Brooks, JL (2012) Counterbalancing for Serial Order Carryover Effects in Experimental Condition Orders. Psychological Methods, 17. 600 - 614.

[img]
Preview
Text
Brooks - Carryover Counterbalancing - Psychological Methods - FINAL.pdf - Accepted Version

Download (626kB) | Preview

Abstract

Reactions of neural, psychological, and social systems are rarely, if ever, independent of previous inputs and states. The potential for serial order carryover effects from one condition to the next in a sequence of experimental trials makes counterbalancing of condition order an essential part of experimental design. Here, a method is proposed for generating counterbalanced sequences for repeated-measures designs including those with multiple observations of each condition on one participant and self-adjacencies of conditions. Condition ordering is reframed as a graph theory problem. Experimental conditions are represented as vertices in a graph and directed edges between them represent temporal relationships between conditions. A counterbalanced trial order results from traversing an Euler circuit through such a graph in which each edge is traversed exactly once. This method can be generalized to counterbalance for higher order serial order carryover effects as well as to create intentional serial order biases. Modern graph theory provides tools for finding other types of paths through such graph representations, providing a tool for generating experimental condition sequences with useful properties.

Item Type: Article
Uncontrolled Keywords: carryover effects; condition order; counterbalancing; experimental design; graph theory; repeated measures
Subjects: B Philosophy. Psychology. Religion > BF Psychology
Divisions: Faculty of Natural Sciences > School of Psychology
Depositing User: Symplectic
Date Deposited: 30 Oct 2017 09:47
Last Modified: 30 May 2019 13:13
URI: https://eprints.keele.ac.uk/id/eprint/4141

Actions (login required)

View Item View Item