Task difficulty, probability, and inter-stimulus interval as determinants of P300 from auditory stimuli

Electroencephalography and clinical Neurophysiology, 1987, 68:311-320 311 Elsevier Scientific Publishers Ireland, Ltd. EEG 03277 Task difficulty, probability, and inter-stimulus interval as determinants of P300 from auditory stimuli 1 J ohn PoUch Division of Preclinical Neuroscience and Endocrinology, Scripps Clinic and Research Foundation, La Jolla, CA 92037 (U.S.A.) (Accepted for publication: 31 October, 1986) Summary The P300 event-related potential was elicited with auditory stimuli in 4 experiments which manipulated combina- tions of stimulus target probability (10% vs. 30%), task difficulty (easy vs. hard), and inter-stimulus interval (5 sec vs. 2 sec). P300 amplitude was smaller and peak latency longer for the more difficult relative to the easier tasks across experiments. Increases in stimulus target probability generally diminished P300 amplitude and shortened peak latency more for the easy relative to difficult task conditions. Increasing the number of non-target stimulus tones decreased P300 amplitude reliably, but increased latency only slightly. Task difficulty did not interact with variations in inter-stimulus interval which produced generally weak effects for both amplitude and latency. These findings suggest that P300 amplitude and latency obtained from auditory discrimination paradigms reflect processing difficulty independently of stimulus target probability unless differences in task requirements affect stimulus encoding. Key words: P300; Task difficulty; Probability; Inter-stimulus interval The effects of stimulus probability on the am- plitude of the P300 or P3 component of the event-related brain potential (ERP) are well known: decreases in the probability of the target stimulus produce increases in P3 amplitude (Tuet- ing et al 1971; Duncan-Johnson and Donchin 1977). In addition, stimulus probability .also changes P3 latency such that more frequently pre- sented target stimuli demonstrate shorter latencies than less frequently presented target stimuli (Duncan-Johnson and Donchin 1982; Brookhuis et al. 1983; Kramer et al. 1986). Despite probabil- ity's hoa .ry history as a major determinant of the i This work was supported by NIMH Postdoctoral Traineeship in Neuropsychobiology H15128-002, NIAAA Grant AA6420, and the J.M. McDonald Foundation. This is Publication Num- ber 4489BCR from the Research Institute of Scripps Clinic. Correspondence to: John Polich, Preclinical Neuroscience - BCR1, Scripps Clinic and Research Foundation, 10666 N. Torrey Pines Road, La Jolla, CA 92037 (U.S.A.). P3 ERP component (Sutton 1979; Donchin 1981), it generally has not been studied in the context of other variables which also affect P3 size and tim- ing. For example, P3 amplitudes are.often reduced in task situations that require more perceptual processing relative to less demanding tasks (Isreal et al. 1980a; Kok and Looren de Jong 1980; • Wickens et al. 1983; Hoffman et al. 1985), while P3 latency is increased when the discrimination between relevant stimulus items is made difficult with manipulations of encoding and categoriza- tion parameters (Pfefferbaum et al. 1983; Ritter et al. 1983; Magliero et al. 1984). Thus, a variety of factors can contribute to amplitude and latency changes of the P3 ERP, but how such variables might be affected by stimulus probability manipu- lations is uncertain. Of the reports which have combined these fac- tors, almost all employed visual stimuli that varied greatly in stimulus parameters and processing re- quirements. A few studies using auditory stimuli have manipulated target probability and task dif- 0168-5597/87/$03.50 © 1987 Elsevier Scientific Publishers Ireland, Ltd. 312 J. POLICH ficulty variables separately, but most have been designed so that the possible interaction of these factors could not be