May 2007
Volume 48, Issue 13
ARVO Annual Meeting Abstract  |   May 2007
Maze-Tracing Performance Measure of Visual Motor Ability in Low Vision
Author Affiliations & Notes
  • R. W. Massof
    Ophthalmology, Johns Hopkins Wilmer Eye Inst, Baltimore, Maryland
  • G. D. Barnett
    Ophthalmology, Johns Hopkins Wilmer Eye Inst, Baltimore, Maryland
  • Footnotes
    Commercial Relationships R.W. Massof, None; G.D. Barnett, None.
  • Footnotes
    Support NIH Grant EY012045
Investigative Ophthalmology & Visual Science May 2007, Vol.48, 2350. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      R. W. Massof, G. D. Barnett; Maze-Tracing Performance Measure of Visual Motor Ability in Low Vision. Invest. Ophthalmol. Vis. Sci. 2007;48(13):2350.

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

  • Supplements

Develop and validate a performance measure of visual motor abilityin visually impaired patients.


Two hundred low vision patients traced 20 different mazes ona touch-screen. Each maze differed in path length, used 1 of3 track widths, and used 1 of 4 line thicknesses. Tracing performancefor each maze was summarized with the log cumulative area betweenthe trace and the wall of the maze that fell outside the track.The difficulty of each maze, the maze-tracing ability of eachpatient, and the validity and reliability of the performancemeasures were estimated with Rasch analysis. The same patientsrated the difficulty of subsets of 173 visual motor daily livingtasks in the Activity Inventory. Self-perceived visual motorability for each patient was estimated from Rasch analysis ofthe difficulty ratings.


The estimated difficulty of each maze was linear with path lengthand track width (see Figure 1), maze difficulty was independentof line thickness. Between subject variability scaled with trackwidth, which manifested as apparent violations of unidimensionalityof maze difficulty. However, estimates of maze-tracing abilityof patients were consistent with unidimensional measures. Personmeasure separation reliability was 0.94 (separation index =4.03), indicating that only 6% of the observed variance canbe attributed to estimation error. As shown in Figure 2, maze-tracingability is linear with self-perceived visual motor ability estimatedfrom difficulty ratings of visual motor tasks.


Maze-tracing provides valid and reliable estimates of visualmotor ability in low vision patients that concur with patients’self-perceived ability to perform visual motor tasks. 


Keywords: low vision • vision and action • aging: visual performance 

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.