G. S. R. systems which AID data analysis

Conclusions

The system of extracting the responses without the use of a series capacitor operates reasonably successfully. The time constant of the delay/low pass filter is a compromise—if it is lowered below nine seconds some responses will have a reading reduced by a few per cent, if raised above nine seconds the trailing edge of a response is lengthened and a greater time elapses before equilibrium is reached. Response percentages can be checked using the basal trace which also gives a record of long term changes in subject resistance. The timing of the leading edge of a response is accurate, facilitating latency (Edelberg, 1967) and response rise time measurements. Any delay through the divider can be considered at most as a fewmicroseconds.

System 3 has been used successfully and is less expensive to make. Over 15 per cent responses can be scaled for non-linearity although a recorder with more deflection capability is necessary than with system 1.