Design of Instrument Dials for Maximum Legibility. Part 3. Some Data on the Difficulty of Quantitative Reading in Different Parts of a Dial
Abstract
This is the report of an analysis to determine whether the likelihood of error in quantitative reading is greater in one part of a dial than in another. The error data which are examined are based on over 45,000 dial readings made in the course of two experiments involving a total of 28 subjects. No evidence is found which would indicate that local scale reading errors (errors of rounding, interpolation, and the like) vary with dial sector. The frequency of systematic scale reading errors depended on dial sector for dials graduated from 0 to 50 or from 0 to 100 but not for dials with scale ranges of 0 to 200, 0 to 400, or 0 to 600. The 50's and 100's dials were similar in that they were numbered by ten unit steps. The dominant error made in reading them was an error of reporting a scale value too great by ten units. This error was more prevalent on the right dial halves than on the left and was especially frequent in the scale region from 0 to 9. Thus, while sector has no consistent effect on either local errors or systematic errors for many dials, it may influence the occurrence of specific systematic errors on certain scales. As more is learned regarding scale design factors related to systematic errors, the easier it will be to account for the effect of sector on the ten units error here reported or to identify other dial designs which may produce error patterns varying with sector.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1950
- Accession Number
- ADA800878
Entities
People
- William A. Smith
- William E. Kappauf
Organizations
- Princeton University