Citation
Psychophysiological monitoring: possibilities and prospects

Material Information

Title:
Psychophysiological monitoring: possibilities and prospects
Creator:
Donchin, Emanuel
Publisher:
U.S. Congress. Office of Technology Assessment
Publication Date:
Language:
English
Physical Description:
47 pages.

Subjects

Subjects / Keywords:
Supervision of employees -- Technological innovations ( LCSH )
Supervision of employees -- Psychological aspects ( LCSH )
Psychology, Industrial -- United States ( LCSH )
Psychophysiology -- United States ( LCSH )
Genre:
federal government publication ( marcgt )

Notes

General Note:
This report describes the present, near, and long term the evolving capabilities for monitoring and interpreting psychophysiological signals and in particular the Event Related Brain Potential (ERP) with particular emphasis on such ERP components as the P300.

Record Information

Source Institution:
University of North Texas
Holding Location:
University of North Texas
Rights Management:
This item is a work of the U.S. federal government and not subject to copyright pursuant to 17 U.S.C. §105.
Classification:
Y 3.T 22/2:2 El 2/7/v.2/psych. ( sudocs )

Aggregation Information

IUF:
University of Florida
OTA:
Office of Technology Assessment

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

PSYOIOPHYSIOLOGICAL MONITORING: POSSIBILITIES AND PROSPECTS Thia document waa prepared for the OTA aaaeaaaent The Electronic Supervisor: Hew Technolo17, New Tensions, Septeaber 1987. by Eel Donchin. Pfl.D. Department of Psycho logy Un1vers1ty of Illinois Ch111P11gn. IL 61820 Sub111tted to: Charles Wilk, Project Director Congress of the United States Office of Technology Assessment Washington, DC 20510 Thia docunt was prepared by an outaide contractor u an input to an on,oing OTA aant. It doea not necaarily reflect the lytical findings of OTA,_ the Advieory Panel, or the technology Aant Board.

PAGE 2

1.0 INTRODUCTION 1.1 The use of psychophysiology 1n mon1tor1ng 1.1.1 The pr1nc1pal theae of the report 6l2-E1111r1.1el Donch1n Psychophys1olog1cal Mon1tor1"9 This report w111 describe for the present, near, and long tenn (10-20 years) t~e evolving caplb111t1es for 110nitoring and interpr~ting psychophysiological signals and in particular the Event Related Brain Potential (ERP) with particular phasis on such ERP components as the P3OO. Of special interest are the theoretical capabilities and constraints that will dete~ine the utility of this field of science as the foundation for 1110n1tor1ng and inferring conclusions concerning office workers and workers in comparable environments. 1.1.2 Motivation for the report Both technological and theoretical developments have greatly amplified the potential for. and the interest in, the use of psychophysiolog1ca1 signals as a means for monitoring cognitive and affectiv~ processes and responses in huaans. Such mnitor1ng 1s not novel. Inferences fr0111 psychophys1ological signals are a routine matter in daily 11fe. Whenever a companion's blush leads one to t~e one action rather than another a psychophysiological signal is used to monitor a human mind . The blush is an externally observed response of a bodily system and we know that there is a At. ti.I) ea O t t t. L C 4. C. P., I. -L : ,; 0, r 0 I I i I [ ( I. I I 0

PAGE 3

'" -tt:n th cn:tn e&ottcst SN &1is::>tthn1fteorr c ztcsCOt+ ,,, >do1s -., Hsnt' ) . ,,,?ti eci' -.. ft ... o\=> Enuel Donchin Psychophys1ologica1 Monitor1~ relatively stable relationship b~tween a person's state of mind and the appearance and intensity of the blush. The use of the blush 1n this manner illustrates wll the paradigm characterizing all the monitoring discussed in this report. It involves: a. A ans for aak1ng external observations on the activity of a bodily syst. b. A knowledge base that 1nfonllS us about the relationship between the bodily syst and cognitive and affective processes and responses. (This knowledge base 111.Y include statements that describe empirically deterained correlations, e.g. People blush when they are llllbarrassec1. The_ knowledge base 111.Y be more theoretical. e.g. Ellbarrassaent causes system X to lead gland Y to secrete hor110ne Z. This hormone causes vasodilation.) c. There exist rules, derived from the knowledge base, that allow inferences regarding psychological entities from observations of physiological function. 1.2 Factors that increased the power of monitoring The power of this paradigm has been much amplified by the following developaents: The detection of the activity of bodily systems depends on the ava11ab111ty of systems that can be used to detect and record the signals. & .. $ $~. A.QI, . l. 41 4 .I .I I 4 ... ,, .. q # a ( . r $ .. 4 )!AW>,;

PAGE 4

"3\LI ,-,el Donchin Psychophysiological Monitoring This requires: a. Sensitive transducers for transfor1111ng the chical or the physical aan1festat1ons of the activity of bodily organs into electrical signals that can be recorded and analyzed. b. A highly reliable signal cond1tion1ng and. recording systa whose noise level is lowr than the necessary resolution of the signals. c. Data acqu1s1t1on and processing system that can analyze 1n real t1 large sses of data. d. A developaent of psychological theory that could supply the concepts and tools necessary to provide the psychophys1olog1st with a source.of constructs and 110dels. e. Reliable observations on the empirical relationships between psychological and physiological observations. f. A developaent of the theory, and the enpirical observation, of the biological systs under observation. It tums out that all of these conditions have been aet by technological advances in the last couple of decades. These have increased the sensitivity of transducers and aide it possible to record a 1111ch broader range of signals. This in turn has led to a trendous expansion 1n the repertoire of uures available to the Psychophysiologist. The explosive development of Cognitive Psychology and of Neuroscience provided extended and very useful conceptual foundations for the field as well as a large body of data that can be integrated with the Psychophys1olog1cal data. Thus, the range of phenomena that can be 110nitored and the strength of the inferences they permit is large and increasing. These capabi11ties S C1U ,C 55 S I a P .. A J. I t CJ i. Cf a P 5. 2 ).. I I 4 0 -..... u. 4

PAGE 5

I .......... ,,, Emar1.1el Donchin Psychophysiological Monitoring focused the attention of developers on the possible applications of these 110n1toring technologies in a wide var1e~y of da111ins. Applications range from the traditional use of Psychophysiology in clinical diagnosis through needs for aonitor1ng that arise in various caaerc1al and govemaental da111ins. The 110st extensive application of ERPs hu beende 1n aed1cal diagnosis in particular in the usess111nt of sensory function (Halliday, 1982). A procedure that hu assuaed the status of a standard d1~1 test utilizes ERPs with very short latencies known as the Brain St Evoked Potentials. These ERPs appHr within 3-5 ailliseconds of the occurrence of an auditory stiaulus and they last for about 5 asec. It tums out that the appearance of these waves 1n a sequence and with tia1ng that deviates fr011 the noraal is an excellent indication that the brain st hU been SOlllhoW d1Mged (Bod1s-Wolner, 1982). Medical applications are being developed also in 111>re cognitive domains. These, however, are 110t yet in widespread use. It has been reported, for exaaple, that a particular brain response (the P300) 1s 111rkedly slowed in people sufferng Senile Dellentia (Goodin, Squires and Starr, 1978). This prolongation Q serve to distinguish patients suffering from depression from those who are senile. The literature on this application is of many opinions, {for a review see Donchin, Miller and Farwell, in press). However, it 1s likely that several such app11cat1ons w111 eaerge in the next decade (a review of the application$ of ERPs 1n Psychiatric research is available in Pritchard, 1986). The aain eleaents of these developments will be: P . C 4 id a. An expanding understanding of the neuroanat011ical origin of the c011ponents is 1 ikely to erge fran studies of ERPs using indwelling I U .I A .. S ( .. I #-4. .a. 4 a a . I $ ;;

PAGE 6

m rt .... 't : c t -crt n w ~ t ten e .... --:h .. at db" :1 ab: ts1 e tt,; vwei G:Nritt: ...... rt l .-c_~ ~Lb E11a,-,el Donch1n Psychophysiological Monitoring electrodes in human patients (Allison, Wood, and McCarthy, 1986); froa studies of ERPs 1n aniMls (Vaughin, 1982) and from studies of the neuro111gnetic analog of the ERP (Okada, 1983). b. An expanding understanding of the functional significance of the potentials is develapi11g u studies of the ERP are conducted with inc~u1ng attention to theories and aodels of cognitive psychology. (Hillyard and Kutu, 1983; Ga111ard and Ritter, 1983; Donchin, 1984) c. The decreuing costs of' caaputing power and the increasing saphist1cation of graphic displays is making it possible for 1nvest1gators to utilize .,ch larger data bases than were possible in the 1-diate past. -bus, the study of multiple electrodes and the analysis and display of two, and even three dimensional, data patterns greatly increase the power of the analyses (see for example Morstyn, Duffy and Mccarley, 1983). Of course, the above exples derive from only one of the many psychophysiological techniques currently under intensive developments. The application of blood flow uureaents, Magnetic Resonance Imagery and siailar approaches allow a detenaination in vivo of patterns of brain activity with an increuing spatial and temporal resolution. All of these techniques share the property that they provide 1nfonaation on the degree to which a subset of brain regions is active at any given tiae. The activation is sured ~Y assuring, using chelllical labeling techniques, that brain activity would be Mnifested through the skull because the chemical, or the biophysical. label 1s detectable by extra cranial probes. These data are often analyzed by the thods of t011ography to provide three dimensional 0 0 -14 U .US & 5 i.. S. . t S ~I T 5 C I I L i I 5 a I Q. ). !IQ 14: ~-CIC ... i, .4.. JQl SP