observed (N. Squires et al. 1977; Johnson and Donchin 1978; Campbell et al. 1979; Banquet et al. 1981; Goodin et al. 1983; Polich 1986). Because many applications of the P3 as an assessment tool for cognitive function in clinical and applied situations have used auditory stimuli with a variety of probability levels and task situations (Brown et al. 1983; Lai et al. 1983; Pfefferbaum et al. 1984a, b; Finley et al. 1985; Howard and Polich 1985; Polich et al. 1986), it is important that the relationship between these fac- tors be explored in this modality since some of the differences in findings may stem from variations in task demands (cf., Goodin et al. 1978; Roth et al. 1980; K.C. Squires et al. 1980; Baribeau-Braun et al. 1983; Picton et al. 1984; Polich et al. 1985a). Without a comprehensive understanding of how task difficulty and probability may affect both P3 amplitude and latency, a clear interpretation of changes in component values within and between various subject populations cannot be obtained. The present experiments were designed to pro- vide additional information about the relationship between task difficulty and stimulus probability on the P3 component obtained from auditory discrimination tasks. The underlying rationale guiding the choice of experimental variables stemmed from the application of the P3 to study individual differences in normal and cognitively variable populations. Task difficulty was varied by manipulating ease of perceptual encoding or by varying the number of items necessary for stimu- lus categorization (Johnson and Donchin 1980; Pfefferbaum et al. 1980; Goodin et al. 1983; Polich et al. 1985b). Probability was varied by manipulat- ing the frequency with which the target stimulus occurred or the time between stimulus events since inter-stimulus interval also can affect P3 ampli- tude (Woods et al. 1980; Fitzgerald and Picton 1981; Perrault and Picton 1984). Several experi- ments factorially combined these variables so that different types of processing difficulty could be assessed across different types of stimulus prob- ability conditions. Experiment 1 compared a typical auditory frequency discrimination task with relatively easy and hard intensity discrimination tasks to de- termine first how the different auditory conditions might affect P3 values. Experiment 2 manipulated discrimination difficulty and target stimulus prob- ability to examine how these factors might dif- ferentially influence the P3. Experiment 3 used a frequency discrimination task and manipulated the number of non-target stimulus items and probability to determine if P3 values would be affected by the memory requirements induced with different numbers of non-target alternatives. Ex- periment 4 compared the easy and hard intensity discrimination tasks over two different inter- stimulus intervals to examine how the time be- tween stimulus events might contribute to P3 re- sults. Me~s Subjects Different groups of 16 undergraduate students (mean age = 21.2, S.D. = 2.6 years) from the Uni- versity of California, San Diego or San Diego State University were employed in each experi- ment. All subjects reported no neurological or psychiatric problems, were naive to electrophysi- ology studies, and received course credit for their participation. Equal numbers of each sex were used in each study. Recording conditions Electroencephalographic (EEG) activity was re- corded at the Fz, Cz and Pz electrode sites of the 10-20 system using gold-plated electrodes affixed with electrode paste and tape, referred to linked earlobes with a forehead ground and impedance at 10 k12 or less. The filter bandpass was 0.5-30 Hz (3 dB down, 12 dB octave/slope). The EEG was digitized at 1.5 msec per point for 768 msec with a 75 msec prestimulus baseline. Wave forms were averaged on-line by a commercial apparatus which also controlled the stimulus presentation and artifact