Research Skills for Academic Writing

The most common research mistake is starting with standard Google rather than Google Scholar (scholar.google.com). Standard Google indexes everything — news, Wikipedia, commercial content, academic papers, and marketing material — with no quality filter. Google Scholar indexes only academic materials: peer-reviewed journals, books, conference papers, theses, and grey literature. Using Scholar as your default search is the single easiest upgrade to research quality.

Beyond Google Scholar, your university library provides access to subject-specific databases that go further: JSTOR for humanities and social sciences, PubMed for medicine and biology, PsycINFO for psychology, Web of Science and Scopus for sciences, and Westlaw or LexisNexis for law. These databases have more targeted search tools, controlled vocabularies (MeSH in PubMed, for example), and institutional access to full-text articles behind paywalls.

Search strategy matters as much as the platform. Many students search in plain English ("how does stress affect memory") and get general results. Academic literature uses precise technical vocabulary ("cortisol working memory capacity interaction"). Finding one good paper, reading its keywords and abstract, and using those terms as your next search terms — a technique called forward chaining — is consistently more efficient than iterative broad searches.

Backward chaining is even more reliable: find one strong, relevant paper and look at its reference list. The papers it cites are likely to be relevant to your topic. Find those papers, check their references, and within two or three iterations you will have located the key sources in the area. Combine backward chaining (finding older foundational work) with forward chaining — Google Scholar's "Cited by" function — to find recent papers that build on the foundational work.

For papers behind paywalls, several legitimate routes exist: your institutional library (most universities provide off-campus access via credentials), Unpaywall (a browser extension that finds legal open-access versions automatically), author pages on ResearchGate or personal websites (academics frequently post their own papers), and direct email to the corresponding author (who will almost always respond positively to a student's request). Paying for individual articles should be a last resort.

Finding sources is the easy part. Evaluating whether they are credible and appropriate for your argument requires a different set of skills. The CRAAP test — Currency, Relevance, Authority, Accuracy, Purpose — provides a systematic framework for assessing any source type, from peer-reviewed articles to government reports.

Currency assesses how recent the source is. In sciences, five years is often the effective horizon for empirical findings; social sciences can typically go back ten years; humanities may draw on decades-old foundational texts. The question is not whether the source is old but whether the field has moved on. A highly cited classic from 1965 may be more relevant to your argument than a poorly constructed 2022 paper.

Authority evaluates who produced the source and whether they are credible. For journal articles, check the author's institutional affiliation (universities and research institutes are more credible than commercial organisations), their prior publications in the field (Google Scholar author profile), and whether the journal is peer-reviewed (check the journal's editorial policy). For books, look for publication by a university press (Oxford, Cambridge, Harvard, MIT, Routledge). For institutional reports, identify the publishing organisation and any known funding conflicts.

Accuracy asks whether the information is supported by evidence. Peer-reviewed articles describe their methodology and data; policy reports cite their sources; academic books provide references. Sources that make strong claims without evidence, that rely on anecdote rather than research, or that present contested findings as established fact should be used with caution — and their limitations noted in your analysis.

Predatory journals are a growing concern in academic research. These are pay-to-publish outlets that claim peer-review status but conduct no meaningful review. Warning signs include: unsolicited email invitations to submit, promises of very fast review (days rather than months), very broad scope claiming all fields, and absence of indexing in Web of Science or Scopus. Check unfamiliar journals using the SHERPA RoMEO database or Ulrichsweb before citing them.

One of the most persistent misconceptions about academic research is that you need to read every source from beginning to end. This is rarely true and often counterproductive. Strategic reading — knowing which parts of a text to prioritise based on your research question — is a core academic literacy skill that most students are never explicitly taught.

The structure of academic papers is designed for selective reading. The abstract summarises the entire paper, including the findings. The introduction identifies the research question and its significance. The conclusion synthesises the argument and states implications. For many purposes — particularly when you are surveying a field to identify relevant sources — reading the abstract and conclusion is sufficient. You only need to read the methods and results in detail if you are evaluating methodological quality or using specific data.

For journal articles, a useful reading sequence is: abstract → conclusion → introduction → results → methods. Read in this order, not linearly from start to end. The abstract tells you if the paper is relevant. If it is, the conclusion tells you what was found and argued. The introduction contextualises this within the field. The results provide the specific data. The methods allow methodological evaluation. This order takes the same time but gives you the most decision-relevant information first.

Annotation is the technique that converts passive reading into active research. Mark claims that directly support or challenge your argument. Note limitations the author acknowledges. Flag passages you may want to quote directly. Record page numbers as you go — tracking down citations retroactively is one of the most time-wasting aspects of student research. Digital annotation tools (PDF highlights, Zotero notes) allow you to review your annotations systematically during the writing phase.

Active reading also means reading with a specific question in mind. Before opening a paper, write your research question at the top of a blank page. As you read, look for what this paper says about that question. This focus prevents the most common reading problem: reading everything the paper says rather than what it says about your question. A paper on working memory and stress may contain three pages on the neuroscience of cortisol response that are entirely irrelevant to your essay on student revision strategies — active reading lets you skip those sections confidently.

The most common note-taking error in academic research is organising notes by source. When you have read four papers and written four pages of notes — one page per paper — you have four source summaries. The problem is that when you write the essay, you will write it as a series of source summaries ("Smith argues X. Jones argues Y. Brown found Z.") because that is how your notes are organised. This produces descriptive, summary-style essays rather than analytical ones.

