Research Methods and Materials in English

Overview

Research begins before the library, with a question that is narrow enough to answer and open enough to matter. A topic such as Shakespeare is not a question; a question such as how soliloquies in one play shift from public rhetoric to private doubt is a question, because it names a text, a feature, and a change to be traced. Test a question by asking whether it can be answered with available materials in available time, whether it has more than one plausible answer, and whether answering it would change what is known. Then turn the question into an aim, a set of objectives, and a tentative hypothesis or, for exploratory work, a set of guiding questions. A hypothesis in literary research is not a prediction to be proved at all costs; it is a proposition to be tested against evidence and revised when evidence resists it. Keeping a research journal from the first day, recording questions, searches, dead ends, and decisions, is not bureaucracy; it is the record that makes later writing honest and revisable.

Research Methods and Materials in English: questions, materials, and methods

Materials in English research fall into three kinds, and confusing them is a common error. Primary materials are the objects of study themselves: literary texts, manuscripts, letters, recordings, learner writing, classroom transcripts, and the editions in which texts circulate. A primary text is not raw or unmediated: editions differ, manuscripts show revision, and a modernised text may hide original spelling and punctuation, so the researcher must state which edition or version is used and why. Secondary materials are scholarship about primary materials: books, journal articles, critical editions with introductions and notes, and theses that argue interpretations. Tertiary materials are reference works that organise knowledge: bibliographies, catalogues, dictionaries, encyclopaedias, and databases that point to primary and secondary sources. A dictionary definition is not an argument, and a tertiary summary is not a substitute for reading the secondary argument itself. The literature review, or survey of scholarship, is not a list of summaries. It compares arguments, identifies agreements and disagreements, traces how a debate developed, and finds a gap, a question previous work leaves open, a method not yet tried, or a text not yet read in a particular way. A gap must be genuine: claiming that nobody has studied a famous text is rarely true, while claiming that nobody has studied its reception in a particular region or its language in a particular corpus may be defensible if the search is documented. Methods must fit questions. Qualitative methods explore meaning in depth: close reading of language and form, thematic analysis of interviews or texts, case study of a classroom or a writer, ethnographic observation of reading or language use, and archival research in manuscripts and records. Quantitative methods count and compare: corpus linguistics counts patterns across large text collections, surveys measure attitudes or practices across a sample, and experiments compare groups under controlled conditions. Mixed methods combine the two, for example using a survey to map a pattern and interviews to explain it. Corpus methods require care with representativeness: a corpus is a sample, not the language itself, and its composition, size, and balance shape every finding. Archival methods require care with provenance: who kept a document, why, and what was discarded shapes what survives. Interview methods require care with questions: leading questions produce leading answers, and open questions, follow-ups, and attention to silence produce richer data. Survey methods require care with sampling: a convenience sample of friends cannot speak for a population, and a low response rate may bias results toward those already interested. Close reading remains a method, not merely a skill: it must be systematic, with a stated focus, a set of passages, and criteria for what counts as evidence, so that another reader can check whether the reading holds. Analysis is where data becomes argument. Thematic analysis codes data into recurring patterns, but codes must be derived from data and question, not imposed. Discourse analysis examines how language constructs identities and power in texts and talk. Statistical analysis in language research must report not only averages but variation, sample size, and limits, and must not claim causation from correlation alone. Triangulation, using more than one method, source, or analyst, strengthens a finding when the different angles converge, and it complicates a finding honestly when they diverge. Ethics runs through all of this: participants must consent freely, understand the study, and be able to withdraw; identities must be protected where promised; vulnerable participants need extra care; and data must be stored securely. Plagiarism is not only copying words; it includes copying ideas, structure, or data without attribution, paraphrasing too closely, and presenting another's argument as one's own. Quotation, paraphrase, and summary each have a place: quotation for exact language that matters, paraphrase for restating an idea in one's own structure, summary for compressing an argument, and all three require citation. Citation styles differ, but the principle does not: give enough information for a reader to find the source, distinguish your voice from your sources, and be consistent. A bibliography lists what was consulted or cited according to the style in use, and it must be checked against in-text citations so that every cited work appears and every listed work is used. Validity asks whether a study measures or explores what it claims; reliability asks whether its procedures would give consistent results if repeated; generalisability asks whether findings from a sample or case can speak to a wider population, and qualitative case studies often aim for transferability of insight rather than statistical generalisation. Every study has limits, and stating them is a strength, not a confession of failure: a small sample, a single text, a short observation period, or a corpus bias should be named, with their consequences for the conclusion stated plainly.

Materials: kind, example, use

Methods matched to questions