rejection. Trials on which the EEG exceeded + 45/~V were automatically rejected. All experimental conditions were recorded with eyes closed (Polich 1986). Rest periods were provided between task conditions as appropriate. P3OO, PROBABILITY, TASK DIFFICULTY 313 Procedures Experiment 1. Three auditory discrimination tasks were performed by all subjects. The frequency task employed binaural 1000 (standard) and 2000 (target) Hz tones presented at 60 dB SPL. The easy intensity task employed binaural 1000 Hz tones presented at 40 (standard) and 60 (target) dB SPL, while the hard intensity task employed 1000 Hz tones presented at 40 (standard) and 45 (target) dB SPL. All stimulus tones had 9.9 msec rise/fall and 50 msec plateau times and occurred in a random series once every 2 sec, with the probability of the target tone at 20%. Subjects received the frequency discrimination task first, with the easy and hard tasks presented in a coun- terbalanced order across subjects. Subjects were instructed to move the index finger of their right hand whenever a target tone was detected in all conditions of each experiment• Experiment 2. The easy and hard intensity task conditions used in experiment 1 were pre- sented with the target stimulus probability at 10% or 30% to produce 4 different task conditions• These were counterbalanced over subjects. Experiment 3. The frequency discrimination task of experiment 1 (1000 vs. 2000 Hz) and a second frequency discrimination task in which the standard tones were either 250, 500, 750, or 1000 Hz with a 2000 Hz target tone stimulus were employed. Probability of target stimulus was 10% or 30% to produce 4 different task conditions which were counterbalanced over subjects. The non-target tone stimuli of the 4-standard dis- crimination task could occur with equal probabil- ity (25%) on any standard tone trial. Experiment 4. The easy and hard intensity discrimination tasks of experiment 1 were pre- sented with either a 2 or 5 sec inter-stimulus interval (ISI). The target stimulus tone had a constant probability of 20%. The 4 task conditions were presented in a counterbalanced fashion over subjects. Results P3 measurement All subjects were highly cooperative and pro- Pz Cz g / \ : 1 I / • I Fz "T 10 V Frequency - - - Easy Intensity . . . . . . . . . . Hard Intensity / m l I I I 0 150 300 450 600 Latency (msec) Fig. 1. Grand averaged ERPs (n = 16) illustrating the P3 com- ponents from the Fz, Cz, and Pz electrode sites for each discrimination task (frequency: 1000 Hz vs. 2000 Hz; easy: 40 vs• 60 dB; hard: 40 vs. 45 dB) condition (experiment 1). 314 J. POLICH duced very few artifacts. Task performance was virtually perfect for the frequency and easy inten- sity tasks with an average of 0.04 misperceived targets across experiments; the hard intensity task yielded an average of 1.4 misperceived targets, with no significant performance differences observed within or between the hard task condi- tions of the experiments. Wave forms from each electrode for each condition in each experiment were analyzed in the same fashion: the largest positive-going peak occurring for all 3 electrode sites after the N1-P2-N2 complex between 250 and 400 msec was designated as the P3 compo- nent. Amplitude was measured relative to the pre- stimulus baseline with peak latency defined as the time point of maximum positive amplitude. Analyses of variance on the ERP values were performed with and without the Geisser-Green- house procedures (Jennings and Wood 1976). Since no major differences between the two methods were obtained, the uncorrected statistical values are reported. In addition, only the P3 was analyzed in detail since it has been the major focus for clinical and applied studies even though some of the effects can be observed at other components as well. Experiment l The ERP wave forms for each task condition and recording position averaged