The alternative is to organise notes by theme or argument, not by source. Before you start reading, decide what questions your essay needs to answer — your sub-claims. As you read, assign each piece of information to one of those sub-claims. If a finding is relevant to sub-claim 1, add it to the sub-claim 1 notes, regardless of which paper it comes from. When you write the essay, each body paragraph will already have its evidence gathered and organised — and you will write from the argument, not from the source list.

A note that is usable in an essay typically contains: the claim or finding in your own words, the citation (author, year, page), a brief note on methodology or limitations if relevant, and a tag for which sub-claim it supports. A note that is not usable is a verbatim transcription of a paragraph with no context about why it matters. Transcription is not note-taking — it is copying. The paraphrase-and-cite form forces you to understand the source well enough to restate it, which is exactly the comprehension you will need to write the essay.

The Cornell notes system is particularly well-suited to research note-taking. The main column captures the content; the cue column on the left holds keywords, questions, or argument tags; the summary at the bottom captures the key takeaway for that source. When reviewing notes before writing, the cue column allows rapid scanning to find relevant material. The Cornell Notes Builder tool on this platform provides a digital version of this system.

Recording page numbers as you read is a discipline that saves significant time later. The temptation is to note the argument without noting the page, intending to go back and find it. In practice, retroactively tracking down a specific sentence in a dense academic paper can take ten minutes or more. Note the page number at the moment of reading — it takes three seconds and saves many minutes later.

Synthesis is the intellectual skill that distinguishes undergraduate from postgraduate writing — and, at undergraduate level, separates first-class from second-class essays. Synthesis means identifying patterns, relationships, and tensions across multiple sources and drawing conclusions from those relationships that no individual source makes on its own. It requires you to treat sources as participants in a conversation, not as isolated authorities.

The most common failure of synthesis is what academic writing researchers call the "patchwork" approach: sources are treated independently, each given a sentence or two, and then a new source is introduced. The result looks like synthesis because it references many sources, but it functions as a list of summaries. The reader cannot see any relationship between the sources because the writer has not identified one.

Genuine synthesis begins before writing — in note-taking. If your notes are organised by sub-claim (as covered in Lesson 4), you will naturally notice patterns when you look at the evidence collected for each sub-claim: multiple sources agreeing, sources disagreeing, sources addressing the same question with different methods, and sources whose combined scope covers more ground than any individual study. These observations are the raw material of synthesis.

The three core synthesis moves are: convergence (multiple sources agree, and their combined weight strengthens the claim), divergence (sources disagree, and the disagreement reveals a significant tension or debate), and qualification (sources agree in general but one or more sources limit the scope of the claim under specific conditions). Each move produces different analytical language: convergence calls for "similarly", "likewise", "consistent with"; divergence calls for "in contrast", "however", "challenges"; qualification calls for "though", "with the exception that", "under conditions of".

The synthesis paragraph structure reverses the source-led structure. Instead of "Smith argues X. Jones argues Y. Both agree on Z," write "X is well-established in the literature, with support from both Smith (2021) and Jones (2020), though each study demonstrates this under different conditions — Smith in laboratory settings, Jones in classroom contexts. This methodological difference suggests that X may be more domain-general than previously assumed." The paragraph leads with the analytical claim, uses sources as evidence, and draws a synthesis conclusion at the end.

Academic plagiarism exists on a spectrum. At one end is intentional fraud — submitting someone else's work as your own, buying an essay, or fabricating sources. At the other end is unintentional plagiarism — paraphrasing without citing, reproducing an argument structure without attribution, or forgetting to cite a fact you learned from a specific source. Most students who face plagiarism findings are at the unintentional end, and preventing unintentional plagiarism is a matter of understanding where the boundaries lie.

The test for whether something needs a citation is simple: is this my original idea, or did I learn it from a source? If you learned it from a source — a fact, a finding, an argument, an interpretation — it needs a citation, even if you have paraphrased it extensively. The only exception is genuinely common knowledge: widely accepted facts that are not controversial and that appear in multiple standard sources without attribution (the earth orbits the sun; Shakespeare lived in the 16th century). When in doubt, cite.

Patchwriting is the most common form of unintentional plagiarism. It involves taking a source's sentence and substituting synonyms while keeping the same structure. Even if every individual word is changed, the intellectual architecture — the way ideas are organised and sequenced — belongs to the original author, and that constitutes plagiarism. The solution to patchwriting is not better synonym selection: it is genuine comprehension. Read the source, close it, write what it said from memory, then compare your version with the original.

AI-generated text creates new plagiarism risks. Using an AI to generate paragraphs you submit as your own is, at most institutions, academic misconduct — regardless of whether AI detection tools flag it. The more subtle risk is using AI summaries of research papers: if the AI has hallucinated citations or misrepresented a study's findings, and you cite those findings as your own independent research, you are both misrepresenting the source and presenting inaccurate information. Always verify AI summaries against the original source.

Self-plagiarism — submitting work you have already submitted for another assignment — is treated as a form of academic misconduct at many institutions, even though the work is genuinely yours. Check your institution's policy before reusing any portion of a previously submitted essay. Different institutions have different thresholds, and some permit reuse with disclosure.