294 IMPROVING THE ACCURACY OF TIMBER ESTIMATES 



But where strips are run at intervals of | mile and the timber is 

 open and large, and especially in coniferous stands which have a fair 

 degree of uniformity of sizes, although varying materially in density, 

 it is possible to view the remaining timber without counting it or caliper- 

 ing. If there were tmie for additional measurements, these would be 

 made. The application of a correction factor is based on the assumption 

 that the per cent actually measured is the maximum possible under 

 the luniting conditions. Where an error would evidently be incuned 

 unless the mechanical average is corrected, this correction should alwaj s 

 be made. 



The method of applying this sort of a correction in the past has 

 been as unsystematic as the ocular estimation of timber itself. The 

 estimate from sample plots or strips was arbitrarily raised or lowered 

 according to unpressions obtained by the cruiser. This system may 

 be greatly improved and a much higher per cent of accuracy obtained 

 by observing the following principles: 



1. The comparison sought is not an absolute estimate of the volume 

 per acre on the remaining area, but a percentage relation between this 

 stand and the strip which is measured, by which the estimate on this 

 remaining area may be obtained by increasing or diminishing that on the 

 strip. 



2. The correction is an average for the whole area to be corrected, 

 in the form of a per cent of total volume. Single observations must 

 therefore be carefully weighted to obtain average results. 



3. The correction actually applies only to the area lying outside 

 the strip and not measured. If applied to the entire area of the unit, 

 the estimate on the strip itvself is arbitarily raised by the same per- 

 centage as applied to the residual area and this factor cannot be neglected 

 in arriving at the proper per cent. 



To illustrate the last point, assume that 50 per cent of a tract has 

 been estimated. By observation, the correction factor on the remainder 

 is assumed as + 10 per cent. The estimate is 100,000 board feet on the 

 strip. The correct estimate on the remaining area is therefore 110,000 

 board feet and the total, 210,000 board feet. If 10 per cent is applied 

 to the results obtained for the forty, the process would be, 10Q,000 

 times 2 gives the uncorrected estimate for the area, or 200,000 board 

 feet. A correction of 10 per cent gives 220,000 board feet, which is 

 an error of 4.8 per cent in the estimate.^ 



• This multiple, which in this illustration is 2, is sometimes termed the correction 

 factor, but assumes no correction. It is merely the extension of the mechanical 

 average over the entire area. For a 25 per cent estimate, the multiple is 4; for 

 20 per cent, it is 5, etc. A method of applying the correction factor is in use, by which 

 this multiple is raised or lowered. Where the multiple is 4, a +25 per cent correc- 

 tion calls for 5; +12^ per cent requires 4^, etc. 



