Treatment Diffusion

Design of Experiments >

What is Treatment Diffusion?

Treatment diffusion is where the control group is affected by the treatment. This could happen because individuals in the control groups and treatment groups talk to each other about the treatment. As such, this is usually an issue in research involving training or informational programs (Borg, 1984). For example, let’s say a set of teachers was exposed to a new teaching technique, and it greatly increased test scores. Teachers in the experimental classes might tell their control group peers about the new teaching tool.

“Secondary” diffusion happens when participants in the treatment group change their behavior in some way, demonstrating to the control group a new behavior, habit, or technique. Indirectly, control group participants behavior is changed because of viewing the altered behaviors in the treatment group.

Issues and Avoidance

Treatment diffusion can make it challenging for you to figure out:

  • If study results are due to a particular intervention.
  • How effective the intervention is.

This could lead to incorrect statistical conclusions, biased treatment effect estimates, and increased Type II errors.

To avoid treatment diffusion, consider placing the control groups and treatment groups in different locations (Borg & Ascione, 1979). For example, choose two different schools, businesses, or geographic locales. However, care should be taken to ensure that you don’t run into other issues with validity. For example, even though you choose two different schools, they should be identical in every way to ensure that other factors (like socioeconomic status, length of school day, or better teachers) aren’t affecting your study results. Another issue with placing experimental and control groups in different locations is that random assignment isn’t possible, which means your experiment will be quasi experimental instead of experimental (Borg, 1984).

References

Borg, W. (1984). Dealing with Threats to Internal Validity That Randomization Does not Rule Out. Retrieved December 28, 2018 from: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.905.2578&rep=rep1&type=pdf
Borg, W.R., & Ascione, F.R. (1979). Changing on-task, off-task and disruptive pupil behavior in elementary mainstreaming classrooms. Journal of Educational Research, 72, 243-252
Danga, L. A., & Korb, K. A. (2014). The effect of treatment diffusion on educational experimental designs. Benin Journal of Educational Studies (BJES) 23, 29-37
Harvard School of Education. Threats to validity in intervention studies; Potential problems, Issues to consider in planning. Retrieved December 28, 2018 from: https://canvas.harvard.edu/courses/5058/files/1180679/download?wrap=1


Comments? Need to post a correction? Please Contact Us.

Leave a Comment