Ratio Estimation fCorrected Registration Information) 



To generate more accurate estimates of use from raw data computed from completed 

 registration forms, the data can be corrected using an expansion factor based on the 

 estimated registration rate; in statistical terms, this provides a ratio estimate.^ 

 Estimates of use with their associated standard errors, corrected for registration 

 rate, are shown in table 5 for the key variables. 



Raw registration totals were multiplied by the following correction factors: 

 number of visits, 1.35; number of visitor-hours, 1.63; and number of groups, 1.24. 



Table 5 . --Estimates of recreation use on the Mission Mountains Primitive 

 Area^ Montana, based on ratio estimation ( aorreated 

 registration information) , summer- fall, 1968 



Strata 



Visits 



Total 



use ' 



Group 



„ ^ . ^ : Standard 

 Estimate i 

 : error ^ 



; Estimate 



: Standard : 

 : error ■'■ : 



„ ^ . ^ : Standard 

 Estimate i 

 : error ^ 





Number 



Percent 



Hours 



Percent 



Number 



Perc 



Summer 



3,652 



6.3 



49,332 



14.7 



843 



8.1 



Fall 



478 



do. 



7,379 



do . 



161 



do. 



Summer + Fall 



4,130 



do. 



56,711 



do. 



1,004 



do. 



1 



888 



do. 



10,613 



do. 



206 



do. 



2 



806 



do. 



7,389 



do. 



210 



do. 



3 



770 



do. 



6,409 



do. 



170 



do. 



4 



1,188 



do. 



24,921 



do. 



258 



do. 



5 



189 



do. . 



2,277 



do. 



63 



do. 



6 



184 



do . 



3,871 



do. 



66 



do. 



7 



105 



do . 



1,231 



do. 



32 



do. 



■■•Standard error, expressed as percentage of the estimate. Standard errors do not 

 vary from stratum to stratum because in ratio estimation variance is unconditional. 



We used the "combined ratio estimate" from Cochran (1963, section 6.11). 



25 



