How to Analyse Quantitative Data: Statistical Analysis Explained

Written By

Edward Fletcher

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

How to Analyse Quantitative Data: Statistical Analysis Explained

Quantitative analysis often intimidates student researchers. Statistics feels technical. Statistical software feels complicated. But quantitative analysis answers a straightforward question: what patterns exist in your data?

Your data consists of numbers. Quantitative analysis finds patterns in those numbers. It describes them. It tests whether those patterns are likely due to chance or whether they reflect genuine effects in your population.

Start With Descriptive Statistics

Before hypothesis testing or complex statistics, describe your data. What are the basic patterns?

If you measured anxiety on a scale of 1 to 10, what was the average anxiety level? What was the range? Did most people cluster around the middle or spread across the scale?

If you surveyed people about remote working practices, what percentage reported communication difficulties? What percentage reported positive experiences?

These descriptive statistics help you and your reader understand what your data looks like before you do anything more sophisticated with it.

Calculate means, medians, and ranges for your variables. Calculate percentages for categorical variables. Create frequency tables showing how many people fell into each category.

Your analytical framework should be chosen because it helps you see your data in a way that other frameworks would not, and explaining this choice clearly in your methodology shows your examiner that you understand its value.

Describe your data accurately before you interpret it.

The Difference Between Descriptive and Inferential Statistics

Descriptive statistics describe what you found. Inferential statistics test whether those findings are likely true beyond your specific sample.

If you surveyed 200 remote workers and found that 65% reported positive experiences, that's descriptive. You're describing what your 200 respondents reported.

Inferential statistics ask: would we see similar patterns if we surveyed a different group of 200 remote workers? Or was the 65% figure just luck? Or an artefact of who happened to respond?

Inferential statistics use hypothesis testing to answer these questions.

Hypothesis Testing Basics

Hypothesis testing asks: is the pattern I found in my data likely to occur by chance alone?

You start with a null hypothesis. The null hypothesis typically states that there's no relationship or no difference. There's no relationship between remote working and team cohesion. There's no difference in anxiety between first year and second year students.

Then you collect data. You calculate a test statistic. You compare that test statistic to what you'd expect if the null hypothesis were true.

The result is a p value. The p value tells you the probability of getting a result this extreme if the null hypothesis were actually true.

If the p value is less than 0.05 (your significance level), you reject the null hypothesis. You conclude that the pattern you found is unlikely to be due to chance. It probably reflects a genuine effect.

If the p value is greater than 0.05, you fail to reject the null hypothesis. The pattern you found could easily have occurred by chance.

A well-structured dissertation requires careful attention to the relationship between each chapter, ensuring that your argument develops logically from the introduction through to the conclusion. Students who invest time in planning their chapter structure before writing tend to produce more coherent and persuasive pieces of academic work, as the narrative flows naturally from one section to the next. Your literature review should not simply summarise existing research but instead position your work within the broader academic conversation, identifying gaps that your study is designed to address. The methodology chapter is particularly important because it demonstrates your understanding of research design and justifies the choices you have made in collecting and analysing your data.

The formatting of your dissertation is not a trivial matter but a reflection of your professionalism and attention to detail, both of which your examiner will notice before they have even begun to read your argument.

Common Statistical Tests and When to Use Them

Use a t test to compare the means of two groups. Are first year students more anxious than second year students? A t test answers this.

Use ANOVA (analysis of variance) to compare means across three or more groups. Is anxiety different across first year, second year, and third year students?

Use a correlation to examine relationships between two continuous variables. Is there a relationship between hours spent remote working and reported team cohesion?

Use chi square tests to examine relationships between categorical variables. Is there a relationship between gender and comfort with remote working?

Use regression to predict one variable from other variables. Can you predict anxiety levels from the number of assessment tasks and the number of available support resources?

Choose your test based on what question you're asking and what type of data you have.

Effect Sizes Matter

A statistically considerable result doesn't automatically mean a large or important effect. A very large sample might find a tiny effect that's technically considerable but practically meaningless.

Effect sizes tell you how large an effect actually is. They're reported alongside p values.

Cohen's d compares effect sizes between groups. An effect size of 0.2 is small. An effect size of 0.5 is medium. An effect size of 0.8 is large.

R squared tells you how much of the variance in one variable is explained by another variable. An R squared of 0.4 means that 40% of the variation is explained. An R squared of 0.05 means only 5% is explained.

Report both p values and effect sizes. A result can be statistically considerable but practically small.

Checking Assumptions

Different statistical tests make assumptions about your data. T tests assume your data are approximately normally distributed. Correlations assume linear relationships. Regression has multiple assumptions about residuals.

Check whether your data meet the assumptions of the test you're using. If they don't, you might need a different test or a data transformation.

Software like SPSS, R, or Python can test assumptions for you. Many provide diagnostic plots that show whether assumptions are met.

Presenting Your Analysis

Present your analysis step by step. First, describe your data. Then, state what test you used and why. Then, report the results of that test including the test statistic, the p value, and the effect size.

Tables make quantitative results clearer than paragraph descriptions. A table showing means and standard deviations for different groups is easier to understand than sentences describing each group's mean.

Interpretation

Once you've completed statistical tests, interpret them. What do the results mean?

Effective use of quotations in your literature review means selecting short, precise extracts that illustrate a specific point and then explaining in your own words why that quotation matters for your argument.

If your t test shows that first year students are more anxious than second year students (t equals 2.4, p equals 0.02, d equals 0.6), you might interpret this as: first year students reported higher anxiety than second year students, representing a medium effect size. This suggests that moving beyond the first year is associated with reduced anxiety.

If your correlation shows no considerable relationship between hours spent remote working and team cohesion (r equals 0.12, p equals 0.29), you might interpret this as: the amount of time spent working remotely was not related to perceived team cohesion in this sample.

Interpretation connects your results to your research questions and to broader understanding.

The importance of choosing appropriate and reliable sources for your literature review cannot be overstated, because the quality of your analysis is directly affected by the quality of the evidence on which it is based.

Common Mistakes

Don't confuse statistical significance with practical significance. A result can be considerable but small.

Don't use hypothesis tests to test the only pattern in your data. If you calculate 50 tests, some will be considerable by chance alone. This is the multiple comparisons problem.

Don't assume causation from correlations. A correlation shows relationship. It doesn't show that one variable causes the other.

Don't exclude data because it doesn't match your expectations. Report all your data.

Don't use tests designed for normally distributed data with non normal data without justification.

If you're uncertain which statistical test to use or how to interpret your results, professional services like dissertationhomework.com can help you choose appropriate analyses and interpret your quantitative results accurately.

===

Preparing for your dissertation viva, or oral examination, requires a different kind of preparation from the written examination revision that most students are more familiar with from their earlier studies. In a viva, you will be expected to defend the choices you have made in your dissertation, explain your reasoning, and respond thoughtfully to challenges or questions from the examiners without the safety net of notes or prepared answers. The best preparation for a viva is to know your dissertation thoroughly, to be able to articulate clearly why you made the key decisions you did, and to have thought carefully about the limitations of your research and how you would address them if you were to conduct the study again. Many students find it helpful to conduct a mock viva with their supervisor or with a group of fellow students, as the experience of responding to questions about your work in real time is something that is very difficult to prepare for through solitary study alone.

Need Expert Help With Your Dissertation?

Our UK based experts are ready to assist you with your academic writing needs.

Order Now