Writing a Quantitative Methodology Chapter

Written By

John Miller

✔️ 97% Satisfaction | ⏰ 97% On Time | ⚡ 8+ Hour Delivery

Quantitative methodology chapters terrify students. All those terms, variables, validity, reliability, sampling, it feels impenetrable. But actually, a quantitative methodology chapter's explaining what you measured, how you measured it, who you measured it in, and how you analysed it. Demystify it and it's far more manageable.

Your methodology chapter's where you establish that you've designed appropriate research to answer your question and that you've executed it properly. Examiners assess: does this approach actually answer the research question? Are the measurements valid? Is the sample appropriate? Has the analysis been conducted correctly?

Establishing Your Research Design

Start by identifying your design. Are you conducting a survey? An experiment? A longitudinal study examining change over time? A comparative study examining differences between groups?

Survey research typically aims to describe populations or examine relationships. You might survey students about their study habits, examining whether study hours correlate with grades. You're not manipulating anything; you're observing what exists.

Experimental research involves manipulating variables. You might randomly assign students to receive peer tutoring or not, then compare their grades. The tutoring's the manipulated variable. You're testing causation.

Quasi-experimental research resembles experiments but lacks some experimental features. You might compare grades between a class that received peer tutoring and a class that didn't, but students weren't randomly assigned to groups. It's more rigorous than pure observation but less rigorous than true experiments.

Longitudinal research follows people over time. You might survey first-year students annually across their degree, examining how engagement changes.

Describe your design and justify it. Why does this design answer your research question?

Identifying Your Variables

Variables are things you measure. Some vary between people (age, study hours, anxiety levels). Some are what you're predicting (grades, employment, wellbeing).

Identify your variables explicitly. Name them. Define them operationally (how you actually measured them). If your variable's "academic achievement," does that mean exam grades? Course grades? GPA? How do you calculate it from raw data?

Distinguish between independent variables (things you predict will influence outcomes) and dependent variables (what you're trying to predict). If you're examining whether hours of study influence grades, hours of study's your independent variable and grades are your dependent variable.

Some variables are categorical (group membership, male/female, UK/international). Some are continuous (measured numerically, age, study hours, test scores). Some are nominal (categories without meaningful order, university name, subject studied). Some are ordinal (categories with order, agreement on a Likert scale). Different analyses suit different variable types. Make sure you've correctly identified your variables' types because that shapes your analysis.

Describing Your Sample

The process of synthesising multiple sources into a coherent argument is at the heart of what makes dissertation writing different from other forms of academic assessment that you may have encountered during your studies.

Who did you measure? Describe your sample thoroughly.

How many people? Sample size matters statistically. Too small, and results are unreliable. Your power analysis (a statistical calculation) determines adequate sample size based on your design and analysis. Conduct power analysis before data collection.

Who are they? Age, gender, background, educational level, whatever's relevant. A sample of 100 university lecturers differs from a sample of 100 first-year students.

How did you recruit them? Did you access school rosters and select randomly? Did you recruit volunteers through online forums? Your recruitment method shapes who participates and whether results might generalise.

Students who begin their writing early in the academic year give themselves the time they need to produce multiple drafts and refine their argument through careful iteration rather than rushing to meet a single deadline.

Is your sample representative? If you're studying first-year students nationally but you've only recruited from universities in the southeast, findings might not generalise. Be honest about sample limitations.

Describe your inclusion and exclusion criteria. What characteristics qualified people for participation? Were there groups you deliberately excluded? Why?

Measurement: How You Actually Measured Things

For each variable, explain how you measured it.

If you used an existing questionnaire or assessment, describe it. What does it measure? How valid and reliable is it? Citation the source.

If you developed your own questionnaire, describe the questions. How many questions measured each construct? What response format (yes/no, Likert scales, open-ended)? Did you pilot-test it? Did you refine it based on pilot results?

If you're measuring existing data (exam grades from university records, absence records), explain what you measured and how you accessed it.

Define reliability and validity for each measure. Reliability means consistency, would the measure yield similar results if you measured the same person twice? Validity means accuracy, does the measure actually measure what you claim?

