Randomization (random allocation) (September 8, 2020 via “Cochrane Austria”): 14-min video.

• Summary: Overviews the principle of randomization (by contrasting results of an observational study vs. a randomized trial) and unpacks various randomization schemes or techniques, i.e., simple (a “coin flip”), block, urn, stratified, and covariate-adaptive randomization. The “non-simple” randomization techniques exist to overcome issues that may arise when randomizing small groups of individuals, which may result in unequal group sizes, and imbalances in baseline characteristics between the intervention and control groups.

Kang M, et al. Issues in outcomes research: an overview of randomization techniques for clinical trials. J Athl Train. 2008 Apr-Jun;43(2):215-21. (7-page paper)

• Summary: Describes the advantages and disadvantages of various randomization schemes (techniques) in clinical trials, including simple, block, stratified, and covariate-adaptive randomization. A flowchart is provided to help researchers select an appropriate randomization scheme for their study.

Randomized Controlled Trials (July 21, 2020 via “Cochrane Austria”): 9-min video.

• Summary: Introduces the concept of randomization as a powerful tool in health research, and visualizes various randomization designs (e.g., parallel group, crossover, and factorial designs, as well as cluster and adaptive designs).

The European Patients’ Academy (EUPATI) - Clinical trial designs (Medicines R&D) (Last updated: November 23, 2015): 1-page website.

• Summary: Discusses parallel group, crossover, and factorial randomization designs, with visual depictions of the different approaches. Matched-pair designs, “withdrawal trials,” cluster trials, as well as terminology such as superiority, equivalence, and non-inferiority, are also introduced.

NIH Pragmatic Trials Collaboratory - Understanding Clustering in Cluster Randomized Trials (June 19, 2020): 17-min webinar (24-slide presentation; slides 6-29)

• Summary: Dr. Elizabeth Turner discusses reasons to randomize clusters instead of individuals and introduces the intraclass correlation coefficient (ICC) and the concept of an effective sample size. Further information on the ICC and it’s role in power calculations can be found in the NIH Collaboratory’s “Intraclass Correlation Coefficient Cheat Sheet” (March 15, 2020).

How Factorial Design Works | NEJM Evidence (August 29, 2024 via “NEJM Group”): 5-min video.

• Summary: Short animated video discussing factorial trial designs, which allow a researcher to test the effect of more than one intervention at once, and to potentially assess interactions among interventions.

Hemming K, Taljaard M. Key considerations for designing, conducting and analysing a cluster randomized trial. Int J Epidemiol. 2023 Oct 5;52(5):1648-1658. (11-page paper)

• Summary: Cluster randomized trials require a larger sample size than individually randomized trials and they face additional complexities. This paper provides simple guidelines to help researchers conduct cluster trials in a way that minimizes potential biases and maximizes statistical efficiency.