Name the three allocation methods for treatment assignment and identify which is most prone to selection bias.

• A. Non-random, Random, Stratified; non-random is most prone to selection bias.
• B. Random, Stratified, Block; random allocation is most prone to selection bias.
• C. Stratified, Cluster, Systematic; cluster allocation is most prone to selection bias.
• D. Non-random, Random, Stratified; stratified allocation is most prone to selection bias.

Answer: (A)

The main idea here is how the way participants are assigned to treatment affects selection bias. There are three common allocation approaches: nonrandom allocation, random allocation, and stratified random allocation. Nonrandom allocation depends on the investigator’s or participant’s choices or on predictable factors, so groups can end up systematically different before any treatment is given. That systematic difference is selection bias, which can distort the estimated effect of the treatment.

Random allocation uses a chance mechanism to assign treatment, so each participant has an equal probability of receiving any option. This randomizes both known and unknown confounders across groups on average, greatly reducing selection bias.

Stratified random allocation adds a step: before randomizing, you group participants into strata based on key characteristics (like disease severity or age) and then randomize within each stratum. This helps ensure balance on important factors while preserving the benefits of randomization, further mitigating bias.

Among these approaches, nonrandom allocation is the most prone to selection bias because the assignment process itself can be influenced by factors related to prognosis, leading to unequal groups before any outcomes are measured.