Strengths: Field experiments generally yield results with higher ecological validity than laboratory experiments, as the natural settings will relate to real life.

Demand characteristics are less of an issue with field experiments than laboratory experiments (i.e. participants are less likely to adjust their natural behaviour according to their interpretation of the study’s purpose, as they might not know they are in a study).

Weaknesses: Extraneous variables could confound results due to the reduced control experimenters have over them in non-artificial environments, which makes it difficult to find truly causal effects between independent and dependent variables.

Ethical principles have to be considered, such as the lack of informed consent; if participants are not made aware of their participation in an experiment, privacy must be respected during observations and participants must be debriefed appropriately when observations come to an end.

Precise replication of the natural environment of field experiments is understandably difficult, so they have poor reliability, unlike laboratory experiments where the exact conditions can be recreated.

Field experiments are more susceptible to sample bias, as participants are often not randomly allocated to experimental conditions (i.e. participants’ groups are already pre-set rather than randomly assigned).

Before constructing an experiment research there are various aspects to consider.

(1) Planning: A good planning always ensures that the research is carried out properly and in proper conditions with appropriate tools and measures.

(2) Sampling: One of the best ways to ensure that the research is conducted systematically and appropriately is to have a proper selection of sample. Sampling is taking any portion of a population or universe as representative of that population or universe. Sample can be classified into probability and non probability sample. Probability samples use some form of random sampling in one or more of their stages. Non probability samples do not use random sampling; they thus lack the virtues being discussed. Still, they are often necessary and unavoidable. The probability sampling includes stratified, cluster, systematic and random sampling method. The non probability sampling includes quota, purposive and accidental sampling method.

(3) Research design: Every research requires a blueprint of the research work that will be carried out. Where the experiment will be conducted, that is the setting, who will be the subjects, that is the sample, how it will be conducted, what instruments will be used, what will be manipulated, what will be measured etc. The experimental design must also provide for the number of subjects that will be in the experiment and the number of subjects who will be considered as the control group.

(4) Tools of data collection: What are the tools that will be used, how the results will be measured, and what statistical tools will be used etc.

(5) Procedure: Once the subjects have been identified and setting has been decided where the experiment will be conducted, the next step is to get permission from authorities to use the setting. Having obtained the permission let us say from school authorities to conduct field experiment regarding which method of teaching leads to better academic performance, the subjects will be selected from a certain class. Let us say we choose children from class 5 all sections. Let us say there are 200 children. We need only 100 children and so from each of the 4 sections we take 25 out of 50 children randomly. From these 100 children, we again take 50 for control group and another 50 for experiment. This again we select randomly. Both the groups children are tested for academic performance and their scores are recorded. Then, to the experimental group of children we give instruction through lecture method and to the other group through tutorial method. After training for 1 month, the academic performance of both the groups are retested. Now the difference in the second testing for the two groups will indicate which method is more effective. Within the group also the pre and post test performance could be measured and the difference noted as improvement or decrease in academic performance.

(6) Statistical analysis: Appropriate statistics such as the test will be used to find if the differences obtained between the two groups as well as between the pre and post tests are statistically significant. The above 6 steps are the ways in which the field experiment is conducted.

From MPC-005 Research Methods – IGNOU