142 



DESIGN OF CONTROLS 



engines (systems), the results will be equiv- 

 ocal. On the other hand, if differences in 

 the characteristics of the several types of 

 engines are considered, then meaningful 

 statements concerning the relationship be- 

 tween octane rating and engine performance 

 become possible. There is nothing con- 

 tradictory in the fact that the best octane 

 rating for airplane engine fuel is different 

 from the best rating for fuel used in the 

 average passenger car. 



The effects of manipulating a variable in 

 one system are not indicative of the effects 

 to be obtained with similar operations on 

 another system unless (1) the two systems 

 are quite similar in respect to their essential 

 characteristics or (2) a known relationship 

 obtains between the variable and the char- 

 acteristics in which the two systems differ. 

 Control systems may contain essentially 

 the same physical elements and yet differ 

 considerably relative to the control task and 

 relative to the role of the operator in the 

 system. The lack of specific reference to 

 the operator in traditional designs has 

 obscured the apparent need to analyze these 

 differences in the study of design variables. 



In addition to these criticisms of the gen- 

 eral approach employed in most control 

 design studies there are others which relate 

 to deficiencies in the application of scientific 

 procedure. These are: (1) data based on 

 too few subjects, (2) data based on subjects 

 with widely varying practice histories, (3) 

 incomplete statistics, and (4) faulty experi- 

 mental design. As a result the available 

 data are limited in their usefulness and must 

 certainly be apphed with caution. At best 

 they may be used as hypotheses for further 

 research and as substitutes for blind hunches 

 in cases where equipment modifications are 

 contemplated. Their application should 

 be considered wdth full knowledge of the 

 conditions under which they were obtained. 

 The results of their application should be 

 tested by controlled experiments using 

 mock-ups or working models, includiag as 

 much of the total task as possible, and giving 



due attention to the problems of all three 

 classes. 



Many of the pressures and conditions 

 which led to deficiencies in procedure in the 

 earUer studies are no longer operative. 

 Likewise, recent studies dealing with Class 

 III problems have indicated the adoption 

 of a new approach similar to that presented 

 in this chapter. In general, the recognition 

 of the designer's need to include the opera- 

 tor in the design of control systems has 

 been accompanied by a recognition of the 

 need for a different sort of information than 

 is currently available. Attention is now 

 directed toward system analysis and toward 

 analysis of the operator's role in the different 

 systems. A brief orientation to this point 

 of view is presented in the following section. 



System Variables and The Role 



OF the Operator in Manually 



Operated Control Systems 



Throughout the previous sections of this 

 chapter we have considered the operator as 

 one of the elements in a control system. In 

 a very general fashion his role has been de- 

 scribed as the transmission of control data 

 to the rest of the system by means of 

 the manipulative devices provided him. 

 Actually the details of his role differ from 

 system to system, and, conversely, the de- 

 tails of the operator's role may determine 

 the classification of the system. It is 

 necessary therefore, to analyze the ways in 

 which his roles may differ before effective 

 use can be made of his characteristics, before 

 system differences may be accurately ap- 

 praised, and before results of system modi- 

 fications may be properly interpreted. Be- 

 cause of the relative novelty of the present 

 approach, a few general statements con- 

 cerning transmission systems prior to be- 

 ginning the discussion of system variables 

 may help to clarify the terminology. 



Most transmission systems consist of a 

 number of units linked together. Although 

 each unit is a transmission system in itself, 

 to avoid confusion we shall call these units 