over all subjects from the first experiment are illustrated in Fig. 1. The mean amplitudes and latencies + 1 standard 20- _i_ - - 15 - :>. Frequency Easy Hard Discrimination Task Frequency Easy Hard Fig. 2. Mean P3 amplitude and latency + 1 standard error from the Pz electrode site for each discrimination task condi- tion (experiment 1). error taken from the Pz electrode site are pre- sented in Fig. 2. Task discrimination difficulty affected both P3 amplitude and latency with the largest differences obtained for the hard intensity task compared to the frequency and easy intensity discrimination tasks. These effects appeared to occur in absence of significant latency jitter since the morphological peakedness for each task situa- tion did not differ dramatically between condi- tions. A 2-factor (task X electrode site) analysis of variance performed on the amplitude data con- firmed these impressions, with a significant effect for task, F (2, 30) = 23.8, P < 0.001, and electrode site, F (2, 30) = 10.9, P < 0.001. Separate one-way analyses comparing the task difficulty effects for each electrode supported this conclusion (P < 0.001). A 2-factor analysis of variance performed on the peak latency data also obtained strong effects for task difficulty, F (2, 30)= 20.1, P < 0.001, and electrode site, F (2, 30) = 5.1, P < 0.02. One-way analyses assessing the task difficulty manipulation for the Pz electrode site (see Fig. 2) demonstrated a significant effect for P3 ampli- tude, F (2, 30) = 16.7, P < 0.001, and latency, F (2, 30) = 22.7, P < 0.001. However, similar com- parisons of just the frequency and easy intensity tasks indicated that neither P3 amplitude nor latency differed significantly between these two stimulus conditions. Thus, the major effects o f task difficulty were obtained from the hard inten- sity task relative to the frequency and easy inten- sity discrimination conditions. Experiment 2 The ERP wave forms for each task difficulty condition, probability level, and recording site averaged over all subjects from the second experi- ment are illustrated in Fig. 3. The mean ampli- tudes and latencies from the Pz electrode site _+ 1 standard error are presented in Fig.:4. Task diffi- culty again reduced P3 amplitude and increased its latency, but these effects were differentially affected by target tone probability. The well- formed peaks of the grand averages suggest that P3 amplitude differences across conditions were caused by variations in task difficulty rather than latency variability. P300, PROBABILITY, TASK DIFFICULTY 315 Pz Cz ;°°% ,. °~ .°~ & Fz 10/aV Easy Hard 10% . . . . . . . . . . . . 30% I I I 0 150 300 450 600 Latency (msec) Fig. 3. Grand averaged ERPs (n = 16) illustrating the P3 com- ponents from the Fz, Cz, and Pz electrode sites for each task difficulty and target stimulus probability condition (experi- ment 2). A 3-factor analysis of variance (task difficulty x probability x electrode site) was performed on the P3 amplitude data obtained from each subject. Target stimulus probability produced a consistent decline in P3 amplitude for the easy but not the hard task condition which resulted in a significant interaction between the task difficulty and target probability variables, F (1, 15) = 11.7, P < 0.01. The overall effects of task difficulty, F (1, 15)= 19.5, P < 0.001, as well as electrode site, F (2, 30) = 23.1, P < 0.001, were also significant, as was their interaction, F (2, 30) = 5.5, P < 0.01. A 2- factor (task difficulty × probability) analysis of variance performed on the amplitude data from the Pz electrode site (see Fig. 4) confirmed the strong interaction between task difficulty and probability, F (1, 15) = 12.9, P < 0.005. The overall decline of P3 amplitude with increased task difficulty also was significant, with F (1, 15) = 26.5, P < 0.001. A marginal overall effect of stimulus probability obtained, F (1, 15) = 