Quantitative data analysis in a dissertation uses statistics to test relationships or differences between numerical variables, while qualitative data analysis interprets non-numerical data — such as interview transcripts — to understand meaning, experience, and context. The right choice depends on your research question, not on personal preference for numbers or words.
Quantitative vs. Qualitative Data Analysis At a Glance
| Factor | Quantitative | Qualitative |
|---|---|---|
| Main purpose | Measure variables, test relationships or differences | Understand meaning, experience, and context |
| Data type | Numerical (counts, scores, measurements) | Non-numerical (text, narratives, observations) |
| Typical research questions | How much? How often? Is there a relationship or difference? | Why? How? What is the experience of...? |
| Common data sources | Surveys, experiments, structured questionnaires, existing datasets | Interviews, focus groups, observations, documents |
| Common analysis methods | Descriptive/inferential statistics, correlation, regression, t-tests, ANOVA | Thematic analysis, content analysis, narrative analysis, grounded theory |
| Sample characteristics | Typically larger, aiming for statistical power | Typically smaller, aiming for depth and information richness |
| Role of statistics | Central to hypothesis testing and generalisation | Rarely used, or used descriptively alongside coding |
| Role of researcher interpretation | Present but constrained by statistical procedure | Central to identifying and naming patterns |
| Common software | SPSS, R, Stata, Excel | NVivo, ATLAS.ti, MAXQDA |
| Typical outputs | Tables, p-values, effect sizes, statistical models | Themes, categories, illustrative participant quotations |
| Main strength | Generalisability and precision when assumptions hold | Depth, context, and insight into "why" |
| Main limitation | May miss context behind the numbers | Findings are typically not statistically generalisable |
| Best suited dissertation questions | Testing predictors, comparing groups, measuring extent | Exploring lived experience, perceptions, processes |
Main purpose
Quantitative
Measure variables, test relationships or differences
Qualitative
Understand meaning, experience, and context
Data type
Quantitative
Numerical (counts, scores, measurements)
Qualitative
Non-numerical (text, narratives, observations)
Typical research questions
Quantitative
How much? How often? Is there a relationship or difference?
Qualitative
Why? How? What is the experience of...?
Common data sources
Quantitative
Surveys, experiments, structured questionnaires, existing datasets
Qualitative
Interviews, focus groups, observations, documents
Common analysis methods
Quantitative
Descriptive/inferential statistics, correlation, regression, t-tests, ANOVA
Qualitative
Thematic analysis, content analysis, narrative analysis, grounded theory
Sample characteristics
Quantitative
Typically larger, aiming for statistical power
Qualitative
Typically smaller, aiming for depth and information richness
Role of statistics
Quantitative
Central to hypothesis testing and generalisation
Qualitative
Rarely used, or used descriptively alongside coding
Role of researcher interpretation
Quantitative
Present but constrained by statistical procedure
Qualitative
Central to identifying and naming patterns
Common software
Quantitative
SPSS, R, Stata, Excel
Qualitative
NVivo, ATLAS.ti, MAXQDA
Typical outputs
Quantitative
Tables, p-values, effect sizes, statistical models
Qualitative
Themes, categories, illustrative participant quotations
Main strength
Quantitative
Generalisability and precision when assumptions hold
Qualitative
Depth, context, and insight into "why"
Main limitation
Quantitative
May miss context behind the numbers
Qualitative
Findings are typically not statistically generalisable
Best suited dissertation questions
Quantitative
Testing predictors, comparing groups, measuring extent
Qualitative
Exploring lived experience, perceptions, processes
What Is Quantitative Data Analysis in a Dissertation?
Quantitative data analysis is the process of applying statistical procedures to numerical data in order to describe patterns, test hypotheses, or examine relationships between variables. It is grounded in measurement: every variable in a quantitative dissertation is operationalised so it can be recorded as a number, score, or category that can be counted.
In a dissertation, this usually entails defining variables beforehand (independent, dependent, and occasionally control or moderating variables), gathering data using tools like surveys, scales, or experiments, and then using descriptive statistics (means, frequencies, standard deviations) and inferential statistics (tests that permit conclusions to extend beyond the immediate sample, within specified limits). Inferential procedures — correlation, regression, t-tests, ANOVA, chi-square — are used to test hypotheses about relationships or differences between groups.
Although it is not always the case, generalization is a typical objective of quantitative research. Sampling approach, sample size, response rate, and the degree to which the study design accounts for confounding variables all have a role. Rather than presuming that statistical significance equates to broad applicability, a dissertation should expressly disclose these restrictions.
