THE PERFORMANCE OF INFORMATION PROCESSING EQUIPMENT OR AN INDUSTRIAL APPLICATION OF THE ANALYSIS OF VARIANCE TO DATA ON THE PERFORMANCE OF THREE BRANDS OF LONG-WEAR MAGNETIC TAPE AND TWO CHANNELS AND FOUR TAPE DRIVES OF THE I. B. M. 7090
Abstract
Three brands of long-wear magnetic tape were tested. The criterion for evaluating performance was the number of error stops encountered in testing the entire tape on each reel during the first run. The analysis of variance with four criteria of classification and equal frequencies within cells revealed that there is a significant difference in performance among brands. Although the evidence of this difference is reliable, the classification of reels by brands is the least important by the four used. The use of a guaranteed reel rather than a standard stock reel accounted for the greatest amount of variance, approximately 13 times as much variance as the brands. The two types of error stops accounted for over 7 times as much variance as the brands. Two groups of tape drives accounted for almost 3 times as much variance as the brands. The statistical analysis of variance revealed that the ranking of the brand depends heavily upon the other factors. The significant two-way and three-way interactions were interpreted, and three-dimensional illustrations were presented to give a visual, geometrical interpretation of these interactions or joint effects of two and three factors in the analysis of variance. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 1961
- Accession Number
- AD0269555
Entities
People
- H.p. Hartkemeier
Organizations
- Lockheed Martin Missiles and Space