3.9, P = 0.07. The differential pattern of amplitude ef- fects for task difficulty and probability were relia- ble such that P3 amplitude from the easy dis- crimination task decreased with increases in target probability, while the hard intensity task showed little change as a function of probability. A 3-factor analysis of variance performed on the P3 latency data did not find a significant interaction between task difficulty and probabil- ity, although the hard task produced substantially longer latencies than the easy task, F (1, 15) = 64.0, P < 0.001. Latency also became shorter from the 15 5 0 I I 10 30 36(] 34O ¥ g 320 5 300 280 I ]0 J 3O Probability of Target Fig. 4. Mean P3 amplitude and latency 5:1 standard error from the Pz electrode site for each task difficulty and target stimulus probability condition (experiment 2). 316 J. POLICH f I t I Cz Pz Fz 10 V 1 Stan 4 Stan 10% ........ 30% I I I I 0 150 300 450 600 Latency (msec) Fig. 5. Grand averaged ERPs (n = 16) illustrating the P3 com- ponents from the Fz, Cz, and Pz electrode sites for 1- and 4-tone standard and target stimulus probability conditions (experiment 3). frontal to the parietal recording sites, F (2, 30) = 3.4, P < 0.05. A 2-factor analysis performed on the latency values from just the Pz electrode site demonstrated a marginally significant interaction between task difficulty and probability, F (1, 15) = 3.2, P < 0.10, with a strong main effect for task difficulty obtained, F (1, 15) = 67.4, P < 0.001. Thus, the interactions between task difficulty and target stimulus probability portrayed in Fig. 4 were reliable for the P3 amplitude data but only suggestive for the latency data. Experiment 3 The ERP wave forms for each task condition and probability level for each electrode site are illustrated in Fig. 5. The mean P3 amplitude and latency data + 1 standard error from the Pz elec- trode site are presented in Fig. 6. The effects obtained for variations in the number of standard stimulus items were small but consistent: condi- tions which contained 4 standard tones produced smaller P3 components than conditions which contained only 1 standard tone. P3 peak latency was not reliably changed by the number of stan- dard tones. Neither the amplitude nor latency data interacted with stimulus probability. Again, the uniform P3 peaks suggest that latency variabil- ity did not contribute significantly to the observed differences in P3 amplitude. A 3-factor (number of standards x probability x electrode) analysis of variance was performed on the amplitude values obtained from each sub- ject for each experimental condition. P3 amplitude was diminished significantly when 4 standard tones occurred relative to the 1 standard tone condi- tions, F (1, 15) = 12.2, P < 0.005, but only tended to decline with increases in target stimulus probability (F (1, 15) = 2.6, P > 0.10). No interac- tion between stimulus target probability and num- ber of standard stimuli was found (F "(17 luatu!~odxo) uo]l!puoo [eA.lalU~ snlnm!ls-aa~u ! pue gllnaUl!p ~ISgl qoga ~oj sal!s opoalaala z d pug 'z 3 'z d aql tuo~j sluouod -tuoa ~d OR1 guD~lsnlI ! (9[ = u) Sd~I~ po~e~a^e pue~ D "L "~H (aastu) /~aualel 009 oqt~ oo~ 0~t o ISI aas [ ............ I$1 aog 9 -- plw Ase:l AY/O t + .¢ ,..,. .., ,a~.. zq ,.! z 3 ..-.. z d -de S~A~ s~oloel osoql uoo~loq uo~.log~olu~ ug jo lu.~q ou '~OAO~O H "luouodmoo £d oql aoJ £auole I u! soseoaau.~ pug opnl.qdtug u! sosgo~aop Su.rplo.~£ u~gge ~Isel uo~.leu.tm~os!p lino~.jj~p o~otu oql ql!~ 'ISI oos g oql ueql so!ouole I xoSuo I pue sopnu.Idtue ~og~g I oonpo~d ot popuol ISI aos g oq~L "8 "$.H u~ poluosoxd oag sloo.~qns sso~og polndmoo ~o~o p~gp -u~ls [ T SOnlga £OUalt3I pug opnl.qdtm3 £d u13o~ oq& "L "$!d u! pol~snII ! o~g ISI pug uoD.ipuoa £1Ino.u~!p ~ISm qago soj smJoj OAE~X cl~El Oq~L p luam!.tadx~t "£1![.~qeqoad sn[ntm.ls loflam ql.