Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The effects of task complexity and response probability on response latency Ryan, Mark William John 1972

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1972_A7_5 R93.pdf [ 2.92MB ]
Metadata
JSON: 831-1.0077218.json
JSON-LD: 831-1.0077218-ld.json
RDF/XML (Pretty): 831-1.0077218-rdf.xml
RDF/JSON: 831-1.0077218-rdf.json
Turtle: 831-1.0077218-turtle.txt
N-Triples: 831-1.0077218-rdf-ntriples.txt
Original Record: 831-1.0077218-source.json
Full Text
831-1.0077218-fulltext.txt
Citation
831-1.0077218.ris

Full Text

THE EFFECTS OF TASK COMPLEXITY AND RESPONSE PROBABILITY ON RESPONSE LATENCY BY MARK WILLIAM JOHN RYAN B.P.E., The University of B r i t i s h Columbia A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF PHYSICAL EDUCATION In the School of Physical Education and Recreation We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA June, 1972 In present ing th i s thes i s in p a r t i a l f u l f i lment o f the requirements fo r an advanced degree at the Un iver s i t y of B r i t i s h Columbia, I agree that the L ibrary sha l l make i t f r ee l y ava i l ab le for reference and study. I f u r ther agree that permission for extens ive copying of th i s thes i s fo r s cho la r l y purposes may be granted by the Head of my Department or by his representat ives . It is understood that copying or pub l i c a t i on of th i s thes i s fo r f i nanc i a l gain sha l l not be allowed without my wr i t ten permiss ion. Department The Un ivers i ty of B r i t i s h Columbia Vancouver 8, Canada ABSTRACT The purpose of the investigation was to study the joint effects of response probability and task complexity on response latency i n simple and choice reaction time tasks. Sixteen, volunteer, University of B r i t i s h Columbia. Physical Education students performed a l l four experimental conditions, one simple reaction time task and three choice reaction time tasks. Response latencies for 800 t r i a l s were obtained from each subject. Analysis of variance for a repeated measures design was used to analyse the data, with Harter and Newman-Keuls post-hoc multiple comparisons performed to test for significant differences between conditions. The empirical findings were generally not in agree-ment with the predicted hypotheses. It was found that very fast response latencies were made to complex tasks combined with low probability of occurrence. Subjects were also better prepared to i n i t i a t e a response to an easy task with a low response probability than an easy task with a high response probability. - It appears that subjects adopted a defensive type of strategy and that low response prob-a b i l i t y has more effect than high response probability on response latencies of choice reaction time tasks of unequal complexity. i i i TABLE OF CONTENTS Chapter Page 1 STATEMENT OF THE PROBLEM 1 Introduction 1 The Problem 3 Definition of Terms . . . . . 3 Delimitations . . . . . . . 4 Limitations 5 Assumption 5 H y p o t h e s e s 5 Significance . . . . 6 2 REVIEW OF RELATED LITERATURE 7 Introduction 7 Empirical Factors Affecting Re- . . . . 8 sponse Latency Task Complexity . 8 Response Probability 10 Other Factors Affecting CRT . . . . 11 Models of Information Processing . . . 13 Welford 13 Falmagne and Theios 13 Schutz ; . . 14 Theoretical Rationale for Hypotheses . 18 Hypothesis 1 18 Hypothesis 2 . . . . . 18 Hypothesis 3 19 3 METHODS AND PROCEDURES 20 Design 20 Subjects 20 Apparatus 21 Experimental Conditions . . . . . . . 21 Condition A .21 Condition B . . . . . . . . . . . 21 Condition € e 21 Condition D . . . . . 21 Procedure . . . . . . . . 23 S t a t i s t i c a l Analysis . . . 24 i v Chapter ' Page 4- RESULTS AND DISCUSSION 26 Results 26 Discussion . . . 3 3 Condition A 3 3 Condition B 36 Condition C and D 3 8 Summary 4-2 5 SUMMARY AND CONCLUSIONS 4-3 Summary 4-3 Conclusions 4-4-Suggestions for Further Research . . . 4-5 BIBLIOGRAPHY 4-6 V LIST OF TABLES Table Page I Mean Response Latencies, Standard Deviations and Coefficients of Variation (in msec.) for a l l Conditions 27 II Means and Differences (in msec.) for Main Effects and Interactions Arranged According to Experimental Conditions . . 27 III Analysis of Variance for the Effect of Experimental Conditions and Task Complexity on Response Latencies . . . . 28 IV Newman-Keuls Post-Hoc Multiple Com-parisons on Condition Means . . . . . . 28 V Hafter Post-Hoc Analysis on Cell Means 29 VI Response Probability of Experimental Conditions 30 VII Cell Means, Grand Means and Difference (in msec.) Arranged According to Response Probability 31 VIII Analysis of Variance for Conditions and Task Complexity Arranged by Response Probability 31 IX Harter Post-Hoc Analysis on Inter-action Means 32 X Hypothetical Time Components of Total Response Latency 35 v i LIST 01 FIGURES Figure Page 1 Hypothesized Task Complexity by Response Probability Interaction 6 2 Schutz' Model of Information Pro-cessing Components i n a Choice Reaction Time Task . 1 7 3 Experimental Design . . . . 2© 4- Arrangement of Stimulus Lights, Response Key, Warning Light, Obstacle and Targets 22 5 Task Complexity by Experimental Condition Interaction 37 6 Task Complexity by Response Probability Interaction 4-1 v i i ACKNOWLEDGMENT To the members of my committee, Dr. R. W. Schutz, Dr. R. G. Marteniuk, Dr. K. D. Coutts, and Dr. W. G. Davenport, I would l i k e to express my appreciation for their guidance throughout the preparation of this study. A very special thanks to my advisor, Dr. Schutz, for his continued assistance and encouragement during my academic program. CHAPTER 1 STATEMENT OP THE PROBLEM Introduction In the l a s t ten years there has been a great deal of research done on man's cen t r a l mechanisms, but to date l i t t l e i s known about the exact nature of these inform-ation processing mechanisms or stages. Most of the s k i l l s man performs can be thought of as information processing s k i l l s . Information i s received from the environment and then coded by man's sense organs i n t o patterns.of neural e x c i t a t i o n which are then stored. A f t e r a number of r e p e t i t i o n s , depending upon the nature of the task, l e a r n i n g occurs and r e s u l t s i n patterns of overt behavior. An information processing s k i l l frequently employed i n studying memory i s the choice r e a c t i o n time task, i n which the delay between the occurrence of a stimulus and the i n i t i -a t ion of a response to the stimulus i s c a l l e d the response latency. A l l tasks which involve man's taking information from the environment and responding to i t show some f i n i t e response latency. During t h i s f i n i t e period of time the sub-j e c t must search through various memory storage areas, r e-t r i e v e the desired response and i n i t i a t e i t . A considerable part of t h i s time i s occupied by the memory search process. In order to gain i n s i g h t i n t o the process of memory search -1-and re t r i e v a l , investigators have u t i l i z e d a number of ex-perimental variables. Some of the factors which affect response latency are: stimulus-response compatibility, v a r i a b i l i t y of the foreperiod, i n t e r - t r i a l interval, number of alternatives, practice, repetition of stimulus or re-sponse, stimulus-response probability and task complexity. The influence each of these has on response latency i s not clearly understood. To date l i t t l e work has been done i n the area of task complexity and response probability and their effect on response latency. It i s generally accepted that a complex task has a longer simple reaction time (SRT) than an easy task. However, i n a choice reaction time (CRT) situation with equal response probabilities the difference i n response latencies between a complex and an easy task decreases (Schutz, 1 9 7 2 ) . This difference i s apparently due to the joint effects of task complexity and response probability but as yet there i s no empirical evidence to support or refute this claim and the effects are not clearly understood. The form i n which the storage and retrieval of learned motor tasks take place i s very controversial. Most research-ers feel that memory storage i s divided into two areas; a short term storage area of limited capacity but high access-i b i l i t y , and a long term store of unlimited capacity but less accessibility (Atkinson, 1 9 7 0 ) . Others feel that man's memory i s composed of three storage areas (Falmagne and Theios, 1968 ; Schutz., 1 9 7 2 ) . These models propose a short - 3 -term and a long term storage area as well as an immediate memory or s e l e c t i v e attention store. This l a t t e r storage area has a capacity of only one response program, which i s us u a l l y that program which the i n d i v i d u a l : f e e l s w i l l he h i s next required response. The reason f o r t h i s l a c k of consensus on the nature of motor memory storage and re-t r i e v a l may be p a r t l y due to the d i v e r s i f i e d methods of i n -v e s t i g a t i o n used i n studying t h i s phenomena. A number of theories and models have been proposed on response latency and the facto r s a f f e c t i n g response latency, but none of these have been strongly supported with empirical findings (Henry and Rogers, I960; Falmagne , and Theios, 1968; Schutz, 1972). I f a theory can be proven tenable then t h i s w i l l a i d i n understanding the process of learning and performance as well as give d i r e c t i o n to further research. This paper w i l l attempt to add support to a current theory of information processing and response r e t r i e v a l , by examining the degree to which i t can account f o r the j o i n t e f f e c t s of response p r o b a b i l i t y and task complexity. The Problem The purpose of t h i s i n v e s t i g a t i o n was to draw i n f e r -ences about the motor memory search process by studying the j o i n t e f f e c t s of response p r o b a b i l i t y and task complexity on response latency i n simple and choice reaction time tasks. D e f i n i t i o n of Terms (1) Response Latency (RL) - analogous to the term reaction time. I t i s the time between the presentation of a stimulus and the i n i t i a t i o n of a response. i ) Simple reaction time (SRT) - response l a t -ency to a task involving one stimulus and one response, i i ) Choice reaction time (CRT) - response l a t -ency to a task involving more than one stimulus and more than one response. (2) Task Complexity (TC) - a subjective measure of the d i f f i c u l t y i n i n i t i a t i n g a response to a specific task. i ) Type I task - a task that requires a simple response program involving few subprograms, i i ) Type II task - a task that requires a complex response program (complex relative to a Type I task) involving many subprograms. ( 3 ) Response Probability (RP) - the preset, objective probability of the occurrence of a specific response. Delimitations (1) The results can only be generalized to the two levels of task complexity studied, as i t i s impossible to cl a s s i f y other tasks on this complexity scale without RLs for them. (2) The results can only be generalized to a two-choice condition of choice reaction time. (3 ) The subjects used i n this study are available volunteers from summer school Physical Education classes, and thus inferences to a larger population must be made on log-i c a l , and not s t a t i s t i c a l grounds. Limitations (1) The study i s limited to the accuracy of the equipment involved. ( 2 ) It i s not known how errors w i l l influence the results. Assumption (1) The study assumes a model of information pro-cessing as outlined below. Hypotheses (1) Simple reaction time i s greater for a Type II task than for a Type I task. (2) In a choice reaction time situation with equal response probabilities, response latency i s greater for a Type II task than for a Type I task, this difference being less than i n the simple reaction time situation (i . e . , c<a, Figure 1). (3) In a choice reaction time situation with varying response probabilities, the difference i n choice reaction times due to task complexity i s an increasing function of response probability. (See Figure 1 for the specific nature of this relationship, i.e., d<c<b<a). R E S P 0 N S E A T . bO fl •H CQ cd o H E (msec.) N C Y -6-.25 .50^  .75 RESPONSE PROBABILITY II I 1.0 Figure 1: Hypothesised Response P r o b a b i l i t y by Task Complexity I n t e r a c t i o n S i g n i f i c a n c e In order to understand motor behavior and to advance the knowledge of how motor acts are i n i t i a t e d i t i s necessary to e s t a b l i s h t e s t a b l e t h e o r i e s . One such theory of inform-ation processing and memory r e t r i e v a l i s that proposed by Schutz (1972). This theory does not f u l l y explain the j o i n t e f f e c t s of RP x TC i n t e r a c t i o n . The s i g n i f i c a n c e of t h i s study i s to confirm or r e j e c t Schutz' fin d i n g s on RP x TC i n t e r a c t i o n and examine t h i s r e l a t i o n s h i p i n greater depth. A confirmation of Schutz' fin d i n g s w i l l point out the need f o r t h e o r e t i c a l r e v i s i o n , by providing support f o r h i s t h e o r e t i c a l l y i n v a l i d r e s u l t s . CHAPTER II REVIEW GF RELATED LITERATURE Introduction The earliest attempt at a theory of CRT i s embodied in the assumptions underlying the subtraction method of Donders (1868). He assumed that total response latency i s composed of a number of successive, p a r t i a l l y independent processes, with the time for each process being additive. The subtraction method involves comparing the reaction times from two different tasks, the difference between the time for the two tasks being a measure of the time taken for the cognitive process involved i n one task which was not part of the other task. It i s assumed that the task with the longer reaction time requires a l l the processes of the other, plus some additional process. Donders took the difference between CRT and SRT and assumed i t to be the time required to discriminate the stimulus and choose the correct response. I n i t i a l l y his results were encouraging, however, around the turn of the century part of his work was attacked and to a great extent discredited due to a lack of empirical support. In the last 10 to 15 years there has been a renewed interest in memory, reaction time and information processing, and i t i s now perhaps the most researched topic i n experimental psychology. -8-Boyko (1964-) has done a very extensive review on experimental variables which affect RT. His review delved into the type of stimulus used: auditory, visual or t a c t i l e ; the intensity of the stimulus; the effect of alcohol, drugs and lack of sleep; the effect of varying foreperiods; the effect of repeated stimuli; the effect of choice (number of alternatives); the effect of age and the effect of i n -structions on RL. Recently such variables as S-R compat-i b i l i t y , motivation, fatigue, anxiety, i n t e r - t r i a l interval (ITI), practice, S-R probability and task complexity, have been shown to be important. There i s an obvious inter-dependence between the variables cited above, but due to the enormous amount of research done i n this