Example 1: A master's dissertation testing whether remote-work frequency predicts self-reported job satisfaction among 250 employees, using multiple regression.
Example 2:A PhD dissertation comparing anxiety scores between two teaching interventions using a pre-test/post-test design and an independent-samples t-test.
What Is Qualitative Data Analysis in a Dissertation?
Qualitative data analysis is the systematic process of examining non-numerical data — such as interview transcripts, focus group recordings, field notes, or documents — to identify patterns of meaning and produce an interpretive account of a phenomenon. Rather than measuring variables, qualitative analysis works with the language participants use to describe their experiences, perceptions, and the context surrounding them.
This procedure involves more than just reading transcripts and summarizing what was said. It involves structured coding: labelling meaningful segments of text, grouping codes into categories, and developing broader themes that are then reviewed, refined, and connected back to the research question. The researcher's interpretation is central to this process, which is why qualitative dissertations typically discuss credibility, dependability, and reflexivity rather than statistical validity.
Example 1: A dissertation exploring how first-generation university students describe their sense of belonging, based on 15 semi-structured interviews analysed using thematic analysis.
Example 2: A doctoral study examining how hospital nurses construct meaning around workplace burnout, using grounded theory to develop a conceptual model from interview and observation data.
Quantitative vs. Qualitative Data: What Is the Difference?
The fundamental difference between the two traditions is not that one is "objective" and the other is "subjective"; rather, both incorporate researcher judgment and strive to generate reliable, defendable results utilizing various evidence logics.
-
Nature of data: Quantitative data is numerical and standardised across participants; qualitative data is textual, visual, or narrative and often varies in form from one participant to the next.
-
Structure: Quantitative instruments (scales, structured questionnaires) are typically fixed before data collection; qualitative instruments (interview guides) are often semi-structured and can evolve as the study progresses.
-
Research purpose: Quantitative analysis tests pre-specified hypotheses or measures the extent of a relationship; qualitative analysis explores and interprets meaning, often without a fixed hypothesis.
-
Sample and analysis logic: Quantitative dissertations generally prioritise sample size and representativeness for statistical power; qualitative dissertations prioritise information-rich cases and depth over breadth.
-
Researcher role: In quantitative work the researcher applies predetermined statistical procedures; in qualitative work the researcher actively constructs categories and themes from the data, which is why reflexivity is reported as part of rigour.
-
Output and generalisation: Quantitative findings are typically expressed as statistics with an explicit generalisation claim (or its limits); qualitative findings are typically expressed as themes supported by evidence, with transferability — not statistical generalisability — as the relevant standard.
How Do You Know Which Data Analysis Your Dissertation Needs?
Instead of starting with your personal preference for data or interviews, start with your research question, objectives, hypotheses, and general study design. Your question should naturally lead to the analytical method. This is the core of the decision framework: research question → research design → data type → analysis method → interpretation.
If your research question asks:
- How many? How much? How often? To what extent?
- Is there a relationship between X and Y?
- Is there a difference between groups?
- Does X predict Y?
→ Quantitative analysis may be appropriate.
If your research question asks:
- Why does this happen?
- How do participants experience...?
- What does this mean to participants?
- How do people perceive or make sense of...?
- What are the lived experiences of...?
→ Qualitative analysis may be appropriate.
Heuristics, not strict laws
These are not strict laws, but rather heuristics. Certain professions have their own traditions regarding which method is required for specific situations, and some questions actually call for both views. Before deciding on a methodology, always compare your question to your research design and the methodological requirements of your department or supervisor.
Research Question → Likely Analysis Method
| Example dissertation question | Likely approach | Likely data | Possible analysis |
|---|---|---|---|
| Does social media use predict academic performance? | Quantitative | Numerical survey data | Correlation/regression |
| Is there a difference in stress scores between two groups? | Quantitative | Scale scores | t-test |
| What factors are associated with employee turnover? | Quantitative | Numerical/categorical variables | Regression/other appropriate tests |
| How do nurses experience workplace burnout? | Qualitative | Interview transcripts | Thematic analysis |
| Why do postgraduate students struggle with research writing? | Qualitative | Interviews/open responses | Thematic analysis |
| How do first-generation students describe their university experience? | Qualitative | Narratives/interviews | Thematic/narrative analysis |
Likely approach
Quantitative
Likely data
Numerical survey data
Possible analysis
Correlation/regression
Likely approach
Quantitative
Likely data
Scale scores
Possible analysis
t-test
Likely approach
Quantitative
Likely data
Numerical/categorical variables
Possible analysis
Regression/other appropriate tests
Likely approach
Qualitative
Likely data
Interview transcripts
Possible analysis
Thematic analysis
Likely approach
Qualitative
Likely data
Interviews/open responses
Possible analysis
Thematic analysis
Likely approach
Qualitative
Likely data
Narratives/interviews
Possible analysis
Thematic/narrative analysis
How to Analyze Quantitative Data in a Dissertation
The general workflow below is adapted to fit each design, statistical test, and discipline; not every study will need every step.
