Scientific Research Methods: How to Write the Methods Section

What are scientific research methods in a thesis?

When you open your proposal and see the heading “Scientific Research Methods,” you need to tell the reader how you conducted the study, where the data came from, how you selected the respondents, and which techniques you used to answer the research questions. This section is not a list of technical terms. The reader should be able to follow the process from the research model and questionnaire to the sample, data file, and analysis results.

In an economics, management, marketing, or finance thesis, the methods chapter is often about 8 to 15 pages, depending on your department’s requirements and the complexity of the model. If your study uses a questionnaire and .sav or .csv data, the focus is usually quantitative research: developing scales, collecting data, cleaning the data, and running Cronbach's Alpha, EFA, regression, or SmartPLS. You can also refer to research methods to distinguish broad methodology from a specific research method.

You need to separate two concepts that are often mixed together in thesis writing. Scientific research methodology is the orientation and logic behind the study, such as a quantitative approach, hypothesis development, and testing relationships between variables. Research methods are the concrete procedures, such as a Likert questionnaire survey, convenience sampling, running EFA in SPSS, and testing the model with regression.

Standard structure of a scientific research methods section

You can adjust the subsection titles to match your university’s template. The table below gives you a practical structure for a quantitative thesis. Each section should answer a question your supervisor or the committee could ask.

If your university requires Chapter 3 to be titled “Research Methods,” you can use this order. For SmartPLS, present the measurement model assessment and structural model assessment separately. For SPSS, state the order of Cronbach's Alpha, EFA, correlation, regression, and hypothesis testing if that is the logic you actually used.

You can also refer to scientific research methods textbook pdf, but do not copy a generic process into your thesis. The methods chapter must match your own model, questionnaire, and data file.

Steps for writing the scientific research methods section

Step 1: Finalize the research questions and the data you need

Write a short answer to this question: to test your hypotheses, which variables do you need to observe, and how will you measure them? If you want to test the effect of service quality on satisfaction, you need data on the observed variables for service quality, satisfaction, and the descriptive characteristics of the sample.

Quantitative research is usually suitable for a study that measures the degree of influence or the relationship between variables. You can read more about quantitative research methods to compare the objectives, questionnaire, sample size, and analysis techniques.

Step 2: Choose the research design and explain why

State clearly whether your study is quantitative, cross-sectional or longitudinal, descriptive or explanatory. A study surveying customers at one point in time can be described as cross-sectional quantitative research designed to test effects between variables in the model.

Your rationale must connect to the research question. You can explain that numerical data allow you to measure variables with scales, assess reliability, and estimate the degree of influence between them. Avoid absolute claims about accuracy when you do not have evidence to support them.

If your study is quantitative, keep the methods chapter focused on the questionnaire, numerical data, and the corresponding analysis techniques. The article on qualitative research methods can help you identify differences in data types and objectives, but do not put a method in Chapter 3 if you do not have the corresponding data and procedure.

Step 3: Present the model, variables, and hypotheses

Name each variable exactly as it appears in your research diagram. For example, “Information Quality” and “Trust” may be independent variables, while “Continuance Intention” is the dependent variable. Each arrow in the model needs a corresponding hypothesis, such as H1: Information Quality has a positive effect on Trust.

Do not insert the model diagram and then move directly to the scales. Before each hypothesis, explain its theoretical basis and how the relationship will be tested with survey data. If your model includes a mediator or moderator, state that variable’s role and the planned testing technique.

Step 4: Develop the scales and questionnaire

Create a table containing the variable code, item wording, reference source, and the way the wording was adapted. If you use a 5-point Likert scale, state how levels 1 and 5 are explained in the questionnaire. Likert is the foundation of this type of scale, as presented in (Likert, 1932).

When translating or adapting items, check their meaning against the research context. Do not call a scale “standardized” when you have only translated it from English-language material. In Chapter 3, describe it as a set of reference items adapted to the context of your study.

Step 5: Define the population, sample, and data collection procedure

State who belongs to the research population, what the screening criteria are, and why those respondents fit the research question. Then record the sampling method, planned sample size, collection period, and number of valid questionnaires used in the analysis.

If there are m independent variables in the regression, you can present the reference sample-size rule as n of at least 50 + 8m according to (Tabachnick and Fidell, 2013). For factor analysis, the rule of 5 to 10 observations per item is stated in (Hair et al., 2010). These are reference grounds for planning. They are not a reason to keep a poor-quality sample.

Step 6: Describe data cleaning and analysis

Before running SPSS, check for questionnaires with many unanswered items, one response pattern repeated across the entire questionnaire, out-of-range values, and reverse-coded items. Record the number of questionnaires excluded at each step so that the figures in Chapter 4 match the methods section.

A common process is descriptive statistics, Cronbach's Alpha, EFA, Pearson correlation, linear regression, and hypothesis testing. If you use SmartPLS, describe outer loading, CR, AVE, HTMT, VIF, R², f², Q², and the path coefficient according to the PLS-SEM model.

You can refer to the scientific research process to compare the overall sequence. Do not add CFI, TLI, or RMSEA to a SmartPLS section simply because you saw them in SEM material. Those indices belong to CB-SEM, not to the default assessment criteria for PLS-SEM.

