removed for fecundity studies of selected species. 



The initial aspects of data processing deal with the completed trawl 

 log. After the log was coded for machine processing, all information was 

 scanned for errors of omission, inconsistencies, or mistakes in calculations. 



Machine processing involves the production of five data record types to 

 facilitate subsequent computer analysis and auditing for gross errors. There 

 are five different data record types: type one contains the number and weight 

 of a particular species taken at a particular station; type two contains the 

 corresponding length frequency data for the previous catch data record; type 

 three contains age-length data by species which are entered into the system 

 after age and growth samples were processed at the laboratory; type four is a 

 summation of the total weight and number for all species combined at a station 

 (produced by the computer by summing the type one records for a station); and 

 type five contains detailed station data. The result is that for each species 

 at each station, there will be a type one and a type two record in the data 

 set. There will be only one type four and type five record for each station, 

 and there will only be type three records if age and growth samples were 

 taken. Record types one, two, and five were used in the production of this 

 report. 



Auditing basically consisted of cross-checking common values between 

 record types; the totals for each species on both type one and type two 

 records; and the observed weight and number for specified species, with the 

 predicted weight and number after a length-weight equation has been applied. 

 In addition, omission errors and gross latitude and longitude errors can be 

 detected. 



After audits were completed ana errors corrected, the data were then 

 stored on magnetic tape for later use. 



