Experimental research design
An experiment can be defined as:
“A rigorous controlled investigation of an activity where key factors are identified and manipulated to document their effects on an outcome” (Fenton and Pfleeger 1996).
“A procedure adopted or operation carried out to make a discovery, observation or test” (Oxford English Dictionary 2006).
This constellation aims to introduce some simple concepts associated with the design of an experiment. Designing an experiment would often take up a full year of study for an undergraduate university student. Here, we hope to introduce you to:
why you might want to conduct an experiment
some basic types of experimental design
generating research questions and hypotheses
variables and randomisation
It is crucial to design an experiment before you begin, otherwise you might find that you do not produce valid, reliable results and that you may not be able to answer your research question. Your experiment will require the adoption of exact procedures which entail manipulating variables in the study, collecting precise measurements and analysing these using established statistical procedures. The data is likely to be quantitative and the statistical procedures are likely to place requirements on the way the research is conducted. You must have more than one piece of data - you can't say that something is better or faster if you have only tested it once!
Experiments are not just for scientists! Experiments are used in the fields of market research, education and psychology to name but a few.
Authors: Deborah Street and Olivia Billingham with contributions from Liz Falconer and Paul White, University of the West of England, Bristol.