Step 7: Check the method against the data file before submission

Open the data file and check whether every variable in Chapter 3 has a corresponding column. Codes such as SAT1, SAT2, and SAT3 must match the scale table. If Chapter 3 says you have 250 valid questionnaires but the file contains only 237 rows after cleaning, you need to correct one of the two places and explain why.

Keep an analysis log containing the run date, the variables removed, the reason for removal, and the result after each round. It is rare for EFA to give a clean matrix on the first pass. Removing items, running the analysis again, and recording each round is normal, as long as you have a clear methodological reason and are not removing items only to force the results above a threshold.

Usable sentence templates

The sentences below are frameworks that you can replace with your actual information. Do not leave the illustrative figures in your thesis.

How to write this in your thesis

You can adapt the following sentence directly: “The study used a structured questionnaire to collect data from [respondents] in [location] during [period]. After the data were cleaned, they were analyzed using [SPSS or SmartPLS] to assess [scale reliability and validity] and test the hypotheses in the model.”

A shortened methods-section example

Suppose the topic is “Factors Affecting Online Cosmetic Purchase Intention among Students in Ho Chi Minh City.” The model includes three independent variables, Perceived Usefulness, Trust, and Information Quality, and one dependent variable, Purchase Intention. Respondents must be students who have purchased cosmetics online at least once in the past six months.

The methods section could read as follows: “The study used a cross-sectional quantitative design. Data were collected with a 5-point Likert questionnaire from students in Ho Chi Minh City. After questionnaires that failed the screening criteria and questionnaires with missing data were removed, the data were processed in SPSS. The analysis process included descriptive statistics, Cronbach's Alpha assessment, EFA, correlation analysis, and multiple linear regression.”

Replace the sample-size section with your actual information. If the questionnaire contains 20 items and you obtain 230 valid responses, state the number of questionnaires distributed, returned, excluded, and retained. Do not write “230 samples were collected” if the file does not contain 230 valid rows.

The results section must connect back to the methods. If Chapter 3 says you ran EFA, Chapter 4 needs to include KMO and Bartlett's Test, Total Variance Explained, and Rotated Component Matrix. If you say you ran regression, include Model Summary, ANOVA, and Coefficients, together with the appropriate assumption checks.

Common errors that make the committee ask follow-up questions

The first error is writing too generally: “The study uses a quantitative method and a questionnaire survey.” This sentence does not say who was surveyed, how large the sample was, where the scales came from, or how the data were analyzed.

The second error is mixing research methods with methodology. You can describe the methodology as quantitative, but you still need to state how you selected the sample, designed the questionnaire, ran Cronbach's Alpha, conducted EFA, and performed regression.

The third error is using a term without a corresponding procedure. If you write “convergent validity was tested,” identify the indicator used. In PLS-SEM, CR from 0.7 and AVE from 0.5 are commonly used as reference criteria according to (Fornell and Larcker, 1981). HTMT below 0.85, or below 0.90 for closely related constructs, is stated in (Henseler et al., 2015).

The fourth error is copying the sample size, location, or collection period from an old proposal. The committee often asks about these figures because they must match the data table and appendices. Update Chapter 3 after data collection is complete instead of keeping the original plan unchanged.

The fifth error is presenting thresholds as absolute rules. Cronbach's Alpha of 0.7 or above is commonly accepted according to (Nunnally, 1978), while 0.6 may be considered for a new or exploratory scale according to (Hair et al., 2010). You still need to consider the number of items, scale content, Corrected Item-Total Correlation, and the study context.

If you are choosing a topic, you can review the Undergraduate Thesis topics to check whether data collection is feasible before finalizing the model.

Frequently asked questions

What is the difference between scientific research methods and scientific research methodology?

Methodology is the system of perspectives and logic that guides the study. Research methods are the specific ways you collect, process, and analyze data. In a thesis, you can choose a quantitative methodology and implement it through a questionnaire survey, sampling, SPSS, or SmartPLS.

What are the common scientific research methods in economics?

For an economics thesis, common choices include quantitative questionnaire research, secondary data analysis, regression, EFA, SEM, and PLS-SEM. Choose the method based on your research questions, variable types, ability to collect data, and your supervisor’s requirements.

Should I include both qualitative and quantitative methods in Chapter 3?

Include both only when the study actually has the corresponding design and data for both. If you only have a questionnaire and an .sav file, state clearly that the study is quantitative. Adding a method without the corresponding data or procedure will lead the committee to ask about the scope of the study.

How large should the sample be for quantitative research?

There is no single number that works for every study. Consider the number of items, the number of independent variables, the analysis technique, and your department’s requirements. You can present the basis of 5 to 10 observations per item according to (Hair et al., 2010), or n of at least 50 + 8m according to (Tabachnick and Fidell, 2013) when it fits the regression analysis.

What should I write in the methods section if the SPSS results do not meet the threshold?

Keep the method you actually used and report the real results. Check the data, variable coding, reverse-coded items, and questionnaire exclusion criteria before running the analysis again. If you remove an item, record the reason and its effect on the scale. Do not alter the data or ignore an unsatisfactory result just to make the hypotheses look better.

Your next step is to open the data file, create a cross-check table linking the model, items, and data columns, and then revise the methods chapter using your actual information. If you need to check the analysis on your own .sav or .csv file, you can use DoThesis M4 analysis.