~t~ poloe~olu! io ~oj sloajjo Suo~ls pal~lsuomap oanseatu Ed ~aR1DN "palaajjgun £[agr.lglaI se~ £aualg I £d lnq £iqe.qaa paseaaaap se~ apm.qdtue £d '[ ol pa~ed -moa 17 se~ sauol p~gpuels 'laSael-uou jo ~aqmnu aql UatlA ~ "luva!j.m$!s a~aAx slaajJa aaqlo o N "1;g0"0 = d '17"17 = (gI 7) d 'suo!upuoa ssoaae aaualgl £d u! sosea~aap aanpo~d ol popual snlntu!ls la$~el aql jo ~l!i!qgqo~d u! sos~o~ou.t leql polga!pu! ol.~s apo~lao[a z d aql tuo~j elep £ouolgI oql uo pomJoj -~od aOUeUgA JO s.~sK[gug ~oloeJ- g V "luga.U.mg!s se~ uo.~lag~olu! ~oqlo o N "I0"0 > d 'L'g = (0£ 'g) at" 'uo.~l.[sod opoalaOlO ql.~ £[lugorj.tu$ls poloeJolu~ £l![.~qgqo~d snlntu.~ls log~gl qgnoqlI e '£1.q~.qgqoad ~o stuol.~ snlntm.ls jo aoqtunu ~oql!o aoj sloojjo u~m olqe.qoJ ou polgJlsuotuop elgp gouole I ~d oql uo patu~oj~ad aauguea lo s.~sKleue ~olaej-g V "punoj se~ salqe.ueA o~1 asaql uaa~laq uoDae~alu ! ue ~o Kl.q.~qeqo~d snlntu!ls la~gl ~oj laajja [Ie~aAO OU "(~ luatu!~odxa) suo!l!puoa ~l~[!qeqo2d snlntu!ls laSJel pug p~puels auol-t; pu~ -I aq~ ~oj alSs apo~laala Zd aql tuo~j Joppa plepuels l + p~z~ daualgi pu~ apnl!ldtu~ ~d U~al~ "9 "$!d laOJe I ~o A~!l!qeqm d o~ Ot O~ I 08~ I 00~ OZg ~ Ot [ 0 ~ K gt 09~ OZ L~[ XI]FIDIzlJIO 2'ISVi 'X~LI'-II~[VflOHd '0Old 318 J. POLICH 20 15 A m P300, PROBABILITY, TASK DIFFICULTY 319 (Donchin 1981), P3 amplitude is reduced since the system's processing capacity is depleted by the increased numbers of target items presented in the same amount of time. In the same fashion, task situations which manipulate central processing op- erations, such as memory requirements or the amount of time between stimulus presentations (Woods et al. 1980; Fitzgerald and Picton 1981; Kramer et al. 1986), also change P3 amplitude but are independent of changes in target probability since such factors do not affect perceptual processing capacity. Thus, only when task diffi- culty influences the perceptual quality but not the central, memory-based processing of stimulus in- formation does target probability interact with task difficulty. The major implication of these findings for the application of P3 ERP paradigms in the study of individual variation is that different task situa- tions can interact with processing capacity de- mands in specific ways. Variables which affect stimulus encoding in addition to target stimulus probability may yield population differences re- flecting input rather than central mental processes. Hence, comparison of populations across such tasks should take into account processing dimen- sions which may differentially affect both P3 am- plitude and latency. This problem has been en- countered previously in the ERP aging literature wherein stimulus encoding manipulations have produced larger P3 latency changes for aged com- pared to young subjects (K.C. Squires et al. 1980). In contrast, sequence effects from the presentation order of standard and target stimuli - - stemming from primary memory processes (K. Squires et al. 1976) - - have not shown any distinct differences for P3 amplitude between aged and young sub- ject~s (Ford et al. 1982). These two findings can be reconciled by assumming that perceptual encoding and central mnestic events affect age-related changes for P3 values differently. Hence, P3 am- plitude and latency may reflect population dif- ferences to the degree that they are determined by the nature of the task parameters in conjunction with stimulus probability. The interaction of these factors, however, is an important consideration when designing task situations which may be ap- plied to differentiate subject groups since the source of the specific P3 effects appears to originate at different information processing loci. References Banquet, J.P., Renault, B. and Lesrvre, N. Effects of task and stimulus probability on evoked potentials. Biol. Psychol., 1981, 13: 203-214. Baribeau-Braun, J., Picton, T.W. and Gosselin, J-Y. Schizo- phrenia: a neurophysiological evaluation