysis, and findings all connect in an unbroken chain. Your committee may notice breaks in the chain while assessing the scholarship; School Review checks alignment as part of a different test — whether the document meets institutional requirements. Misalignment signals confused thinking rather than merely untidy writing. Tip: write your problem, purpose, questions, and method on a single page and read them in sequence. If any link feels strained on one page, it will be indefensible across two hundred.
• Delimitations – The boundaries you deliberately set on your study: the population you chose to examine, the timeframe, the geographic scope, the theoretical lens. Delimitations are choices. This distinguishes them from limitations, which are constraints imposed on you. Committees frequently ask candidates to separate the two clearly, and conflating them is a common revision request.
• Limitations – Constraints on your study that you did not choose and could not eliminate: sample size, resource and time limits, self-report bias, instrument reliability, access restrictions. Every study has them. Identifying them honestly strengthens your work; pretending they do not exist weakens it.
• Saturation (data saturation) – The point in qualitative data collection at which new data stops producing new codes, categories, or themes. It is the most common justification for a qualitative sample size, and committees will ask you to demonstrate you reached it rather than simply assert it. Tip: document your saturation reasoning as you go — how many interviews before new codes stopped appearing. Reconstructing it afterwards is much harder and reads as an afterthought.
• Triangulation – Using multiple methods, data sources, investigators, or theoretical perspectives to examine the same question, so that convergent evidence increases confidence in your findings. It is one of the strongest available defenses of qualitative credibility and can repair weaknesses a single approach leaves exposed.
• Validity – Whether your study actually measures or examines what it claims to. Internal validity concerns whether your conclusions genuinely follow from your data within the study. External validity concerns whether your findings extend beyond your specific sample and setting. Qualitative work generally speaks of credibility and trustworthiness rather than validity, but committees vary in the vocabulary they expect.
• Reliability – Whether your measurement or procedure would produce consistent results if repeated. A reliable instrument produces stable results; an unreliable one produces noise you may mistake for findings. Published, validated instruments carry established reliability data — one of several reasons universities prefer you use one rather than build your own.
• Generalizability – The extent to which your findings apply beyond your specific sample. Quantitative studies pursue it through sampling strategy and sample size; most qualitative studies do not claim it at all, and should not be criticized for failing to. Overclaiming generalizability is a classic Chapter 5 error.
• Population – The entire group your research question is about. The gap between your population and your sample is where most sampling criticism lives.
• Psychometric properties. Published evidence that an instrument measures what it claims to measure, and measures it consistently. Usually reported as validity — does it measure the right thing — and reliability — does it give the same answer twice. An instrument without published psychometric properties has not been shown to work, however sensible it looks.
• Sample – The people or cases you actually collected data from. Your population is the whole group your conclusion is about; your sample is the part of it you could reach. A reader weighing whether your findings generalize is weighing exactly that gap.
• Purposive sampling – Selecting participants deliberately because they have the specific knowledge or experience your study requires. Standard in qualitative research, and the correct answer when a committee asks why you did not sample randomly.
• Convenience sampling – Selecting participants because they are accessible. Common, sometimes unavoidable, and always a limitation you must state plainly.
• Snowball sampling – Asking participants to refer others who meet your criteria. Useful for hard-to-reach populations; introduces network bias you should acknowledge.
• Survey instrument – The specific, structured tool used to collect quantitative data. Tip: many online universities require you to use an existing validated instrument rather than writing your own, and require documented permission from its author. Establish both before your proposal, not after — see our methodology page.
• Likert scale – A response format offering a graded range of options, commonly five or seven points from strong disagreement to strong agreement. Most survey-based doctoral research uses one. The scale you choose constrains the statistics available to you, so decide it alongside your analytical approach, not afterwards.
• Hypothesis – A testable statement predicting a relationship between variables. The null hypothesis states there is no relationship; the alternative hypothesis states there is. Quantitative studies test the null and report whether it can be rejected.
• Statistical significance (p-value) – The probability of observing your result if the null hypothesis were true. A conventional threshold is 0.05, but conventions vary by field and the threshold is not a verdict on importance. A statistically significant result can be practically trivial.
• Effect size – How large an observed relationship or difference actually is, independent of whether it is statistically significant. Increasingly expected alongside p-values, because significance alone tells a reader nothing about magnitude.
• Descriptive statistics – Statistics that summarize your data: means, medians, standard deviations, frequencies. Inferential statistics – statistics that support conclusions about a wider population from your sample: regression, ANOVA, t-tests, structural equation modeling and others. Most quantitative dissertations report both.
• Regression analysis – A family of statistical techniques estimating how changes in one or more independent variables predict changes in a dependent variable. Among the most common analytical approaches in doctoral quantitative work.
• Structural Equation Modeling (SEM) – A statistical approach modeling relationships among multiple variables simultaneously, including latent constructs that are not directly measured. Powerful, and considerably more demanding than regression. Tip: if SEM is in your proposal, budget for statistical support. An hour of expert time early is far less expensive than discovering a misfit model after data collection.
• Confounding variable – A variable you did not account for that influences both your independent and dependent variables, producing a relationship that is real in your data but misleading as to cause.
• Case study – A research design examining a bounded case — an organization, a program, an event — in depth, typically using multiple data sources to triangulate. Requires clear definition of what is inside and outside the case boundary.
• Ethnography – A qualitative approach studying a culture or community through extended immersion and observation. Rarely feasible within online doctoral program timelines and resource limits.
• Action research – A design in which the researcher works collaboratively with an organization to address a real problem or shape practice, rather than observing from outside.
• Delphi technique – A structured method using successive rounds of expert input to reach consensus on a question, often about best practice or future developments. Useful where empirical data does not yet exist.
• Open coding – The first pass of qualitative coding, generating codes directly from the data without imposing a prior scheme. Axial coding – the subsequent stage, identifying relationships among codes and grouping them into categories. Both are core to grounded theory and used more widely.
• Member checking – Returning your interpretations to participants to confirm you have represented their meaning accurately. A recognized credibility strategy in qualitative research, and one committees like to see.
• Reflexivity – Explicit, documented examination of how your own position, assumptions, and involvement shape the research. Expected in most qualitative traditions. Bracketing (epoché) – the related phenomenological practice of deliberately setting aside your preconceptions in order to encounter the phenomenon as described.