Describing Your Procedure

What did you actually do? How did you collect data?

Did you conduct a questionnaire in person? Online? By post? This affects response rates and data quality. Online questionnaires have lower response rates. In-person questionnaires yield better response but are more labour-intensive.

How long did data collection take? Did you collect all data at once or across time?

How did you ensure participants understood and consented? Did you provide written information? Obtain consent?

Students who develop the habit of writing regularly throughout their final semester rather than leaving everything for the final few weeks tend to produce work that demonstrates more careful thought, stronger structure, and a more confident academic voice than those who resort to last-minute marathon sessions.

What were the conditions of data collection? Did all participants complete the questionnaire in similar conditions? Or did conditions vary? Consistency strengthens research.

Statistical Analysis: What You Actually Did

Describe your analysis in detail. Don't assume readers know standard procedures.

Did you explore data first? Descriptive statistics show how data distributes. You might report means, standard deviations, ranges. These give readers a sense of your data before you conduct inferential analyses.

Did you examine relationships? Correlations show whether variables covary. If study hours and grades correlate positively, more study time associates with higher grades. Specify the correlation coefficient and whether it's statistically considerable.

Did you compare groups? T-tests compare means between two groups. ANOVA compares means across multiple groups. Describe which test you used and why.

Did you examine prediction? Regression analysis examines how well one variable predicts another. Multiple regression examines how multiple variables together predict an outcome.

Be clear about which tests you conducted, why, and what results you found. Include effect sizes (how large differences are), not just p-values (whether differences are statistically considerable). A difference can be statistically considerable but so small it's meaningless practically.

Addressing Validity and Potential Limitations

Discuss threats to validity. Internal validity concerns whether your design actually tested your hypothesis. External validity concerns whether findings would generalise beyond your sample. Statistical conclusion validity concerns whether statistical conclusions are accurate.

If you used a questionnaire, does it actually measure what you claim? Could unmeasured variables explain your findings?

If you compared groups, were groups comparable before your intervention? If they differed initially, conclusions about intervention effects are ambiguous.

Discuss your sample's limitations. If your sample was small, convenience-recruited, or particular to a certain context, findings might not generalise.

The challenge of balancing breadth and depth in your dissertation is one that every student faces, and the best approach is to focus on depth in your analysis while providing enough context for the reader to follow.

Acknowledge limitations honestly. Every study has limitations. Acknowledging them demonstrates understanding of research standards. It doesn't weaken your work; it strengthens it.

Ethical Considerations

Describe ethical approval. Did you obtain formal approval from your university's ethics committee? Describe the approval process and what was approved.

Planning your time effectively across the dissertation period means breaking down the overall task into manageable weekly goals and building in extra time for the unexpected delays that inevitably arise during research.

How did you ensure informed consent? How did you protect confidentiality? How is data stored securely?

If your research involved any risk or discomfort, how did you minimise harm? If participants could withdraw, how did you communicate this option?

Frequently Asked Questions

Q: What sample size do I need? A: Conduct power analysis. This statistical calculation determines adequate sample size based on your hypothesis, your analysis, and your desired statistical power (typically 0.80, meaning 80 per cent chance of detecting a real effect). Power analysis requires technical knowledge; your supervisor or a statistician can help. Generally, for survey research, 100-200 participants suffices for many analyses. For experiments, 30-50 per group is often adequate. These are rough guides. Power analysis gives you the specific answer for your research.

Q: Can I use existing data from my workplace? A: Maybe. If using existing data (absence records, grades, employment outcomes), you still need ethical approval. Using data without people's knowledge or consent raises ethical issues. If data's anonymised and you're not identifying individuals, ethics issues lessen. But check with your ethics committee. They'll advise whether formal approval's needed.

Q: Should I use parametric or nonparametric statistics? A: Parametric tests (t-tests, ANOVA, Pearson correlation) assume your data meets certain conditions (normally distributed, equal variances). Nonparametric tests (Mann-Whitney U, Kruskal-Wallis, Spearman correlation) don't assume these. If your data violates assumptions, nonparametric tests are more appropriate. Check your data and consult statistics textbooks or your supervisor. Different tests suit different data characteristics.