area i t w i l l be necessary to r e s t r i c t this review to the variables being studied i n this paper. This review w i l l therefore be con-cerned with recent literature on task complexity, response probability and several information processing models. Empirical Factors Affecting Response Latency Task complexity. The effects of task complexity on response latency have been studied by a number of researchers but their findings are not a l l i n agreement. Henry and Rogers (I960) hypothesized a longer reaction latency for a complicated movement than for a simpler move-ment. This i s because a more comprehensive program, i . e . a larger amount of stored information w i l l be needed, and thus the neural impulses w i l l require more time for co-ordination and direction into the eventual motor neurons - 9 -and muscles. A complicated movement necessarily involves several muscle groups and several specific areas of neuromotor co-ordination centers; thus more extensive use of learned and stored neuromotor patterns are required to i n i t i a t e the overt motor response. They used three move-ments, A, B and C i n order of increasing complexity. They found that response latencies increased as the movement became more complex. The reaction preceding movement B was about 20% slower than the RT for A ( 195 msec, and 163 msec, respectively) and the RT preceding movement C (208 msec.) was about 7% slower than the RT for B. They found these differences to be significant. However, in a subsequent study Henry ( 1961) found that movement complexity had no influence on RT under four conditions of.movement studied. Sidowski et a l . (1958) also found that increased complexity of the task resulted i n longer RTs. They found a significant interaction effect between type of reaction and task complexity. They stated that the CRT was more affected by response complexity than was the SRT. However, this statement does not agree with their finding. The graph clearly indicates the SRT i s more affected by task complexity. Griew ( 1959 ) found that the complexity of response had no significant effects on a group of young subjects (20-26 years of age), but did have a significant effect on a group of older subjects ( 5 0 - 59 years of age). He said this may be due to a breakdown with age in the a b i l i t y to ' monitor one movement while preparing for another. - 1 0 -B l a i r ( 1 9 6 8 ) using three conditions of movement com-plexity also found that movement complexity had no sig-nificant effect on RT. His RTs did increase as complexity of movement increased ( 3 7 2 msec . -393 msec.) but the change was declared non-significant. He says that i t i s possible that his study lacks sensitivity and that a Type II error was made. Response probability. The majority of studies done on probability have examined stimulus probability {P ( S ) ] whereas this paper i s concerned with response probability {p(R)j . However, as P ( S ) = P(R) i n most cases, this i s not too important. There i s general agreement of the papers reviewed on response probability that the higher the response probability the faster the reaction time (Bertelson, 1 9 6 1 ; Laberge and Tweedy, 1 9 6 4 - ; Bertelson, 1 9 6 5 ; Bertelson and Barzeele, 1 9 6 5 ; Bertelson and Tisseyre, 1 9 6 6 ; Leonard et a l . , 1 9 6 6 ; Kornblum, 1 9 6 7 ; and Remington, 1 9 6 9 ) . It should be mentioned that i n some of these studies i t i s d i f f i c u l t to t e l l which i n -creased probability led to the faster RT - the stimulus or the response. It i s obvious that an increased stimulus probability i n a 1 :1 mapping of S-R pairs w i l l result i n an increased response probability. Only a few researchers have used unequal S and R probabilities i n an attempt to determine which probability has a greater effect. Laberge and Tweedy (1964- ) used 3 stimuli and 2 re-sponses i n a 1 : 1 , 2 : 1 S-R mapping experiment i n an attempt -11-to separate stimulus-response probability. The 1:1 mapping gave an RP of .4, and the 2:1 mapping gave a RP of ^6 - each stimulus with a stimulus probability of .33. They concluded that the response component had a stronger effect than the stimulus component. Bertelson (1965) also used a task where more than one signal i s associated with each response. In such a situation, the relationship of one t r i a l to the preceding t r i a l can be one of: identity (same signal - same response), equivalence (different signal but same response), or d i f -ference (different signal - different response). He found that the repetition of the signal, apart from that of the response, can exert some effect. However, i n this situation the main effect i s linked to the repetition of the response. Bertelson and Tisseyre (1966) found just the opposite results to those of Laberge and Tweedy (1964) and Bertelson (1965). They used four stimuli of relative frequencies of .55 - . 1 5 - .15 - .15 and two responses, each corresponding to two stimuli. They said in this situation the relative frequency of the response does not affect the RT, and that the only c r i t i c a l variable i s the relative frequency of the stimulus. Other factors affecting CRT. There does not appear to be any literature to date related to the TC by RP inter-action. However, as this i s of primary concern i n this study, i t i s necessary to discuss a few factors which may affect this interaction. -12-The repetition effect (RE) refers to the fact that i n a se r i a l choice reaction time task, responses to a signal are faster when the signal i s a repetition of the preceding signal than when the signal changes. Thus by increasing the RP of a given response, this w i l l necessarily involve more repetitions of the response which could i n turn he responsible for a decreased RT. The exact mechanism of the RE i s unknown. Thus i t i s possible that the RE could i n -fluence different levels of task complexity much i n the same way i t affects different levels of S-R compatibility, i . e . , non-compatible S-R pairs show a stronger RE than compatible S-R pairs.(Bertelson, 1961). Norrie (1967) showed that movements to complex tasks show a greater practice effect or shortening of PL's than do simple tasks. She found that the complex movement group continued to show improvement in RT throughout the experiment while the performance curve for the simple movement group leveled off during the f i r s t 20 t r i a l s . Thus i f Type II had a greater R'L, i t i s possible that there may be greater practice effects on II than on I... -Sidowski et a l . (1958) i n studying the influence of task complexity and instructions, found that instructions of either making as fast a RT as possible (group I) or making as fast a movement time as possible (group II) had no sig-nificant difference on mean RTs for the two groups. It i s well known that instructions emphasizing speed rather than accuracy w i l l decrease RT but increase the number of errors - 1 3 -and vice versa. It may be possible that a greater subjective emphasis on accuracy would be placed on Type II task because of i t s complexity, thus increasing the RT. Models of Information Processing Welford. Welford (i960) proposed a single channel hypothesis of information processing. He states that the central mechanisms do not act as a single whole, but as a chain with at least three links. Sense organs receive stimuli from the stimulus display and convert them into patterns of nerve impulses. These are relayed to a perceptual  mechanism i n which integration and iden t i f i c a t i o n take place. The next l i n k i s termed the translation mechanism which i s concerned with the choice of action i n relation to what i