Institutional & Process

• ABD (All But Dissertation) – A candidate who has completed coursework and comprehensive exams but not the dissertation. Widely used, informally. It is also the point at which most doctoral attrition happens.
• Candidacy – Formal status as a doctoral candidate, usually conferred after comprehensive exams and before dissertation work begins. Universities differ on what candidacy requires and what it entitles you to.
• Prospectus – At many institutions, a synonym for the dissertation proposal; at others, a shorter document preceding the full proposal. Check which your program means, because the two are not interchangeable in scope.
• IRB (Institutional Review Board) – The committee that reviews research involving human subjects for ethical compliance before data collection may begin. Also called an ethics committee or REB in some jurisdictions. Tip: IRB approval takes longer than most candidates expect, and you may not collect a single piece of data before you have it. Build the timeline in early — an IRB delay at the wrong moment can cost you an entire term.
• Informed consent – Documented agreement from participants who understand what the study involves, what will happen to their data, and that they may withdraw. A standard IRB requirement and a standard source of submission delays when handled casually.
• School Review (academic quality review) – An institutional review stage, separate from your committee. Your committee assessed whether the scholarship is sound; School Review is a different test — whether the document meets institutional requirements (internal consistency, alignment, conclusions that follow from findings, claims that do not outrun evidence, methodology properly evidenced, citation integrity, and template/format compliance). Passing your committee does not guarantee passing School Review. Some institutions also run a separate editorial check after the defense. See what School Review actually checks.
• Defense (viva voce) – The oral examination in which you present and defend your completed dissertation. Called the viva in the UK and much of the Commonwealth, and the two are genuinely different events.
  In North America you face your full committee, usually four or five people who have supervised or read the work throughout. The result is normally a vote, and outcomes run from pass to pass-with-revisions. By the time it happens your Chair has generally decided you are ready, which is why the defense is often more ceremony than examination.
  In the UK and much of the Commonwealth you face two examiners — one internal, one external to your institution — and often not your supervisor at all. The external examiner may never have met you. It can run for hours. Outcomes are formally graded, from no corrections through minor and major corrections to resubmission, and major corrections is common and not a disgrace.
  The practical difference: a North American defense usually confirms a decision already reached. A viva can still change it.
• Milestone – A program-defined checkpoint — proposal approval, IRB approval, data collection complete, chapters submitted — used to track progress and, at many institutions, to determine whether you remain in good standing.
• Committee – The group of faculty, chaired by your Chair, responsible for approving your proposal and dissertation. Members frequently hold differing expectations, and reconciling them is a recognized part of the work rather than a sign anything has gone wrong.
• Time-to-degree limit – The maximum period your institution allows for completing the doctorate before you are dismissed from the program. Extensions exist but are neither automatic nor guaranteed. If this limit is approaching, see our Panic page.
• Template – The institution's required document format for the dissertation: headings, margins, front matter, citation style. Templates change, sometimes mid-candidature, and a template change can require substantial reformatting of finished chapters.

