For a university student of English linguistics, the phrase “scientific research” might initially seem out of place. Linguistics is not about test tubes or microscopes, but it is a science nonetheless—the systematic study of language structure, use, acquisition, and change. Conducting scientific research in English linguistics means moving beyond personal intuition or grammar-book rules and instead adopting a methodical, evidence-based approach to understanding how English works in the real world. Whether you are investigating how first-year students use hedges in academic writing, analyzing vowel shifts in local English dialects, or testing whether explicit grammar instruction improves second-language writing, the same core scientific principles apply: ask a clear question, gather data, analyze systematically, and draw conclusions with caution.

 1. From Curiosity to a Researchable Question

All scientific research in linguistics begins with a genuine puzzle. You might notice that your peers use “like” as a filler far more often in casual conversation than in classroom presentations. Or you might observe that some English past-tense forms (“texted” instead of “text”) seem to be changing. Transform these observations into a focused, testable question. Avoid questions that are too broad (“How do people use English?”) or purely subjective (“Is British English better than American English?”). Instead, aim for precision: “To what extent do native and non-native English-speaking university students differ in their use of discourse markers such as ‘well,’ ‘so,’ and ‘like’ in informal interviews?” A good research question in linguistics identifies the variables (e.g., native vs. non-native status, type of discourse marker), the population (university students), and the context (informal interviews).

 2. Grounding Your Work: The Literature Review

Before you collect a single word of data, immerse yourself in previous research. Use academic databases such as the Linguistics and Language Behavior Abstracts (LLBA), Google Scholar, or your university library’s journal portal. Search for keywords related to your topic—for example, “discourse markers + L2 English” or “sociolinguistic variation + discourse markers.” Read not only findings but also methods: How did other researchers record and transcribe speech? How did they quantify “native-like” use? A thorough literature review helps you refine your question, avoid known pitfalls, and position your study within the ongoing academic conversation. It also teaches you the theoretical frameworks (e.g., variationist sociolinguistics, corpus linguistics, psycholinguistics) that will shape your own analysis.

 3. Research Design and Data Collection

The heart of scientific research is systematic, replicable data collection. In English linguistics, data typically come in one of three forms:

- Corpus data: Using existing collections of spoken or written English (e.g., the Corpus of Contemporary American English, the British National Corpus). This is ideal for large-scale pattern analysis. You might search for occurrences of a specific syntactic structure across different genres.

- Experimental data: Designing tasks to elicit specific language behaviors. For example, you could show participants pictures and ask them to describe them, measuring reaction times or error rates in producing passive sentences.

- Naturalistic observation: Recording and transcribing authentic conversation, interviews, or classroom talk. This requires ethical approval and informed consent, especially if recording human subjects.

Whichever method you choose, be rigorous about documentation. Record your search parameters, your coding scheme for grammatical features, or your transcription conventions. Without such transparency, your research loses its scientific credibility.

 4. Data Analysis: Moving from Examples to Evidence

It is tempting to pick a few vivid examples that illustrate your idea. That is not science. Scientific analysis requires systematic handling of all your data. If you are studying hesitation markers in student presentations, you must count every occurrence across all speakers, not just the most dramatic cases. In quantitative approaches, use descriptive statistics (frequencies, averages) and, if appropriate, inferential tests (chi-square, t-tests) to determine whether observed differences are likely due to chance. In qualitative approaches, such as conversation analysis, you need to examine each instance of a phenomenon (e.g., overlapping speech) and develop a rule-based explanation that accounts for all cases, not just the convenient ones.

For example, if you hypothesize that female students use more tag questions (“It’s a nice day, isn’t it?”) than male students, you must transcribe equal amounts of speech from each group, define exactly what counts as a tag question, calculate frequencies per thousand words, and then test whether the difference is statistically significant. Software like AntConc (for corpus analysis), ELAN (for transcription), or even Excel can become your laboratory tools.

 5. Interpretation and the Art of Skepticism

Once you have results, resist the urge to overclaim. Science in linguistics demands humility. If you find that non-native speakers use “so” less frequently than native speakers, consider alternative explanations: maybe they were more nervous in interviews, or maybe the interview topic favored certain discourse patterns. Acknowledge the limitations of your sample (only 20 students from one university) and your methods (short interviews may not capture natural speech). Compare your findings to the literature: Do they support previous studies or challenge them? If they challenge, what might explain the discrepancy?

Also, be honest about unexpected results. Did you predict that formal instruction would reduce the use of “like,” but instead it increased? That is not a failure; it is a discovery. Science progresses not by proving our hypotheses right but by testing them against reality.

 6. Writing and Peer Review: Sharing Your Work

Your research is not complete until you share it. Write a clear, structured paper following the IMRaD format (Introduction, Methods, Results, and Discussion), which is standard in linguistics journals. In the Methods section, provide enough detail that another student could replicate your study exactly. In the Results section, present tables or figures, but do not interpret them yet—just state what you found. Save interpretation for the Discussion. Use the linguistic terminology correctly, but define any specialized terms (e.g., “pro-drop,” “copula deletion”) for a broader academic audience.

Finally, seek feedback. Present your findings to a class or a student research conference. Submit to an undergraduate research journal. Listen to criticism as a gift: a peer who asks “How did you decide which fillers to count?” is helping you strengthen your next study.

 In conclusion, conducting scientific research in English linguistics as a university student is not about possessing innate genius or expensive equipment. It is about learning a disciplined way of asking questions about human language and letting the data—not your intuitions or opinions—provide the answers. You will face frustrations: ambiguous recordings, contradictory statistics, and the daunting blank page of a research proposal. But you will also experience the quiet thrill of noticing a pattern no one has described before, of testing a claim and watching it hold, or even of watching it shatter and thereby opening a new line of inquiry. That is the scientific spirit. And it begins wherever you are—in the library, at a café with a recorder, or scrolling through a corpus—with a single, well-formed question about the English language.