-
Define research questions and hypotheses.
-
Identify and operationalise variables.
-
Prepare and code data for entry.
-
Clean the dataset (duplicates, outliers, formatting errors).
-
Check for missing values and handle them appropriately.
-
Assess statistical assumptions relevant to the planned test (e.g., normality, homogeneity of variance).
-
Conduct descriptive analysis (means, frequencies, distributions).
-
Select the inferential test appropriate to the research question and data type.
-
Run the analysis using statistical software.
-
Interpret the results in relation to the hypotheses.
-
Present findings in tables and figures.
-
Relate findings back to the research questions and the existing literature.
How to Analyze Qualitative Data in a Dissertation
Qualitative analysis is not simply reading interviews and writing a summary — it is an iterative, documented process.
-
Prepare and transcribe raw data (interviews, recordings, field notes).
-
Read and re-read the data for familiarisation.
-
Generate initial codes across the dataset.
-
Group codes into candidate categories or patterns.
-
Identify broader themes across categories.
-
Review and refine themes against the coded data and the full dataset.
-
Define and name each theme clearly.
-
Interpret what the themes mean in relation to the research question.
-
Support interpretations with direct evidence (e.g., participant excerpts).
-
Connect findings to the research question and relevant literature.
-
Write the findings/analysis chapter, structured around the themes.
Data Type Does Not Determine Analysis Method
Having qualitative data does not automatically mean you must use one specific qualitative analysis method, and having quantitative data does not automatically determine which statistical test to run.
Two dissertations can both use interview transcripts yet require different analytical approaches — one might call for thematic analysis, another for narrative analysis, another for grounded theory — depending on the research question and theoretical framework. Similarly, two dissertations can both use numerical survey data yet require different statistical tests depending on the number of variables, their measurement level, the study design, and the specific hypothesis being tested. The data type tells you the general analytical family; it does not select the specific method for you.
Common Quantitative Data Analysis Methods
Descriptive statistics summarise the sample (means, frequencies, standard deviations) without testing hypotheses.
Inferential statistics allow conclusions about a population to be drawn from sample data, within stated confidence levels.
Correlationexamines the strength and direction of association between two numerical variables.
Regressionmodels how one or more predictor variables relate to an outcome variable.
t-tests compare means between two groups or two conditions.
ANOVA compares means across three or more groups.
Chi-square tests association between categorical variables.
Non-parametric tests (e.g., Mann-Whitney U, Kruskal-Wallis) are used when data do not meet the assumptions required for parametric tests.
Not every quantitative dissertation needs every test. The correct choice depends on your research question, the type and measurement level of your variables, your study design, the assumptions each test requires, your sample characteristics, the distribution of your data, and your specific analytical objective.
Common Qualitative Data Analysis Methods
Thematic Analysis
Identifies, analyses, and reports patterns (themes) across a dataset; it is broadly applicable and widely used in dissertations.
Content Analysis
Systematically categorises textual data, often with a more structured or quantifiable coding frame than thematic analysis.
Narrative Analysis
Focuses on how participants structure and tell stories about their experiences, attending to sequence and voice.
Grounded Theory
Aims to develop a theoretical model or explanation grounded directly in the collected data, through iterative coding and theoretical sampling.
Framework Analysis
Uses a structured matrix of cases and themes, often favoured in applied policy or health research.
These are distinct methods with different aims and procedures — qualitative analysis is not synonymous with thematic analysis, and choosing among them should be justified by your research question and theoretical approach, not convenience.
