Skip to content
Resource Methods Toolkit wp6-tuni Published: 12 May 2022


An experiment is a method that is planned to enable causal explanations (Cook & Campbell, 1979). In an experiment, the researcher is probing whether a certain event (a task, a program, use of a new app etc.) is causing a certain result. The expected causal relation between a certain event and a result is called a hypothesis.  An example of a hypothesis is “Children who are exposed to the anti-cyberbullying intervention will develop a higher perceived behavioural control to intervene in cyberbullying situations” (Vlaanderen et al., 2020). Thus, the event does not necessarily always culminate in the development of the result (for example the development of a change in behavior occurring in every participant).  The causal relation is established when the result occurs for most of the participants; in other words, the result occurs with certain probability. The probability of the occurrence of the result of interest is compared to the probability of occurrence of the result in case of a neutral (control) event. 

Designing experiments

Every experiment is planned in accordance with a theory. For example, the ‘Theory of Planned Behaviour’ (Ajzen, 1991) is used in our example to study a concerning anti-cyberbullying intervention (Vlaanderen et al., 2020)). The theory is used to define the expected relationship between the causal event and the result. The theory is also used to define the important, causal aspects of the intervention. 

The intervention is named an independent variable because the participant cannot change or choose this variable. The study’s independent variable is predetermined before the participants are involved in the experiment; therefore, it is independent from the participant. The dependent variable(s) are also predefined before the experiment. The dependent variables are the measures of behaviour of interest. Dependent variables are thus named because they are expected to be dependent on independent variables. For example, when perceived behavioural control change is the result under study, the change can be measured with pre- and post-intervention questionnaires. Questions assess cyberbullying knowledge, awareness of cyberbullying, and empathy toward victims on a six-point likert scale (1=totally disagree, 6=totally agree). In this example, the independent variable (anti-bullying intervention) is expected to cause a change in the dependent variable (in the answers to the perceived behavioural control questionnaire). Confounding variables (place an unknown influence of the participant’s measured behaviour) can adversely influence the result of the intervention. Therefore, it is important to find out about the participants’ lives and activities to take those variables under control. Another option is to make sure that the confounding variables occur randomly, meaning that it has a random value in each participant’s case, and all the values of the confounding variable occur in approximately equal amounts in the study group. For example, in the case of anti-bullying intervention, the study may coincide with the premiere of a popular movie that is also about cyber-bullying. It becomes a confounding variable when some of the participants in the study go to see it, and others don’t. Taking this influence of the dependent variable under control means that one could add a question in the questionnaire about whether a participant has seen the movie or not, or take all the participants to see the movie to equally expose them to this variable.  

It is important to design a control condition that is similar to the independent variable condition but differs in content and does not affect the studied behaviour. Experimental design means that participants of the study are assigned to study groups randomly (when different groups are used).


  • It is unlike any other method, it enables researchers to study causality.

  • It enables to research a problem or phenomenon step-by-step (with consecutive experiments).


It may be complex 

  • come up with independent variables, (interventions or aspects of interventions) which are truly representative of the phenomenon and influence the dependent variable

  • determine the best dependent variables

  • design a control condition that shares all important aspects with the intervention but has no relevant influence on the studied behaviour

  • avoid the known or unknown influence of confounding variables (like researcher bias, flaws in the procedure, expectations of the participants, sampling problems), due to the confounding variables the expected causality may not be observed in the experiment.

  1. Ajzen, I. (1991). The Theory of Planned Behavior. In ORGANIZATIONAL BEHAVIOR AND HUMAN DECISION PROCESSES (Vol. 50).

  2. Bleize, D. N. M., Anschütz, D. J., Tanis, M., & Buijzen, M. (2021). The effects of group centrality and accountability on conformity to cyber aggressive norms: Two messaging app experiments. Computers in Human Behavior, 120, 106754.

  3. Cook, T. D., & Campbell, D. T. (1979). Quasi-Experimentation: Design and Analysis Issues for Field Settings. Rand McNally.

  4. Slater, A., Halliwell, E., Jarman, H., & Gaskin, E. (2017). More than Just Child’s Play?: An Experimental Investigation of the Impact of an Appearance-Focused Internet Game on Body Image and Career Aspirations of Young Girls. Journal of Youth and Adolescence 2017 46:9, 46(9), 2047–2059.

  5. Vlaanderen, A., Bevelander, K. E., & Kleemans, M. (2020). Empowering digital citizenship: An anti-cyberbullying intervention to increase children’s intentions to intervene on behalf of the victim. Computers in Human Behavior, 112, 106459.

Share this post:


Team member, CO:RE at TUNI

Iiris Tuvi

Iiris Tuvi, PhD, is a post-doctoral Research Fellow a the Faculty of Information Technology and Communication Sciences of Tampere University. She has long experience in experimental psychology and methods involving psychometrics and data analysis. Currently working on methods used to research children in digital environments.

Tampere University
Tampere University

The team at the Faculty of Information Technology and Communication of TUNI identifies, develops and provides access to resources on qualitative, quantitative and mixed research methods together with evaluating their validity in research practice. These resources are collated in the CO:RE methods toolkit that cross-references resources from the evidence base, the compass for research ethics, and the theory toolkit, to give users tools to apply to their individual research contexts.

Leave a comment

Required fields are marked with a *
Your email address will not be published.

Cookie preferences

We use cookies on our website. Some of them are essential, while others help us to improve this website and your experience.