Jul 2, 2018

Quantitative Research Methods: Surveys, Experiments & Statistical Analysis

Jonat N.
Published by:Jonat N.
Quantitative Research Methods: Surveys, Experiments & Statistical Analysis

Did you know that according to publications from the American Psychological Association (APA), around 75% of published empirical research in the social sciences relies on quantitative methods? You have a love for numbers, but have you ever wondered how researchers can turn those numbers into a compelling conclusion?

Whether you are writing a university dissertation, preparing a project proposal, or reading a research paper from an academic journal for the first time, quantitative methods can feel overwhelming. Many students find the math, formulas, and academic terms confusing.

This guide by Cheap Essay Writing UK strips away unnecessary academic jargon. Below is a clear, step-by-step walkthrough of quantitative research methods,  from choosing your design to running your surveys, picking your sample, and running the right statistical tests. 

What is Quantitative Research?

At its simplest, quantitative research is about collecting and analysing numerical data to spot patterns, test theories, and make reliable, generalisable claims. While qualitative research explores human experience and meaning using words, interviews, and open-ended observations, quantitative research relies on numbers, scales, and structured data that you can measure, compare, and test statistically. 

Qualitative Question: "How do international students feel when settling into British university culture?" (Answered through open conversation and personal stories).

Quantitative Question: "What is the relationship between weekly lecture attendance and final exam grades among first-year undergraduates?" (Answered with attendance records, test marks, and statistical correlations).

It's easy to explain: you make a hypothesis, gather measurable data, do some statistical math, and either prove or disprove your hypothesis. This method is appreciated in research and industry environments for its objective, reproducible findings. Learn more about the role of replication studies in quantitative research.

It has been applied in various disciplines, such as the social sciences, healthcare, business, education, and engineering, all of which are quantitative in nature. 

The Four Levels of Measurement (The NOIR Framework)

Before you collect data or open statistical software, you must know what kind of numbers you are actually handling. In quantitative research, numbers are divided into four levels, easily remembered with the acronym NOIR:

Figure 1: Four levels of measurement, NOIR framework

Figure 1: Four levels of measurement, NOIR framework

Nominal (Naming Categories)

Data that falls into separate categories with no natural order or numerical value. Numbers are used only as tracking labels.

  • Examples: University department (Business, Engineering, Humanities), nationality, residential accommodation type.
  • What you can calculate: Frequencies, counts, percentages, and mode. (You cannot calculate an average nationality).

Ordinal (Ordered Rankings)

 Categories that have a clear rank or logical order, but the physical distance between the points is not mathematically equal.

  • Examples: Likert agreement scales (Strongly Disagree, Disagree, Neutral, Agree, Strongly Agree), socioeconomic status (Low, Medium, High), or degree classifications (1st, 2:1, 2:2, 3rd).
  • What you can calculate: Medians, quartiles, and rank correlations (such as Spearman's ).

Interval (Equal Gaps, No Absolute Zero)

Continuous numerical data where the distance between each point is identical and measurable, but there is no true zero point. Zero does not mean the variable is absent.

  • Examples: Temperature measured in Celsius (0°C is cold, but it does not mean there is "zero temperature"), standardised IQ scores.
  • What you can calculate: Means, standard deviations, and addition/subtraction. You cannot make ratio statements (for example, 30°C is not "twice as hot" as 15°C).

Ratio (Equal Gaps, Absolute True Zero)

The highest level of measurement. It has equal mathematical intervals and a true, non-arbitrary zero, meaning zero equals the total absence of that property.

  • Examples: Weekly study hours, student loan balance, reaction time in milliseconds, age.
  • What you can calculate: All arithmetic operations. You can reliably state that 20 hours of study is twice as long as 10 hours

Key Quantitative Research Designs

Your research design is your master plan. Choosing the right quantitative research design is required before gathering any data. How you will collect evidence and what conclusions your methodology can defend depend on your design. Figure 2: Quantitative Research Designs

Figure 2: Quantitative Research Designs

Descriptive Quantitative Research

Descriptive quantitative research tries to describe a population or phenomenon as it exists in its natural state. Does not attempt to create cause and effect; maps out what is there.