PAGE 7

b .. I I t rs\i ?is) +,,,, ditrtceit rs ):idi:rhn:'f 1 r ;: .. ,,-, ..,aeab lWl'dr:!11ntriilf'or ,;,,,., ..__.. .. "61, E111l'l.lel Donchin Psychophysiolog1cal Monitoring 1aages of considerable resolution. While currently available techniques have a telllporal resolution uured 1n seconds 1t 1s likely that over the next decade the resolution w111 be 1aproved.su~stant1a11y. As w111 be seen below, the range of app11cat1ons extends beyond IN!d1cal diagnosis and mn1tor1ng. The app11cat1ons 1n the area of Engineeri_ng Psychology, for exaple, are beg1nn1ng to mve froa the laboratory to field testing. Thus, AFAMRL awarded 1n the early 801es a substantial contract to the General Dynu11cs Corporation to test the feas1b111t) of a Psycho~hys1olog1cal Battery developed at the Cognitive Psychophysiolog1cal Laboratory at the University of I111no1s for the sureaent of Mental Workload (See Heffley et al 1985; Otto et al, 1985; Sopher and Danch1n, 1n press). The technological developaents greatly reduced the cost and siaplif1ed the operation of the psychophysiological recording equ1paent. With the wide availab11ity of such devices 1t is easy to see how they may get into hands less responsible than those found in the laboratory. The potential for abuse is substantial. 1.2.1 Plan of the report This report will provide an overview of currently available Psychophys1olog1cal aon1toring techniques. The general principles underlying these procedures w111 be sketched and illustrated, largely by reference to Interrogative Polygraphy, C011110nly known as Lie Detection, an unfortunate misn0111er. We shall then focus on one class of Psychophysiolog1cal signals: the Event Related Brain Potentials (ERP). These will be introduced and their .. 4 \I S I 4 t .j 40. I J 4 I s . :: a s x a : o a .4

PAGE 8

s ., ... ~-a z>-a Ni t' re,' f ,. t <,,'Mt: = k ... :r ...... ,r&oar+fo' X er e ; ... ,e, tr e1'> .. ... :. -:.,p a ; .. 5I~ EnJel Donchin Psychophysiolog1ca1 Monitoring scope and range w111 be enumerated. The utility of one such component 1n aonitoring studies w111 be described and future trends w111 be assessed. A d1scuss1on of the potential f~r coercia11zat1on. that ~uld lack sc1ent1f1c basis. will be considered. 2.0 AN ILLUSTRATION OF PSYCHOPHYSIOLOGICAL tl>NITORIN&: LIE DETECTION. 2.0.1 An overview of traditional polygraphY It 1s useful to dwell briefly on one of the 110st coaon d01111ns in which Psychophys1ological signals are used for aon1toring. This is the . dolllain of Lie Detection. Or, as we shall call it here. I~terrogative Polygraphy (IP). All current approaches to IP are based on the assuapt1on . that a deceptive action on the part of the subject will be accmpanied by an affective (i.e emt1ona1) reaction. The tera affective rather than emtiona1 will be used here because the tera eaotion 1s often used to include the perceived C0111POnent of the reaction. However, subjects aay react affectively to a situation without necessarily being aware of the fact. In other wrcts. it 1s possible to react affectively without I concoa1tant emotional reaction. This reaction. 1n turn. is associated w1th activity 1n the Autonoaic Nenous Syst (ANS. an activity that 1s manifested by an ensemble of actions generated within a variety of effectors driven by the ANS). SOiie of these actions can be detected by 110nitoring several organs whose activity is mdulated by the ANS. The aost c01110nly used ANS driven responses used 1n IP are the acceleration, or deceleration, of the heart 0 () u .. .. 44. ;_ I 4 CJ.! a .& .. I Z Z. 0 PI 14 D P

PAGE 9

t t:i' . '; tzf a te . 'atii' 0btnOtt.i,;t:;a,f 'b r,'-,,; )r4 ......... .... ..,. -ri:i, 'l'i' di'! ... Q' ',i .,., ._ f " ........ ._,.,, 9c: '. fi:Hb:tr ~-,.-,; tCzC& H Er1.1el Donch1n Psychophysiolog1cal Monitoring rate; changes 1n the depth and the rate of respiration and the activity of sweat glands as nifested by the &alvan1c Skin Response. The Lie Detector 1s a polygraph that 1s used to 110nitor these three responses. The Interrogator presents the subject with a sequence of events (words. questions, sentences, pictures and such like) and the degree to which an event e 11 cits a response in one. or a 11, of the 110nitored responses 1 s recorded. The pattern of ANS act1v1ty 1s then used to determine which, 1f any, of the stiau11 e11c1ted a supra threshold affective response. Different investigators tend to use different thresholds 1n detecting a response. There are also substantial differences 1n the thod for developing and pre~ent1ng the st1au11. Hawver, 1n the n we can ignore these details and view the standard, ANS based, IP as an att111Pt to determine 1f the critical 1ts do, or do not, elicit an affective response fr011 the subject. In the 110re c01110n technique the affective response to the items regarding which deception 1~ under 1nvestigation are cmpared to the response to control its. The control 1tellS are questions that are not relevant to the subject of the 1nvest1gat1on but which are likely to e11c1t an affective reaction. A question like have you ever aasturbatec1? Q serve as such a control question as 1t 1s assud to elicit a defensive reaction in all subjects. 2.0.2 The character of Psychophys1olog1ca1 Monitoring The nature of IP illustrates well so of the critical aspects of Polygraphy which aust be kept in aind in assessing all forms of 110nitoring discussed 1n this report. The most important point 1s that the instrument ,. I

PAGE 10

3GO En.ae 1 Donch in Psychophysiological Monitoring used by the Polygrapher does not provide for the direct observation of a psychological process. Rather, the 1nstnaent asures a variety of organisaic variables. The relationship to psychological variables 1s 1nferent1al and its va11d1ty 1s not assured by the re11ab111ty and va11d1ty with which the physiological functions are aeasured. The technical propriety of the -sureant is a necessary but not a sufficient condition for the psychological validity of the uures. A critical eleaent in the interpretation of these asurents in the psychological doin is the situation 1n which the measurements are Mde. Any stimulus presented to the subjects is interpreted by that subject within the fraework of the perceived situation and the ensellble af 111110ries, expectat1~ns, and attitudes that are brought into the asureaent environ1ent. These psychological factors are 1nextr1cable froa the technological aspects of the aeasurent. It is important to understand, in evaluating the technology, that its success does not depend solely on technological factors. The effectiveness with which the test environment is set up and the subject's partic1pat1on 1n the creation of the psychological context w111 be critical (See ~kken, 1981). 2.1 D1fficult1es Associated with ANS-Based IP In this section we review a number of the probl111s presented by the application of psychophysiolog1ca1 measurements in IP as an illustration of psychophys1olog1cal 110nitor1ng. 0 .,,-... ;' \ V

PAGE 11

,.,,,.~,-.,, ttxee el ~-'. -2.1.1 Introduction 3~l EMr11el Donchin Psychophysiological Monitoring While the use of affective responses in the detection of deception has ny advocates and the technique is widely eaployed by government agencies and by private practitioner$. it suffers a nullber of inherent difficulties. In any detection proceclun the uti11ty depends in part on the ratio of False Negatives to False Positives yielded by the ,thnique and 1n the behavior of this ratio as a function of changes in the decision criterion. I will ignore. in this discussion. the difficulties encountered in validating IP that derive froa the paucity of independently verifiable criteria for the deceptiveness or verifilbility of the subjects. This aatter has been discussed in great detail by protagonists of the Polygraphy debate such as Lykken and Rukin; for an accurate and accessible review of the issues see EICllan (1985). I w111 also ignore, for the 11>nt, the sensitiv1ty of the test to the arousal level of the individual. Let me just eniaerate briefly s01111 of the featuns inherent in the use of affective responses as indicators of a specific eaotion and as specific indicators of deception. 2.1.2 The Non-Specific Nature of ANS Activity Perhaps most serious is the relatively non-specific nature of affective responses. This non-specificity is double-barreled. It operates at the level of defining the direction and cognitive content of the emotion and it operates at the level of n1festation of the affect via ANS activity.

PAGE 12

32.,.z_ En,el Donch1n Psychophysiological Monitoring The difficulty in inferring 1110tions froa their ANS correlates 1s well known. As has been reported in IIIIIIY studies of IP, False Pos1t1ves are often the consequence of an affective response to a question, or to the presentation of an itell. for reasons that have little or nothing to do with deception. Conversely. False Negatives often result fna the absence, or the low 111Plitucle1 of an ANS 111111festation of an affect. 2.1.3 The Non-Affective Aspects of ANS Activity It 1s also the cue that only a portion of the variance 1n the ANS activity is driven by affective processes. The ANS is charged with the regulation of soae of the mst vital of the body's functions and changes 1n heart-rate. blood pressure, respiration or sweat gland activity reflect a variety of non-affective deaands by the syst for increased supply of blood, of variations in the need for oxygen and for IIOdulations of body temperature. Fr011 the point of view of IP the ANS is quite a noisy syst. The signal, that is, the ANS activity driven by affect, 111.Y be swuped or 111sked by the noise, the ANS activity that is related to the vegetative and.energetic functions of the systell. . ') The factors e111r1ted above are, in part, responsible for the ab111ty of individuals to produce, or to suppress, .ANS activity u their circuastances require. The suscept1b11ity of the ANS for control using biofeedback 111kes it even aore vulnerable to voluntary control by the subject, especially for appropriately trained 1nd1v1duals. Control techniques may range from the voluntary generation of affect by cognitive \., \ u

PAGE 13

-; tee n tt , a> r u e s .. to ,, ?tinsttt'dn re wtea t+crc:nttttt '1,; tJto ,, '! r c .,.,. ,e EMN!el Donchin Psychophysiological Monitoring control to the activation of bodily parts so as to create energy needs that w111 be reflected 1n ANS responses at critical points in the interrogatory sequence. 2.1.4 The utility of psychoph.Ys1olog1ca1 aon1tor1ng The discussion of the caaplex of probls that aust be solved 1n order to 1ap1-nt an effective psychophys1olog1cal aon1tor1ng system 1s not intended to 1aply that such aonitoring 1s without value. Quite the contrary is the cue. There 1s a large, and 1ncreas1ng, arrQ of thods that perm1t effective aon1tor1ng 1n a large n,lll)er of dalllins. My intent is to eaphas1ze that the application of psychophysiology to mn1tor1ng of hu111ns will require auch care. One of the dangers presented by psychophys1ological uasures is that they a" relatively easy to aploy 1n a shoddy manner. The equ1pnt 1s generally siaple and euy to attach and 1ts output has an air of veris1ai11tude that can be quite deceptive. The r1111bers1 and the curves, project an authority to which they have a right only 1f mployed by a psychologically sophisticated investigator. Unless the setting 1s right and the mn1tor1ng 1s conducted within very strict boundaries the output of the uur1ng 1nstruaent uy be data, but these data certainly do not provide 1nforaat1on. I will return to these rarks when I co to discuss the coaerc1a11zat1on of psychophys1ological su"nts. With these caveats 1n a1nd we can exu1ne same of the monitoring procedures that have erged over the last several years as powerful _.... ___ ._... ......... 11'9"!1111 ... -------~-.... ------------.... --.... ~...--........................... ...... "!l'l'TI -.

