1. Summary of Progress A comprehensive series of runs involving a variety of electrodermal and other physiological measures has been completed on a population of 26 subjects under a variety of task conditions and contrived psychological "sets" to assess the relative sensitivity and reliability of various electrodermal measures, including the recovery limb time constant. Assessment o2f impulsivity and anxiety by psychological tests as well-as a psychomotor test of fnhi ition have been obtained on the subjects. Data analysis is still in progress, but some preliminary conclusions have already been drawn and are reported here as an interim progress note. Runs on a separate por)ulation of twenty subjects under a -different set of stimulus concutions have been used to examine the 2dependence of recovery limb time constant upon base conductance level and response amplitude. The initial study of the vascular compcnent in the sl-,In potential response has been completed. Develot)ment of the automatic measurement of recovery limb time constant has progressed to the point at which a system is now available, using first derivatives, which overcome many of the di2fficulties of the earlier appro@, ch in N,vhich second derivatives were employed. The new system has been used to print out the time constants for 22 subjects whose data had been stored on magnetic tape. These data are now being subjected to statistical evaluation. 11. '@'tudv of Characteristics of Recoverv Limb in Rela,'Uion to Oth--r Electrodermal h,leasures a. Der)--nde,,ice of Recovery Limb Time Constant on Base Level. To determine whether the apparent relatl-ior of recover@,- limb to nature of the stimulus condition was to be explained by diff--rences in base resistance level, compe-risons Approved for Releas3 Date 2 N,;ere made between base level and recover, .; lirb time constan" for 21 subjects in two conditions, each of which caused activation and consequent decrease in base resistance, but with different degrees of goal oriented task demands. 2 Mean,,, for the two situations were as follonvs: BackNvard Countiag Condition ven.-,) Cold Pressor Average time constan@.' (scale units) 17. 1 21.7 Average base resist.,.ace level 29. 5 K 27.6 K 7bese de-ta indicate tl,.at despi-Le the fact that the cold pressor task was, If anythin.,-,, -cli,,Yh*,Iy more activating (dffference in level not statistically si,,mLficant) the time constants v.,ere sig,2-.ifican,,Iy differeat (p < .01., paired comporison), tvie difference is therefore not r-ttributable to difference in base level. It- ic to be expected, however, th.,,t subjects v.,ho shom -,reater activatioa for p--r.lormance of a task migh,, show a refle--',,ioa of this'mob,.I2ization both in base level and in recovery limb time constant. AccordLi.-ly the co@--elation betyieer. base cor-- ductance and rime constant was determined for this same grcup ol' subjects during the bact,-ward countilg tas!c, and it showed a signii@ica,,-t rela-tiorship (--r = -.49. r, 05). -7or the cold pressor si;'Iiation where the --c-.ivation is presumed to be of a different quality, the co-rreL--tion was consideralily lesn (r = -.4'3, @'.S.) 3 b. Dependence of r-,ecovery Limb Time Constant on Response Amplitude Although the time constant of an exponential decay should, if properly measured, be independent of amplitude of the curve, there are many conceivable reasons for suspecting that such may not be the case for the electrodermal response. Accordingl2y for each of 12 subjects, time constants were compared with peak amplitudes (con- ductance change) for 30 sequential responses talen across four experimenw.1 conditions (rest, perceptual task, cognitive task, perceptual-motor task). The Pearson's correlation--, are shown below: Subiect Pearsor.'s r Si-ibiects Pearson's r 2 1 .29 7 -.04 2 .29 8 .14 3 .06 9 -.04 4 43 10 .03 5 -.25 11 2 -.06 6 .15 12 .00 Tne inderendence of these two measures is well demonstrated. Some n@-ga#,-ive I correlations might be expected because demandiag tasks could produce higher responses and also, 2because of the nature of the activation, shorter time constants. For most of these subje--ts, 1-ioviever, such was clearly not the case. c. Denendence of Recoverv Limb Time Cons,-,ant on Ouall@tv of Task A study ivas u@-idertal-,en to answer t%vo cuestions. 1) IDoes the recovc--y limb vary as a funct'lon of the p--rceptual, cognitive, 4 or motor aspects of a task. 2) If it does so, how does its specificity and sensitivity compare vtith those of more conventional methods. A. Method Twenty-six subjects were used in an experiment in which three tasks were performed, alon,,-, with associated rest periods.2 In order to exclude the variance known to be due to the operator's mode of instruction and his relation to the subject, the tasks were programmed on video tape. 1. Cognitive Each of a series of 12 Ravens Matrices Problems was presented for 30 seconds followed by a 15 second rest period. Tne subject was to choose the pattern which properly completed the matrix2 and was to report only after the given problem dir,- appeared from tiie screen. Responses were chosen only from thd period in which he was solving the problem. Obviously, contLmL-i-Lints such as visual perception, frustra- tioa, etc., were present, but it was hoped ttiat the group result as a whole would be heavily wei&,Iited by the cognitive performance factor. 