  • Ideal for baseline studies and needs assessments
  • Findings of this kind are very common in public health surveys and in market research.
  • Generates frequency counts, averages, and percentage distributions

The study, which recorded the average screen time of UK university students, provides good quantitative research examples of a descriptive study. It describes what is occurring, but not an explanation.

Correlational Quantitative Research

Correlational research explores the relationship between two or more variables, but does not manipulate any variables. Both variables are naturally measured, and it is determined whether they are related.

  • If both variables increase, it indicates a positive correlation.
  • If there is a negative correlation, one goes up as the other goes down
  • Correlation does not imply causation. This is critical to remember

One possible research question for a correlational study is the relationship between hours of sleep and examination performance among students. The data could indicate a connection, but cannot establish cause and effect.

Experimental Quantitative Research

An experiment is the best way to establish cause and effect. You can actively change a variable (independent) and observe how it affects another variable (dependent), while holding all other variables constant.

  • True experiments involve random assignment of experimental and control groups.
  • Quasi-experiments involve pre-existing groups and don't involve randomisation.
  • Laboratory experiments offer greater control and realism than field experiments.

A researcher who is interested in determining whether a new teaching method is effective in enhancing test scores would randomly assign students to either the experimental group (new teaching method) or the control group (traditional teaching method). The distinction in results is due to the approach taken in practice in experimental quantitative research. For a detailed comparison, see our guide on observational vs. experimental methods.

Quantitative Data Collection Methods

As important as your design decision is, your choice of data collection method is equally important. Each of the common quantitative research techniques has its own advantages and disadvantages.

Survey Research

Surveys are by far the most popular quantitative tool in UK higher education and industry research. They allow you to gather structured information across hundreds of participants efficiently, especially through online survey platforms like Qualtrics, Jisc Online Surveys, or SurveyMonkey. 

To make survey data statistically usable, rely on closed-ended questions:

  • Multiple Choice: Categorical selections with mutually exclusive answers.
  • Likert Scales: Statements paired with an agreement or intensity scale (e.g., 5-point or 7-point scales ranging from Strongly Disagree to Strongly Agree).
  • Numerical Sliders: Visual continuous scales where participants assign a value between 0 and 100.

Explore our best strategies to work on survey & experimental research for dissertations for deeper insights.

Structured Observation

Instead of asking people to self-report their thoughts, structured observation involves observing and recording behaviour directly according to a set of codes. All behaviours are recorded numerically, making statistical analysis easy. This approach is widely used in child development studies, primary education research, and usability testing to eliminate self-reporting bias. 

Secondary Data Analysis

You do not always have to spend weeks gathering original survey responses. Secondary data analysis is a method in which existing, high-quality, pre-cleaned public datasets (government statistics, institutional records, or published databases) are analysed by running fresh statistical models to address new research questions. Read our comprehensive guide on how to conduct secondary research. 

Excellent open-access UK and international data repositories include:

  • The UK Data Service (UKDS): The largest UK digital repository for social, economic, and population survey data.
  • Office for National Statistics (ONS): Open data covering regional employment figures, public health data, and census records.
  • The European Social Survey (ESS): Standardised cross-national datasets covering European public attitudes and political behaviour.

Quantitative Research Questionnaire Design

A poorly designed questionnaire ruins your study before you run a single test.  Here are 5 ground rules for a clean quantitative questionnaire design:

  • Map every single question to an explicit variable: If an item does not directly measure your independent, dependent, or control variables, cut it. Unnecessary questions increase survey length and lower completion rates. 
  • Use established, validated scales where possible: If you are assessing constructs like workplace burnout, depression, or website usability, avoid inventing novel questions. Use established psychometric scales (such as the PHQ-9 or System Usability Scale) that already have verified academic reliability. 
  • Eliminate double-barrelled questions: Never combine two separate concepts into one question.

Problematic Item: "Do you find this university course engaging and well-organised?" (A respondent might find the course very engaging, but poorly organised).

Correct Revision: Split this into two distinct questions—one for engagement and one for organisation.

