Angular path integration errors are greater for trajectories with full loop crossovers

Does crossing one's path result in greater uncertainty in one's internal sense of direction (path integration)? The present study addressed this issue using immersive virtual reality involving matched paths with and without a crossover. Crossed compared to uncrossed paths showed greater pointing variability to the start location; the location of the crossover also had an effect on pointing variability. A simple vector addition model was not able to capture these results, suggesting the need for additional assumptions. These findings suggest that a looped crossover path incurs more errors in path integration, suggesting important constraints on models of path integration.