PAGE 14

nnari:c &wt'Htu1ftelt+t1ttitx++ t: ic )N; _, ., b tt ect 8 r I .. t,."tribs:ht tttfir ..,.,,, .. tt?!tn:Htt::'J#-)ftf tt& 11 iit .,, ~ &1tbt~ttb c;':rt;;&iiotM "+w:',, .: 3 2.'-{ Emanuel Uonch1n Psychophysiological Monitoring psychophysiological tools. I will begin with an enumeraticn of various techniques spanning the range of available monitoring procedures. I will then focus on the ERP and eraarate the variety of monitoring options provided by this particular class of aeasures. This will lead to a focus. as an 111ustrat1on. on one of the coaponents of the ERP, the PlOO. 2.2 Varieties of Psychophysiolog1cal asurements The repertoire of bodily signals that are available for Psychophys1olog1ca1 asures can be classified according to the natu~e of the signal ~e1ng aon1tored. A classification will be presented in this section. 2.2.1 Motions In this class belong measures that depend on the recording of 1111scular aovents. We exclude, of course. the general class of 111.1scular responses that are considered overt responses (whether voluntary or involuntary). the functions 110nitored reflect largely unconscious. involuntary .,veaents. though in most cases people can gain control over the asures. For exaple1 respiration is nonaally a automatic process. However, people can inflate or deflate their air cavity. within certain limits, if they so wish. a. Respiration Measures of the changing volu of the lungs as air is inspired or expired are obtained with relative ease by encircling the chest with a 0 \ V ------..................................... ___ .... ____ ..,... .................. -....... ~~R'.I'. ..... .,., ...... ..:-,,,,~~-~-I

PAGE 15

,.. ,1 eb %71 "'@ 1 rb ,t~ ntti,. a; rt#t1:riii1"ker:i:::", rt 3<.5 Emanuel Donchin Psychophysiolog1cal Monitoring pne11111tic tube and monitoring changes in its state as the chest inflates or deflates. Affective reactions often involve changes in the period of the respiratory cycle and 1n the plitude of the cycle. These reactions are often amng those to which t~e Interrogative Polygrapher pays auch attenti~. A detailed review is provided by Kaufaan and Schneiderman (1986.) Their-s..-ariz1ng conclusion 1s worth quoting in full as it points to future developments in the use of this index that move it from the category of mechanical indices to that of the chemical 1nd1ces. Although a knowledge of pul110nary voluaes and capacities during behavioral exper1nts can tell us a lot about relationships bet~en breathing activities and behavior, knowledge about the consumption of 02 and the excretion of CO2 provides a more direct picture of the metabolic changes that occur during behavioral situations. Use of an ear oximeter, the Douglas bag, a servo controlled 02 monitor or a blood gas analyzer can provide important 1nfon11tlon about the tabo11c changes occurring during psychophysiological exper1aents. (p 120) b. Eye movements The pattem of eye mvents can be readily measured by placing electrodes above and below the eye (i.e. on the forehead and ianediately under the lower eyelid). There has been many attempts to use the direction of the gaze as a critical dependent variable in psychological experiments. For exple1 McConk1e and his coworkers have made extensive use of an eye tracker in the study of reading. Stark (1983) has extended his pioneering work on the control of eye movements to the study of f. r I. ,. t' I I t ---4l~.Q~4&$-~P~.--..-.. . 1., .... ._. .... _,._~4-.llJ--.-~il~I...,__._,.. .... ._.,_. ..... ..,_, ~-~l 4~4...,._..Pl~.t~41.P~M~M~h~~.f~#-4~K~i~~

PAGE 16

3 Zt:; Enuel Donch1n Psychophys1ologica1 Monitoring psychopathology. Extensive work on eye movents and on blinks as dependent variables was done by John Stern and his colleagues. For a detailed review of. this and related approaches to monitoring focusing on the facial 111scul1ture see Schwartz (1986). c. Facial expressions Paul Ekn has deve~oped a COllll)rehensive thod for coding facial expressions in a nner that does not prejudge during the coding process the emt1o~ that underlies the expression. Thus, rather than code expressions as frowning, joyousN or threatening, the EklDan sche describes the facial expression as a pattern of movement in the facial musculature. With this tool in hand Eklllan and his colleagues studied the relation between facial expression and affect and provided a 1110st effective set of tools. Theirs is probably the single most innovative approach to the study of the psychophysiology of affect in the latter half of this century. For a technical treatment of the subject see Ekman and Friesen (1984). For an excellent description of the application of this approach to Lie Detection see Ekman (1985). d. Pupillary dilations As Beatty (1986) notes 1n a detailed review of the psychophys1olog1cal value of the pup111ary system the primary use of pup111o~ry in psychophysiology is in the study of brain activation and arousal, (p. 43). Pupillary di1111eter can be monitored relatively easily, though the analysis of the dta may be rather laborious. Simply stated, mental effort appears to cause the pupil to dilate. The study of the pupil has a long and venerable history (Hess, 1975; Khaneman, 1973) and modern 0 : -~""9.all!IIIIIIP!l~ .... """-IIIJll!lll .... lllf!ll!l!l!~,... ...................... """'""" ........ ............. ..,,...(IIMN,......,.,.,, ................. ~."""'""''~' .... ,.__;~' QtS !CA Qi J.J .... ,es 1411 I U 4 L C .... A. t O $ i 41 .; PS us i 4 .. 4 4 ) 4 S : .. P t; 4 4 k

PAGE 17

..__ .... Eanuel Donchin Psychophysiolog1ca1 Monitoring pupilloters allow observation of the pup111ary diameter regardless of head 110tion and direction of gaze. Albeit that such unrestricted aon1toring requires at this tiae fairly expensive equipnt (by the standards of psychological research where an 1~strunt costing s10,000.00 1s considered pricey). e. PlethysaographY Theplethysa:,graph aasures cardiovascular activity by monitoring the volume of the extremities, usually the finger or the earlobe. It 1s one of ny psychophys1olog1cal asures of the heart's activity and does not require a special discussion here except as an example of the variety of chan1cal aeans that can be used for monitoring various functions. 2.2.2 Voltages In this class we include measures 1n which the recording instrunt 1s sensitive to voltages generated within the organism as a consequence of the act1v1ty of the bodily organ. a. Electroencephalogram -(EEG) Voltage differences between two points on the scalp that are due to ionic flows within the brain and the fields generated as a consequence of the suat1on of large population of 1ndividua1 neurons, (see Allison, Wood and McCarthy, 1986; a very extensive analysis of the physical foundations of EEG analysis can be found in Nunez, 1981) .................. WJ'91#"""' .... iflllllli4.,._P .... ...... 111111!1.-i&_,.....,,.>A-4 ..... -w.-..1.919'4-.,,.1'91111 ...... 91C(~. ~."""'l ...... ...,_.11111\111,il!IIIIS ...... %~.'91111$:.'( .... 1111!1'15,.,.I .....,, .... s ,...,,...;,..; ~)111111( .... #11!!1'1 ....... 111911!, .... ?~I".""!. "41114ipl!I . Z--~ I' I

PAGE 18

"32~ E111,.,el Donch1n Psychophysiolog1cal Mon1tor1ng ,b. Event Related Brain Potentials -(ERP) A subset of the EEG. The ERP represent that camponent of the electroencephalographic activity that is ti-locked to external or internal events. It 1s extracted by caaputer analysis from conventional EEG. This clus of aeasure 1s discussed in detail in section 3 of this report. c. Electrocardiogr111 -(EKG) Voltages generated 1n the course of action of the heart muscles. The signal is recorded by means of the standard cardiograph1c procedures can be analyzed for a number of parameters ranging from the global analysis of the Heart Rate to the mira,te examination of the wave shape of the EKG which is widely used in clinical d1cine. Even though classical psychophysiology 1s essentially card1ovasa,1ar psychophysiology it wi 11 not be discussed in detail in this report as aach material is readily available. The chapter by Larsen, Schneiderman and Pasin (1986) provides a detailed review. d. Galvanic Skin Response -(GSR) The c01111110n asuraent associated with the GSR is actually a asure of changes of resistance rather than of voltage. However, these changes of sweat gland activity that have played an important part in the history of Psychophysiology has a voltage counterpart, discovered by Fere. For a review see Fowles (1986). U 4 P P4 ., S <.. Ct .. c ,.. '* ; '. Ct. . ( . ,. 5 Q

PAGE 19

tt z:or rn n dt:tn ot:cv,etttr+tri:Vnrt1c .... : t: .. ri-++ : r + h1to ~ctktif:rit"'VW:tsde r ~+,ctr :*. .......... 2.2.3 Other asures Emanuel Donchin Psychophys1olog1cal Monitoring In this section I w111 eN.1rate a number of additional asures that cannot be easily categor1ad. a. Metabolic changes This is probably the mst exotic, and the most rapidly developing, 110de of mn1tor1ng. In general the technique. labels chcals that are utilized 1n a brain process with a rad1oa~ive label and a computer aided aonitor1ng syst tracks the behavior of the coapound over an epoch. It 1s possible to reconstruct fr0111 these data the progress of action 1n different brain structures. When subjects are engaged in various psychological activities 1t 1s possible to determine by these techniques which different brain structures are involved in a specific activity and to what extent. It 1s not clear at this t1ae with what resolution would these techniques serve as 110n1tors of hu111n 1nfonaat1on processing. At this ti they all depend for their success on the 1ntegrat1on of surements taken over periods measures by seconds, or by minutes. Thus they seem more suitable for aon1tor1ng state variables rather than specific responses. However, the developnt 1n these approaches has been, and is, explosive and therefore predictions should be made with great care. b. Magnetic asurents The electrical activity of the brain has, of course, a magnetic counterpart. The very s11111 aap11tude of the magnetic act1v1ty delayed the developnt of the sure. However, 1n the last decade a a.1 a c.a .. :, s : a s. ,.c a a c .: C .I Q PA .z. t .BU .. a:. t I I 4 I 4$ $ i t I