  • Use clear, accessible English: Avoid idiomatic expressions, cultural slang, double negatives ("Do you not disagree that..."), and complex academic phrasing. Clear, direct language prevents misunderstandings across diverse and international participant groups.
  • Pilot test your instrument: Always test your draft survey on a small pilot group of 5 to 10 colleagues or fellow students before running your full study. You will immediately spot broken skip-logic, confusing words, and questions that take too long to complete.

Sampling Techniques in Quantitative Research

Your sample is the smaller group of people you select to represent the wider target population. If your sampling method is biased, your statistical results cannot be trusted, no matter how sophisticated your statistical calculations are. 

Figure 3: Sampling Techniques

Figure 3: Sampling Techniques

In probability sampling techniques, each member of the population has an equal probability of being selected:

Simple random sampling: Every individual is chosen at random.

  • Stratified sampling: The population is divided into subgroups (strata), and random samples are taken from each
  • Cluster sampling: Natural groupings (schools, hospitals) are sampled at random, and all the individuals in that group are then studied.
  • Systematic sampling: Every nth person on a list is selected

Non-probability sampling methods are more readily accepted and less generalisable:

  • Convenience sampling: selecting those people who are most readily available at the time of the study
  • Purposive sampling selects participants based on specific characteristics

In most of the quantitative research in academic settings, probability sampling is preferred because it allows statistical inference (generalising your findings from your sample to your wider population).

Determining Sample Size: Don't Guess, Run a Power Analysis

A common mistake among students and early-career researchers is plucking a participant target out of thin air (e.g., "I'll just survey 100 people and that should do it"). If your sample size is too small, your statistical tests will lack statistical power, meaning you run a serious risk of missing genuine, meaningful effects even when they truly exist—a classic Type II error ().

Most beginner methodology blogs and study guides online skip this step entirely. They often hand out arbitrary rules of thumb—such as "aim for 100 respondents" or "gather at least 10% of the population"—without accounting for study design, variance, or test sensitivity. Following that generic advice is one of the fastest ways to lose marks in a methodology chapter or face rejection in peer review for being underpowered.

Before collecting data, use dedicated, free software like G*Power, or statistical packages in R and Python, to run an a priori power analysis.

A standard power calculation balances four interacting parameters:

  • Alpha (): Your chosen significance threshold, typically set at 0.05 (a 5% accepted risk of a false positive).
  • Statistical Power (1-): The probability of detecting a real effect when one exists. The widely accepted academic baseline is 0.80 (80% power, or a 20% risk of a false negative).
  • Expected Effect Size: How strong a relationship or difference you realistically expect to see based on published literature in your field (e.g., Cohen's d for comparing two means, or Pearson's r for correlations).
  • Result (N): The software outputs the precise minimum sample size you need to achieve statistically sound, defensible results for your specific test.

Planning your sample size this way demonstrates true methodological rigour and protects your research from producing inconclusive, ambiguous findings.

Variables and Hypothesis Testing in Quantitative Research

The 5 Types of Variables

To test relationships, you must define and label your variables clearly:

Variable TypeDefinitionWorking Example
Independent Variable (IV)The factor you manipulate or treat as the cause/predictor.Method of study: Revision App vs. Printed Textbook.
Dependent Variable (DV)The outcome you measure; changes in response to the IV.Final exam mark (0–100%).
Control VariableExtraneous factors kept identical to prevent distorted findings.Giving both groups the same exam, with identical time limits in the same room.
Moderating VariableA factor that alters the strength or direction of the IV-DV relationship.Student native language: the app might work better for native speakers than ESL students.
Confounding VariableAn uncontrolled outside factor that distorts both the IV and DV.Student prior knowledge: one group may have studied the topic previously.

The Hypothesis Testing Process

In Quantitative Research, hypothesis testing is performed.

Hypothesis testing is the systematic way of using statistical evidence to make decisions with respect to your research questions. It is organised in a logical sequence:

  1. The null hypothesis (H₀) is that there is no effect or relationship.
  2. Write your alternative hypothesis (H₁) that predicts the effect or relationship you predict
  3. Choose your level of significance (α), usually 0.05, which means there is a 5% chance that you will reject a true null hypothesis.
  4. Collect data and run your statistical test
  5. Draw the Statistical Conclusion:

If p < 0.05: You reject the null hypothesis. The findings are statistically significant.