Tools and Software for Dissertation Data Analysis
| Analysis Type | Common Tools | Typical Use |
|---|---|---|
| Quantitative | SPSS | Statistical analysis |
| Quantitative | R | Statistical computing and modelling |
| Quantitative | Stata | Statistical analysis |
| Quantitative | Excel | Basic data preparation and descriptive analysis |
| Qualitative | NVivo | Coding and thematic analysis |
| Qualitative | ATLAS.ti | Qualitative coding and analysis |
| Qualitative | MAXQDA | Qualitative and mixed-methods analysis |
Analysis Type
Quantitative
Typical Use
Statistical analysis
Analysis Type
Quantitative
Typical Use
Statistical computing and modelling
Analysis Type
Quantitative
Typical Use
Statistical analysis
Analysis Type
Quantitative
Typical Use
Basic data preparation and descriptive analysis
Analysis Type
Qualitative
Typical Use
Coding and thematic analysis
Analysis Type
Qualitative
Typical Use
Qualitative coding and analysis
Analysis Type
Qualitative
Typical Use
Qualitative and mixed-methods analysis
Software is a tool; the research question and methodological design determine the appropriate analysis, not the software's capabilities. Learning a program does not substitute for understanding the underlying methodology it supports.
Can a Dissertation Use Both Quantitative and Qualitative Analysis?
Yes. Mixed-methods research deliberately combines quantitative and qualitative approaches within a single study when the research problem genuinely calls for both breadth and depth. Researchers combine methods for triangulation (cross-checking findings from different sources) or complementarity (using one approach to explain or extend the other).
Common designs include sequential designs — one phase informs the next, such as a survey followed by interviews — and concurrent designs, where both types of data are collected around the same time and integrated during interpretation.
Example: A dissertation surveys 300 employees to identify a statistical pattern in remote-work satisfaction, then conducts follow-up interviews with a subset of participants to explain why that pattern exists. The quantitative phase identifies the "what"; the qualitative phase explains the "why."
Mixed methods should be justified by the research problem itself — not used simply to make a dissertation appear more sophisticated. A mixed-methods design adds complexity in data collection, analysis, and word count, so it should only be chosen when a single approach genuinely cannot answer the research question.
Is Quantitative or Qualitative Data Analysis Easier for a Dissertation?
Neither approach is universally easier; the difficulty depends on your research design, data volume, statistical or coding expertise, software familiarity, and your department's methodological expectations.
Quantitative analysis often requires stronger upfront statistical planning: choosing the correct test, checking assumptions, and correctly interpreting output can be technically demanding, especially for students without a strong statistics background. Qualitative analysis, by contrast, is frequently time-intensive in a different way — transcription, iterative coding, and interpretation can take considerably longer than running a statistical test, even though the underlying logic may feel more intuitive to some students. In practice, students who are comfortable with numbers may find quantitative work more manageable, while students comfortable with language and interpretation may find qualitative work more natural — but both require rigour and practice to do well.
Dissertation Examples Across Disciplines
The examples below are illustrative of typical dissertation designs in five common disciplines; they do not represent fabricated study findings.
1. Education
Research question
Does the use of gamified learning apps improve maths test scores among secondary school students?
Data type
Pre/post test scores
Analysis
Paired-samples t-test or ANCOVA
Expected result
A statistically reported difference (or lack thereof) in mean scores, with effect size reported.
2. Healthcare/Nursing
Research question
How do intensive care nurses experience moral distress during end-of-life care decisions?
Data type
Semi-structured interview transcripts
Analysis
Thematic analysis
Expected result
A set of named themes describing dimensions of moral distress, supported by participant excerpts.
3. Business/Management
Research question
What is the relationship between transformational leadership style and employee engagement in SMEs?
Data type
Survey scale responses
Analysis
Correlation and multiple regression
Expected result
Reported strength and direction of the relationship, with statistical significance and limitations noted.
4. Psychology/Social Sciences
Research question
How do young adults construct their identity following a career change?
Data type
Narrative interviews
Analysis
Narrative analysis
Expected result
An interpretive account of identity construction organised around participants' storylines
5. Engineering/Technology
Research question
What factors influence user trust in AI-based recommendation systems?
Data type
Survey (trust scale) plus open-ended user interviews
Analysis
Descriptive/inferential statistics for the survey; thematic analysis for interview data
Expected result
A statistically described trust pattern, explained and contextualised by qualitative themes
Common Mistakes When Choosing a Dissertation Data Analysis Method
-
Choosing the analysis method before clearly defining the research question.
-
Choosing quantitative analysis because it "looks more scientific".
-
Relying on older theses without checking whether newer literature has already closed the gap.
-
Choosing qualitative analysis because statistics seem difficult.
-
Assuming interviews are automatically qualitative data (they can support content analysis with quantifiable coding, for example).
-
Assuming surveys are always quantitative (open-ended survey questions can generate qualitative data).
-
Selecting a statistical test without checking whether its assumptions are met.
-
Confusing data collection methods with data analysis methods.
-
Treating coding as the entirety of qualitative analysis, rather than one step in a longer interpretive process.
