Quantitative Research
What is quantitative research?
Quantitative research methods are frequently applied in health and social care research. They use objective measurements with statistical methods, mathematics, economic studies or computational modelling to enable a systematic, rigorous, empirical investigation.
Why use quantitative research?
Quantitative methods often fall into two distinct categories – descriptive studies (often ‘hypothesis generating’) and hypothesis driven studies.
In applied health research, descriptive studies can be used to examine trends and patterns in health and health care use that can help planning and monitoring, this refers to frequencies, averages and other statistical calculations.
Hypothesis driven studies fall into two broad categories: experimental and observational. Both types are designed to test a hypothesis, usually the association between variables. In a health care scenario, this would typically be the relationship between a treatment (intervention) and an outcome. In epidemiology more broadly, hypotheses often relate to a potential aetiological (causal) factor and a disease.
What study designs use quantitative methods?
Hypothesis-driven study designs attempt to identify the associations relevant to the study. They do this by identifying a specific group of participants and trying to control the effects of other factors that might influence outcomes. These factors are known as ‘confounding’ variables or effect modifiers. For example, a study examining the relationship between pollution and health would have to take into account exercise and other aspects of lifestyle that might affect or ‘confound’ the outcome. A general rule is that if you cannot design out potential ‘confounding’ you must measure it, so that once the results come through, the role of other factors can be examined.
There are a number of experimental study designs available, some of which include:
Randomised Controlled Trials: In randomised controlled trials (RCTs), individuals or study units are randomly assigned to, for example, two experimental groups, one with a new therapy (experimental group) and one with current therapy (control group). The great advantage of well design randomized controlled trials is that it is possible to show causal effect of an intervention.
Cluster Randomised Controlled Trials: This type of randomised controlled trial involves group randomisation (for example a hospital or ward or GP) rather than randomisation of individual subjects. It is an important design that can be used to reduce contamination or treatment dilution between individual subjects.
Cochrane Infectious Diseases Group; A guide to including cluster randomized trials and participant randomized trials in intervention reviews
Crossover Trials: A cross-over trial design can be used where each patient receives both treatments in a defined order. Randomisation is often used to determine the order in which the patient receives each treatment, i.e. treatment A followed by B, or treatment B followed by A. It is important to allow sufficient time (wash out period) between the treatments to ensure that any effects of the first treatment have disappeared completely before the second treatment begins.
For more information:
Senn S. Cross-over Trials in Clinical Research. 2nd edition. Wiley, Chichester, 2002.
Factorial Design: Factorial randomised trials allow investigators to evaluate more than one intervention in a single experiment. For example, in a cardiovascular disease prevention trial, patients are being randomised to receive aspirin vs placebo and then beta-carotene vs placebo.
For more information:
Montgomery AA., Peters, T.J. & Little, P. Design, analysis and presentation of factorial randomised controlled trials, BMC Med Res Methodol 3, 26 (2003).
Parallel Group Design: A parallel group study is a simple and commonly used clinical design which compares two treatments. Usually a test therapy is compared with a standard therapy. The allocation of subjects to groups is usually achieved by randomisation. The groups are typically named the treatment group and the control group.
Non-inferiority Trials: Non-inferiority trials are used to compare a novel treatment to an existing treatment/comparator in order to determine that it is not clinically worse with regards to a specified endpoint. It is assumed that the comparator treatment has been established to have a significant clinical effect (against placebo). These trials are frequently used in situations where new treatment or intervention is compared with standard treatment or usual care.
For more information:
Klein, John P., Moeschberger, Melvin L. Survival Analysis: Techniques for Censored and Truncated Data Springer-Verlag New York, 2003.
Equivalence Trials: Equivalence trials aim to compare a clinical outcome measurement between two interventions. This design is frequently used to demonstrate that the outcome between the two is not different.
For more information:
http://www.bmj.com/content/346/bmj.f184
Meta-analysis: Meta-analysis is a statistical technique for combining the findings from independent studies, it is frequently used to assess the effectiveness of healthcare interventions. Meta-analysis of trials provides a precise estimate of treatment effect, giving due weight to the size of the different studies included. However the validity of the meta-analysis depends on the quality of the systematic review on which it is based and can suffer from publication bias.
For more information:
The Cochrane Collaboration provides Review Manager (RevMan) software that includes performing meta-analyses and presenting the results graphically.
How can the RDS assist you with quantitative research?
The RDS is able to:
- identify a research question;
- provide guidance concerning the most appropriate research design, potential ethical issues and the resources required to undertake the research;
- assist you in identifying the relevant published research in terms of quantitative methodology that may support your application;
- signpost expert statistical support for study design.