If p 0.05: You fail to reject the null hypothesis. You do not have enough evidence to claim a real effect.

Note on Academic Phrasing: Never say a statistical test "proved" your hypothesis. In empirical research, data provides probabilistic evidence that supports or fails to support a theory; it does not provide absolute mathematical proof. 

Statistical Analysis in Quantitative Research: Choosing the Right Test

There are two ways to make sense of your numbers; one is a statistical analysis. The analysis design you employ will vary depending on your research design, the number of variables being examined, and the level of measurement of your data.

Descriptive statistics provide a summary of your data:

  • Means, medians, and modes (measures of centre)
  • Standard deviation, interquartile range, and variance (measures of spread)
  • Frequencies and percentages

Inferential statistics can be used to make inferences beyond the sample:

  • The t-test is used to compare the means of two groups.
  • ANOVA (Analysis of Variance) compares the means between three or more groups
  • Data Analysis: The chi-square test tests relationships between categorical variables.
  • Pearson's r or Spearman's rho assesses the strength of the correlation.
  • Regression analysis can be used to estimate the value of a dependent variable from one or more independent variables.
  • Industry-standard software for conducting these analyses is SPSS, R, and Stata. As a dissertation writer, you will surely require at least one of these tools.

 The Inferential Decision Matrix

Choosing the wrong statistical test is one of the most common reasons dissertation methodologies receive major revisions. Use this decision matrix to match your research question to the correct test:

Research QuestionIndependent Variable (IV)Dependent Variable (DV)Parametric Test (Normal Data)Non-Parametric Alternative
Compare 2 independent groupsCategorical (2 groups, e.g., Commuter vs. Campus Students)Continuous (Interval/Ratio)Independent Samples t-testMann-Whitney U test
Compare the same group at 2 time pointsCategorical (Paired, e.g., Pre-course vs. Post-course score)Continuous (Interval/Ratio)Paired Samples t-testWilcoxon Signed-Rank test
Compare 3 or more independent groupsCategorical (3+ groups, e.g., London vs. Manchester vs. Edinburgh)Continuous (Interval/Ratio)One-Way ANOVAKruskal-Wallis test
Compare the same group at 3 or more time points (or conditions)Categorical with 3+ related levels (e.g., Start of Semester vs. Midterms vs. Finals Week)Continuous (Interval/Ratio) or OrdinalRepeated-Measures ANOVAFriedman test
Test the association between 2 continuous factorsContinuous (e.g., Revision hours)Continuous (e.g., Coursework grade)Pearson Correlation (r)Spearman Rank Correlation
Predict a continuous outcome from 1 or more factorsContinuous or Categorical predictorsContinuous outcome variableMultiple Linear RegressionQuantile / Robust Regression
Test the association between 2 categoriesCategorical (e.g., Passed / Failed)Categorical (e.g., Attended Revision / Skipped Revision)Chi-Square Test (χ²)Fisher's Exact Test

Need assistance choosing or running the right statistical test for your data? Consult with our academic experts for personalised guidance.

Modern Statistical Software Choices

You do not need to calculate these equations manually. Most university projects use one of the following tools:

  • SPSS: The traditional standard across UK social science departments. It relies on a graphical menu interface, though it requires an institutional license.
  • Jamovi & JASP: Modern, free, open-source programs that look and feel like SPSS. Both generate publication-ready tables and are well-suited for university-level research.
  • R & Python: Code-based, open-source environments preferred for advanced econometric modelling, machine learning, and large-scale data analysis.

Quality Control in Quantitative Research: Reliability and Validity

The only way to analyse your data statistically is if it is reliable and valid data.

Reliability is defined as consistency – does your instrument yield the same responses when used under the same conditions? You can test reliability using:

  • Cronbach's alpha is an index of internal consistency of a scale (typically, a value of 0.7 or higher is acceptable)
  • Constructive validity is established through test-retest reliability, which involves administering the same test twice and comparing the results.

Validity is the degree of accuracy–are you measuring what you say you are measuring?