PAGE 20

330 Ema,.,.el Donchin Psychophysiological Monitoring technology based on the use of superconducting probes has developed and 1t allows the recording of mgnetic brain activity both as continual signal analogous to the EEG, and as an evoked responses analogous to the ERP. At present the gnetic recording requires rather expensive and cUllbers0111 equ1paent. Large Squids 1111st be used and a supply of liquid he11ua 1s required. At the mst sophisticated it 1s possible to record from 14 siaultaneous probes. However, a system tha~ supports such a recording montage sells, in 1986, for around one aillion dollars. A n111ber of research programs attpt to develop smaller detectors. Should these beco available the obvious advantages of the magnetic recordings w111 greatly accelerate their use (See Beatty, et. al. 1986). The 110st important aspect of the Magnetic Evoked Field is that it is possible to infer, with s01111 precision, the locus of the intracranial generating source of a magnetic field from the scalp distribution of the field. That is, a aap of the distribution of the strength of the aagnet1c field at different scalp locations provides the data necessary to identify the location of the field within the skull. It is not possible to perfora the equivalent analysis using ERPs because the electrical potentials are volume conducted in the skull and the same scalp distribution can be generated by an 1nf1n1ty of sources. c. Temperature changes It is possible to sense body temperature extemally. Several investigators have proposed that such changes may be useful in 0 '----'' a _, ; 1t 4 .. a ts a t. __ a a ; .! I .S J S 2 5 t .! .J .. C .! I .Qi t .i# >4i. %ii .J

PAGE 21

' .__,, 3.s\ En.ael Donchin Psychophysiological Monitoring Psychophys1ologica1110n1toring. However, these techniques have not been developed to any serious extent (see Schwartz, 1986) 2.3 The Psychophys1olog1cal Signal Criteria for Monitoring As is true of all technologies, their proper application requires careful attention to details. The purchase of a polygraph and the availab111ty of a subject to wholl the polygraph can be attached (or whose brain can be seamed by a CAT scanner) guarantees that the 1nvest1gator w111 have curves, or pictures, to show. However, these w111 be of no value if they -.re obtained without proper attention to thodolog1cal, and logical, aspect.s of the recording. In evaluating mn1tor1ng techniques it 1s important to assure that the following criteria have been properly considered. a. Proper thodology The utility of the surements depends, almost by definition, on the technical quality of the effort. It is important, however, to realize that technical considerations are not exclusiely related to the instruments in use. Of course, one .. st assure that the equipnt 1s properly 111.intained and adequately calibrated. The bandwidth of the . recording must be broad enough to include the signal of interest. To the extent that digjtal analysis is involved such aatters as the digitizing rate and resolution are crit1ca1. However, attention aust also be paid to the biological and the psychological adequacy of the effort. By way of example of the biological cautions that 111Jst be observed we note that the body is a i,?61 ... JQ ,4 (.1 I.( (14.04.@ 1 C SJ.! .. !_ l . $ t 4 Q I 4 .b .I .. A .. ,. C L t . I 4 D I OJI#% _ij $ 4

PAGE 22

. L 332.. E111n,el Donchin Psychophys1olog1cal Monitoring source of aany signals and that sme signals Y appear to the recording 1nstnaent in the shape and fora of the signal of interest. In other words, the body itself aay be a llljor source of aeasureaent artifacts. Ttus, aovaents of the eye bill that ay 1n one context be the signal of interest have a t1 course and a shape that kes if very easy to take oculograph1c signals u 1f they are ERPs. It 1s the investigator's respons1b111ty, whenever ERPs are the target of 11easureaent, to ascertain that the data presented are not confounded by oculographic nose. Several techniques are available (see Gratton, Coles and Donchin, 1983) that sarve this purpose and the extent to which these are used deteraines the value of the ERPs as a aonitor1ng tool. There are of course 1m111111rable such cautions, ny for each of the techniques enumerated. The relevant conclusion for this discussion is that the application of Psychophysiolog1cal monitoring 1s not a matter for amateurs. It ~snot possible now, and will probably not be possible in the future, for anyone to purchase off the shelf instruments and to use these for accurate and valid mn1toring. Extensive training in the finest details of the eng1neer1ng, biological and psychological aspects of the recordings 1s a necessary condition for such applications. b. Construct validity At~eapts to use psychophys1ological signals for the 110n1toring of psychological processes require that the psychological constructs used 1n the analysis be pr~perly validated. That is, if one proposes that, say, blushing is correlated with embarrassment, one is relating a physiological response that can be sured, given the precautions .(. t 4 ' A IC 4 I J W P M 4 J I 0 C)

PAGE 23

l. a,ri-,, Cl re: :, .... ,r ..... th' % L"Mri'i ztt'6ids miW#"rtL:~-t;,e,.itS ,;-;,,.-, ,r rr-e:e,"'dit ,-. btast't:Se ;;;n;,,'a'-1:fc:k6'a0 ictti:t>'" .:,... .:' ,_/ Eman,el Donchin Psychophys1ological Monitoring aentioned above, to_ the psychological construct "barrassment.11 Evidently such a relationship can be of interest only if it has been Mele very clear how >arassaent has been defined and asured and that ttte_psychophys1olog1ca1 usertions undelying the assumption that aeasure X aon1tors construct Y are indeed an1ngful. c. Ulllllbiguous psychophYsiolog1cal relat1onsh1ps The proper v~11dat1on of the constructs is a necessary but not a suff1c1ent cond1t1on for the success of aon1tor1ng. It 1s usually far fraa easy to establish the distinctness of such constructs. Is, for exuple. emarusement d1st1nct froa arousal? How does one tell when subjects are aroused without being lllblrrassed and when they are elllbarrassed without being aroused? These, of course. are very difficult questions that serve as the substance for much of psychological theory. However. they evidently lie at the heart of much of the controversy surrounding psychophys1ologica1 applications. The aost widely known example being the disputes regarding the use of polygraphy 1n Lie Detection where ~he activation of ANS aeasures may be attributed to the operation of several psychological chan1sms only one of which. presubly, indicating that the subject has been deceptive. d. Cost of system should not be larger than the benefit of the infon11t1on On this point not much need be said. In virtually all cases psychophys1olog1ca1110n1tor1ng 1s expensive. Equ11J1111nt is a necessary caaponent. The 110re elaborate the technique the larger the data base that 1s generated Jnd all the data need be stored and analyzed. There I I ..... 'llllllllllll.-.i1!911!1111 .... -ll!ll!lll"""" .... !lllllll ................. (ll!lllllllillll ...................... !1!1119....,!19111111!1!1111111111111 .. ll!lf-3llllfllllli ll!IJ.t IJIIJJ ........... U,.....l.-._!11111'!11.441111f14 9!1'13 1!1111. @. -.a. 11111!.WU__.l A. IS S4. 4 I J I . J J J. .I I .. 1 Q I .

PAGE 24

EMr1.1el Donch1n Psychophys1olog1ca1 Monitoring aust be full access to normative data bases. Furthermore, in many of these techniques the subjects must come to a recording site where the equipaent resides. Thus, the benefit of the gain in information derived froa the use of the equipaent aust be carefully weighed against the cost of obtaining the info111&tion. No general rules can be developed as the issue need be resolved 1n each setting. (Of course, as technology develops and as the de111nd for equipment increases, creating a-substantial market, the size and the cost of the equipment 111y decline. Consider. for exUll')le, the wide availability in the mass 1111rket of tools for the asurent of blood pressure and heart rate as these have beCOIIII a aatter of personal use in the health mnitor1ng and exercise domain, a,ltiple units whose fields are so aligned that they s&a11.te to produce recordable potentials.) With these criteria in mind we proceed to exuine in some detail one psychophys1olog1ca1 index, the Event Related Brain Potential (ERP). We will describe the measure, discuss briefly the recording procedures and 11st so of 1ts c011ponents, each of which can be used in mnitoring. For reasons that have largely to do with the personal interests of the author we will examine 1n much detail one of these components, the P300. 0 --... 0. l -----------------------------~---~...-.------

PAGE 25

.,.;.-e-tie t't X Wt'#HXS 7ft .,-..., ....._/ 4 sss Ellll'llel Donchin Psychophysiolog1cal Monitoring 3.0 The Event Related Brain Potential 3.0.1 An overview of the technique The data we discuss here are obtained by placing electrodes on a person's head and recording electroencephalographic (EEG) activity ~h11e the person is engaged in a tuk. Using the technique of signal averaging, we then extract froa the EEG (a voltage x t1ae function) est1Mtes of the portion of the voltage (the ERP) that 1s t1ae-locked to soae event of interest. We usu that these ERPs represent the synchronized activity of enselllbles of neurons. A distinction is usually Mde betwen exogenous c011ponents, which occur early (within the first 100 11Sec) and endogenous cmponents, which occur later (Donchin, Ritter, & Mctan .. 1978). Exogenous components reflect early neural processing of the features of the sti111111 presented. They are obligatory responses to stimuli whose amplitude and latency are responsive to changes 1n the physical characteristics of the stilus, and whose scalp distribution 1s determined by the sensory syst activated. In exper1nts designed to investigate these early components, subjects typically lie passively while stimuli are presented, often at very high rates (10 t1s/sec 1s CCIIIIIDn 1n auditory parad1gas). In contrast, cognitive psychophys1olog1cal paradigms often challenge the subject with caaplex. tasks, and the subject always has sme task to perform. Several attributes of the task are aanipulated and serve as the independent variables of the study. Syste111tic changes 1n brain activity measured at the PJW. (. Q .. 14 .. J ; 4 I .: .Q .. # iJ . I I 5 I U. J

