
Table 12 - PROGRESS OF RETURN 

Mailing Report No. 
Date Period Deliv. Date Accum. Ret. Cum. 
Oct. 7 7-8-9 627 Oct. 14 151 oh 
18 188 30 
21 201 32 
Oct. 13 10-11-12- 691 Oct. 19 215 31 
13-14 al 2h3 35 
26 275 ho 
Oct. 15 15-16 783 Oct. 20 203 26 
ah 268 34 
27 308 39 

The rate of return varied between strata. The best and fastest return 
came from farms nearest the marsh where the kill is highest. There is a 
natural tendency for people with the most to report to be the most prompt in 
reporting. A total of 781 questionnaires were sent to farmers in townships 
bordering the marsh and 40 percent were returned. 1,320 questionnaires were 
sent to farmers in the townships of the outer ring that do not border on the 
marsh. They averaged only a 36.4 percent return. These ratios yield a 
chi-squared value of 3.462, which lies between probabilities of 0.06 and 0.07. 
A group of 140 cooperators on the first questionnaire were resampled on 
the final questionnaire. The return from these experienced cooperators 
amounted to 48 percent, in the same time required to get a 37 percent return 
from those contacted the first time. The return of both experienced and 
inexperienced might have been greater except for the fact that it was known 
that the season was to be closed before the questionnaires were received. 
Duplicate Reports Between Surveys 
The mailed survey and the ground survey both rely on the cooperation of 
farmers. The close agreement of results might be criticized by arguing that 
both rely on farmers. Some farmers cooperated in both surveys, thus they 
should agree. It is true that some farmers were reporting to both surveys, 
put it is doubtful whether they influenced the estimates too heavily. A 
check of respondents to the mailed survey against those cooperating on the 
ground survey shows approximately 30 farmers reporting on both surveys. The 
smallest number returned on any mailing was 201 questionnaires. All of the 
duplicate reports from the 100-percent check zone were compared. Reports. on 
the mailed survey averaged three more geese each than did the same individual's 
report on the ground survey. Although duplication did exist, it probably did 
not force the close agreement between the two surveys. 
Crippling Loss 
The total crippling loss was not measured by this questionnaire. Coop- 
erators were asked whether any dead or crippled geese had been seen on the 
farm, and if so, how many. This would result in a low figure as it excludes 
26 
