A look at Experimental Design for Advanced Science Projects

Advanced experimental design requires that the individual performing the experiment take testing further than a simple hypothesis and one test. The experimental design requires that overall research between the relationships of independent and dependent variables be illustrated within a conceptual framework and that data is reported.

When using the advanced experimental design, the researcher reflects results that are both valid and reliable. In translation, this means that the final presentation should exemplify measurements that provide consistent and correlating information that is replicable. [1]

A hypothesis consists of a testable idea that will then be followed up with several groups of data in follow-up support.  The hypothesis should be concise and usually grammatically structured as one sentence.  In thesis statements or journal articles, a concisely structured hypothesis allows future researchers to gather and utilize the information in which they seek. [2]

When choosing the conditions, groups or factors to experiment and observe, it is crucial to represent the population in which the hypothesis seeks to inform. This creates a broad effect of results across a real life bell curve statistically.[3]

However, when choosing an experimental group or area, controlling the experiment and narrowing down the selection type for a study must be carefully chosen in order to not introduce extraneous variables that will influence the study and skew the results. [4]

After choosing and setting up the experiment details and population, it is time to separate the clusters. The evidence of support for the hypothesis should be made up of an experimental group and control group. The control group is sometimes known as placebo, and is not subjected to variable conditions. This group should be chosen according to the experimental guidelines and fall within the same ranges that the testing bunch falls into. [5]

The control is used as a standard to rule out extraneous results. For example, if an experiment is questioning whether or not 6-month-old albino male rats are more likely to become addicted to cocaine than pigmented 6-month-old male rats, there would have to be a control group to contrast the results.

In this example with albino rats, control groups would be made up of 6-month-old male albinos subjected to the same conditions to rule out any extraneous results. This second cluster of 6-month-old albinos would be set in the same types of cages, same room temperature, given the same food and water, etc., with the exception of cocaine administration. [6]

Continuing with this illustration, ultimately, there would be a minimum of three groups:  one pigmented, 6-month-old male rats and two albino 6-month-old male rats. There would be two groups that would be given cocaine: one group pigmented, one not. These are the experimental groups. The control group is the other albino pack that receives no cocaine. The independent variable is being manipulated in the experiment and the dependent variable is the measurable outcome of the test.  [7]

Remaining with the albino rat hypothesis example, the advanced design urges the researcher to include several repeated tests for varying conditions.  The scientist may want to test different times of day to inject the cocaine, differing amounts, lower dosing given more frequently, and perhaps a self-administering device. [8]

Eventually, correlations and research data have to be documented in some form. If a presentation is made, peers may document and publish the report. This should be written and presented in a clear, scientific format in which other peers in a related field can critique the work.  Findings may not report causation and cannot be stated as fact.  The research must illustrate that what has been found is quantitative and qualitative.[9]  This means that the information should be statistically replicable, in measurable form, and provide multiple explanations in support for the hypothesis.

[1] Heppner, P.P. Kivlghan, D. M. Jr., & Wampold, B. E. (1999). Research and design in counseling (2nd ed). New York: Brooks/Cole.

[2] Trochim, W. M. K. (2006). Experimental Design. Research Methods Knowledge Base. http://www.socialresearchmethods.net/kb/desexper.php

[3] Palafi, T. & Jankiewicz H. (1997). Drugs and Human Behavior, 2ND Edition. Madison, WI. Brown & Benchmark Publisher.

[4] Athabasca University. Centre for Psychology (2005).  Research Methods and Experimental Design (2260).

[5] Heppner, P.P. Kivlghan, D. M. Jr., & Wampold, B. E. (1999). Research and design in counseling (2nd ed). New York: Brooks/Cole.

[6]  Athabasca University. Centre for Psychology (2005).  Research Methods and Experimental Design (2260).

[7] Palafi, T. & Jankiewicz H. (1997). Drugs and Human Behavior, 2ND Edition. Madison, WI. Brown & Benchmark Publisher.

[8] Palafi, T. & Jankiewicz H. (1997). Drugs and Human Behavior, 2ND Edition. Madison, WI. Brown & Benchmark Publisher.

[9] The Higher Education Academy Psychology Network. (2005-2010). www.psychologypracticals.com