3. Provide a check on the accuracy of the kill as 
reported by the landowner (this would allow a 
correction of the total estimate for inaccurate 
reports, if necessary). 
Observations were to be conducted on opening weekend and then at 4-day 
-intervals. The kill for the first 3 days exceeded expectations, and it 
became necessary to conduct observations each day to obtain the necessary 
data before the quota was reached. 
A check of past records from the State Managed Public Hunting Area 
check stations, together with field observations, indicated that approxi- 
mately 75 percent of the total daily harvest occurs before 9:30 a.m. On the 
strength of this information, observations on kill and crippling were dis- 
continued at 9:30 a.m. each day. The observed kill was assumed to be 75 per- 
cent of the day's total, and for purposes of comparison with the reported 
kill it was simply expanded to 100 percent. 
The 1960 season observations, corrected and reported, as well as crip- 
pling loss, are summarized in table 3. 
The "corrected" total from the 35 days of observations shows a kill of 
LOO geese, while actual reports from the same areas for the same days show 
only 337 geese killed. Then if all assumptions are correct, the total kill 
reported for zone 1 in which the observations were all made can be corrected 
for inaccurate reports by the following procedure: 
“s. 63 (difference between reported kill and observed calcu- 
lated harvest on certain farms under observation) 
4,020 (total reported kill by farmers in zone 1) 
337 (reported kill on farms under observation) 
X (correction necessary for underreporting) 
63 : 337 as X : 4,020 = 751 (correction for underreporting) 
The seme method was followed in the 1961 season with only one important 
change. Observers were instructed to remain on the area they chose to observe 
each day until by their estimate 90 percent of the daily kill had occurred. 
This method has provided a correction in the kill both in 1960 and 1961. 
In both years the corrected kill from the observation areas exceeded the 
reported kill. However, the differences are not satistically significant 
because of high variability of both observed and reported kills. Examina- 
tion of table 3 shown that the reported kill equals or exceeds the corrected 
kill in better than half the cases. The greater share of the difference in 
the totals of reported or corrected kills is due to only a few excessively 
low reports. 
13 