s perceived. The translation mechanism then passes 'orders' to the central effector mechanism which carries out chains of coordinated actions by means of the effector organs. The manner i n which signals pass from one stage to the next i s not as yet understood, but the evidence suggests that a feedback loop from the effector side controls the passage of data from the perceptual to the translation stages, allowing through data from a new signal once the action i n response to the previous signal has begun. Palmagne and Theios. The original work on the model to be used i n this study was done by Palmagne and Theios (1968). They assumed that when a stimulus i s presented, the template corresponding to that stimulus can be i n a selective attention (SA) state, immediate memory (IM) state or i n a long term memory (LTM) state. The difference between SA and IM i s that i n SA there i s exactly one template while i n IM there can be several or none. They state that when a stimulus i s presented, i t takes a time t-p after which an internal representation (IR) of the stimulus i s created. The Sf then checks to see whether the internal representation of the stimulus presented, matches the template i n the selective attention state. This takes a time t2» I f there i s a match, the response appears after an additional motor time, t^. The total response latency i s then: If the IR i s not i n SA, the S looks into immediate memory to see whether the representation of the stimulus matches any of the templates. This takes a time t^. I f there i s a match, the response appears after a time t^, the same motor time as before. The total response latency i s then: t = t-^  + t2 + t^ + t ^ If the IR i s not i n IM, the S looks into LTM. This takes a time t^. With the same motor time as before the total re-sponse latency i s : t — t^ + tg + ... t ^ Schutz. Schutz (1972) modified this model and pro-posed a response oriented theory i n which most of the v a r i -a b i l i t y i n response latency can be accounted for by the retrieval and organization of the memory response engram. His basic assumption i s that there are a number of stages involved i n information processing and response organ-i z a t i o n i n a CRT task, and that these stages act independ-ent l y of each other with the t o t a l CRT being the summation of the time f o r each separate stage. He also assumed that a S learns to make a s p e c i f i c type of response by organ-i z i n g a number of already learned and stored motor sub-programs into one s p e c i f i c response program which can then be stored as a u n i t i t s e l f . The model can be discussed most e f f e c t i v e l y by considering i t i n terms of the three basic processes involved; storage, search and r e t r i e v a l . (a) Memory Storage Areas. Schutz proposed that the search and r e t r i e v a l process examines up to three memory storage areas. The f i r s t area examined i s s e l e c t i v e a t t e n t i o n (SA). I t has a capacity of one engram and has very short r e t e n t i o n . SA u s u a l l y contains an event very r e c e n t l y exper-ienced or a response expected to be c a l l e d upon i n the immediate future. In the usual CRT task a S w i l l place i n SA that program which has the highest subjective p r o b a b i l i t y of being the required response to the next stimulus. The second memory storage area examined i s primary memory (PM). I t has a capacity of about 5 - 1 0 engrams, being a function of the complexity of the tasks. PM has l i m i t e d r e t e n t i o n . The programs stored here are quite accessible but not as accessible as a program i n SA. The t h i r d area i s secondary memory (SM). I t has an unlimited capacity with permanent storage. Programs stored -16-i n SM are less accessible than programs stored i n either SA or PM. (b) Memory Search Process. The memory search pro-cess for a specific response program i n PM i s assumed to be a se r i a l exhaustive search, examining each available re-sponse i n the storage area. On completion of the scanning process the sought after response i s retrieved and executed. (c) Response Retrieval Process i ) SRT. The model of information processing and response retrieval mechanisms i n SRT i s composed of a stimulus perception stage, response search and release stage (which i s re a l l y just a release, because with only one response there i s no search as the response i s i n SA with a probability of 1.0), and a response action stage. This model i s similar to that proposed by Palmagne and Theios (1968) with each stage taking a f i n i t e amount of time and the summation of the stages equalling SRT. i i ) CRT. In a CRT task two additional stages are required i n e l i c i t i n g the correct response to a given stimulus, the stimulus categorization stage and the response selection stage (see Figure 2). The stimulus categorization stage c l a s s i f i e s the neural impulses transmitted from the stimulus perception phase into one of the possible stimuli. Under the condition of equal frequency of stimulus presentation, the time taken for this stage i s constant (t2). \ Stimulus Stimulus Stimulus a — » —> Perception Categorization *1 + t2 -t Response Selection Response ->Search & Release Response Action t ^ + t ^ + t ^ •-Response A CRT Figure 2: Schutz's Model of Information Processing Components i n a Choice Reaction Time Task. i H I The response selection stage receives the name of the stimulus and pairs the appropriate response to i t . The time taken for this stage w i l l be a constant for equally well learned S-R pairs. The response search i s conducted i n the same manner as proposed by Falmagne and Theios, searching f i r s t i n SA, then PM and f i n a l l y SM. Theoretical Rationale for Hypothesis Hypothesis 1 The model accounts for hypothesis 1 by way of a longer readout or release time for the Type II task due to the number of subprograms involved. It i s thought that these subprograms (which are the components of one response program) are read out s e r i a l l y each taking a f i n i t e amount of time. Thus the longer the program, the longer the release time. Failure of the data to support hypothesis 1 would probably suggest a poor construction of the tasks, i . e . , they were not rea l l y of unequal complexity, or that the more complex task was not being processed as a unitary program. . Hypothesis 2 The model predicts i n hypothesis 2 that the d i f f e r -ences:; i n CRT between Type II and Type I tasks w i l l be less than the SRT difference because with equal response prob-a b i l i t i e s the S w i l l place into SA a more complex (Type II) - 1 9 -response program more often than a simple (Type I) response program. This i s "because i t i s easier to execute"! "the simple response program than the complex response program. Rejection of this hypothesis (i . e . , say c = a) would suggest that the only factor governing the relative frequency with which a response i s placed i n SA i s RP. Or, that there i s no such thing as SA, and memory search i s always i n PM. Hypothesis 3 With unequal response probabilities as i n hypothesis 3 , the model predicts that as the task becomes more complex, increases i n response probability w i l l have less effect on CRT. As the response program for a complex task w i l l already be placed i n SA with a f a i r l y high probability, RP w i l l have a re l a t i v e l y small effect. Similarly increased response probability w i l l have a greater influence on the Type I task. I f hypothesis 3 i s rejected and i n fact b = c = :d<a, this suggests the effect of RP i s independent of task com-plexity. But i f these differences (b,c,d) are a l l less than a, then there must be something l i k e SA which accounts for the smaller differences i n CRTs over SRTs ( i . e . , c<a). How-ever, a lack of change i n differences due to task complexity at different levels of RP suggests that although both TC and RP effect RT through some type of response readiness mechanism, their effects act independently of each other. Perhaps i t would be that the larger difference i n the SRT condition i s due to some other mechanism not accounted for by the model. CHAPTER III METHODS AND PROCEDURES Design The research design i s a 2 x 4- complete f a c t o r i a l with repeated measures on both factors. Response latency i s the dependent variable which was measured i n msec. There are two independent variables; (1) Task complexity ( 2 levels) and ( 2 ) Response probability (4- levels) as shown i n Figure 3 » TYPE I TYPE II p 1 . 0 p . 7 5 P * 5 0 P . 2 5 P 1 . 0 p . 7 5 P . 5 0 p . 2 5 S l S 2 S 3 S4 Figure 3-: Experimental Design A 4x4- Latin square was used to control the order i n which each.subject performed each condition since i t has been shown that RT i s affected by practice. Subjects Sixteen male and female volunteer physical education graduate and undergraduate students from the University of B r i t i s h Columbia participated i n the experiment. A l l sub-jects were right handed. - 2 0 --21-Apparatus Figure 4 shows the arrangement of the stimulus lights, response key and warning l i g h t . A reaction time interface and d i g i t a l printer were used to record response latencies i n msecs. The experimenter selected the stimulus l i g h t and foreperiod, and also turned the warning and stimulus lights on by means of a SA600 universal timing module. Experimental conditions Four experimental conditions were tested. Condition A consisted of 100 t r i a l s of both Type I and Type II tasks i n a simple reaction time condition, (P-^ Q for both Type I and Type II tasks). Condition B consisted of 100 t r i a l s of both Type I and Type II tasks i n a CRT condition with equal response probabilities, (P ^ for both Type I and Type II tasks). Condition C consisted of 1 5 0 t r i a l s of Type I task and 50 t r i a l s of Type II task with unequal response prob-a b i l i t i e s , (P and P 2 5 for Type I and Type II respectively). Condition D consisted of 50 t r i a l s of Type I task and 150 t r i a l s of Type II task with unequal response prob-a b i l i t i e s , (P 2 5 P 7 5 for Type I and Type II tasks respectively). Procedure Each subject was asked to respond to a stimulus l i g h t which required one of two possible movements. These move-Warning Light o Target C Stimulus Lights 0 © © Target A o Response. Key Obstacle A Figure 4 : Arrangement of Stimulus Lights, Response Key, Warning Light, Obstacle and Targets - 2 3 -ments are referred to as Task I and Task I I . In both tasks the subject was seated directly i n front of the apparatus and assumed a ready position by placing his index finger on the response key. Task I required, the subject to depress the response key as quickly as possible and h i t target A, 2 inches directly i n front of the response key. Task II required the subject to depress the response key as quickly as possible, move across obstacle A and h i t targets B and G i n rapid succession. Obstacle A was a 1/2 inch raised sur-face situated between the response key and target B, target B being 5 inches to the l e f t of the response key (Figure 4). Target C was located 8 inches directly i n front of target B. No movement time was recorded although subjects were led to believe target C stopped a movement time clock. Four stimulus l i g h t s , each with probability of occur-rence of . 2 5 were located 4- inches directly i n front of the response key. In condition A stimulus li g h t 1 was. used for 8 Ss and stimulus l i g h t 4- for 8 Ss. In condition B, 8 Ss responded with Task I to stimulus lights 1 or 2 and Task II to stimulus lights 3 or 4. The other 8 Ss responded with Task I to stimulus lights 3 or 4 and Task II to stimulus lights 1 or 2 . In conditions C and D one stimulus l i g h t (light 2 and 3 , balanced over Ss) acted as the p' 2 5 stimulus for the low probability of occurrence task, with the remaining three lights providing the . 7 5 stimulus. Ss were told before each testing session exactly what the response probability was for each task and that each stimulus l i g h t had a prob--24-a b i l i t y of occurrence of . 2 5 . The stimulus foreperiod of 1, 2 or 3 seconds was randomized by a table of random numbers. In conditions B, C and B t r i a l s were randomized by a table of random numbers. Each subject performed two conditions per day or about 420 t r i a l s . A sample order consisted of: 50 t r i a l s 1 min. rest 50 t r i a l s 1 min. rest 50 t r i a l s 3 - 5 min. rest Condition II Same. After each response the subject immediately returned to the response key. When the S had returned and was ready, a warning lig h t prepared him for the next stimulus which appeared 1 - 3 seconds l a t e r . A sample sequence consisted of: - placing of index finger on response key - warning signal (on-off flash) - r 1 - 3 second foreperiod - stimulus light (on-off flash) - response - RT recorded by automatic d i g i t a l recorder - manual resetting of foreperiod and stimulus (approximately 2 seconds) Condition I 5 50 practice t r i a l s t r i a l s ( 3 min.) 1 min. rest - 2 5 -- repeat Subjects were asked to notify the experimenter i f they made an error. A l l incorrect responses were retested along with the two preceding t r i a l s with only the corrected t r i a l being recorded. S t a t i s t i c a l Analysis^ Analysis of variance for a repeated measures design was used to determine the s t a t i s t i c a l significance of any differences among mean response latencies. ANOVA (for 16 Ss) Source df Subjects 1 5 a Task 1 b Condition 3 c Task x Condition 3 * Subjects x Task 1 5 * * Subjects x Condition 4 5 *** Subjects x Task x Condition 4 5 1 2 ? * Error for a ** Error for b *** Error for c The F ratio, for the Task main effect provided a test for differences between Task I and Task II, and a planned orthogonal comparison was used to test hypothesis 1 specif-i c a l l y . The F ratio for Task x RP aided i n explaining - 2 5 a -hypotheses 2 and 3 . Newman-Keuls multiple comparison test was used to test for differences between a l l pairs of condition means. Harter's multiple comparison procedures were used as a post-hoc analysis on any significant inter-actions (Harter, 1 9 7 1 ) . CHAPTER 4 RESULTS AND DISCUSSION Results Upon successful completion of data collection the raw data was put on punch cards and the analysis done by > the IBM 360 computer. The empirical findings were sub-jected to s t a t i s t i c a l tests through the use of computer programs, UBC Simcort for means and variances, and BMD02V for the ANOVA. The results are presented below. The means, standard deviations and coefficients of variation were calculated for a l l experimental con-ditions and are shown i n Table I. The means of the experi-mental conditions were plotted and are shown in Pig. 5 (p.3 7 ) . An analysis of variance was performed on the data to test for significance. The results of the analysis of variance for the 2 x 4 repeated measures design are presented i n Table I I I . The analysis of variance shows conditions as being highly significant (<.001). A Newman-Keuls post-hoc multiple comparison was performed on the 4 condition means (as given i n Table I I ) . Condition 1 was significantly different from a l l other means. None of the choice con-ditions were significantly different from each other (see Table IV). -26-- 2 7 -Table I Mean Response Latencies, Standard Deviations and Coefficients of Variation (in msec.) for a l l Conditions Condition Task Mean SD V A B C D I 2 2 1 2 9 . 5 .13 II 2 5 3 5 1 . 0 . 2 0 I 3 3 0 7 1 . 0 . 2 2 II 3 1 5 7 0 . 7 . 2 2 I 323 5 3 . 8 . 1 7 II 286 4 9 . 1 . 1 7 I 3 1 0 7 8 . 7 . 2 5 II 3 2 3 7 0 . 5 . 2 2 Table II Means and Differences (in msec.) for Main Effects and Inter-actions Arranged According to Experimental Conditions Condition 1 B C D X T I 2 2 1 3 3 0 3 2 3 3 1 0 296 A S II 2 5 3 3 1 5 286 3 2 3 K X 2 3 7 3 2 2 304 317 2 9 5 Difference 3 2 - 1 5 - 3 7 13 - 2 28-Table III Analysis of Variance for the Effect of Experimental Con-ditions and Task Complexity on Response Latencies Source df Mean Squares F P Subjects 1 5 24820 .55 Conditions 3 49846.29 33.56 <.001 Subj. X Cond. 45 1485.21 Task Complexity 1 96.43 <1.00 >.05 Subj. X T.C. 1 5 160 . 9 2 Cond. X T.C. 3 7275.10 40.24 <.001 Subj. X Cond. X T.C. 4 5 180 .79 Total 12? Table IV Newman-Keuls Post-Hoc Multiple Comparisons on Condition Means Conditions A C D B A - 8.26 9.73 10. 44 C - 1.47 2. 18 D - 0. 69 B Q 2 . 0 5 = 2 - 8 6 Q 3 . 0 5 = 3.44 Q4 .05 = 5 ' 7 9 - 2 9 -The task complexity effect was found to be non-significant (p> . 0 5 ) , but the difference between Tasks I and II under condition A was highly significant, as shown by the t value of 6 . 7 4 (p< . 0 0 1 ) obtained from the pre-planned comparison. The interaction between conditions and task complexity was found to be highly significant (p ' < . 0 0 1 ) . This significant interaction i s also shown by the non-parallel lines i n Figure 5 . Post-hoc analysis on the difference scores (Table II) between c e l l means using the Harter method ( 1 9 7 1 ) showed significant differences between levels of task complexity amongst a l l pairs of experimental conditions (Table V). Table ? Harter Post-Hoc Analysis on Cell Means Conditions C B D A C 4 . 6 1 8 . 4 6 1 4 . 3 6 ' B - 5 . 9 6 9 . 7 5 D - 3 . 7 9 A -Significant Studentized Ranges Q 2 . 0 5 = 2 . 8 6 Q 3 . 0 5 = 3 . 4 4 Q 4 . 0 5 = 3 . 7 9 - 3 0 -Supplementary to the main analysis, an analysis of variance was performed on response probability rather than experimental conditions as main effect. This was done be-cause response probability was the independent variable of main concern. Table VI shows this transformation. Table VI, Response Probability of Experimental Conditions Task Complexity Experimental Condition Response Prob a b i l i t y A 1 . 0 B . 5 0 I C . 7 5 D . 2 5 A 1 . 0 B . 5 0 II C . 2 5 D . 7 5 Table VII gives the c e l l means, grand means and d i f -ferences arranged according to response probabilities. Figure 6 (p. 4-1) shows these c e l l means. This i s i n contrast to Figure 5 which presents the same means arranged according to experimental conditions. To test the significance of the data arranged by re-sponse probabilities an analysis of variance was performed. -31-The results of: this are presented i n Table VIII. Table VII Ce l l Means, Grand Means and Differences (in msec.) Arranged According to Response Probabilities Response Probability l.G .75 .50 . 2 5 X 221 323 330 310 2 % 2 5 3 323 3 1 5 286 294 K . . > X 2 3 7 3 2 3 3 2 2 298 2 9 5 D 32 0 - 1 5 - 2 4 - 2 Table VIII Analysis of Variance for Conditions and Task Complexity Arranged by Response Probability Source df Mean Squares P P Subjects 15 24820.55 Conditions 3 52414.91 49.04 < .001 Subj. & Cond. 45 1068.76 Task Complexity 1 96.43 <1.00 > . 0 5 Subj. x T.C. x5 160 . 9 2 Cond. x T.C. 3 4706.50 7.88 <.01 Subj. x Cond. x TC 45 597.24 Total 1 2 7 Note that the F ratios are approximately the same as those i n Table I I I . A.Harter post-hoc analysis on the difference scores (Table VII) of interaction means showed only response probability 1 . 0 different it from a l l the other response probabilities (Table IX). Table IX Harter Post-Hoc Analysis on Interaction Means Response probabilities . 2 5 . 5 0 . 7 5 1 . 0 . 2 5 - . 9 8 2 . 7 7 6 . 3 6 . 5 0 - 1 . 7 9 5 . 3 8 . 7 5 - 3 . 5 9 1 . 0 -Significant Studentized Ranges Q 2 . 0 5 = 2.86 Q 3 . 0 5 = 3 . 4 4 -Q 4 . 0 5 = 3 . 7 9 - 3 3 -Mscussion Condition A Condition A, which consisted of both Type I and Type II tasks i n a simple reaction time situation, exhibited a very slow SRT for Task I ( 2 2 1 . 2 msec). This slow time could be due to a small amount of play i n the subjects' re-sponse key (approximately 2 mm.) or possibly unusually slow subjects. However, this slowness would be constant and thus have no effect on the results. It can be seen i n Table IV that Condition A i s significantly different from a l l other conditions. This was to be expected as the other three conditions were CRT tasks and i t i s well documented that SRT i s always less than CRT (Mowbray and Rhodes, 1 9 5 9 ; Smith, 1 9 6 7 ) . The significant difference i n Condition A between Task I and Task II v e r i f i e s the fact that Task II was more complex. This finding also supports the theory that a more complex task requires a longer release time (Schutz, 1 9 7 2 ) . Schutz proposed that a more complex task requires a greater number of already learned and stored motor subprograms requiring more time for organization and thus a longer release time. The difference between Task I and Task II due to this longer release time should remain constant, thus any changes i n the differences i n Conditions B, C or D are due to other factors. Table X shows the magnitude of the difference i n release time between Task I and Task II, Task I being 43% -34-faster. The figures shown i n this table are estimates (derived from the findings of this study) of the com-ponents of Schutz' 1972 model. This table helps to isolate the possible causes of the differences i n time between conditions. The stimulus perception and response action stages are constant throughout conditions. In Condition A no time i s required for stimulus categor-ization or stimulus-response matching whereas i n Con-ditions B, C, and D a constant time of 30 msec, i s re-quired. As mentioned above, the response release time for Task I and Task II i s constant over conditions as this i s variable over levels of task complexity only. It i s now possible to calculate response search time for Conditions B, C, and D, Condition A being zero. The response search time gives an indication of the effect task complexity and response probability have on the re-sponse strategies employed by the Ss. The S w i l l place, into SA, the response program he feels has the highest probability of being required next, after consideration of task complexity and response probability. - 3 5 -Table. X Hypothetical Time Components of Total Response Latency Conditions A B C • D Task I II I II I II I II Stimulus Perception (a) 50 50 50 50 ... 50 50 50 50 Stimulus Categor-ization and (b) X X 30 30 30 30 30 30 Stimulus Re sponse Matching Response Search (c) X X 80 35 75 5 60 4-5 Response Release (d) 70 100 70 100 70 100 70 100 Response Action (e) 100 100 100 100 100 100 100 100 Total Response Latenc; (msec?I 2 2 0 2 5 0 330 3 1 5 3 2 5 285 310 3 2 5 -36-Oondition B Condition B consisted of both Type I and Type II tasks i n a CRT condition with equal response probabilities. The difference of 1 5 msec, between Task I and Task II i s smaller than the difference i n Condition A.(32 msec.) as predicted i n hypothesis 2 . However, a negative d i f -ference, as. shown by the crossing of lines i n Figure 5 was not expected. There i s strong support (Schutz, 1 9 7 2 ) that with equal response probabilities a more complex task i s placed into SA more often than an easy task to compensate for the increased task complexity. A possible reason for this i s that S would rather be prepared to respond to the complex task as he feels this response takes longer to i n i t i a t e . Consequently this readiness reduces the response search time thus decreasing the difference i n RL between tasks (Table X). The effect of placing into SA the response program for Task II was so great as to more than compensate for the 3 0 mese. difference due to response release time thus yielding a faster RL for Task II. Because of the equal response probabilities i n Condition B, the only factor affecting the placing of either response into SA i s task complexity. - 3 7 -A B C D Condition Figure 5'- Task Complexity by Experimental Condition Interaction -38-Condition C and D For the purposes of this discussion i t i s necessary to talk about Conditions C and D together. Condition C consisted of both Type I and Type II tasks with unequal response probabilities, P and P ^ respectively. Con-dition D consisted of both Type I and Type II tasks with unequal response probabilities, P 2 5 p ,^5 respectively. The results of Condition C were quite unexpected. With a high RP (P r,^) and an easy task i t was hypothesized that RL for Task I would be less than for Task II due to the RP effect on search time and to shorter release time for the easy task. As the results i n Table :2I ^ show, just the opposite seems to have occurred and Task II has been placed i n more often. Table X shows a response search time of 7 5 msec, for Task I and only 5 msec, for Task I I . It i s d i f f i c u l t to explain why a S would place a response with low probability of occurrence into SA more often than a task with high probability of occurrence. The strategy employed i s very different than what was anticipated. Pos-sibly the Ss were leary of making a mistake or not being ready for the low probability task and thus placed i t i n SA more often. This type of strategy i s similar to placing the more complex task into SA more often. However, the literature shows this not to be so with typical CRT tasks of equal complexity (Smith, 1% 7 ). Therefore there must be some type of TC x RP interaction which- i s not understood at this time. - 3 9 -It was hypothesized that at high response prob-a b i l i t i e s task complexity would have less effect than at low response probability, consequently the R P factor would affect the strategy by placing the high probability response into S A more often. Because the complex task i s more d i f f i c u l t to i n i t i a t e , the factor of TC affects the strategy by placing the response to the complex task into S A more often than the response to the easy task. In Condtion C the factor of TC seems to have had more effect than the R P factor as the complex task was placed into S A far more often even though i t was the low R P task. The 7 0 msec, d i f -ference i n search time between Tasks I and II more than com-pensated for the negative difference of 3 0 msec, resulting from the longer release time of Task II. The net result i s the 40 msec, difference i n total response latencies as shown in Table X. In Condition D TC also seems to have more effect than R P as the response program to Task II i s placed into S A more often. However, i t can be seen that Task I, P 2 ^ has a faster R L ( 3 1 0 msec). This i s due to a shorter response release time i n Task I ( 3 0 msec, shorter). The response search time for Task II i s 1 5 msec, less than for Task I, accounting for the 1 5 msec, difference i n R L between Task I and Task II. However, i n Condition B, Task II i s the high probability ( P r ^ ) . task and i s placed into S A less often than when i t was the low probability task (in Condition C). Thus Task I, although the low probability ( P p i - ) task, i s -4-0-being placed into SA proportionately more often i n Con-dition D than in Condition C. Possibly when RP differences are so great, i.e. P ^ - P 2 5 * S s overcompensate for the p o s s i b i l i t y of being unprepared for the lov; RP task. This results i n a faster RL for low RP tasks than was anticipated as the subject places the low RP response into SA more frequently. When this factor i s combined with a d i f f i c u l t task as i n Condition C (Task II, P . 2 5 ) s i s even more leary and places this response into SA nearly every time. In employing this strategy S must not be particularly worried about recovering for the easy task even though he knows i t w i l l occur more frequently. Thus i t appears that low:prob-a b i l i t y and high TC have a joint affect on the probability of a response program being placed into SA. It i s impossible to t e l l which has a greater effect. In Condition D i t appears that RP has a greater effect as Task I, P 2 ^ i s placed into SA more often than Task I, P 7 5 i n Condition C. Also, Task II, P 75 i s placed into SA less often than i n Condition C (Task II, P 2 5 ) hut s t i l l more often than i n Task I, P 2 5 * Thus i n Condition D, S has more d i f f i c u l t y i n establishing a definite strategy as he must constantly decide whether to place the more complex task or the low probability task into SA. It appears that the high RP vs the low RP works in reverse to what was orig i n a l l y expected, i . e . high RP does not seem to be as important i n subject strategies as low RP. This can be further i l l u s t r a t e d by Figure 6 . " -41-3 5 0 -Response Latency (msec.) . 2 5 . 5 0 . 7 5 1 . 0 Response Probability Figure 6 : Task Complexity by Response Probability Interaction -r -4-2-It can be&seen at P oc- the R L S S P for both Task I and Task • O II are faster than the RLs& at P ^ or P nr-. Thus low RPsr> seem to affect S strategies to a greater extent than do high EPs when using tasks of different complexities. Summary It i s evident that TC affects RL i n that i t takes a longer release time for a more complex task. This factor, when combined with varying RPs, affects S response strategies. With equal RPs S places the more complex task into SA more often than the easy task. This strategy more than compen-sates for the difference i n RLs due to TC with the RL to the more complex task becoming faster. In the s t r i c t sense hypothesis 2 was ver i f i e d , however, a negative difference was not anticipated. With varying RPs, low RP appears to have more effect than high RPs especially at high TCs. The interaction between TC x RP i s not understood at this time. Hypothesis 3 was also ver i f i e d but an absolute difference rather than an algebraic difference was predicted. CHAPTER 5 SUMMARY AND CONCLUSIONS Summary The purpose of this investigation was to study the joint effects of response probability and task complexity on response latency i n simple and choice reaction time tasks. It i s based on the e a r l i e r work of Schutz (1972). Pour experimental conditions, one simple reaction time task and three choice reaction time tasks, were used to test the hypotheses. Response latencies for 800 t r i a l s were obtained from each of sixteen subjects during two, one-hour, testing sessions. Analysis of variance for a repeated measures design was used to analyse the data. Harter and Newman-Keuls post-hoc multiple comparisons were performed on the data to test for significant differences between conditions. The empirical findings were generally not i n agreement with the predicted hypotheses. It was predicted that with equal RPs, the difference i n RLs due to TC would be reduced due to a shorter search time for the complex task. This occurred to a greater extent than predicted resulting in a negative difference. This difference can only be attributed to the difference i n the complexity of the tasks. It was also predicted that with varying RPs the -43--44-subject would place into SA the response with the highest probability of occurrence. However, i t was found that subjects adopted the opposite strategy and placed the low RP response into SA more frequently. The TC x RP interaction i s d i f f i c u l t to understand. It appears that at high levels of task complexity combined with a low response probability, subjects adopt a very defensive strategy and prepare for the complex task at the low response probability rather than the easy task with a high response probability. Possibly subjects feel they can recover for the easy task occurring more frequently. How-ever, when the complex task with high response probability i s one choice and the easy task with low response probability the other, subjects seem to be undecided as to what strategy to employ. On the one hand they have the d i f f i c u l t task which they l i k e to prepare for and on the other the low response probability which they l i k e to prepare for. Under these conditions i t appeared that subjects preferred to prepare for the d i f f i c u l t task. Conclusions Very fast RLs were made to complex tasks combined with low probability of occurrence, indicating subjects were using a defensive type of strategy. Subjects were also better prepared to i n i t i a t e a response to an easy task with a low response probability than an easy task with a high response probability. Thus low response probability seems -4-5-to have more effect than high response probability on ~ RLs of CRT tasks of unequal complexity. It i s evident that the TC x RP interaction has a profound effect on RL but that further research i s necessary to understand this interaction. Suggestions for Further Research It was suggested earlier that when the difference i n response probabilities i s great, i . e . P 2 5 ~ ^ 7 5 * subjects adopt a defensive type strategy. This may be because subjects do not want to make a mistake or a slow response to the low probability task. A further study i n this area might test response probability differences of: p.10 " p.90' P . 2 5 - P . 7 5 ' P.33~ P . 6 7 ' ^ P.44 " p . 5 6 i n o n e experiment to see what effect the difference i n response probability has on subjects strategy. -46-Bibliography Bertelson, P. "Sequential redundancy and speed i n a seria l two-choice responding task." Quarterly Journal of  Experimental Psychology, 1 3 : 9 0 - 1 0 2 , 1961. Bertelson, P. "Serial choice reaction time as a function of response versus signal-and-response repetition." Nature, 206 : 2 1 7 - 8 , 1 9 6 5 . Bertelson, P., and Barzeele, J. "Interaction of time uncertainty and relative signal frequency i n de-termining choice reaction time." Journal of Experi- mental Psychology, 7 0:448 - 5 1 , 1 9 6 5 . Bertelson, P., and Tisseyre, P. "Choice reaction time as a function of stimulus versus response relative f r e -quency of occurrence." Nature, 212:1069-70, 1966. Bla i r , S. N. The effect of stimulus and movement complexity . upon reaction time and movement time. Paper presented at Washington Symposium, 1 9 6 8 . Boyko, Ye. I. Vremya Reaktsii Cheloveka. Meditsina Pub-lis h i n g House, 1964. Translated by JRRS as Reaction  Time of Man. JRRS : 2 7 , TT:64 - 5 1 9 5 4 , 1964. Broadbent, D. E., and Gregory, M. "On the interaction of S-R compatibility with other variables affecting reaction time." B r i t i s h Journal of Psychology, 5 6:61 - 6 8 , 1 9 6 5 . Donders, P. E. "Die schnelligkeit psychischer processe." Archive Anatomie und Physiologie, 657-681, 1868. Cited by Smith, E. E., I % 7 . Palmagne, J. C , and Theios, John. On attention and memory i n reaction time experiments. Paper presented at the Donders Centenary Symposium on Reaction Time, Eindhoven, Holland, 1 9 6 8 . Griew, S. "Complexity of response and time of i n i t i a t i n g responses i n relation to age." American Journal of Psychology, 7 2 : 8 3 - 8 8 , 1 9 5 9 . Hale, D. J. "Repetition and probability effects in a seria l choice reaction task." Acta Psychologica, 2 9 : 1 6 3 - 1 7 1 , 1 9 6 9 . Harter, H.'L. "Multiple comparison procedures for inter-actions." The American Stat i s t i c i a n, 24 : 3 0 - 3 2 , 1 9 7 0 . _47-Henry, F. M., and Rogers, D. E. "Increased response latency for complicated movements and a 'Memory Drum' theory of neuromotor reaction." Research Quarterly, 31:448-55, I960. Henry, F. M. "Stimulus complexity, movement complexity, age, and sex In relation to reaction latency and speed i n limb movements." Research Quarterly, 32:353-66, 1961. Hick, W. E. "On the rate of gain of information." Quarterly  Journal of Experimental Psychology, 4:11-26, 195 2 . Hyman, R. "Stimulus information as a determinant of re-actiontime." Journal of Experimental Psychology, 45:188-96, 1953. John, I. D. "Mediating processes i n choice reaction tasks." Acta Psychologica, 30:58-64, 1969. Keele, S. W. "Repetition effect: A memory dependent process." Journal of Experimental Psychology, 80:243-248, 1969. Kornblum, S. "Choice reaction time for repetitions and non repetitions." Acta Psychologica, 27:178-87, 1967. Krinchik, E. P. "A study of human information processing i n a choice situation." Soviet Psychology and Psy- chiatry, 1:22-24, 1963. Krinchik, E. R. "The probability of a signal as a determ-inant of reaction time." Acta Psychologica, 30:27-36, 1969. Laberge, D. and Tweedy, J. R. "Presentation probability and choice time." Journal of Experimental Psychology, 68:477-81, 1964. Leonard, C. A., Newman, R. C., Carpenter, A. "On the handling of heavy bias i n a self paced repetitive task." Quarterly Journal of Experimental Psychology, 18:130-41, 1966. Moss, S. M., Engel, S., Faberman, D. "Alternation and repetition reaction times under three schedules of event sequencing." Psychonometric Science, 9:557-58, 1968. Mowbray, G. H., and Rhoades, M. V. "On the reduction of choice reaction times with practice." Quarterly Journal  of Experimental Psychology, 11:16-23, 1959. Norrie, M. L. "Practice effects on reaction latency for simple and complex movements." Research Quarterly, 38:79-85, 1967. -48-Remington, R. J. "Analysis of sequential effects i n choice reaction time." Journal of Experimental Psychology, 80:1-8, 1969. Schutz, R. W. "A theory of motor response organization and retrieval i n choice reaction time tasks." Doctoral Dissertation, University of Wisconsin, 1972. Schvaneveldt, R. W. "Effects of complexity i n simultaneous reaction time tasks." Journal of Experimental Psy-chology, 81:289-96, 196^1 S h i f f r i n , R. M., and Atkinson, R. C. "Storage and retrieval processes in long term memory." Psychological Review, 76:179-83, 1968. Sidowski, J. B., Morgan, R., and Eckstrand, G. "Influence of task complexity and instructions upon simple and dis-crimination reaction times." Journal of Experimental  Psychology, 55:163-66, 1958. Smith, E. E. "Choice reaction time. An analysis of the major theoretical positions.". Unpublished mimeographed materials, Behavioral and c l i n i c a l studies research center, Washington, D. C., 1967. Sternberg, S. "High speed scanning i n human memory." Science, 153:652-664, 1966. Welford, A. T. "Evidence of a single channel decision mechanism limiting performance in a serial reaction task."- Quarterly Journal of Experimental Psychology, 11:193-210, 1959 . Welford, A. T. "The measurement of sensory motor performance: Survey and reappraisal of twelve years' progress." Ergonomics, 3:189 - 2 2 9 , I960. Williams, J. A. "Sequential effects i n disjunctive reaction time. Implications for decision models." Journal of  ExperimentalCPsychology, 71:665-672, 1966. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0077218/manifest

Comment

Related Items