2 2. Percen' ,,lal A fixed pattern remained on the screen for three minutes. At random intervals during this perlod, il@. would disappear, momentarily (40 msec) and for a total of IS times. The subject was to continuously obser,,,e the pattern and after the end of this task, upon command, report the number of disaopearances. All us8eable resoonses d,,lring the observation period were am. l@,zed. 5 3. Reaction Time A series of five reaction time trials was used, cacti having a warning signal arid a 10 second foreperiod. hicasures were taken -at the time of the reaction key press. 4. Other Measures a) Beb-cvior2al. ln v ddition to the ,-bove, two psychological tests were administered to these subjects, the &-irrett Impulsivity Scale and the Spielberger Tr--it Anxiet@,, Scale. A behaviorl,.l test of motor inhibitl-ion task- .@,as a'L--o administered. In this a-irLveatig=r, "-.the subject i.,3 told to turn a knob through 2180 degrees, continuously, but as slowly as he can. Inhibi,6ion is assessed by the total time taken and by the degree to which ,-,-te is accelerated as t@le task progresses. b) Phvsiological. In addition to the skir- conductance measure, skin potential, c@, rdiotachometer recorcling, resdiration (mechanical pr.--u,-ro6-raph2 only) and -frank perspiration were meas=-ed. The last measure was obtained bv a mod:L4icaticn of the photoelectric p--i.-@m method previously described. The device has been improved to eliminate vasomator artifacts and to allow use of a lightweight unit secured to the .inz--r tip. ]During-the-azove-a:uns-;--test-s-were,run2-on-24 of -the subjects toidete--mine t e 4 serisit@v,.' )nal ,,tyo-.I,thq;recov,-ry limb time constant for-detecting sLybtle-c@.ang,-.s-of emoti( test was adopted for th,:-- purpose and c 2e-@- ;-,,,,cs used-undel-@ three@-differe-.it-derr,@-nd-situations-.- The, use of vit-@co tape was again hclp-.@"u'l'ln-"r'ed-u-'ciag, v;az presented.four let:,,ers andasked to 5 - - -------------- 6 choose one and v.,ritc it dov.-n, to be turned in actcr the task- was over. lie was then presented a screen in N,.,hich these four letters appeared in a new order, the group being preceded by three other lettcrs and 2folio%ved by three more. An arrow I ppeared and pointed to the first letter where It paused for 15 seco It then moved a i on to the next in the sequence and again paused for 15 seconds, etc. The subject @mc,,v in advance that this would happen in regular sequence. In the first such test, 2 the subject was to simply observe this sequence without any v--rbal report. In the nd test, the identical sequence seco' La a third series run on 13 of these 24 subjects, the suli.)Jcct was told tli-,t he -,vould earn an extra dollar if he managed2 to successfLlly7 An entirely new series and new .@were used. At the'end of tL-e previous two tests he had been told the operatorts conclusion as to v;h',-r-h'o,,' the letters was This time his se,,-" had been altered2 by his knoviled-e of the operator s previous success and by the added incentive APain he v;as to respond no" to each letter. B. Results to Date Analysis of these data Is still in process, but some early results may be presented. Follow-up progress reports viill be made as calculations are completed. 2 I Data F-aneling The tapes were made primarily so that Lnalysis of these recor&- could be m--de af-,er furtiic;r development of electronic techniques for calculation cf the time const--,nl@-. This development went on as a p--rallel e'kfort durlig completion of the 9 above series of runs and has since been co-@nDleted. The t-,,-i)es have been run off 7 out by usc of first and the e'ach subject N@eim and stand:,rcl deviations of responses for d-.rivativ@-I proccsc.L-I"Y. calcul.,tcd In prei@?.,@-atio',l for t- under each task coiidition are being tests of diff crences betvicen coi,.dltions. To d.,:@te calculations (.,n 12 0; t4c 26 subjccts have been completed. 2. comiriariso,,i BetNt.,ccn Conditionr, No comparigous acro33 subjects cau be m d,,- until all means ard star-&@rd d--vi?-'t'Lon,- have been calct,.L,,ted, but intr,,t-indivi2dual comparisons betvieen t@,sks h-@ve been r:.ade for -Iiose stibiects completed. 'o 11 ',7j2 rossible comparisons, only "d signI.;ica,-cc. L,,, 10 of the 11 su'3j--cts who 3 reached sl&9"-ifir-ance; aopro.,ch sho-,ved reaction time rezpo=,es, the time corstar@l, for this tasi: was the shor@-.e!3t of the -hree tas"-."F-.o conzistonl.- dif@ference appe----rs Letween the 'Cognitive and ,ierccptual rlsks for that part of th,-@ p^,,nula,-aO.