Publishing & Scholarly Communication

• Impact factor – A metric reflecting the average number of citations to articles in a given journal over a set period. Widely used as a proxy for journal prestige, widely criticized as a measure of individual article quality. Verify any impact factor claim against Clarivate or Scopus — invented metrics are a hallmark of predatory journals.
• CiteScore – Elsevier's journal citation metric, calculated from Scopus data. An alternative to the impact factor, and similarly a journal-level rather than article-level measure.
• h-index – A researcher-level metric: your h-index is h if you have published h papers each cited at least h times. Increasingly used in hiring, tenure, and promotion decisions.
• DOI (Digital Object Identifier) – A persistent identifier assigned to a publication, intended to remain valid even when web addresses change. Tip: a DOI is not proof a source is genuine. Fabricated citations can carry plausible-looking DOIs, and real DOIs increasingly fail to resolve as publishers change platforms. See Purging Citation Contamination.
• ORCID – A persistent unique identifier for individual researchers, distinguishing you from others with similar names across your publication record. Increasingly required by journals and funders. Free, and worth registering for before your first submission.
• Open access – Publication model in which articles are freely readable rather than paywalled. Legitimate open-access journals conduct full peer review; the model itself is not a quality signal in either direction.
• APC (Article Processing Charge) – The fee some journals charge authors to publish, common in open-access publishing. Legitimate journals display APCs transparently before submission. Concealed fees revealed only after acceptance are a predatory-publishing red flag — see our predatory journals page.
• Preprint – A manuscript posted publicly before peer review, on a server such as arXiv or SSRN. Useful for establishing priority and gathering feedback; not a substitute for peer-reviewed publication, and not generally acceptable as a source in a literature review without care.
• Desk rejection – Rejection by a journal editor without sending the manuscript out for peer review. Usually triggered by scope mismatch, formatting non-compliance, or presentation quality — which is to say, by things that are entirely preventable.
• Revise and resubmit (R&R) – An editorial decision inviting you to address reviewer comments and submit again. It is not a rejection. Major and minor R&Rs differ substantially in what they demand and how likely eventual acceptance is.
• Corresponding author – The author responsible for communication with the journal and, usually, for the manuscript's integrity. On multi-author papers, this designation carries real accountability.
• Systematic review – A literature review conducted to an explicit, pre-registered protocol, with defined search terms, inclusion and exclusion criteria, and a reproducible screening process. Distinct from the narrative literature review most dissertations require.
• Scoping review – A structured review mapping the extent and nature of research activity in an area, often used to determine whether a full systematic review is feasible.
• PRISMA – The reporting standard for systematic reviews and meta-analyses, including the flow diagram documenting how many records were screened, excluded, and included at each stage. Expected by most journals publishing systematic reviews.
• Grey literature – Material produced outside commercial or academic publishing: government reports, industry research, consultants' reports, working papers, theses. Sometimes necessary in fast-moving fields where scholarship lags practice. Verify with your Chair whether grey literature is acceptable in your program, in what proportion, and in what role. Treat it circumspectly and minimize its use — see Chapter 2.