PAGE 26

c s r r n a .. o ire dt .. 1 a, .. rt t,e: 1 -.. tatr en-rt rFn' n ,-sis mis'st' rrri ii &iS:ify> ti'i ..-: ?>3-6 E11ar-,el Donchin Psychophys1ological Monitoring scalp are considered representative of the engagent of different neural processes. and serve as the dependent variables. The complex brain response is analyzed for patterns of activity that are characterized by a distinct scalp distribut1on and by a consistent relation to task variables. These are defined as the endogenous components of the ERP. Such ccaponents are d1stinguished by the fact that they are not obligatory _responses to physical st1au11; whether they w111 or will not be elicited depends on the nature of the infonu.tion processing required of the subject. The endogenous components that follow a warning st11111lus and precede an iaperative stiaulus, for ex111Ple, are believed to be aanifestat1ons of 1ntracran1al processes that are involved in the subject's processing of the wam1ng st1aulus and preparation for a forthcoming stiaalus. Endogenous coaponents that follow a task-relevant stiaulus depend more on the processing demands imposed by the task than on the physical nature of the eliciting stimulus (Donchin, 1979, 1981). In contrast to the exogenous components, they can also occur long after the st1aulus (often between 500 and 1000 mec). Although the endogenous ccaponents of the ERP can be recorded with reliab111ty, thesaurce of these c011ponents is not known. Evidence has begun to accuaulate that suggests that saae of these components 1IIIY not be generated 1n the cortex (Halgren, Squires, Wilson, Rohrbaugh, Babb, & Crandall, 1980; Wood, Allison, Goff, W1111son, & Spencer, 1980). There 1s a hint from this work that we may be observing neural activity generated, in part, fr011 arch1cortex (the h1ppocapal formation) and/or the subcortical aaygdalo1d nuclear complex. S011e of the more compelling data, obtained in t 8 .! P I .a l IQ I 4 ., -' .~ V

PAGE 27

-"h m n> w, tS2 e:ner' h i' a: .,.i> r a wtmz ter :whhritt biecttMr-to ee trtt n w, w zhtMtt rl ti' n trt .. E111raael Donchin Psychophysiological Monitoring studies of patients with indwlling electrodes. have led Wood, et. al (1980) to conclude that one endogenous C0111Ponent, the P300, uy not reflect the activity of a unitary generator. They suggest that this ccaponent r~resents the see 1d activity of ault1ple sources with ault1ple orientations (see also Okada, K111fllln, I Williaason, 1983). The ERPs provide a rich class of responses that IIQ, within the appropriate research paradigm, allow the study of processes that are not readily accessible to exper1ntal psychologists by other ans. The. key USU111Pt1on of cogn1t1ve i)Sychophysiology 1s that ERP ccaponents are nifestat1om at the scalp of the activity of specific intracran1al processor$. Our reference is not to specific neuroanatollical entities. but rather to specific functional processors. While netwrks of nuclei IIIIY be involved in a dyr1111ic fashion in the activity.represented by each ERP a,mponent, our current understanding of the underlying neuroanat011Y is insufficient to ge"!!rate aeaningful neuroanatomical hypotheses. Yet. the available data, regarding the consistency with which certain c011ponents asured at the scalp behave, perait us to hypothesize that these coaponents do signal the activation of intemal subroutines. The discussion above should not be construed as implying that the electrical activity recorded at the scalp is itself of functional significance. For our purposes, the ERPs 111.Y be due solely to the fortuitous suation of electrical fields that surround active neurons. Although so have argued that EEG fields do have functional significance (especially persuasiveis Freeaan, 1975), we can reaain agnostic on this issue. For our purposes, 1t is sufficient that we elucidate the functional role, in I -

PAGE 28

. h:&' '*180 :tiie n OMX 1' . D P'.h tr' esett ftrttt' 'tbb ,.++f :tft\ i1i:veitt"nic:tte-iif:i'tftt #tktstt cs. ii&et'Jtf!i:l" ,Enuel Donchin Psychophysiological Monitoring 1nforaat1on-processing terms, of the subroutines manifested by the ERP C0111Ponents. It will. of course. be of considerable interest to obtain a detailed neurophysiological description of the neural implementation of the subroutine. However. 1 description of the subroutine as a processing entity 1s of interest independently of the neurophys1ologica1 descript1.on. 3.0.2 The benefits of ERPs 1n 110nitor1ng The principal benefit that may accrue to monitoring fr0111 the introduction of ERPs into the repertoire of the investigator is the more direct contact they provide with cognitive, rather than with affective, activity. As in ANS-based IP, the investigator's task is the creation of n .__ conditions in which knowledge un~que to the guilty will reveal itself by .. J triggering a specific. and uniquely interpretable, response. The ERP responses. however, are n1festations of specific information processing ... activities, not of specific 11110tional responses to stin,li. As far as we can tell. the activation of the appropriate internal subroutines is always acC0111Panied by the appearance of the appropriate ERP c011ponent. To the extent that the subject processes the information 1n a certain manner and the processing aanifests itself 1n an ERP component, the appearance of the C0111Ponent tends to be obligatory. This 1s not to say that any ERP component is an autOlllltic candidate to use 1n 110nitoring. However. there are so very strong indications that the technique can be of considerable utility. .,_ "\ ,, V

PAGE 29

: ~..,,,. ,r t Ht,a er tr:rattr s :":hsst rnbetrH rtonwrt dth 3.1 The Vocabulary of ERP Ccaponents M:ttdi nltti"Oteit!n'H'hsfit: Ii"\ 'tNfraofsirt 'ribiieriiritrr4it: Hi~,. 1 r Emanuel Donchin Psychophysiological Monitoring It 1s important to eaphu1ze that the P300 is not the only ERP component that can be ployed 1n Monitoring. The ensble of infonaat1on processing act1vit1es 1111nifested by the ERPs 1s already quite rich. Additional coaponents are being discovered and deeper understanding 1s being reached of C0111Ponents which have been known since the mid-601s. In principle, 1l1110st all of these cmponents can be ut111zed 1n aonitor1ng. a. NlOO 'Direction of Attention Hillyard and his associates (see Hillyard and Hasen, 1986) have shown that the NlOO c0111ponent (recall that. components are labeled as egative or

os1t1ve, and the number following the character refers to the modal latency of the cmnponent) is affected by the directions of the subjects attention. Events in the focus of attention tend to manifest a sowhat larger NlOO. The effect is reliable. It can be used to a:,nitor changes 1n the directions of the attention, or changes in the attentional level. Thus, the NlOO can plQ a role either in ascerta1n1ng 1f the subject is actually following instructions with regards to the allocation of attention, or to determine if events in the environment have caused the subject to shift attention. In a way, the NlOO 1s sensitive to affective processes as are the ANS responses Yet, 1t is likely to be less susceptible to subject-controlled actions. b. N200 Detection of Mis111tch Considerable evidence exists that the N200 1s elicited by events that violate a subject's expectations, even if they occur entirely outside I I .,, I I I ~' .......... ,, ...... a ..-i _ 'l'll!!f .S .... W ... .... ............... p """' ...... """'o-, ...... ,,.,.., _,......._ ........... :; .-; ........ !19R't .:~ ........ 11"-"1 .... Jll!l"I ~...-..... l!llllll'lllla11t ..... ~--'-.,.. .. .,.._ ""'' ....... A-1!1!1"-""')1 .-,ii; '"""'...,..._. i

PAGE 30

:!>L)O Emal'l.lel Donch1n Psychophys1ological Monitoring the focus of attention. Thus, the N200 seems to be a manifestation of the activation of a 1miS11atch0 detector. This component 1s, it would seea, t~ least susceptible to control by the subject's voluntary actions. The occurrence of any deviation froa regularity, indeed any a1s111tch between an event and 1ts 11111ediate predecessor, elicits an N200. Thus, with the proper setting and with 1dequ1t'9 recording equipnt it may serve u an exceptionally effective tool 1n monitoring (see Naataneen, 1985). c. P300 -Updating of Working Memr,v We have already referred to the P300. It m-.y be useful to note that in addition to its 111Pli~de1 that can play a useful role 1n monitoring, the latency of the P300 has proven an excellent indicator of ntal. processing time. Since the elicitation of the P300 depends on the categor1zat1on of the events in the oddball sequence, then its latency is evidently a function of the duration of the categorization. A variety of scenarios suggest themselves for utilizing the P300 latency 1n IP. d. Readiness Potential Preparation to Respond The Readiness Potential and the Contingent Negative Variation are among several ERP c0111ponents that we call Event-Preceding Negativities (see Ritter, Kelso, Kutas, and Sh1ffrin, 1984). They are quite clearly related to the activation of preparatory, often unconscious, processes by the subject. In recent work we have shown that it is possible to deteraine, frm the extent to which these potentials are larger over one hisphere rather than the other, which response the subject was $ Z,. z. !i4P Al I 4 4 OZ I. t .. W IS 5 I. t 4 a I P,J 41 #-. A ca;: -q C) q

PAGE 31

y-.. ,., wt em ttf :'if"tf :1tlliitt1tsd:{'sr&i'.fr tt S t&ttbttssr t' ee"Nrzs'"tr1r 1ffl'ttSthr::iihks11m nrt r:trt:&Mn ;it -w:rte', 'die..-wrCMifit:triawt1 .. ,. -------~ Em,-.el Donchin Psychophys1ological Monitoring contemplating regardless of the response that has actually been made. Th1s feature of t~ese cmponents 111k1s th a very strong candidate for use in IP as they appear to provide a aeans for assessing conflicts that occur before a response 1s tted. e. N400 Sellllntic Mistch Kutu, 1n an extensive series of studies (see Kutu, Lind111100d1 and Hillyard, 1986), has shown that words.that are 1n some way incongruous, or unexpected, within a discourse elicit an ERP ccnponent that 1s .negative and has a latency of about 400 as,ec. In a way this appears to be a semantic component that 1s e11c1ted by unexpected events. However, it is uniquely associated with linguist1c st1mu11. Its promise 1n IP 1s substantial. 4.0 ILLUSTRATION: THE P300 AND THE ODDBALL PARADIGM In this section we will discuss the aanner in which the P300 component can be used 1n 110nitoring. As is true of all other ERP components the P300 is extracted from the ongoing EEG by signal averaging. Its measurable attributes are the amplitude, the latency and the distribution of the potentials_across the scalp. These can be used in various combinations in domains as disparate as clinical neurology and hun factors. However, 111ch of the use depends on the integration of one basic paradigm, the so called "oddba11 paradigm, into the mn1tor1ng situation. That paradigm will be discussed in the next section. The manner 1n which it 1s employed in a diversity of monitoring situations will then be reviewed. __,.....,._ 5? .... i 4 ---~-# l""LP ..... i_,P(..,L!19'14""'-"14 ,_.;"""" t,"'!111, ......,1111191!1__, ....,.. !11111, ......... .. 4.--...., _. .... C~i ..-.... -... -~, 111114~1 .-.z ....... ; !!'I. ,,.,. ..... 111!'9'. ftlS411!11!'1P.t"""'9il!IIIIIP'II 4 .. 1 ... ill!lllll ..... 111111.tt,.... ..,..44"'1-...,_P ,~,1-.-4 f.