-. completed. Using the orrecti-y idertified tl-C con- Lhe op-arator c le-,ter la the first .ex-ies for IS of the 24 sab;ects. @.Ioreover, Nvhe-. tl,.c subject to 2 ibv. mon--ta--y re,,,:ard, th-, operator i,,Ie,-A,,,if ic d it- co=. ectly In 10.of th2 13 cases. 1,1 5instances, the op-c!ra,,or d failed to identify the letter in the first scr-A.--r-, but, iilien monem. rv--revia.-d v,,2as ad@dcd he correctly ider--,ific-c! tLe r-ev@- fo,@: 4 of these i.L- i c ci--, .Tl,-@., comparison of is note%vor-LIay zhat i:i soi-.,Ie tnsmn c--!: in v;ii@cii Cl.)ii"uc,Lamcc aclivi@7 g-IY-c results, tl-,c GP,-.1cal s-weat recordln,@, sho%,,,e%l@ a 8 4. Psvcholozical and Nvchomotor Tests These tests will later be used as a basis for sorting the population into different types of physiological responders. To date, the individual impulsivity ratings on the Barrett Scale have been compared with tile time taken for 180 2 degree knob rotation. The co2@.-clation was unpredictably a positive one for the 26 subjects (r = .0'7. p <.05), i.e., individuals judged more impulsive by the Barrett Scale took longer to rotate the knob, indicating that the type of inhibition measured by the two tasks is different. Tile Barrett Scale was also demonstrated to give ratings wh2ich were independent of the Spielberger Anxiety ratings. Later calculations will examine the interaction of these scale ratings and physiological findings. III. Study of a Posrible Vascular Comnonent in the Skin Potential Resnonse This study, Involving tile effect of arti@riol or venous cuffs upon skin potential, has been completed on a total population of 242 subjects. Results are highly sin, if I- cant and show tha,. the shifl in potential can be more easily demonstrated upon deflatio.n than upon inflation of the cuff, and that the p--la.-it-,r of this r)ost-deflation re-cponse differs according to the type of pressure cuf.L used. For venous (engorging) cuffs (p = 60 min Hg) the r2esponse positive. For arteriol (occlusive) cuffs (p --- 180 mm Hg) the post-de.flation response is negative. The conclusion that these represent vasomotor adjustments is supported by a follow-up pilot study o-.i the local potential responses reported earlie.-. It had been concluded th,--t the slow component of this positive resr)onse to stre'.ch is o2f vascular origin. If so, it should be demonstrable even whea tile epide--mis is penetrated. M-icroelect.-odes were pushed through the epidermis until the negative porenf*ial fall prec-lpitously indicating penetralion of the dermis. After allowin- time for stnbilization. increase of pressure on the 9 electrode causes a potential variation. These experiments must be refined to eliminate the possibility of the "responses" being only a reflection of variations in tip potential or sweat gland activity. If valid they will offer strong support for a vascular component in the sl,-in potential response. I2V. Development of ALtomatic Svstem Comparison of @-,-I&uer:, for recovery limb time constant obtained by template matching with those obtained by the use of first derivatives indicated the inter- changeability of these two methods. The principle involved in the first derivative method is that the absolute steepness of the recovery limb is a measure of its time constant for responses of equal amplitude. Responses of higher amplitude having the same time constant will have a greater absolute steepness to the recovery limb. Hence the absolute slope of the recovery limb can be used as a measure of time constant if a correction for ampl-,ude is made. In practice the simplest correction is to divide the amp'li-tude of 2the peak slope ('Lirst derivative) of the ascending limb by that of the recovery limb. This ratio, plotted against the temolate readin,.-., gives a good line-ar relationship. To accomplish this electronically, the derivatives are obtained by r differentiation and the output stored on capacitors with diode input so that per..k 05 ,va 10 two frequencies. The output is printed out upon command and the system reset. At the same time as the above occurs, the value of the peal-, first derivative of the ascending limb which is proportional to response amplitude is printed out after measurement by a separate counter. Thus the response amplitude and re2covery limb time coastant are measured in a single operition. It is @resumed, ho%,iever, that most facilities will not have the luxury of t%yo parallel voltage-to-frequency converters and counters and, therefore, an analog division circuit is being tested. In this system the two voltages to be divided a-re fed Into a diode division circuit and the quo2tient, as a voll@.-ave.level is fed into an A -D conversion and printing unit. This circuit is very simple, and as an alternative the output (proportional to time coastant) can be written out on an o-!--dinary @,,@-nalog recorder. The system then approaches practical utility.