ANALYSIS 
Limitations of Method 
Results from the singing ground counts apply in a direct 
way only to the particular set of routes; interpretation beyond this 
must be very cautious. Routes are selected by the cooperators as 
areas where woodcock may be heard and thus may represent something 
better than the average of all potential habitat. Routes are re- 
placed from time to time. Since the routes are not randomly selected, 
the extent to which counts made on them reflect population changes 
of the species is unknown. It is, of course, anticipated that some 
relationship exists. 
It seems unlikely, however, that it ever will be practical 
to sample woodcock habitat in a random manner throughout the wood- 
cock's entire breeding range. Therefore, an important research 
need’ is to evaluate the extent to which changes in counts on the 
survey routes selected in the present manner accurately reflect 
the. changes in the population throughout a large area. With these 
limitations in mind, the data presented here are the best available 
at present. 
This analysis is based on data from those States and Provinces 
with two or more routes comparable in 1959 and 1960. (A minimum of 
two routes is necessary for the estimation of sampling error.) The 
important States of Wisconsin and Minnesota had to be omitted this 
year for lack of sufficient coverage, as did the entire southern 
perimeter of the woodcock's range. 
Weighting Factors 
Weighting of the data is necessary because the number of 
- routes varies from area to area without relation to woodcock popula- 
tions. We lack, as weights, adequate figures on woodcock population 
density and on area of suitable nesting habitat throughout the breed- 
ing range, but have used land area as a reasonable substitute. 
From 1956 through 1959 the weights used were computed on the basis 
of mean woodcock per stop in a period of recent years, and the total 
land area of each State or Province (adjusted in a few cases where 
large portions of a State or Province lay in ecological regions 
where no woodcock habitat occurred). The weights used in 1960 are 
restricted to the area of uncultivated land in the State or Province. 
It is realized that much unfavorable nesting habitat still is in- 
cluded, and it is hoped that future study will make it possible to 
delete other unsuitable habitats so the weighting procedure may be 
refined still further. 
