At the end of every month, each coder mailed to the Center the Code Sheets representing 

 completely coded journal articles (preferably, completely coded journal issues), completely coded 

 sets of Screening Program test reports, etc. , accompanied with a record of the number of lines coded 

 and time spent. Also accompanying the Code Sheets were special "Coder's Comment Sheets" on which 

 the coder explained any problems encountered in coding the particular data, etc. These Comment 

 Sheets are discussed later in Appendix B. 



Checking of Coding; Arbitrating between Checker and Coder : 



The Center treated a shipment of Biology Code Sheets from the coder as a "Set", for convenience 

 in maintaining records. The Set was given a number which, however, was solely for purposes of main- 

 taining its integrity through the several processing steps prior to final filing. Each Code Sheet was 

 at this point assigned its Code Sheet Number and other records were made for it. The Set of Code 

 Sheets was then ready to be checked. 



It was discovered that errors made in interpretation and coding were far too frequent and serious 

 ever to permit coding being punched and the Code Sheets filed without having first been checked once 

 by a second person. Ordinarily, those coders who had developed the most skill and who had demon- 

 strated superior accuracy and comprehension of coding procedures were regularly assigned coding 

 from other coders for checking. However, none of the persons did only checking, but each was always 

 given some coding assignment which was subsequently checked by another of the more experienced 

 coders. 



Checked coding was returned from its checker at the end of each month, accompanied by a 

 record of the number of code lines checked, the number of code lines prepared by the coder which the 

 checker had deleted, the number of code lines added by the checker, and coded Sheets added by the 

 checker. Also accompanying the work were comments (on special Checker's Comment Sheets); these 

 were addressed by the checker to the coder of the Set and pointed out errors made in coding, reasons 

 for deletions and additions, etc. Finally, the checker sent with the Set an explanation of any coding 

 problems he was unable to solve, items that needed to be added to the Code or Key, and questions 

 about procedure for which the Key had no explanation. 



Upon receipt of the checker's work at the Center, the Checker's Comment Sheets were recorded 

 and mailed to the coder of the Set and the Set was processed (any added Code Sheets were given Code 

 Sheet Numbers, etc. ). Each Set was assigned then to one of the resident biology staff members. 



It was the policy of the Center that a resident staff member examine at least superficially each 

 Set of checked coding. If the checker had sent comments indicating difficulties or disagreement with 

 the coder, these comments were studied with the Set. Frequently, it was necessary to obtain the 

 original data in order to arrive at a solution of the difficulty or to make a special coding provision. 

 This examination and correction of the coder's and checker's work the Center referred to as "arbitrating". 

 While this might possibly seem to require a minimum of time, it actually was a very tedious and time- 

 consuming task; the results of any decisions needed to be transmitted to both the checker and coder 

 who pointed out the difficulty (or who erred in coding) and, if additions and changes in the Code 

 and/or Key were necessary, all coders had to be advised of them. An accumulation of such changes 

 was largely responsible for the regular Supplements to the Code and Key, issued to the coders. 



Each Code Sheet was initialed by the coder, checker, and arbitrator. When the staff member 

 completed his arbitration of the Set, it was returned to the clerical staff for notation of its having been 

 arbitrated. 



Processing of Biology Code Sheets by Chemists : 



The arbitrated Code Sheets had all the information about the test compound which the author 

 gave, but only that information. None of the biologists (the coder, checker, or arbitrator) had iden- 

 tified the test compounds with their CBCC Chemical Serial Numbers; the coders and checkers, being 

 non-resident, did not have access to the CBCC chemistry files and the arbitrators, though resident 

 biologists, had no reason to develop the proficiencies in identifying chemicals which the chemical 

 staff was specialized to do. Therefore, the Set of arbitrated Biology Code Sheets was submitted to 

 the Center's chemistry staff. There all the chemical information recorded on the Code Sheets by the 

 biology coder was used to determine the test compound's identity with one of the compounds already 

 entered in the CBCC chemistry files. If the information on the Biology Code Sheet were inadequate 

 to ascertain the identity, the chemists frequently wrote to the author for more information. If the 

 chemical were new to the CBCC chemistry files, it was coded on a Chemistry Code Sheet and index 

 cards were made for it. 



- 213 - 



