Skip to main content
ARS Home » Pacific West Area » Maricopa, Arizona » U.S. Arid Land Agricultural Research Center » Pest Management and Biocontrol Research » Research » Publications at this Location » Publication #360162
Research Project: Ecologically Based Pest Management in Western Crops Such as Cotton

Location: Pest Management and Biocontrol Research

Title: Common statistical mistakes in entomology: pseudoreplication

Author
item Spurgeon, Dale

Submitted to: American Entomologist
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/16/2019
Publication Date: 3/12/2019
Citation: Spurgeon, D.W. 2019. Common statistical mistakes in entomology: pseudoreplication. American Entomologist. 65(1):16-18.

Interpretive Summary: Pseudoreplication occurs when a sample or observation unit is used as if it was a true statistical replication. In some cases of pseudoreplication, experimental treatments are not replicated at all. In other cases the treatments are replicated but the statistical analysis does not correctly distinguish these replicates from the sampling units. This statistical problem is common in the entomological literature, and often results in incorrect conclusions. A key to avoidance of pseudoreplication is recognition of the experimental unit, which is the physical entity to which treatments are assigned and applied. When these experimental units are properly identified in a statistical analysis, valid tests of the treatments are possible. As a rule of thumb, when statistical software reports any test of a treatment in which the error degrees of freedom are much greater than the number of experimental units, pseudoreplication has occurred. This simple guideline will allow scientists to recognize and correct problems with pseudoreplication, and readers to recognize when reported statistical tests are not valid.

Technical Abstract: Pseudoreplication occurs when a sample or observation unit is used as if it was a true statistical replication. In some cases of pseudoreplication, experimental treatments are not replicated at all. In other cases the treatments are replicated but the statistical analysis does not correctly distinguish these replicates from the sampling units. This statistical problem is common in the entomological literature, and often results in incorrect conclusions. A key to avoidance of pseudoreplication is recognition of the experimental unit, which is the physical entity to which treatments are assigned and applied. When these experimental units are properly identified in a statistical analysis, valid tests of the treatments are possible. As a rule of thumb, when statistical software reports any test of a treatment in which the error degrees of freedom are much greater than the number of experimental units, pseudoreplication has occurred. This simple guideline will allow scientists to recognize and correct problems with pseudoreplication, and readers to recognize when reported statistical tests are not valid.