How to plan it?
While planning experimental research, it is important to:
- Identify the problem(s) and question(s) the researcher aims to answer. This should be in the form of an interrogative sentence expressing the relationship between two or more variables, independent and dependent. For example, if one wants to understand the impact of uncertainty about the health consequences of consuming a certain product, over perceptions of that product, the researcher might ask: “What is the effect of risk uncertainty over consumer’s risk perception?”. Once this has been defined, the relevant literature should be reviewed to help understand the methods and procedures through which it can be investigated, and identify and clearly define the variables that can or need to be assessed. This can also aid in better defining the problem to address and associated research questions.
- Formulate hypotheses that identify the expected effects of the independent variable over the dependent variable, in a way that can be tested/measured. The literature review should also allow identifying the effects that can be expected from manipulating the independent variables. This should be in the form of clear and specific predictions about the relationship between the two types of variables and also the expected direction of this relationship. For example, an hypothesis could be formulated as: “Participants presented with information communicating uncertainty about the health consequences of consuming product X, will have a higher risk perception for that product, compared to participants to whom certainty of no health consequences was communicated”.
- At this stage it is important to:
Define how the independent variable(s) will operate and be manipulated. Generally, the manipulation is accomplished by varying its presence vs. absence (e.g., communicating uncertainty vs. comunicationg certainty), the amount and/or the form of independent variable (e.g., communicating different types of uncertainty). It is also important to determine how variation will be accomplished (e.g., by manipulating instructions or events or by measuring the internal states of participants). Following the example given before, one can determine what type of uncertainty the study is investigating, in this case, uncertainty about the health consequences of consuming product X. In order for uncertainty to be manipulated, the researcher should also determine which comparison can allow the demonstration of the effect of this. In this case, this could be achieved by comparing it with a condition in which uncertainty was not present.
Define how the dependent variable(s) will be measured. At this stage it is important to consider and establish the validity of the dependent variable by obtaining evidence that it actually measures the construct (e.g., attitude, behavior, risk or benefit perception) that one wants to measure. One way of ensuring this is by incorporating measures used in previous research and found in the corresponding literature.
- Choose the type of experimental design to be used in the study. This decision should be based on determining which is the most effective way to answer to the research question and hypotheses under study, based on the characteristics of the independent and dependent variables (i.e., how they will be manipulated and measured) and the strengths and weaknesses of each type of experimental research. For example, the comparison between the uncertainty condition and a condition with this absent (control condition), could allow the hypothesis testing. However, the researcher can also determine that this is not enough to avoid potential confounds from other variables and other conditions might be needed for comparison (e.g. testing different types of uncertainty in different conditions).
- Identify possible extraneous and confounding variables and incorporate the appropriate control techniques. In the attempt to identify a causal relationship one must control the influence of extraneous variables. These refer to any variable other than the independent variable(s) that may influence the dependent variable(s). There are many kinds of extraneous variables; however some control techniques can be used to limit their impact.
- Choose sampling strategy and determine the sample size. This decision should depend on the research question and hypotheses, the nature of the population being studied, the complexity of the experimental design and the resources available.
- Develop the materials to be used during the experiment. This should include the preparation of: instructions to participants, describing the purpose (or disguised purpose) of the experiment and the tasks involved; measures of independent and dependent variable; the participant’s consent forms, informing participants right to withdraw from the study at any time and ensuring the confidentiality of the data collected; and the debriefing form, explaining the (true) purpose of the experiment and asking participants not to discuss it with others.
- Pilot-test the experiment. Although not mandatory, this can help evaluate if the instructions and the procedures are clear to the participants. It can also indicate whether the independent variable(s) manipulation produces the intended effect and the sensitivity of the dependent variable(s) measure(s).
- If necessary, revise the experimental design and the procedure in accordance with the pilot-test’s feedback and results.