PAGE 32

4.0.1 Description of the Oddball Paradigm ~42, Emanuel Donchin Psychophysiological Monitoring The procedures for using P300 IIIY serve as an illustration of the conceptual issues described thus far. We nor11ally study the P300 in the context of the so-called oddball paradigm. In this procedure the subject is presented with a series of events. The cr1t1ca1 attribute that defines a series as an oddball paradip is that there is a classif1cation rule that categorizes 111 the events in the series into two categories. The class1f1cat1on rule can range fr011 the concrete to the abstract (i.e., frequency differences of tones, exemplars of two different categories). The events can be quite diverse and the categorization may depend on an ensemble of properties of the events. Another critical attribute of the paradigm 1s that the events from each of the two categories are chosen at random and that one of the categories is selected much less frequently than the other. Finally, it is necessary that the subject be assigned some task that requires active processing of the events. When these cond1t1ons are satisfied, then events that belong in the rare category elicit a large P300, whose amplitude is inversely proportional to the probability of the eliciting event-category and directly proportional to the relevance of the event to the subject's task. It appears from research that we, and others, have conducted 1n the last 15 years, that the P300 1s a n1festation of the activation of processes t~at are involved in the 111.intenance of working 11111110ry. The activation of these processes is nonconsc1ous, except in the sense that the subject does consciously monitor the o-, ... --, u. !-~. I r I -~_.___.....~_,...,,._~--..,,-~----------l

PAGE 33

E~el Donch1n Psychophys1olog1ca1 Mon1tor11'1Q events. Furthermore, as far as we can tell, the P300 1s an ~b11gatory response to the event, 1f the subject 1s 1n fact processing it. The ut111ty of the parad191111n mn1tor1ng environ1111nts 1s that 1t provides I rather standlrd1zed r1ng tool that can be easily ported froa one situation to another. The following attributes of the P300 have been ut111zed 1n various mdels of aon1tor1ng: a. As the P300 is coaonly elicited by rare, task relevant, events 1t can be used to infer which events 1n a series to which the subject attends were perceived as being rare events. In other words the P300 can be used to detera1ne 1f an event has surprised the subject. . b. The lower the problb111ty of the event, the larger would be the P300. Thus, 111 things being equal, the uap11tude of the P300 can be used as an estimate of subjective probab111ty assigned by a subject to events. c. The more attention the subject directs to a task the 1110re likely are sti11111 associated with that task to e11c1t a P300. These task relevant stimu11 generally have to be rare to e11c1t a PlOO. However, 1f the subjective probability is held constant and the subject attention drifts away from st1mu11 the amplitude of the P300 w111 decline. Th1s characteristic underlies the use of P300 1n the monitoring of ntal workload. d. The latency of the P300 is proport1~na1 to ntal processing ti. Therefore 1t can be used as a means for monitoring ntal timing 1n a manner that 1s not cantmninated t,y motor activity. [, I F t '""""*''C?t lll!IIIJ_ll!a_lll!IIIJlll!IIII: 111!1.P .... Cl!lll,JOtlll!'l .. e IIJl!!II -~ s 1!1111. PiJIIIC.1111, ... 4 t !!111-... 1111111!11.t .......... ._.._.... .. ___ ; .. ,...-... z~t.111! .... !!11115 Sll!l!lltl~Alll!1191194__.IIIIIJM$@...-,.._llll!l_!llll,Q.-,;,-q;_--.i ....... Zllft-

PAGE 34

Emanuel Donch1n Psychophys1olog1cal Monitoring 4.0.2 The use of P300 1n the assessment of Mental Workload It 1s possible to use the P300 as a tr1c for Mental Workload. The technique is s0111Wh1t indirect, but 1t is in keeping with the general tradition of Workload uuraent. We have adopted a technique called ~he secondary task procedure. An ex111Ple uy help. Suppose one is driving a car on a highway and carrying out a conversation with a companion. Driving the car is the pr1aary task and the conversation 1s a secondary task. (While 1t 1s perfectly possible to think of a situation 1n which the conversation 111y be 110re important than the driving, in our example this is not the case.) tbJ. as long as the traffic is light and the weather 1s good and one knows the way . the conversation flows naturally. It is easy to follow a companion's train of thought and one can articulate one's views with great eloquence. Then. the traffic becomes heavy; the driver is in the left lane and 1111st pull to the right to exit on the very next exit. The driving is still smooth. There 1s no drop in speed. The lane change is executed in a t1mly and efficient mer. In other words. the performance of the primary task has been protected. But, th1s 1s done at the expense of the secondary task as is evident from the fact that the driver stopped attending to the conversation. The answers become laconic and, in the end, irrelevant. Thus, the pressure experienced in connection with the primary task which did not reflect itself 1r the driving, did cause a deterioration 1n the performance of the secondary task. If we had some measure of secondary task performance we could determine the difficulty of the primary task by suring the deterioration in the performance of the secondary task. The 0 0 0, -~~""" ..... """",..~""""'l!illllllla ........... """"'-'....,. ....................... ~llllllll!l!lll~""""' ..... !911.!111l!Jlll'll!l'll'~ ...... """""" ...... ............................... -.......... t : 1 J. tit ( 0$. t I S -. E S C . .J . 3 . P i $ 4 .. li U . t I 4 I 4 ii i IPA Ji.

PAGE 35

..... 3t-1S Emanuel Donchin Psychophysiological Monitoring trick is to find secondary tasks that provided a usefully gradable measure of performance and which can be c011bined with pr1ry tasks without affecting in tum the pr1111ry task. Research conducted priaar11y at the Cogn1 ti ve Psychophys1olog1cal Laboratory at the University of Illinois has shown that the oddba 11 parad1911 can serve as such a secondary tuk and that the aap11tude of P300s e11c1ted by st111111 associated with such secondar-y tasks w111 vary inversely with the workload associated with the primary task (Donch1n, Krmner, & Wickens, 1986). 4.0.3 The use of P300 at a ntal prosthesis A project currently under way at the Cognitive Psychophysiology Lab at the University of Illinois uses the P300 1n a tool designed to allow a person who has lost 111 control of his musculature, and has~ speech, to ca11111Unicate words to a camputer. This 1s achieved by converting a matrix in which 36 different item are turned into an oddball paradigm. The latter the subject is thinking is the mode to elicit a P300. Note this is an example of a technique which appears to be the ultimte in mechanical mind reading but which in effect 1s no more than a trick which allows the subject to tum a switch with his brain waves rather than with a finger. The key to success, as 1n all such aonitoring, lies 1n the setting of the situation in which the data are acquired rather than in the data acqu1s1t1on approches a.. I.. # C P .. .i.4 ;p 241 (QS@. S# ii a. ,0 $, I # 4 41 ,. 0.. t 4 C JS! I UP IA !4.. . 0. :.a . J

PAGE 36

5.0 PSYOfOPHYSIOLOGICAL MONITORING AND PRIVACY E111n1el Donchin Psychophysiological Monitoring If we can indeed Mke such powerful inferences fran the ERP. we d_o seem to have the capacity to read -the a'lnct. We have on hand a psychophys1olog1ca1 tool that gives its user what Y appear to be quite a disturbing power. The wrries have indeed surfaced. Take for ex~le a fairly lengthy article carried on June 3, i984 by the Washington Post. The article was written by correspondent Michael Schrage and was entitled Technology Could let Bosses Read Minds. The article continued on the following page under the headline Privacy Veil May Block Brain Watchers View. Schrage reports that Researchers in both Acadia and industry say it is now possible to envision a marketable product that could instantaneously assess whether sployees are concentrating on their jobs by analyzing their brain waves as they work He tells us that Westinghouse Research and Development Center in Philadelphia is exploring the use of brain wave analysis particularly the brain wave known as the P300 -as a means of determining the individual's level of attention and cognitive processing. A Westinghouse man is quoted as predicting that within the next 10 years Westinghouse could rket a camplete systa capable of 110nitoring the mental processing effort of employees as they worked This discussion inevitably leads Schrage to discussions with legal scholars -and labor leaders regarding the invasion of privacy implied by these '. developments. While the present report does assert the feasibility of the enterprise described by Westinghouse, I less inclined to worry about 0 . t) \, ....,. invasion of privacy. The most likely aisuse of psychophysiology will be the 0 ~:s11111_zi...-iz._.tc,...a..-.a .. ,..,UJ.~. 1111a.a ..... a.u-...... 11!".--IP.t -..tt!llllft".111111 ..... 2., ...,..f ... C.llll.!Z ....... ,...ll!"II. 1!111111. ,........, ..,.._1....-~IPl!l!: .. ipa, ,... ..,., ........ 1!'1'11. ..... ,., .. ,..,.......,s~.'!'1!111 ...... .:-l.

PAGE 37

;":,,~ raqs t ... r .z "i-rmt -crnt ,. s : o in' ti' 'Ws.tw it:t'iife cf,e,-dtit"f' ". ci!:iHrci:i"rii~')C,s'}i~--,k CAii.~ .. l '3'-/7 Emaratel Donchin Psychophysiological Monitoring Polygraphyzation of the research conducted with the ERP over the past 20 years. By Polygraphyzation I an the ldoption of scientific discoveries fer ~rcjal ut11ization without adequate attention to the effectiveness and ut111ty of the procedure. Polygraphyzation occurs when the c011111erci1l development 1s done without an anchor 1n the scientific coaunity. Actions are taken to assure the profitabi11ty of the product and caution and control beCOIIII less critical~ Training 1s a1n1 and tends to be highly professionalized. I phas1ze that all this 1s done well within the law. But, i_t raains the case that it is quite possible to have what appears to be an iapressive instn1111ent that is essentially worthless. 5.0.1 The issues presented by the P300 data Let us take a few mil'lltes to consider the issues raised by the P300 data. It 1s clear that in l<'boratory conditions we can "read the mincl.11 One can tt:tink of any number of applications for such a procedure, and few have been described above. However, 1f one is offered a box in which a pointer 1s driven to the right or to the left by the magnitude of the P300 as a function of so mental process two classes of questions arise. There is the question of utility. Does the technique really work? How general is the utility. To date there has not been a test of the P300 based on procedures in valid ecological testing; with the exception of some of the more clinical applications. We know the procedures work well in very complex lab arrangements. Yet there has never been enough support to allow the sort I i I I, L ,. I 1 il .......... !!!II . "'' --...,,.__ :~.'!Ill .... ...,._,_ ........ """"'"-'!. ...... ,11111'911!111!1 ______ ...__..,1!1111. s--....!"!! . ,..... ...... 4 111!111. .,..,: ... '""'' .... a....-i . ~:1111a.,..,...-i ... s ..,..; .. ,11111cP.111. ......

