Chapter 1. Understanding Quantitative Research
1.1 What quantitative research is
Quantitative research uses structured observations or measurements represented numerically to describe patterns, estimate quantities, compare groups, test relationships, evaluate interventions, or assess theoretical propositions. It relies on explicit procedures for selecting cases, defining variables, measuring constructs, managing uncertainty, and reporting results. Quantitative evidence can be collected directly through surveys, tests, observations, experiments, sensors, or organisational records, or obtained from existing databases.
A quantitative study is not automatically rigorous because it includes numbers. Rigour depends on whether the research question is answerable with the design, whether measures represent the concepts, whether the sample supports the intended inference, whether assumptions are assessed, whether missingness and bias are addressed, and whether conclusions remain within the evidence.
1.2 Common purposes
| Purpose | Typical question | Illustrative business example |
|---|---|---|
| Description | What level or distribution is observed? | What is the average employee engagement score in a company? |
| Comparison | Do groups differ? | Do engagement scores differ between remote and office-based staff? |
| Association | Are variables related? | Is perceived supervisor support associated with engagement? |
| Prediction | How well can variables predict an outcome? | Which factors predict customer churn? |
| Causal evaluation | Did an intervention cause a change? | Did a training programme improve performance relative to a comparison group? |
| Measurement validation | Does an instrument behave as expected? | Does a customer-experience scale show acceptable reliability and validity? |
1.3 Core vocabulary
A population is the set of units about which the researcher wants to draw conclusions. A sample is the subset actually observed. A variable is a recorded characteristic that can vary, such as tenure or satisfaction. A construct is a theoretical concept, such as trust or organisational commitment, that may require several indicators. A parameter describes a population; a statistic is calculated from a sample. An estimator is a rule for estimating a parameter, while a confidence interval communicates uncertainty around an estimate.
An independent variable is often used to represent a predictor or explanatory variable; a dependent variable is the outcome. These labels do not, by themselves, prove causality. In a cross-sectional survey, the terms predictor and outcome are often safer because temporal order and confounding may be unresolved.
1.4 The research chain
- Define a consequential and researchable problem.
- Establish what is already known and what remains uncertain.
- State an aim, objectives, questions and—where justified—hypotheses.
- Build a theoretical or conceptual model.
- Select a design that can answer the questions.
- Define the population, sampling strategy, measures and procedures.
- Obtain required ethical and organisational approvals before recruitment or data collection.
- Collect, document, secure, clean and analyse the data according to a planned protocol.
- Interpret estimates in relation to theory, prior studies, uncertainty, limitations and context.
- Report transparently so that readers can judge the evidence and reproduce the workflow where possible.
1.5 Strengths and limits
Strengths can include consistent measurement, explicit comparisons, estimation of prevalence or association, hypothesis testing, reproducibility, and analysis of large datasets. Limitations can include measurement error, nonresponse, restricted coverage, confounding, dependence on model assumptions, poor contextual depth, and false confidence created by statistical significance. A large sample cannot correct a biased sampling frame or a badly designed instrument.
1.6 When quantitative methods fit
Quantitative methods fit when the question requires measurement, frequency estimates, comparisons, prediction, or tests of specified relationships. They are less suitable when the principal aim is to understand meanings or experiences that have not yet been adequately conceptualised. In such cases, preliminary exploratory work may be needed before a quantitative instrument can be designed. This book concentrates on quantitative procedures once the research concepts can be operationalised.