  • Does the measure include all aspects of the concept?
  • Construct validity: Does the measure relate to other variables in theoretically expected ways?
  • Can external validity findings be generalised to other populations and settings?

Reliability and validity are necessary for research submitted for peer review and should always be reported in the methodology chapter of the dissertation.

Mixed Methods: Combining Quantitative and Qualitative Work

While quantitative methods provide statistical precision, they can strip away personal context, lived experience, and cultural nuance. As a result, UK researchers often adopt a mixed-methods approach that combines quantitative data and qualitative insight.

  • Sequential Explanatory (QUANqual): You run a broad quantitative survey first to uncover statistical patterns and demographic trends, then conduct follow-up qualitative interviews to explain why those patterns occur.
  • Sequential Exploratory (QUALquan): You conduct qualitative interviews first to explore an under-researched topic, using those insights to build an accurate survey to test on a broader population.
  • Convergent Design (QUAN+QUAL): You collect numerical survey data and in-depth qualitative interview responses concurrently, comparing the two datasets during discussion to see if the findings support one another.

Quantitative Methodology Checklist

Before submitting your research methodology chapter or proposal, verify that your project addresses these foundational steps:

  • Design Alignment: Does your choice of descriptive, correlational, or experimental design match your research question?
  • Measurement Scale: Have you identified whether each variable is Nominal, Ordinal, Interval, or Ratio (NOIR)?
  • Instrument Quality: Have you eliminated double-barrelled questions, adopted validated scales, and run a pilot test?
  • Sampling & Power: Have you justified your probability or non-probability sampling strategy and run a power analysis for sample size?
  • Statistical Assumptions: Have you checked your data for normal distribution, outliers, and variance before choosing between parametric and non-parametric tests?
  • Transparent Reporting: Have you prepared to report exact p-values, effect sizes, confidence intervals, and reliability scores (e.g., Cronbach’s )?

If you are struggling with your academic writing or worried about falling behind, read our guide on how to save yourself from failure. 

Conclusion

Quantitative research methods provide you with the tools to ask questions and receive answers. It's not just about selecting a research design and conducting experiments or surveys; it's about every step that you've taken from the research design to collecting data, defining variables, and running statistical analysis.

The findings are that:

  • Relate your design (descriptive, correlational, experimental) to your research question
  • Apply the correct sampling method to make the results generalisable
  • Carefully craft unambiguous questions to ensure the data gathered can be used.
  • Always test reliability and validity before making conclusions
  • Select the appropriate statistical test for a set of variables and the type of data.

These are the basics that you can use for either a dissertation methodology chapter or your first research proposal. 

Master Your Quantitative Dissertation Today

Ready to transform your numbers into a distinction-worthy dissertation? Connect with our expert team for dedicated quantitative dissertation help or chat with us on WhatsApp today!

Frequently Asked Questions

What methods are used in quantitative research?

The main methods include surveys, structured observations, experiments (true and quasi), and secondary data analysis. Each collects numerical data that can be statistically analysed.

What is the difference between qualitative and quantitative research?

Quantitative research deals with numbers, measurements, and statistical analysis to test hypotheses. Qualitative research explores meaning, experience, and context through words, interviews, and observation.

How is data collected in quantitative research?

Data is collected through surveys (online or paper), experiments, structured observations, standardised tests, and secondary datasets from government or institutional sources.

What are variables in quantitative research?

Variables are measurable characteristics that can take different values. The independent variable is what you manipulate; the dependent variable is what you measure. Control variables are kept constant to avoid distorting results.

Why is statistical analysis important in quantitative research?

Statistical analysis allows you to identify patterns, test hypotheses, measure relationships, and determine whether your findings are statistically significant, meaning they're unlikely to have occurred by chance.

Cite this article

If you want to cite this source, you can copy and paste the citation or click the Cite this article button to automatically add the citation to our free Citation Generator.

Jonat N. (July 2, 2018).
Quantitative Research Methods: Surveys, Experiments & Statistical Analysis to Improve Grades and Save Time Retrieved November 13, 2026, from https://www.cheap-essay-writing.co.uk/blog/2018/07/all-you-need-to-know-about-quantitative-research