AI & Digital Tools

• Generative AI – Software that produces text, images, or other output in response to prompts, based on statistical patterns learned from large training corpora. It predicts plausible continuations; it does not know things, and it does not reason toward truth. Understanding that distinction is the foundation of using it safely.
• LLM (Large Language Model) – The class of AI system underlying most current text-generation tools. An LLM predicts the next most probable text given what came before. Fluency is what it optimizes for. Accuracy is not.
• AI hallucination – Output that is fluent, confident, and false. Hallucinations are not malfunctions; they are a direct consequence of how these systems work. They are most dangerous where they are most plausible, which in academic work means citations, statistics, and summaries of studies you have not read. See AI: Use & Abuse.
• Ghost citation (phantom citation) – A reference to a source that does not exist, generated by an AI. Typically well-formatted, with a believable title, a real author name, a plausible journal and year, and sometimes a DOI. Formatting gives no warning. Only opening the actual source catches these — see Purging Citation Contamination.
• Citation contamination – The presence of fabricated or unverified AI-generated references in a body of work, and their spread as those works are themselves cited. Documented rates in published literature have risen sharply since 2023.
• AI detector – Software claiming to identify machine-generated text by analyzing statistical patterns in sentence structure and word choice. Detectors produce probability estimates, not evidence. False positives are common, and are documented as substantially more common for non-native English writers. See Navigating AI Detection Paranoia.
• Perplexity (AI-detection sense) – A measure of how predictable a text is to a language model. Low perplexity means highly predictable, which detectors treat as a signal of machine authorship. Note the trap: clear, concise, formulaic scholarly prose — exactly what your Chair asks for — is low-perplexity by nature. Writing well can raise your detector score.
• Sycophancy (AI) – The documented tendency of AI systems to agree with users, validate their reasoning, and affirm their plans, including when those plans are flawed. Research published in 2026 found leading models markedly more agreeable than humans in equivalent situations. For a stalled dissertation, this is the opposite of useful — see Human-in-the-Loop Structural Refinement.
• Human-in-the-loop – A workflow in which AI performs bounded, mechanical, or diagnostic tasks while a competent human retains judgment and final decisions. The practical model for using these tools in doctoral work without surrendering the thinking your committee is actually assessing.
• Closed-loop prompting – A technique constraining an AI to answer only from text you supply, with explicit instructions to state when information is not present rather than inferring or supplying outside knowledge. Substantially reduces hallucination when analyzing a specific document. See AI: Use & Abuse for the exact prompt structure.
• Reverse outlining – Extracting an outline from a chapter you have already written, listing what each paragraph actually argues rather than what you intended it to argue. One of the strongest available diagnostics for structural problems, and a task AI performs genuinely well. See our fast thesis writing page for the method.
• Context window – The amount of text an AI system can consider at one time. Large windows allow whole chapters or documents to be processed together, but capacity is not comprehension: models still misread and flatten nuance across long documents.
• Prompt engineering – Constructing instructions to obtain more reliable output from an AI system. Useful, and frequently oversold — no prompt eliminates the underlying tendency toward fluent invention.
• RAG (Retrieval-Augmented Generation) – An architecture in which an AI retrieves relevant material from a defined document set before generating an answer, rather than relying on training data alone. This is what tools such as NotebookLM use, and it is why a closed corpus of your own vetted sources is safer than open-ended questioning.
• AI disclosure statement – A declaration, increasingly required by journals and some institutions, stating whether and how AI tools were used in preparing a manuscript. Requirements vary widely and are changing quickly. Check your specific target venue's current policy before submission.

Legal terms — UK and Commonwealth usage

Doctoral processes, and the legal vocabulary surrounding them, differ across the Anglosphere. The terms below follow UK and Commonwealth usage.

They are not only for our UK and Commonwealth clients. United States law is a common law system derived from English law, and American scholars working comparatively — or reading colonial and early republic cases — will meet these terms in their own primary sources. Where usage has since diverged, that divergence is often precisely what is under study.

