This Experimental Design Essay example is published for educational and informational purposes only. If you need a custom essay or research paper on this topic, please use our writing services. EssayEmpire.com offers reliable custom essay writing services that can help you to receive high grades and impress your professors with the quality of each essay or research paper you hand in.
An experimental design is a plan for assigning experimental units to treatment levels and the statistical analysis associated with the plan (Kirk 1995: 1). An experimental design identifies the: (1) independent and dependent variables, (2) extraneous conditions that must be controlled (nuisance variables), and (3) indicates the way in which the randomization and statistical analysis of an experiment are to be carried out. Carefully designed and executed experiments are one of science’s most powerful methods for discovering causal relationships.
The seminal ideas about experimental design can be traced to Ronald A. Fisher, a statistician, biologist, and geneticist. Fisher vigorously championed three key principles of experimental design: randomization, replication, and blocking or local control. Random assignment has three important benefits. First, it helps to distribute the idiosyncratic characteristics of experimental units over the treatment levels so that they do not selectively bias the outcome of an experiment. Second, random assignment permits the researcher to compute an unbiased estimate of error effects – those effects not attributable to the manipulation of the independent variable. Finally, random assignment helps to ensure that the error effects are statistically independent. Through random assignment, a researcher creates two or more groups of experimental units that at the time of assignment are probabilistically similar on the average.
Replication is the observation of two or more experimental units under the same conditions. Replication enables a researcher to estimate error effects and obtain a more precise estimate of treatment effects. Blocking or local control removes variation attributable to a nuisance variable from the error effects. By removing a nuisance variable from the numerator and denominator of the test statistic, a researcher is rewarded with a more powerful test of a false null hypothesis. Two simple experimental designs that illustrate these principles are described next.
One of the simplest experimental designs is the randomization and analysis plan that is used with a t statistic for independent samples. Consider an experiment to compare the effectiveness of two ways of presenting nutritional information – newspapers and television – in getting obese teenage boys to follow a more nutritious diet. The dependent variable is a measure of improvement in each boy’s diet 1 month after the presentation. Assume that 30 boys are available to participate in the experiment. The researcher randomly assigns the boys to the 2 treatment levels with the restriction that 15 boys are assigned to each level. An independent samples t statistic is used to test the researcher’s statistical hypothesis.
In this experimental design, the nuisance variable of gender is held constant: only boys are used. Other nuisance variables such as initial obesity and age are probabilistically controlled by random assignment. The design described next uses an additional procedure, blocking or local control, to remove the nuisance variable of preexisting differences in obesity from the error effects.
It is reasonable to assume that responsiveness to nutritional information is related to the amount by which a boy is overweight. The design of the experiment can be improved by isolating this nuisance variable and removing it from the error effects. This can be accomplished by using a dependent samples t statistic design. Suppose that instead of randomly assigning 30 boys to the treatment levels, the researcher formed 15 blocks each containing two boys who are overweight by about the same amount. After all the blocks have been formed, the two boys in a block are randomly assigned to the media presentations. This design is more likely to detect any differences in diet improvement than the independent samples t statistic design.
- Kirk, R. E. (1995) Experimental Design: Procedures for the Behavioral Sciences, 3rd edn. Brooks/Cole, Pacific Grove, CA.
- Keppel, G. (1991) Design and Analysis: A Researcher’s Handbook, 3rd edn. Prentice Hall, Upper Saddle River, NJ.
- Maxwell, S. E. & Delaney, H. D. (2004) Designing Experiments and Analyzing Data: A Model Comparison Perspective, 2nd edn. Lawrence Erlbaum, Mahwah, NJ.