How to describe research results in a diploma thesis?

First, prepare and check your data

Before analysis, remove duplicates, check for missing values, incorrect values ​​and the way the responses are coded. Write down your rules for excluding records so that you can later accurately describe the final sample size.

Variable and category names should be consistent. A separate code dictionary makes work easier, especially when multiple-choice questions or open answers are divided into several columns.

Structure by research problems

Start with the sample characteristics and then move through the research questions. For each part, recall the analyzed problem, show the most important data and end with a short answer resulting from the results.

The order of questions in the form serves the convenience of the respondent, not necessarily the logic of the chapter. Combine results from different items if together they describe the same construct or issue.

Table, chart or text description?

A table is good for exact values ​​and multiple categories, a chart for showing a trend or comparison, and text for a single simple result. Don't present the same data in three ways at the same time.

Each object should have a number, an informative title, a source, and an in-text reference. Describe the pattern, major differences, and values ​​relevant to the question rather than typing out all the cells.

Qualitative results and citations

In qualitative research, organize the material into categories and topics. Explain how they were identified, and then describe their meaning, similarities, differences, and outliers.

Quotes from interviews are intended to illustrate the conclusion, not to replace analysis. Select short, representative fragments, anonymize interlocutors and mark statements in accordance with the adopted system.

Interpretation, discussion and limitations

Interpretation answers the question of what the result might mean in the context of the problem and theory. Compare results with previous studies, but do not hide discrepancies. They often lead to interesting conclusions.

The scope of the proposals must correspond to the method of sampling and the quality of the tool. With a convenience sample, the results should not be automatically generalized to the entire population. Present the limitations factually, along with a proposal for further research.

Organise the chapter by research questions, not column order

The results chapter should guide readers through the research problems. The survey questions may have followed a different order for participant convenience. Before writing, assign the necessary tables, charts and qualitative excerpts to each problem. Remove displays that answer no question and do not help describe the sample. A software export alone is not a logically organised empirical chapter.

Within each part, use a clear sequence: recall the problem, present the data, describe the main pattern and draw a cautious conclusion. For extensive analyses, distinguish descriptive findings from tests or modelling. Do not make readers guess why another table appears. The introductory sentence should explain its purpose, and the following sentence should identify information relevant to the thesis objective.

An example of describing percentages with the correct denominator

Educational example: of the 80 people who answered an item, 48 selected a particular category. That is 60% of valid responses to that item. If the study had 90 participants overall, disclose the missing answers rather than presenting the percentage without its base. For a multiple-response question, percentages may total more than 100%. Readers need to know whether the percentage refers to people or to all selections.

The description should not repeat every number in the table. Highlight the dominant pattern, relevant differences and exceptions needing comment. Do not replace data with judgements such as very many when a clear proportion can be given. Round consistently and do not imply greater precision than the measurement supports. The example figures only illustrate the calculation; use your own verified data in an actual thesis.

A descriptive result is not yet evidence of an association

A difference between group means or percentages is initially an observation in the collected sample. Calling it statistically significant requires an appropriate analysis and valid conditions for inference. Statistical significance, in turn, does not establish practical importance. When reporting a test, provide the information needed to assess the method, effect size and uncertainty, in accordance with the chosen reporting approach.

Avoid the word impact when the design only assesses an association or a reported opinion. If participants say a solution helped them, you are reporting their assessment, not automatically objectively established effectiveness. Similarly, no clear difference does not prove the groups are identical. Precise language keeps data, method and conclusions aligned. When uncertain, return to the question that the analysis was meant to answer.

How to describe qualitative material

Present interview findings through categories and patterns related to the research problem. Explain what connects the accounts, how they differ and under which conditions exceptions occur. Brief quotations may illustrate interpretation but should not replace analysis. Use participant labels consistently and ensure that contextual details do not unnecessarily reveal identities. A selected quotation must preserve the speaker's meaning.

Do not automatically equate how often a theme appears with its importance. A rarely mentioned experience may be crucial to the mechanism being studied. If using counts, explain exactly what was counted and why. Do not generalise a few interviews to an entire population. Qualitative reporting should also present a case that does not fit the dominant pattern and discuss how it affects interpretation, rather than removing it from the narrative.

Results, discussion and conclusion serve different purposes

Results present what was obtained from the material. Discussion relates findings to the literature, considers explanations and identifies limitations. The conclusion returns to the objective and synthesises the answer. Chapter arrangements can vary, but these functions should remain distinct. If you combine results and discussion, signal the transition from data to interpretation. Do not present your own conjecture as information directly established by a table.

Before submitting the chapter, check agreement between tables and the dataset, text and tables, and conclusions and questions. Identify sample-size changes between analyses and explain why they occurred. All results should come from the same agreed dataset version. Retain calculation files and the data-preparation rules. This makes discrepancies and questions from your supervisor or reviewer easier to address without reconstructing the whole process from memory.

Summary

A good results chapter takes the reader from the research question to a data-driven answer. Tables and graphs help you see patterns, but it's the text that explains their meaning and limitations. Select only information needed to achieve your goal, be consistent in your presentation, and do not draw broader conclusions than the sample allows. Thanks to this, the analysis remains clear and reliable.