PAGE 38

Er11el Donchin Psychophys1olog1cal Monitoring of thorough exper1aentat1on in highly nonnative si~uations. This issue is, to a large extent, open. Then there is the issue of Privacy. What is the degree to which mn1tor1ng 1ap1nges on ind1v1dua1 rights? Does 1t go beyond currently accepted 1nterpreta~1on of the rules? The worry here must be tempered by an understanding of what the 111ch1ne is actually doing. It 1s useful to review briefly the constraints on the aach1ne. Also we need to d1st1ngu1sh between two aodes of privacy 1nvu1on. The p_opular notion is that 1t wi-11 be possible to achieve the technical feat of king audible by chanical (or electrical) ans those thoughts which constitute our 1ntema1 speech. Thus, the taphor driving the worry 1_s that of eavesdropping. But, this has never proven possible. Eavesdropping on the a1nd is unlikely. It would only be possible if the signals we can record extemally carry within tha the richness and the variety available in mental life. One can fantasize, of course, that new technologies will increase the range of the 110n1toring. Indeed, given the trends in increased computing power and reduced size and cost, same super ainicomputer 1mplented on biological principles a1ght have the power and the savvy to interpret the signals in a depth that aatches the profundity of the task. However, 1t 1s ay belief, that even in that case psychophysiolog1cal eavesdropping wuld not be possible. To some extent the constraints that could not be eliminated come from the fact that there 1s too auch noise to develop a useful eavesdropping technique. There are too any processes all working 1n parallel and furiously 0 ._ C) interacting with each other for there to be a possibility that an external .:) A $2 Q Q ii b, ) a a -Z a U tt .. Uh. Q. ; C QC. . S JU QS:. . C C c. :a

PAGE 39

~,. .. ritit:?(ztt tt< #arr a Wk't ; ., t1t:r'httt7X C T E111n.1el Donchin Psychophysiolog1ca1 Monitoring nifestat1on of any one of these processes w111 talk to a computer in the s-language 1t talks to 1ts counterparts 1n the syst. This tter of the anguage of the brain 1s crucial here. However thought processes are tapleaented, their 1ap1-ntat1on 1s ult1aately a tter of neurons talking to neurons. Indeed~ 1t 1s aost likely a tter of a1111ons of neurons talking to a1111ons of other neurons. These neurons converse, of course, wtth each other by whatever 1 anguage is used for such coaun1cat1on. There 1s not yet a full consensus as to the nature of .that language. It 1s clear that 1n the n neurons affect other neurons by secreting ttny doses of chicals (the neurotransaitters). These secretions are the consequence of the conduction of neuronal impulses across axons and the integrative activities of the dendrit1c IIUlbranes. We, who study ERPs, benefit froa the fact that when occurring in the ss, and 1n a highly synchronous nner, these inter-neuronal actions nifest themselves on the scalp in the form of large integrated fields of potentials recorded as the EEG. However, this largely epiphenoaenal activity while valuable as an index for the ti course and level of neuronal action is unlikely to serve as a source of 1nfonaat1on on the specific nature of the vast exchanges 1n the neuron's own language that have given rise to these psychophysiolog1ca1 signals. Thus, 1t 1s only the second sense of aonitortng that one can assu psychophysiolog1cal signals to be of use. We are not, and will not, eavesdrop on the aind. Rather we are observing the consequences of neural action and by judicious construction of the situation we mQ be able to pose questions to which the psychophysiological signals aay give an answer. This metaphor . .... 7 i

PAGE 40

rz snhr tHt 2en w em s ._.. se ,. t' tt r a oms r ,e,r s :tlttttt tfrt--x nw11;,; r:frdef#frct sfrttiisftfie'tfatitit,s.,,;,,; 3so EnJel Donchin Psychophysiological Monitoring describing psychophys1ological monitoring as a process of seeking answers to spec1f1c questions 1s very 1aportant because it underlines the pr1nc1pal condition to the success of such aon1tor1ng. The value of the answer will depend on the sagacity with 1m1ch the question hu been put. In other words the key to the utility of these approaches is the ability to pose useful questions rather than the procureaent of ye~ another asur1ng 1nstnaent. Proper application of psychophys1o1ogica1110n1toring requires that one realize that what 1s being aon1tored 1s the activity of bodily system. These systas are driven for physiological reasons by the deMnds of the system. These bodily organs serve 110re than one function and therefore their activity cannot be presuaed to be uniquely related to any psychological construct. The s1gnalswe record Ike sense, therefore. only in teras of situation. There is no deception wave a wave that no tter when and under what circumstances it has been recorded indicates that the person has lied. There is not even a wave that indicates unequivocally that any e110tion occurred. Rather, the change in the physical signal indicates the activation of a certain processor, or processors. If. and only 1f. this is uniquely interpretable w1th1n the context of the recording 1t 1s possible to aake psychological inferences fro the data. Thus the degree to which psychophysiolog1cal signals can have psychological an1ng depends aore on the degree to which the syst is set to be driven in a unique fashion by the psychological variables. For example: The workload assessaent techniques employing the P300 we described above depend on the estab11shllent of a very sensitive relationship between the conditions of aeasurent and the subject's ability to perceive the demands. 0 ,. _,~z-""llllz _.... ... ,.,.. ....-.--_lllll!~a .,,_,.....~""'-""'' .... .,..-i,1191'!'.~1111111111 .. fPl'!t ............. ...,.....,..,....,.a ...... --..tu..,....~--,~-----~-... !".--. ~--

PAGE 41

ts n to n*astt rs r Wt r 11 o n id : t>> tt: t rl rt es :en "' +:tie oae en tntn t et1 tr >IT d t> ......_ ~SI E.-...el Donchin Psychophysiological Monitoring Unless the subject is performing the two tasks assigned in the dual task setting the asuraents of the P300 have no value. Furtheraore intense va11dat1on and reliab111ty assessaent are st111 needed. In other -,rc1s. active. part1cipat1on of the subject is a condition for the success of psychophys1ologica1110nitoring. It 1s unlikely that it ~uld be possible to apply an unobtrusive probe that will intrude on the subject without at scae level the subject accepting the structuring of the situation that constitutes the question addressed to the psychophysiological system. The above r1111rks should not be construed as casting doubt on the utility of psychophysiological 110nitor1ng. As I indicated above the increasing depth with which these signals are understood and the increasing sophistication of cognitive models when coupled with the spectacular developments 1n a1n1aturiz1tion1 sensor technology and data analysis open a very broad scope for such aonitoring. However, this will only be accomplished within the constraints of good thodology. Furthen10re1 whatever monitoring. can be done w111 be constrained by the nature of the biological and psychological systs involved. One aust steer clear of the extravagant claim, and avoid the unnecessary fears that these aight invoke. 6.0 THE DANGERS OF eotltERCIALIZATION It .. st be emphasized that the remarks 111de in this report, and 1n particular 1n section 61 pertain to the scientifically valid use of the techniques. It is I sad, but obvious~ fact that the limits science and nature impose on feasib111ty do not always serve as constraints on the huckstering I I

PAGE 42

tr r +: r:t'tt rt.a :ri ntoa ""t ea oz St1t nilrn are r ', m c-11 k' tt> tc7' wartt w -Wr:e' tbtri:in:stlt,.Yttr it :t :tr x En.1el Donchin Psychophysiological Monitoring of technologic~l pseudo-rvels. That reason, and proper scientific analysis, suggest that eavesdropping on the aind 1s unlikely does not, unfortunately. 1aply that the public or elected officials will not be persuaded that a technical 111racle his been achieved by saaeone with a gadget and good rket1ng technology. The cues are innumerable in which the gu111b11ity of the 11arket plac~ has served to enrich those who peddle wrthless solutions to serious problem. ~r011 Snake 011 to water D1vin1ng to, perhaps, Strategic Defense, the public (and its elected and appointed representatives) can be induced to believe in the efficacy of scae technique despite the caveats of informed scientists (w.o, adlll1ttedly, can also err). TIiis, this report is addressing solely the issue of the scientific feas1b111ty and the va11d1ty of psychophysiolog1cal mnitor1ng. Whatever conclusions -. draw w111 not address the degree to which there is a possib111ty that Industry, or Government, will be induced to believe that a specific technique is indeed valid and that using it is the best one can do in a difficult situation. Under such c1rcuastances vast systs for 110nitor1ng 1nd1v1duals IIIQ be iapl-nted. That th~y will be of no scientific value will not reduce their social impact. However, the consideration of public decision Mk1ng and the sorry 1nteract1on between the gullible and the unscrupulous, or the naive, is a Mtter for Sociologists or Political Scientists not Psychophsy1ologists. I L I C t t w t 4 .l 4 PI I SJ t 0 (> ._;J

PAGE 43

.' ., rwro nrf1tt a :d:ttf+r a r itt ne,;, T. 7.0 SlltMRY Emanuel Donchin Psychophysiolog1ca1 Monitoring In th1s report I surveyed the aanner in which psychophysiolog1ca1 approaches IIIY be used, at present and 1n the near future, to mn1tor 1nd1v1dua1s, with particular attention to the mnitoring of affective and cogn1t1ve responses. A general survey of various indices was followed by a aore detailed discussion of the use of Event Related Brain Potentials and 1n particular the P300 C0111Panent u a. tool 1n the study of mental functions. It 1s concluded froa the review that there is indeed a large and increasing potential for the use of psychophys1ological techniques. However, we also note that the use of these techniques requires considerable care and that their application requires training and sophistication in the technological, biological and psychological aspects of the work. We emphasize that the monitored data can be interprted only within a detailed understanding of the context of the measurents. We note that it is unlikely that such monitoring would allow intrusive reading of a person's mind. Vet, 1t is quite possible to arrange recording circumstances that, with the subject's cooperation, could serve to illuminate the content and form of a person's mental activity. Thus, the worries regarding the violation of privacy by such psychophysiological means are exaggerated when couched in alana1st terms, yet these worries are not to be entirely discarded. A particular worry that rooted in the sociology of technological development rather than 1n psychophysiology is that a superficial application of psychophys1olog1ca1 findings Y lead to widespread employment of devices .--.. .:c CA ue;a: a. uz : ; 1.1;. 4 t : ; a s ; : t p ... : :. ; a : z .. ts a t u .au : s.a.

