Both age and sex determinations are believed to have been 
accurate for mallards, American widgeons, shovelers, pintails, 
canvasbacks, redheads, and lesser scaup. Sex determinations for 
blue-winged teals and wood ducks were accurate, but age determinations 
were known to be accurate only through the month of October, because 
known-age material fram later months had not been available for study. 
Age determinations probably were accurate for both black ducks and 
ringnecks, but at that time the sex of neither could be determined. 
The age and sex of both buffleheads and goldeneyes may have been 
accurate, but no quantitative analysis of known material had been 
made. Age and sex determinations of gadwalls, greenwings and 
greater scaup were of unknown accuracy. 
The wings were discarded after they had been examined and 
verified. The greater part of the wing envelope was trimmed away 
with a paper-cutter, leaving a slip approximately 5" x 8" contain- 
ing the data. State and county codes were added to these slips, 
which then were taken to the Patuxent Wildlife Research Center, 
where all data were transferred onto data cards. 
Weighting Procedures 
There were marked differences among the States in the proportion 
of the hunters that were sent requests for wings. For example, approxi- 
mately one of every 80 Minnesota hunters but only one of every 310 
Indiana hunters was asked to send wingse As a result, the wings 
received from the various States represented different fractions of 
the total kill. Therefore, whenever data from two or more States were 
combined they were weighted to give each State the proper influence 
in the over-all average. 
For future wing collections, it may be desirable also to weight 
data within States, but for the present we lack the information that 
would make it possiblee For this year, when we believed certain 
State samples were either too small or distributed too poorly we 
simply excluded data from these States from certain averages (in each 
such instance, this fact is noted on the table concerned). States 
where the sample was judged adequate in size and distribution were 
Illinois, Iowa, Michigan, Minnesota, Missouri, Ohio, and Wisconsin. 
There are several possible methods for weighting the data, but none 
seems entirely satisfactory. Our method is discussed below It is 
described for mallard sex ratios (males per female),but the general 
Statements refer to other ratios as well. We first weighted the 
number of wings in each category separately (as, male mallard wings 
and female mallard wings), then calculated the ratio of these two 
shy 