• Actus Reus – A guilty act, the physical act of a crime (e.g., the physical act of using a knife to stab another, would constitute the actus reus of murder)
• Agent – someone acting legally on another's behalf, or behalf of a company.
• Alternative dispute resolution (ADR) – a method of solving a civil legal dispute privately, without going to court and obtaining a legal judgment.
• Appellant – One who appeals a judgment or decision.
• Arbitration – a form of ADR, where an objective, mutually appointed arbitrator decides the outcome and comes to a decision for the parties.
• Asylum – residency in a country due to risk of serious harm or persecution in one's own country of origin/citizenship, resulting in the need to flee.
• Bankruptcy – When a person or company cannot pay debts as they fall due. Assets are sold to meet what is owed, and the court usually writes off whatever remains.
• Barrister – A court advocate lawyer in England and Wales, as well as other commonwealth jurisdictions. Barristers argue in court, write legal opinions, and draft court documents like Particulars of Claim and Defense & Counterclaim's.
• Beneficiary – someone who receives the benefits of a trust, will, insurance policy, or other legal pay out mechanism.
• Bequest – Personal property given via will.
• Burden of proof – The requirement to prove or disprove an occurrence or fact.
• Chambers – The building or offices used communally by a group of Barristers, each working as a self-employed practitioner.
• Chattels – Private property that is not land or real property (e.g., a car).
• Claimant – A person who is making a claim in civil proceedings.
• Conciliation – ADR method whereby a conciliator meets with each disputing party in private, and together, in an attempt to help them resolve the dispute.
• Conveyancing – the transference of real property ownership from one person (or entity) to another.
• Counsel – One's legal representative (a Barrister in the UK & commonwealth).
• Creditor – The person or entity to whom money is owed to.
• Culpable – Deserving of blame for a wrongdoing or crime.
• Damages – A compensatory monetary amount owed to someone as a result of losses they incurred, resulting from another person's breach of contract, tortious acts or omissions, or fraudulent actions.
• Disbursement – A payment from a public fund.
• Executor – The person who is legally responsible for carrying out the instruction of a will.
• Fiduciary – a duty of loyalty to act in the best interests of a beneficiary in your professional capacity and powers concerned.
• Forfeiture – Being legally compelled to relinquish control of an asset in bankruptcy proceedings or as a penalty for wrongdoing.
• Frustration – When a contract cannot be performed or its purpose no longer exists, for reasons outside of either party's control.
• Incorporated company – A business that is a separate legal entity from the people who formed it and control it.
• Indemnity – A contractual duty to compensate another party for any certain losses or types of loss incurred.
• Intellectual Property – the ownership of non-physical ideas (e.g. an invention, a book, music, formula or recipe, trademarks etc.).
• Intestate – Not having a will before death.
• Legal Ombudsman – A sub-body of the Bar where complaints against lawyers can be heard and investigated.
• Legal professional privilege – The concept of confidentiality of communications between a client and his legal counsel.
• Liability – Being legally responsible for something.
• Limited liability partnership – A partnership where some or all partners have limited liability.
• Litigation – The process of a legal dispute being heard by the court, starting from the moment the original claim form is deemed served, up to when the final judgment is sealed.
• Litigant in person – Someone who chooses to represent themselves in court, as opposed to counsel on their behalf.
• Magistrate – Someone who is not a judge, but who decides upon low level cases, such as summary criminal offences in the Magistrates court.
• Mediation – An ADR method where an independent mediator attempts to facilitate a productive negotiation between the disputing parties so that they may come to an agreement or settlement.
• Mens Rea – “Guilty Mind”, the mental aspect of guilt, usually intent to commit a certain offense.
• Notary – An impartial witness to the signing of a legal document.
• Obiter dictum – "Other things said." Remarks a judge makes in a judgment that give context or raise further considerations. They are not part of the decision itself, and so do not bind future courts.
• Patent – A legal document that protects ownership of intellectual property.
• Probate – The judicial process where a will is proved.
• Prima facie – “at first face” or “at first appearance”.
• Pro bono – The provision on free legal services, advice, or representation, on a charitable basis.

Lexicon/glossary Links - by discipline

• Education
• Social sciences

Links - by method

• Qualitative Analysis Methods

Links - how to

General academic skills

• How to Write a Bibliography

Abstracts

Systematic reviews

• How to do a systematic review/08%3A_Data_Collection/8.06%3A_Systematic_Literature_Reviews_Essays_and_Theoretical_Theses)

Content Analysis [not endorsed nor ranked by us]

• Content Analysis | Guide, Methods & Examples
• Content Analysis - Colombia U. online resource
• A Hands-on Guide to Doing Content Analysis - article
• Demystifying Content Analysis - NIH
• Content Analysis: A Flexible Methodology - article / PDF
• Content Analysis - book