of abnormal in- formation processing. Science, 1983, 219: 874-876. Brookhuis, K.A., Mulder, G., Mulder, L.J.M. and Gloerich, A.B.M. The P3 complexes as an index of information processing: the effects of response probability. Biol. Psy- chol., 1983, 17: 277-296. Brown, W., Marsh, J. and LaRue, A. Exponential electrophysi- ological aging: P300 latency. Electroenceph. clin. Neuro- physiol., 1983, 55: 277-285. Campbell, K.B., Courchesne, E., Picton, T.W. and Squires, K.C. Evoked potential correlates of human information processing. Biol. Psychol., 1979, 8: 45-68. Donchin, E. Surprise!... Surprise?. Psychophysiology, 1981, 18: 493-513. Duncan-Johnson, C.C. and Donchin, E. On quantifying surprise: the variation in event-related potentials with sub- jective probability. Psychophysiology, 1977, 14: 456-467. Duncan-Johnson, C.C. and Donchin, E. The P300 component of the event-related brain potential as an index of informa- tion processing. Biol. Psychol., 1982, 14: 1-52. Finley, W.W., Faux, S.F., Hutcheson, J. and Amstutz, L. Long-latency event-related potentials in the evaluation of cognitive function in children. Neurology (NY), 1985, 35: 323-327. Fitzgerald, P.G. and Picton, T.W. Temporal and sequential probability in evoked potential studies, Canad. J. Psychol., 1981, 35: 188-200. Ford, J.M., Duncan-Johnson, C.C., Pfefferbaum, A. and Kopell, B.S. Expectancy for events in old age: stimulus sequence effects in P300 and reaction time. J. Gerontol., 1982, 37: 696-704. Goodin, D., Squires, K.C., Henderson, B. and Start, A. Age-re- lated variations in evoked potentials to auditory stimuli in normal human subjects. Electroenceph. clin. Neurophysiol., 1978, 44: 447-478. Goodin, D.S., Squires, K.C. and Start, A. Variations in early and late event-related components of the auditory evoked potential with task difficulty. Electroenceph. clin. Neuro- physiol., 1983, 55: 680-686. Hoffman, J.E., Houck, M., MacMillan, F. and Simons, R.F. Event-related potentials elicited by automatic targets: a dual-task analysis. J. exp. Psychol.: Hum. Percept. Perform., 1985, 11: 50-61. Howard, L. and Polich, J. P300 latency and memory span development. Develop. Psychol., 1985, 21: 283-289. Isreal, J.B., Chesney, G.L., Wickens, C.D. and Donchin, E. 320 J. POLICH P300 and tracking difficulty. Evidence for multiple re- sources in dual task performance. Psychophysiology, 1980a, 17: 5770. Isreal, J.B., Wickens, C.D., Chesney, G. and Donchin, E. The event-related brain potential as an index of display-moni- toring workload. Hum. Fact., 1980b, 22: 211-224. Jennings, J.R. and Wood, C.C. The c-adjustment procedure for repeated measures analysis of variance. Psychophysiology, 1976, 13: 277-278. Johnson, R. and Donchin, E. On how P300 amplitude varies with the utility of the eliciting stimuli. Electroenceph. clin. Neurophysiol., 1978, 44: 424-437. Johnson, R. and Donchin, E. P300 stimulus categorization: two plus one is not so different from one plus two. Psycho- physiology, 1980, 17: 167-178. Kok, A. and Looren de Jong, H. Components of the event-re- lated potential following degraded and undegraded visual stimuli. Biol. Psychol., 1980, 11: 117-133. Kramer, A., Wickens, C.D. and Donchin, E. Processing prop- erties: evidence for dual-task integrality. J. exp. Psychol.: Hum. Percept. Perform., 1985, 11: 393-408. Kramer, A., Schneider, W., Fisk, A. and Donchin, E. The effects of practice and task structure on components of the event-related brain potential. Psychophysiology, 1986, 23: 33-42. Lai, J.A., Brown, W.S., Marsh, J.T. and LaRue, A. Covariation of P3 latency and mini-mental state scores in geriatric patients. Psychophysiology, 1983, 20: 455. Magliero, A., Bashore, T.R., Coles, M.G.H. and Donchin, E. On the dependence of P300 latency on stimulus evaluation processes. Psychophysiology, 1984, 21: 121-186. Miller, J. Cognitive influences in perceptual processing. J. exp. Psychol.: Hum. Percept. Perform., 1979, 5: 546-562. Miller, J. and Pachella, R. On the locus of the stimulus probability effect. J. exp. Psychol., 1973, 101: 227-231. Pachella, R. and Miller, J. Stimulus probability and same-dif- ferent classification. Percept. Psychophys., 1976, 19: 29-34. Perrault, N. and Picton, T.W. Event-related potentials re- corded from the scalp and nasopharynx. II. N2, P3, and slow wave. Electroenceph. clin. Neurophysiol., 1984, 59: 261 --278. Pfefferbaum, A., Ford, J.M., Roth, W.T. and Kopell, B.S. Age-related changes in auditory event-related potentials. Electroenceph. clin. Neurophysiol., 1980, 49: 266-276. Pfefferbaum, A., Ford, J., Johnson, R., Wenegrat, B. and Kopell, B.S. Manipulation of P3 latency: speed or accuracy instructions. Electroenceph. clin. Neurophysiol., 1983, 55: 188-197. Pfefferbaum, A., Ford, J.M., Wenegrat, B.G., Roth, W.T. and Kopell, B.C. Clinical applications of P3. I. Normal aging. Electroenceph. clin. Neurophysiol., 1984a, 59: 85-103. Pfefferbaum, A., Wenegrat, B.G., Ford, J.M., Roth, W.T. and Kopell, B.S. Clinical application of the P3 component of event-related potentials. 1I. Dementia, depression, and schizophrenia. Electroenceph. clin. Neurophysiol., 1984b, 59: 104-124. Picton, T.W., Stuss, D.T., Champagne, S.C. and Nelson, R.F. The effects of age on human event-related potentials. Psy- chophysiology, 1984, 21: 312-326. Polich, J. Attention, probability, and task demands as determi- nants of P300 latency from auditory stimuli. Electroenceph. clin. Neurophysiol., 1986, 63: 251-259. Polich, J., Howard, L. and Starr, A. Effects of aging on the P300 component of the event-related potential from audi- tory stimuli: peak definition, variation, and measurement. J. Gerontol., 1985a, 40: 721-726. Polich, J., Howard, L. and Start, A. Stimulus frequency and masking as determinants of P300 latency in event-related potentials from auditory stimuli. Biol. Psychol., 1985b, 21: 309-318. Polich, J., Ehlers, C.L., Otis, S., Mandell, A.J. and Bloom, F.E. P300 latency reflects the degree of cognitive decline in dementing illness. Electroenceph. clin. Neurophysiol., 1986, 63: 138-144. Ritter, W., Simson, R. and Vaughan, H.G. Event-related potential correlates of two stages of information processing in physical and semantic discrimination tasks. Psychophysi- ology, 1983, 20: 168-179. Roth, W.T., Pfefferbaum, A., Horvath, T.B., Berger, P.A. and Kopell, B.S. P3 reduction in auditory evoked potentials of schizophrenics. Electroenceph. clin. Neurophysiol., 1980, 49: 447-505. Squires, K.C., Wickens, C., Squires, N. and Donchin, E. The effect of stimulus sequence on the waveform of the cortical event-related potential. Science, 1976, 193: 1142-1146. Squires, K.C., Chippendale, T.J., Wrege, K.S., Goodin, D.S. and Starr, A. Electrophysiological assessment of mental function in aging and dementia. In: L.W. Poon (Ed.), Aging in the 1980's: Selected Contemporary Issues in the Psychology of Aging. American Psychological Association, Washington, DC, 1980: 125-134. Squires, N., Donchin, E., Squires, K. and Grossberg, S. Bi- sensory stimulation: inferring decision related processes from the P300 component. J. exp. Psychol.: Hum. Percept. Perform., 1977, 3: 229-315. Sutton, S. P300 - - thirteen years later. In: H. Begleiter (Ed.), Evoked Brain Potentials and Behavior. Plenum Press, New York, 1979: 107-126. Tueting, P., Sutton, S. and Zubin, J. Quantitative evoked potential correlates of the probability of events. Psycho- physiology, 1971, 7: 385-394. Wickens, C., Kxamer, A., Vanesse, L. and Donchin, E. The performance of concurrent tasks: a psychophysiological analysis of the reciprocity of information processing re- sources. Science, 1983, 221: 1080-1082. Woods, D.L., Courchesne, E., Hillyard, S.A. and Galambos, R. Recovery cycles of event-related potentials in multiple de- tection tasks. Electroenceph. clin. Neurophysiol., 1980, 50: 335-347.