-
Using software (SPSS, NVivo) without understanding the underlying methodology it is meant to support.
-
Overclaiming generalisability from a qualitative or small quantitative sample.
-
Reporting statistical or thematic results without genuinely interpreting what they mean.
-
Failing to align the chosen analysis with the stated research objectives.
How to Present Quantitative or Qualitative Analysis in a Dissertation
Quantitative dissertations generally need to report: the variables studied, sample characteristics, the statistical methods used, relevant assumption checks, descriptive results, inferential results, and — where appropriate to the discipline — effect sizes, confidence intervals, and p-values. Findings are usually presented in tables and figures, followed by an interpretation connecting the statistics back to the research questions.
Qualitative dissertations generally need to report: the analytical approach used (e.g., thematic analysis), the coding process, the themes or categories that emerged, supporting evidence such as participant excerpts, an interpretation of what the themes mean, and — where relevant — a discussion of credibility, dependability, or reflexivity. There is no single universal reporting standard; expectations vary by discipline and by university guidelines, so checking your department's requirements is essential.
Results, Findings, Discussion, and Interpretation: What's the Difference?
These terms are often used inconsistently, which confuses many postgraduate writers.
1. Analysis is the process applied to the data (running a statistical test, coding transcripts).
2. Results/findings report what the analysis produced (the statistics obtained, or the themes identified) with minimal commentary.
3. Interpretationexplains what those results mean in relation to the research question.
4. Discussionsituates the interpreted findings within the wider literature, addresses their implications, and considers limitations.
Terminology and chapter structure vary by discipline and institution — some universities combine results and discussion into a single chapter, others separate them strictly. Always follow your specific programme's guidelines rather than assuming one universal structure applies.
Getting the Right Support for Your Dissertation Analysis
One of the most important methodological choices in a dissertation is deciding between quantitative and qualitative data analysis and then carrying it out meticulously. If you are still working out which approach fits your research question, need help interpreting statistical output, or want a second opinion on your coding and thematic structure, working with an experienced academic research advisor can help you clarify your methodology and strengthen how you analyse, interpret, and present your findings. HelpForThesis supports postgraduate researchers in understanding and applying the right analytical approach for their own dissertation work.
Frequently Asked Questions
Quantitative data analysis applies statistical methods to numerical data to test relationships or differences, while qualitative data analysis interprets non-numerical data, such as text, to understand meaning and context. The choice depends on what the research question is actually asking.
Look at your research question first. Questions about frequency, extent, relationships, or differences generally point toward quantitative analysis; questions about experience, meaning, or perception generally point toward qualitative analysis. Your research design and supervisor's guidance should confirm the fit.
It is the process of applying statistical procedures — descriptive and inferential — to numerical data collected through instruments such as surveys or experiments, in order to test hypotheses or describe patterns relevant to the research question.
It is the systematic interpretation of non-numerical data, such as interview transcripts, through coding and thematic development, to understand meaning, experience, or context relevant to the research question.
Neither is inherently easier. Quantitative analysis demands statistical planning and correct test selection; qualitative analysis is often more time-intensive due to transcription, coding, and iterative interpretation. Difficulty depends on your background, data volume, and research design.
Examples include testing whether study hours predict exam scores using regression, comparing average satisfaction scores between two groups with a t-test, or examining the association between two categorical variables using chi-square.
Examples include using thematic analysis to identify patterns in interview transcripts about workplace burnout, or using narrative analysis to understand how participants tell the story of a major life transition.
Yes, this is called mixed-methods research. It is appropriate when your research problem genuinely requires both statistical breadth and interpretive depth — for example, a survey to identify a pattern, followed by interviews to explain it.
For quantitative analysis, SPSS, R, Stata, and Excel are common choices. For qualitative analysis, NVivo, ATLAS.ti, and MAXQDA are widely used. The right tool depends on your methodology and your institution's available licenses — the software does not determine the methodology itself.
Test selection depends on your research question, the number and type of variables, their level of measurement, your study design, and whether your data meet the assumptions required by a given test (such as normality). A statistics consultant or supervisor can help confirm the right choice.
Transcribe the interviews, become familiar with the content through repeated reading, generate initial codes, group codes into categories, develop and refine themes, and then interpret those themes in relation to your research question, supported by participant excerpts.
Clean and code the data, run descriptive statistics to summarise responses, check any assumptions relevant to your planned test, then apply the appropriate inferential statistic (such as correlation, regression, t-test, or ANOVA) to address your research question.
Leave a Reply