SIZE OF AREA AND PER CENT OF AREA ESTIMATED 263 



itself as increasing or decreasing the probability of accuracy for a given 

 method of sampling. 



The first of these problems is influenced b}^ the size of the tract. 

 In any method of estimating based upon measuring a part of the area, 

 the system employed must be that of strips or plots spaced at regular 

 intervals. Otherwise the element of judgment in selection introduces 

 a difficult factor which will improve the average obtained only when 

 accompanied by considerable individual skill. With plots or strips 

 at fixed intervals, the number of such strips depends upon the dimensions 

 of the tract. 



The choice between plots and strips does not affect this principle. 

 Plots, when substituted for strips and taken along compass courses at 

 regular mechanical intervals, serve to reduce the per cent of total 

 area covered. Since the distribution of the sample areas is more evenly 

 diffused on the basis of the per cent covered, by plots, than it is by 

 strips, the loss in accuracy by substituting plots for strips is not in 

 proportion to the reduction in per cent of area covered, but is consider- 

 ably less, thus resulting in a material saving where the use of plots 

 permits of the reduction in size of crew (§ 224). 



The size of the separate units of area on which accurate estimates 

 are desired — as for instance, when owners require the estimates sepa- 

 rately by " forties " (§ 8), is the basis for determining the effect of the 

 spacing of these strips. If the estimate must be accurate only for the 

 entire tract, a quite different problem is presented from that when the 

 same degree of accuracy is required for smaller subdivisions. Assuming 

 that the tract is in the form of a square, the coefficient of accuracy bears 

 a close relation to the number of strips run across this area, rather 

 than to the distance between these strips. This may be expressed as 

 follows: 



The per cent covered by strips wiU be the product of the number 

 of strips and width of each strip, divided by width of the area. With 

 strips of a uniform width, e.g., 8 rods or 132 feet, run at intervals of 

 i mile, the per cent of area covered is A or 10 per cent, whether the tract 

 be 40 acres, 1 square mile or 25 square miles. But the probability of 

 accuracy in securing an average stand is not in the same proportion 

 for each tract, but increases with the size of the tract. The reason is 

 that, regarded as a unit, the larger tract is more uniformly sampled, 

 and with reference to its total area, the strips or plots are more 

 thoroughly distributed than on the smaller areas. The relative accu- 

 racy is m proportion to the distribution of the sampled or estimated 

 strips with respect to this total unit, which for large tracts tends to 

 reduce the per cent of area required to obtain a given standard of accu- 

 racy. 



