Validation of a recording system for computer pointing device activity

This study reports on the development of the pointing device activity monitoring software, Mlog, which is used to assess elements of computer use such as pointing device use time. The Mlog monitoring program was written with Virtual C++ and is 1.59 MB in size. It is applicable for computers equipped with computer processing units (CPUs) of 1 GHz or faster, and records pointing device activity at the millisecond level. The accuracy, precision, and reliability of the Mlog monitoring program were tested by comparing Mlog-recorded activity time intervals, for example, the time interval between pressing and releasing the pointing device key, against oscilloscope-recorded pointing device electric signals. For most tested time intervals, the Mlog monitoring program's margin of error for accuracy was within +/-0.50%, and it boasted precision error rate of less than 1.00%. Correlation coefficients between the Mlog recordings and the references were all greater than 0.99905, with the highest values observed for dragging actions. For most tested time intervals, the results of the reliability test showed that their intraclass correlation coefficients were higher than 0.95, with few less than 0.90 for those with relatively longer tested time intervals. Linear regression analyses indicated that the pointing device clicking action time intervals recorded by the Mlog monitoring program were very close to the reference values and not affected by CPU speed and memory capacity, computer system loadings of software programs in operation, or types of pointing device key action. In conclusion, the Mlog monitoring program was validated as an accurate tool for monitoring pointing device activity during computer work. Future study is warranted to examine the influences of idle time threshold for electronic activity monitoring, passive computer tasks, and job type on the performance of the program in field studies. Relevance to industry: The Mlog monitoring program was validated as accurate in measuring pointing device activity during computer work, providing an indirect and supplemental exposure indices for those traditional ergonomic measurements for physical loading assessment. (C) 2009 Elsevier B.V. All rights reserved.