Laboratory Experimental Designs 

Essential questions 

What differentiates experimental study from other types of research?

What are the typical steps in preparing experimental design? 

Experimental research aims at discovering causal relationship between the treatment and its outcome. It is not simply that one causes another but the certainty that nothing else causes the effect that is ultimately important in experiments. This requires a researcher to maintain control over all known factors that can affect the outcome of the experiment.

Besides discovering relationship between cause and effect, a researcher is interested in learning how much cause is necessary to produce certain effect. That is how the independent variable affects the dependent variable. This may include testing the varying degrees of independent variable or removing it altogether.

The following video illustrates these points:

https://www.youtube.com/watch?v=UN206cSaF0k

Experimental design is a conceptual framework for the experiment. It is a plan of the procedures that enable us to test a hypothesis by conducting experiment and reaching valid conclusions based on the relationship between independent and dependent variables.

In experimental study, a researcher is expected:

• Identify and define a problem and goal of study
• Formulate the testable idea (hypothesis) related to the problem, and deduce its consequences
• Create an experimental design to test this hypothesis. It should include all elements of the system, conditions, and known relations of the elements
• Conduct the experiment
• Record raw data, reduce it to usable form and analyze
• Use the appropriate statistical test of significance.

Experimental design is created in advance of conducting the actual experiments, and typically it includes the following steps:

• Selection of subject species and sample(s)
• Selection of groups of subjects
• Identification of non-experimental factors and their control
• Selection or construction of appropriate instruments that measure experimental outcomes, and their validation
• Determination of location, time, and duration of experiments.

Unlike any other types, true experimental research design relies on the random assignment of subjects to interventions/treatments. This eliminates spurious associations that can frequently impact less powerful research designs.

The following types of research designs are most common in experimental settings:

Randomized two-group design: subjects are randomly assigned to one of the two groups: control (not subject to intervention) and experimental (subject to intervention). The results of intervention are interpreted based on the statistical difference between the measured responses of experimental and control group of subjects.

Example: liquid culture of bacterial cells was divided in two samples that contained equal number of cells. One sample was subjected to a single dose of X-rays (treatment that is hypothesized to have mutagenic effect) while another was subjected to the same exposure of visible light (that does not have mutagenic effect). Mutation rate was determined in the two samples to determine the mutagenic effect of this particular dose of X-rays on this bacterial species.  

Pre-test/post-test control group design (or four-cell experimental design, or before-and-after two-group design): subjects are randomly assigned to one of the two groups: control (not subject to intervention) and experimental (subject to intervention). The experimental study parameter is tested in both groups prior to intervention, then experimental intervention is applied to experimental group, and both groups are subsequently re-tested to obtain the measurements of experimental parameter in question after the conclusion of experiment. Experimental outcomes of intervention are evaluated based on the difference of pre-experimental and post-experimental measured outcomes in control group and experimental group. These differences are attributed to the experimental intervention. 

Example: height of plants of the same species and age was recorded before the experimental intervention in both groups. Experimental intervention consisted of application of a single dose of ammonium nitrate at the same time in all plants of experimental group, while the control group was treated with the same volume of water. All other experimental parameters were exactly the same in both groups of plants. Plant height was measured in both groups two weeks after the date of treatment. The differences between the overage height of plants in experimental group and control group are attributed to the effect of that dose of ammonium nitrate on plant metabolism and growth.

Factorial design: subjects are randomly assigned to one of the two groups: control (not subject to intervention) and experimental (subject to intervention). The effect of two or more experimental treatments (independent variables) on the experimental group are studied. The following experimental conditions can be applied: each independent variable alone, pair-wise combination of independent variables, more than two independent variables in combination. This design is complex and allows to test hypotheses about possible outcomes of interaction between various experimental treatments on the subject. The results are interpreted based on the statistically significant differences between the measured outcomes in control group, and each experimental treatment in the experimental group [4].

Example: one DNA sample was divided in three that contain the same quantity of DNA. The first sample (experimental A) was subjected to digestion by the specified volume of restriction enzyme 1. The second sample (experimental B) was subjected to digestion by the specified volume of restriction enzyme 2. The third sample (control C) was subjected to the addition of the same specified volume of water. The same volume of sample was taken from samples A, B, and C, and loaded on agarose gel to determine the size of DNA fragments. The results were analyzed, and the restriction map was produced for enzyme 1 and enzyme 2 separately, based on the differences between the size of DNA fragments in samples A and C (enzyme 1), and samples B and C (enzyme 2).

Subsequently, specified volume of enzyme 2 was added to sample A, and the same volume of enzyme 1 was added to sample B, while the same volume of water was added to sample C. After incubation under the same experimental conditions, all three samples were loaded on agarose gel to determine the size of DNA fragments. The results were analyzed, and the restriction map was produced for enzyme 1 and enzyme 2 combined. It was based on the experimental determination of size of DNA fragments in samples A and B (they were hypothesized to be the same), and sample C (it was hypothesized to be different).

If human subjects are used in laboratory experiments, the proposed research project should be subject of approval by the institutional review board (IRB). This process ensures that the methods proposed for research are ethical, and appropriate with the view of protecting the rights and welfare of human subjects of experimental research.

Suggested readings

 Zacks S. (1996) "Adaptive Designs for Parametric Models". In: Ghosh, S. and Rao, C. R., (Eds). "Design and Analysis of Experiments," Handbook of Statistics, Volume 13. North-Holland. 

Creswell J.W. (2008) Educational research: Planning, conducting, and evaluating quantitative and qualitative research (3rd edition), Upper Saddle River, NJ: Prentice Hall.

Festing M. F. W. (2003) Principles: The need for better experimental design. Trends in Pharmacological Sciences. – 24 (7), p. 341-345.

Collins, L. M., Dziak, J. J., Kugler, K. C., & Trail, J. B. (2014). Factorial experiments: Efficient tools for evaluation of intervention components. American Journal of Preventive Medicine, 47, 498-504.