PAGE 44

~~:i,c. ft flt IM >tldW-5 ... er ti MhlT # t t ...... ,. 1r :tetttffCOJ":ittec e+ t#t ... ca; tts+ttl\'',n,;ra,ef : ttib. i' 0ta-rtf t-"Jit't~rtatfir#~'1,tif1+,"e -. Emanuel Donch1n Psychophys1olog1ca1 Mon1tor1ng purporting to allow monitoring. The fact that these devices are so error prone and ambigous may not affect their widespread use 1n a manner that would affect decisions regarding individuals' fate. The manner 1n which. the Polygraph is used 1n persomel screening 1s a case in point. Thus, even though the real power of the techniques Y be constrained and benevolent, the rriage of greed and gu111b11ity that so ~ften determines the action of public bodies does leave r0011 for concern. NOTE: In parts this report draws on material in Donchin, Kramer and Wickens (1986) and in Donchin (in press). The research in the CPL referred to in this report has ~en supported by AFOSR (F49220-85-C-0041, 1-5-25288) and DARPA (MDA903-83-C-0017, 1-5-24742). .41 I l.@. .. Q $. I J.JQ z.;en D.J U CL ;:: "' . t. I t S t I : Jt ( a 3 I 4 #.. 4,.p 4 .SQ I I#. I I Ph ,. 0 -() ,. ::J

PAGE 45

. ~'; 'tr t :ri rdis ae: : n: tt 131 : *' st r1b u tbs t r : a'k r wr :su ti '-:rstnz t+::ttf't s-&"t f "ttiri:tt#r fr, r: -cir' h ;,REFERENCES 355 Emar1.1el Donchin Psychophysiological Monitoring Allison, T Wood, C. C. and Mccarthy, G. The Central Nervous System. In Coles, M. G. H., Donchin, E. and Porges, s. (Eds.) Psycho,hysiology: Sflteas, Processes, and Applications. New York: The 6u1 ford Press, 1 6, p. 5-25. Beatty, J., Barth, D.S., Richer, F., and Johnson, R. A. Neuro111gnetometry, Psychophysiology, In Coles, M.G.H., Donch1n, E., and Porges, s.w . (Eds.) Psycho,hys1ology: 51sts, Processes, and Applications. New York: The &u1 ford Press, 1 86 p. 26-40. Bod1s-Wolner, I. (Ed.) Evoked Potentials. New York: New York Academy of Sciences, 1982. Donchin, E., Miller, G. A., & Farwell, L.A. The endogenous components of the event-related potential --A diagnostic Tool? In E. Fliers (Ed.), Progress 1n Brain Research. Amsterdam: Elsevier. Donchin, E. Event-related brain potentials: A tool 1n the study of human infor111t1on processing. In H. Geble1ter (Ed.), Evoked potentials and ,--. behavior. New York: Plenum Press, 1979, pp. 13-75. -~' Donchin, E. Can the Mind be read in the Brain Waves? Presented as a Science Public Policy Sa1nar, sponsored by the Federation of Behavioral, Psychological and Cognitive Sciences with support from the Carnegie Corporation of New York. Washington, DC: May 17, 1985. Donchin, E. Surprise! Surprise? Psychophysiology. 1981, .!, 493-513. Donchin, E., Kramer, A., & Wickens, C. App11cat1ons of event-related brain potentials to probls 1n engineering psychology. In M.G.H. Coles, E. Donch1n, & S. Porges (Eds.), PsychoJhys101091: Systems, Processes, and Applications. New York: Guilford ress, 19 6, p. 70216. Donchin, E., Ritter, w . &McCallum, C? Cognitive psychophysiology: The endogenous components of the ERP. In E. Callaway, P. Tuet1ng, & Koslow (Eds.), Brain events-related potentials in man. New York: Academic Press, 1978, pp. 349-441. Donch1n, E. (Ed.) Cognitive Psychophysiology: Event-related potentials and the study of cognition. Hillsdale, New Jersey: L. Erlbaum Associates, 1984, p. 428. Politics and Marriage. 'l!.PPJ 4 4 u . ti Pi JES lb .. ewe ;Fa ?-4.L t .. a t1 a s.. z ii t. t. ..... .9 4 C aa P CS C 5 .4 #1. ( I CZ, $. *'

PAGE 46

tCtt st ,-'t 011 r .,... r t :a,, r ... n: 'tbttmfht > e:t 11 ti@rt lilt: re 't>'cenY' -Wfferx YSz e+<:dlt t+tn hd'+::e:: Emanuel Donchin Psychophys1olog1ca1 Monitoring Fowles, D.C. The Eccrine Syst and Electrodermal Activity. In Coles, M.G.H., Donch1n, E. and Porges, s.w. (Eds.). Psychoph!siology: sutw1 Processes, and Applications. New York: fheuiiford Press, i 6, p. 5l-96. Freeun, w. Mass Action 1n the Nervous System, New York: Academic Press, 1975. Gaillard, A. W. K., & Ritter, w. (Eds.). Tutorials in event related yotent1al research: Endogenous co.,,onents. Aisterdu: North Holland, 983. Goodin, p. s., Squires, K.C. & Starr, A. (1978b) Long latency event-related ccaponents of the auditory evoked potential in dementia. Brain, 101, p. 635-648. Gopher, E. & Donch1n, E. Workload -An exainat1on of the concept. In K. Boff & L. Kaufman (Eds.), Handbook of Perception and Human Performance. New York: Wiley & Sons, 1986. Gratton, E., Coles, M. -G. H.1 & Donchin, E. A new thod for off-11ne r1110vaJ of ocular artifact. Electroencephalography and Clinical Neurophysiology, 1983, ,a. 468-484. Halgren, E.1 Squires, N. K., Wilson, C. L., Rohrbaugh, J. w., Babb, T. L. & Crandall, P.H. Endogenous potentials generated 1n tbe human h1ppocampal formation and amygdala by infrequent events. Science, 1980. 210 t 803-805 Halliday A. M. (Ed.) Evoked Potentials in Clinical Testing. Churchill Livingstone, Edinburgh, 1982, p. 575. Heffley, E., Foote, B., Mui, T., & Donchin E. PEARL II: Portable laboratory C0111PUter syst for psychophysiological assessment using event related brain potentials. Neurobehavioral Toxicology and Teratology. 1985, 7, 399-407 -Hillyard, s. A., & Kutas, M. Electrophysio1ogy of cognitive processing. Annual Review of Psychology, 1983, _li, 33-61. Hillyard, S.A., & Hansen, J.C. Attention: E1ecrophys1olog1ca1 Approaches. In Coles, M.6.H., Donch1n, E. and Porges, s.w. (Eds.). Ps~chophys1ology: Systems Process, and Applications. New York: The Gu lford Press, 1986, p. 227-243. Kaufn, M. P., & Schneiderman, N. Physiological Bases of Respiratory Psychophysiology. In Coles, M. G. H., Donch1n, E., and Porges, S. w. (Eds.). PslfihoJhysiology: 51stems, Processes and Applications. New York: The 11 ord Press, 1 86 p. l07-l2l. .,. oa au q; a .: ... .; I. a : a z. :ca_ ca:.e_-4,.1 .; U. i ... a. ;. ii .)UAQL P.,.J ;qqc 0 ;

PAGE 47

.. r ., 35, Emanuel Donch1n Psychophys1o1ogica1 Monitoring Kutas, M., L1nda1110od, T.E. and Hillyard, S.A. Word Expectancy and EventRelated Brain Potentials during Sentence Processing. Ins. N. Kornblum and J. Requin (Eds.), Preparatory States and Processes. Hillsdale, NJ: Erlbaum Associates, 1984, p. 217-237. Lykken, D. T. A Treaor 1n the Blood: Uses and Abuses of the Lie Detector. New York: McGraw H1ll Book cciiipany, l98l, p. 317. Morstyn, R., Duffy, F.H. and McCarley, R.W. Altered P300 topography in. sehizophrenia. Archives of General Psychiatry. 1983, .!Q, p. 729-734. Naatanen, R. Processing Negativity: An evoked potential reflection of selective attention. Psychological Bulletin, 1982, !l, 605-640. Okada, Y. Neurogenes1s of evoked 111gnetic fields. Biomagneti: An 1nterd1sc1p11na'l arroach. In Rman1, G.L., Kaufllan, (. and Modena, I. (Eds.) New York: P eraa, 1983. Okada, Y.C., Kaufn, L. and W1111amson, S.J. The hippocmapal formation as a source of slow endogenous potentials. Electroencephalography and C11n1cal NeurophYsiology, 1983, ,ll, 417-426. Otto, o. Application of a Portable Microprocessor Based System for Electrophysiological Field Testing of Neurotoxicity. Neurobehavioral Toxicology, l, 1985, p. 409-414. Ritter, w., Kelso, S., Kutas, M. and Sh1ffrin, R. Report of Panel III: Preparatory Processes. Cc;gn1tive Psychophysiology. In Donchin, E. (Ed.). Hillsdale, New Jersey: Lawrence Erlbaum Associates, 1984, p. 179-219. Vaughan, H. G. Jr. The neural origins of human event related potentials. In I. Bodis-Wolner, (Ed.) Evoked Potentials, 1!!, 1982, p. 125-137. Wood, C. C., Allison T., Goff, w. R., Williamson, R., & Spencer, D. C. On the neural orgin of P300 1n man. Progress 1n Brain Research, 1980, fi, 51-56 ; L ,, I I t f, l ........ ........ ,a.ec .... a .. e19111149--u: .....-.... .: .... J.-1111!; -......... s.-. ..... ,-. ....---., ... 111111 ........... .,., ,.., ....... .... ,...,.. -....."11!~!11111!1!' .,........,...,......., ..., ............. ,.., .................. ,...,.~


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E5ZUH8YPW_ARWQSP INGEST_TIME 2017-06-05T16:32:18Z PACKAGE AA00055673_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES