c A COMPARATIVE ANALYSIS OF THE INFORMATION PROCESSING RATE OF THE DECISION MECHANISM OF RETARDATES AND NORMALS by JANIS ELAINE LINDSAY B.P.E., The University of B r i t i s h Columbia, 1 9 7 0 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF PHYSICAL EDUCATION i n the School of Physical Education and Recreation We accept this thesis as conforming to the required standard The University of B r i t i s h Columbia August, 1 9 7 2 In present ing th i s thes i s in p a r t i a l f u l f i lment of the requirements for 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 ib ra ry sha l l make i t f r e e l y ava i l ab le for reference and study. I fu r ther agree that permission for extensive 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 h i s representat ives . It is understood that copying or pub l i c a t i on o f t h 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 of The Un ivers i ty of B r i t i s h Columbia Vancouver 8, Canada i ABSTRACT Eight male retarded Ss and eight male normal Ss took part i n a choice reaction time experiment involving four levels of information load. The experiment required Ss to choose between two, four and six alternatives, de-pending on the condition being tested, and to respond by pressing the correct response button for each t r i a l . Simple reaction time of both groups was also tested. The results obtained from the choice conditions of the experi-ment were tested by an analysis of variance test and a t-test was used to test the difference i n simple reaction time between the two groups. The results indicated that there was a significant difference between the simple reaction of the two groups and that as information load was increased there was a significant increase i n the difference i n reaction time between the two groups. Also differences were found between the group v a r i a b i l i t y and intra-individual v a r i a b i l i t y of the two groups. The results were discussed i n relation to information theory and other studies which have dealt with the reaction time differences of normals and retardates. I t was concluded that retarded Ss have a slower simple reaction time than normal Ss, and that retardates process information in* their decision mechan-ism, at a slower rate than normals. The decision mechanism of retarded Ss was cited a p a r t i a l source of the delay in reaction time of retarded Ss. I t was f e l t that further investigation i s necessary to determine to what degree the decision mechanism i s responsible for the slower than normal reaction times of retarded Ss. ACKNOWLEDGEMENT I would l i k e to thank the members of my committee, Dr. R. Marteniuk, Dr. R. Schutz, Dr. C. David, Miss A . T i l l e y , and Dr. D. Whittle for their guidance and support during the preparation of this thesis. I would especially l i k e to thank my chairman, Dr. Marteniuk, who has so freely given his time and knowledge to assist me i n a l l phases of this thesis,, i v TABLE OF CONTENTS CHAPTER PAGE I Statement of the Problem 1 - Introduction 1 - Purpose of the Study 6 - Hypotheses 6 - Limitations ~ 7 - Delimitations 7 - Definition of Terms 8 II Review of Literature 9 - Differences in RT Between Retardates and Normals 9 - Relationship Between RT and I n t e l l i -gence 9 - A Model of Human Motor Response 10 - The Perceptual Mechanism - Suggested Source of Delay i n RT of Retardates 12 - The Effector Mechanism - Suggested Source of Delay i n RT of Retardates 17 - The Decision Mechanism - Suggested Source of Delay i n RT of Retardates 1 9 III Methods and Procedures 22 - Subjects 22 - Apparatus 22 - Position of Subject (S) and Experimenter (E) at Apparatus 2 5 - Experimental Design 26 - Procedures 27 - Analysis of Data 28 V CHAPTER PAGE IV Results and Discussion 31 - Results 31 - Difference Between Group RT over a l l Four Conditions 31 - Group Va r i a b i l i t y 36 - Intra-Individual V a r i a b i l i t y 38 - Discussion 40a - Simple RT, Retardates and Normals 4-1 - CRT, Retardates and Normals 42 - Group V a r i a b i l i t y 48 - Intra-individual V a r i a b i l i t y 49 V Summary and Conclusions 5 1 - Recommendations 5 3 References 54 Appendices A. Results from P i l o t Study 5 7 B. Instructions used i n P i l o t Study and Study 60 C. Individual Score Sheets 62 LIST OF TABLES Mean RT For Normals and Retardates Under Various Conditions Of Intensity And Warning Mean RT and Variance Of Retardates And Normals For Conditions 1, 2, 3 and 4 Analysis Of Variance Of RT Over Conditions 1, 2, 3 and 4 T-Test And Planned Comparison Analysis Of Difference In Mean RT For Condition 1 Intra-Individual Variance Of Retardates And Normals For Conditions 1, 2, 3 , and 4 Comparative Analysis Of Simple RT Of Retardates And Normals In The Present Study With The Results Of Two Other Studies V l l LIST OF FIGURES PAGE 1 Frequency Polygons of RTs for Normals and Retardates 3 2 The Central Mechanisms as Described by Welford's Model ( 1 9 6 5 ) of Human Behavior 1 1 3 RTs of E.M.R. and Normal Ss on a Gross Motor-Perceptual Time Task (Austin 1 9 6 9 ) 16 4 Apparatus and Position of Subject and Experimenter at Apparatus 2 3 5 Mean RT as a Function of Information Load 3 2 6 Group V a r i a b i l i t y - Individual Mean RTs as a Function of Information Load 3 7 7 Mean Intra-Individual V a r i a b i l i t y as a Function of Information Load 3 9 8 Mean Times and Regression of Time on Information (Annett 1 9 5 7 ) 46 CHAPTER 1 STATEMENT OF THE PROBLEM Introduction In studying the human organism i t has been found that there i s a f i n i t e elapse of time between the present-ation of a stimulus and the i n i t i a t i o n of a response. This characteristic of human behavior i s known as reaction time (RT). To help understand the process underlying RT, models of human performance have been used to break i t down into i t s component processes. Welford (1965; 1968), has suggested a model which looks at behavior i n terms of three central mechanisms; a perceptual mechanism which organizes and c l a s s i f i e s incoming information; a decision mechanism which selects the appropriate response; and an effector mechanism which organizes and executes the re-sponse. Several studies, which have compared the RT of normals with that of retardates, have found a number of differences i n RT between these two populations (Berkson 1960a, 1960b, 1960c; Berkson and Baumeister, 1967; Baumeister and Kellas, 1968; Jones and Hinkle, 1 9 7 0 ) . Berkson and Baumeister ( 1 9 6 7 ) found that retarded subjects were generally slower i n RT than normals and that the -2-retarded subjects were more variable both between and within subjects. Baumeister and Kellas (1968) also found retardates to have slower than normal RTs. Their results showed that the distribution of the retardates' responses tended to be more variable* platykurtie and symmetric, while the normals demonstrated typical leptokurtic dis-tributions, skewed to the right (Figure 1). Jones and Hinkle (1970), provided further confirmation of both the Baumeister and Berkson (1967) and the Baumeistersfand Kellas (1968) studies, with their finding that the RTs of re-tardates were significantly more variable than the RTs of normals and that retardates had a slower RT than normals on both simple and choice RT tasks. The finding of differences i n RT between retardates and normals has prompted researchers to study each of the component processes of performance in an attempt to de-termine which of these processes i s causing the slower than normal RT i n retardates. Urquhart, Beedle and Smith (1964) looked at the perceptual component of RT by studying the effect of stimulus intensity on retardates and normals. They concluded that retardates were rela t i v e l y more sus-ceptible to decreases i n the intensity of the RT stimulus than normals. Contrary to these findings Berkson (1960a) found that there was no difference i n the duration threshold and consequently speed of perception of retardates and normals. REACT ION TIME "Cor SEC.) FIGURE U FREQUENCY POLYGONS OF RTs FOR NORMALS AND RETARDATES. (BAUMEISTER AND KELLAS 1968) -4-Other studies were designed to determine i f the effector mechanism was causing the delay i n the RT of retardates. An example of research done along this line i s a study by Groden ( 1 9 6 9 ) , which found that the relationship between RT and mental age (r = . 71 ) was eliminated when the complex perceptual-motor tasks (multiple key press tasks), were held constant but not when the simple s k i l l s (finger strength measured by a Lafayette dynamometer and finger o s c i l l a t i o n ) , were held constant. From this he concluded that the complex s k i l l s involved something over and above what was required i n the performance of a simple s k i l l . John Annett ( 1957) looked at the decision mechanism as a possible source of the delay i n RT in retardates. He found a significant difference i n the information pro-cessing rate of the decision mechanism (the time taken by the S to choose a correct response to match incoming i n -formation) of retardates depending on their level of re-tardation. He found that moderately retarded Ss process information faster than severely retarded Ss. Berkson ( 1960c ) did a study i n which he found no difference i n the informational processing rate, of the decision mechanism, of retardates and normals. However, a number of faults were evident i n the design of Berkson's study, as late r pointed out in a study by Hawkins, Baumeister, Koenigs-knecht and Kellas ( 1 9 6 5 ) . In 1952 Hick developed a method of measuring the - 5 -rate of information processing of the decision mechanism in bits/seconds after refining the ear l i e r work of Merkel ( 1885 )• He found that i n a choice reaction time (CRT) experiment with an increasing number of alternatives, RT increased l i n e a r l y as information load was increased. Further, he postulated that the reciprocal of the slope of this line represented the rate of information pro-cessing i n bits/seconds. This method of determining the rate of information processing of the decision mechanism i s dependent on the assumption that i n a CRT experiment the roles of the perceptual and effector mechanisms are minimized. This assumption i s known to be correct i n re-lation to the effector mechanism since the motor response involved i n a CRT experiment i s usually a very simple key press task. However, i t i s questionable whether the role of the perceptual mechanism, i s truly minimized, since there i s a slight increase i n the size of the perceptual f i e l d i n a CRT experiment. Fortunately, a study by Hyman (1953) has proven that the role of the perceptual mechanism i s definitely minimized i n the typical CRT experiment. Hyman used three different methods of placing an increasing i n -formational load on his Ss. He increased the amount of information by; a) increasing the number of stimuli, b) by varying the probability of occurrence of each stimuli, and c) by varying the sequential probability of occurrence of the stimuli. He found that for a l l three methods the regression of RT on information was the same. Since the last two methods, which didn't increase the perceptual f i e l d , produced the same effect as the f i r s t method, which did s l i g h t l y increase the perceptual f i e l d , Hyman's study supports the viewpoint that a slight increase i n number of stimuli doesn't increase perception time.-Purpose of Study The purpose@of this investigation i s to compare the information processing a b i l i t y of retardates and normals i n terms of their a b i l i t y to process information i n a CRT experiment. Hypotheses 1. That the simple RT of retarded Ss i s slower than the simple RT of normal Ss. 2. That retarded Ss process information, i n their decision mechanism, at a slower rate than normal Ss. This should be evidenced by retarded Ss displaying a slower CRT than normal Ss at a r e l a t i v e l y low information input level with this difference becoming progressively larger as input information i s increased. It has been found by a number of studies that re-tarded Ss have a slower simple RT than normal Ss. Also, some of the literature reviewed i n the present study has indicated that retardates possibly have a lower channel capacity, of the decision mechanism, than normal Ss. On -7-the basis of the literature reviewed, i t i s expected that the present study w i l l show retarded Ss have a slower than normal simple RT and that retardates process inform-ation at a slower rate than normals. Limitations The conclusions of this study are limited to: 1 . The sample size of eight Ss taken from each population. 2 . The methods and procedures used i n investig-ating the problem. Delimitations This investigation i s delimited to the study of: 1 . Normal right-handed males estimated by th e i r teachers on the basis of classroom achievement to be of average intelligence ( 9 5 - 1 0 5 I.Q.), between 1 0 to 1 2 years of age, from McBride Elementary School, Vancouver. 2 . Retarded right-handed males, estimated by their teachers on the basis of classroom achievement to be within the 4 5 to 5 4 I.Q. range, between 1 0 to 1 2 years of age, from Oakridge School, Vancouver. 3. The information processing rate of the decision mechanism i n CRT. 4 . The groups of Ss defined by their teachers' estim-ates to be within the 9 5 - 1 0 5 (normals) and the 4 5 - 5 4 (retardates) intelligence quotient range. - 8 -5 . A description of human performance as hypothesized by Welford's model ( 1 9 6 5 ; 1 9 6 8 ) . Definition of Terms Normals - have been operationally defined for the purpose of this study as those individuals estimated by their-school teachers to be within the 9 5 - 1 0 5 I.Q. range. Retardates - have been operationally defined for the purpose of this study as those individuals estimated by their school teachers to be within the 4 0 - 5 4 I . Q . range. Informational Processing Rate - i s the amount of transmitted information per response, divided by the time i t takes to make the response. It i s measured i n b i t s per second. Information theory defines one " b i t " of information as the amount of information needed to make a decision between two equally l i k e l y alternatives. The number of bit s i s computed by the following formula: "Bits" = l o g 2 N where N equals the number of alternatives and where the alternatives were equally probable. CHAPTER II REVIEW OP LITERATURE Differences i n RT Between Normals and Retardates In the la s t decade many differences have been found between the RTs of normals and retardates„ The major d i f f e r -ence that has been observed i s that retardates, as a group, have si g n i f i c a n t l y slower RTs than normals (Berkson, 1960a, 1960b, 1960c; T e r r e l l and E l l i s , 1964; and Jones and Hinkle, 1970). In addition to being slower than normals, Baumeister and Berkson (1967) found that the RTs of retardates are much more variable between and within Ss, than those of normals. Also, Baumeister and Kellas (1968) found that the distribution of retardate responses are platykurtic and symmetric, while normals display leptokurtic distributions skewed to the right. The studies that have been reviewed i n this investig-ation have a l l used moderately retarded subjects unless other-wise stated. However over the l a s t three decades, i n which most of the lit e r a t u r e reviewed has been done, the def i n i t i o n of moderately retarded has varied greatly. Because of this i t i s recognized by the investigator that some of the dis-crepancies of results betv/een studies may be p a r t i a l l y due to differences i n I.Q. leve l s . The presently accepted range of moderate retardation i s between 40 and 54 I.Q. points. Relationship Between RT and Intelligence Since a difference i n RT has been found between retarded and normal Ss, i t has been hypothesized by some researchers - 9 --10-that there may be a strong relationship between intelligence and RT. E l l i s and Sloan ( 1957 ) did a study entitled "Relationship Between Intelligence and Simple RT i n Mental Defectives". In this study they found a negative correlation of -.54, between RT and mental a b i l i t y . This correlation was significant beyond the .01 level of confidence. A neg-ative relationship was also found by Bensberg and Cantor ( 1957 ) and by Dingman and Sil v e r s t e i n (1964). Dingman and Sil v e r s t e i n obtained a negative correlation of - . 1 5 5 * which was significant at the . 05 level of confidence but not at the .01 l e v e l . Although these correlations reached the .01 and . 05 levels of confidence, respectively, they are s t i l l very low correlations, and do not indicate a strong relationship betv/een RT and i n t e l l i -gence. When these correlations are interpreted i n terms of explained and unexplained variance ( i . e . r x 100), i t becomes apparent that differences i n RTs only account for 25$ and 2.5$ respectively, of the variance i n intelligence. On the basis of this closer analysis i t must be concluded that RT cannot accurately predict a person's mental a b i l i t y . However, these studies, which have attempted to relate RT and intelligence have shown that there i s some relationship be-tv/een these two variables, and has led researchers to wonder, which of the processes involved i n RT i s causing the dis-crepancy i n RT betv/een normals and retardates. A Model of Human Motor Response Welford (1965; 1968) suggested a model of human performance which outlines the processes involved i n the -11-i n i t i a t i o n and execution of a response. Welford feels that i n order for information to be meaningfully exchanged from the environment to an individual's behavioral re-sponses, i t must pass through a chain of mechanisms which he has termed the perceptual mechanism (P.M.), the decision mechanism (D.M.), and the effector mechanism (E.M.). The sequential organization of these mechanisms i s shown i n Figure 2 below. -> s e n s -> e s ~ ) -> M \ u / s c 1 .. e -> s -> Perceptual Mechanism Decision Mechanism Effector Mechanism Figure 2: The Central Mechanisms as described by Welford's (1965) Model of Human Behavior. The P.M. i s responsible for perceiving external stimulation and encoding this information into a usable form. The D.M. receives information from the P.M. i n the form of an information signal and i s responsible for -12-selecting and calling-out from the E.M. the appropriate response to match the incoming information. Once the D.M. has selected and retrieved the appropriate response the E.M. receives an information signal informing i t of the decision. The E.M. i s then responsible for controlling the response throughout i t s execution. The Perceptual Mechanism - Suggested Source of the Delay i n RT of Retardates Some researchers have suggested that the difference in RT of normals and retardates i s caused by deficiencies i n the P.M. of retardates. Baumeister, TJrquhart, Beedle and Smith (1964), hypothesized that, "The d i f f e r e n t i a l reactions of these two groups are partly a function of the intensity of the stimulus to respond, and further, that decreases i n intensity are more deleterious to the performance of the defective subjects." This study showed that the predicted Intelligence Groups by Stimulus Condition interaction was significant at the .05 l e v e l . Prom this Baumeister et al concluded that, " i t tends to support the hypothesis that re-tardates are re l a t i v e l y more susceptible to decreases i n the RT stimulus than normals". However, many weaknesses can be found i n the design of this study. Baumeister et al f a i l e d to control for threshold difference between the two populations, and used a constant preparatory interval. A second study was done by Baumeister, Hawkins and Kellas (1965), i n order to correct these faults. This study also found a significant interaction (Intelligence -13-by Stimulus Condition), and therefore Baumeister et al concluded that the difference i n RT of retardates and normals i s somewhat caused by deficiencies i n the P.M. of retardates. Although a significant interaction was found (.©5 level) i t i s noted by the author that the decrease i n the mean RT of the retarded Ss for the three different \ • preparatory intervals, at the greatest intensity l e v e l , was only 50 milliseconds more than the mean decrease of the normal Ss. The results of this study are shown i n Table 1 below. TABLE 1 MEAN RT FOR NORMALS AND RETARDATES UNDER VARIOUS CONDITIONS OP INTENSITY AND WARNING Group Intensity 25 db P.I. 50 db P.I. 75 db P.I. 4 sec 5 sec 6 sec 4 sec 5 sec 6 sec 4 sec 5 6 sec sec Normals .311 .290 .300 .279 .276 .268 .247 .244 .238 Retard-ates .470 .459 .462 .422 .415 .422 .360 .349 .361 Also, i t i s f e l t by the author that differences i n the per-ceptual capacities of retardates and normals explain only a small portion of the difference i n RT of these two groups -14-since the RT of the retardated group, at the highest i n -tensity l e v e l , was s t i l l 122 2/3 milliseconds slower than the RT of the normal group. Contrary to the findings of Baumeister et al (1964; 1965), Berkson (1960a) found that there was no difference i n the duration thresholds, and consequently speeds of perception, of retardates and normals, and thus concluded that intelligence isn't related to speed of per-ception. Berkson used a t-test to compare the mean duration thresholds of the retarded and normal groups. He found no significant mean difference at the .05 level of confidence. Research by Goldiamond i n I960, also showed no significant threshold difference between retardates and normals. However, contrary to both Berkson's and Goldia-mond' s studies, and congruent with earlier studies by Baumeister et al (1964; 1965), Spitz (196?) found that retardates, as a group, had a lower channel capacity i n their perceptual mechanisms than did normals subjects. Spitz tested three groups of Ss; retardates, and equal mental age and equal chronological age normals, on an absol-ute judgement task i n which they -were required to judge a position of a pointer on a horizontal l i n e . After each judgement they received feedback from a previously hidden scale. Spitz found that the channel capacities of the three groups were 2.68, 3 .13 and 3.43 bi t s , respectively. A lat e r study by Spitz (1969) also showed that retardates were more negatively effected by reduction of stimulus i n -formation i n a visual search task, than were normal Ss of - 1 5 -the same age. Spitz used three groups of subjects; adol-escent retardates and fourth and seventh grade normals. In analyzing the results of the three groups Spitz found that the loss of information had an equally negative effect on the retardates and the fourth graders, but i t had a much lesser effect on the seventh grade group. Spitz applied Weisser's ( 196 ? ) two-level hypothesis of mechanisms used i n visual search, to his results and concluded that the retardates and fourth graders were having d i f f i c u l t i e s i n the pre-attentive phase (organism scans the material), rather than i n the focal attention phase (organism con-centrates on and analyzes the chosen object). Spitz con-cluded that as non-retarded people mature they develop additional resources at the pre-attentive stage of alert-ness. He f e l t that these resources are lacking i n retarded Ss. A recent study by P a t r i c i a L. Austin ( 1969) found evidence which lends support to Spitz's ( 196? ) findings. Austin found that the difference i n RT between normals and retardates was greatly increased as the complexity of the perceptual task increased. These results are shown i n Figure 5 . Austin concluded that Educably Mentally Retarded (E.M.R.) Ss process information, i n their perceptual mechan-ism, at a slower rate than normals. However, i t i s f e l t by the author, that the results of this study are confounded since the complexity of the task was increased by increasing both the number of stimulus alternatives and possible re--16-1*84 O 1-6-[j 1-4* LU 1-24 10-•8-• 6 -RETARDATES NORMALS C O N D I T I O N S F I G U R E 3« RTs OF E.M.R. AND NORMAL Ss ON A GROSS MOTOR - PERCEPTUAL TIME TASK (AUSTIN I969>. - 1 7 -sponses, and by increasing the complexity of the stimulus patterns. It i s f e l t that the results do not clearly indicate whether the delay i n information processing, demonstrated by the retardates, i s occurring i n the per-ceptual mechanism or the decision mechanisms. Also i t i s the opinion of the author, that the results of Austin's study were further confounded by the fact that as com-plexity was increased i n the second condition, stimulus response compatibility was greatly decreased. Therefore, the author feels i t i s impossible to conclude whether the retardates are processing information of a perceptual nature, at a slower rate or whether their rate of learning i s slower. Effector Mechanism - Suggested Source of Delay i n RT of Retardates Compared to the amount of research that has been done on the P.M., very l i t t l e research has been done i n which the E.M. has been studied as the possible cause of the slow RTs manifested by retardates. Two recent studies have indirectly looked at the E.M. by trying to correlate intelligence with motor a b i l i t y of retardates. Dingman and Silverstein (1964) measured 265 retardates on RT and two tests of motor a b i l i t y (tapping and steadiness). They found that when the effects of tapping and steadiness were held constant the significant correlation between intelligence and reaction time, which they had obtained, disappeared. And when intelligence and steadiness were held constant the -18-significant correlation between reaction time and tapping s t i l l remained. A later study by Knights, Atkinson and Hyman (1967) found contrary to Dingman and Silverstein, that holding tapping constant did not significantly affect the correlation found between intelligence and RT. However, i t was noted by Groden (1969) that the tapping tasks used by Dingman and Silverstein, and Knights, Atkinson and Hyman were quite different. The task used by the former authors required quite a b i t more perceptual motor co-ordination than the one used by the l a t t e r authors. With this i n mind Groden designed a study i n which he included both a simple and a complex motor task. He found that the re-lationship between reaction time and mental age (r = -.710) was eliminated when the complex, but not the simple, motor s k i l l s were held constant. From his results Groden con-cluded: "Apparently i t i s not simple motor a b i l i t y which may be fundamental to the relationship between intelligence and reaction time since removal of effects of motor strength, finger o s c i l l a t i o n and motor a b i l i t i e s did not cause the relationship to disappear. In other words, motor a b i l i t y per se, seems to be irrelevant. Both reaction time and com-plex, perceptual-motor task, key press, re-quire something over and above what i s re-quired by such simple motor tasks as motor strength and finger o s c i l l a t i o n . " _ 1 9 -The Decision Mechanism - Suggested Source of Belay i n RT of Retardates A few researchers have regarded the functioning of the perceptual and effector mechanisms secondary, and instead looked towards the decision mechanism as the cause of the discrepancy i n the RTs of normals and re-tardates. In 1957 John Annett did a study entitled ".The Information Capacity of Young Mental Defectives i n an Assembly Task". Annett used three levels of retardates, low, mid and high grade, as his Ss. He had them perform an assembly task made up of four sub-components: reach, grasp, carry and assemble. Reaction time was measured as the time taken to perform the f i r s t component ( i . e . contact the peg). The informational load was increased from one to three bits of information. An analysis of variance on the pooled scores of the low, mid and high grades showed that the effect of the informational load was significant and the slopes of the graphed information processing lines of the three groups were significantly different. It i s noted by the author that the task used i n Annett's study did not minimize the role of the perceptual and effector mechanisms. In fact a great deal of perceptual and motor a b i l i t y was involved i n the task, but these were held constant while the information load was increased, and therefore i t i s questionable whether or not this affected the v a l i d i t y of the results. Berkson (1966c) did an interesting experiment i n -20-which he varied separately i n different conditions, the complexity of the stimulus and the complexity of the re-sponse. He found that there was no interaction between intelligence and task on conditions where stimulus com-plexity was increased, and therefore concluded that i t wasn't a difference in the decision mechanism of these two intelligence groups that was producing the difference i n RT. He did, however, find a significant interaction between task and intelligence group for the conditions i n which re-sponse complexity was progressively increased. From these results he concluded that the difference i n RT of retard-ates and normals i s caused by the functions involved i n tie i n i t i a t i o n of performance, ( i . e . effector mechanism). However, i t i s important to note that i t was late r pointed out i n a study by Hawkins, Baumeister, Koenigsknecht and Kellas ( 1965) that, "One of the d i f f i c u l t i e s with Berkson's studies was the confounding of the type of response with complexity of the dis-crimination. The response required i n simple reaction time was not the same as that employed i n the disjunctive task." With the faults of Berkson's study i n mind, Hawkins, Baumeister, Koenigsknecht, and Kellas ( 1965 ) designed a study to again look at the differences or si m i l a r i t i e s of the decision mechanism of normals and retardates. They increased the complexity of the response stimulus from one to two stimulus l i g h t s . Contrary to their hypothesis there wasn't a significant interaction between intelligence groups and complexity. It i s f e l t by the author that a -21-significant interaction wasn't found since the increase in complexity, from one to two stimulus l i g h t s , wasn't large enough to produce differentiation i n the rate of inform-ation processing of retardates and normals. In conclusion, a review of the literature to date studying the slower than normal RTs of retardates, shows conflicting and incomplete results. It points to a need for further research i n this area i n order to determine which component process of RT i s the cause of the slower than normal RT of retardates. More sp e c i f i c a l l y i t points to a need for a test of the decision mechanism since i t i s f e l t that to date there hasn't been an adequate test of the decision mechanism i n relation to the delay i n RT of retarded subjects. The literature strongly suggests that the delay in RT of retardates could be caused by deficiencies i n a l l three of the component processes of human behavior. Each of the three processes have been implicated by evidence found in different studies. However, i t i s f e l t by the author, important that each of these mechanisms be studied separately i n relation to the delay RT of retardates, so that the results are not confounded and that a clear picture regarding the RT of retarded subjects w i l l eventually emerge. CHAPTER I I I METHODS AND PROCEDURES Subjects Eight right-handed males, 10 to 12 years of age, estimated by t h e i r school teachers to be within the 95 -105 I.Q. range, from McBride School, were used as representatives of the normal population. Eight right-handed males, 10 to 12 years of age, estimated by their school teachers to be within the 40-54 I.Q. range, from Oakridge School, were used as represent-atives of the mentally retarded population. No subjects with a medical history of organic brain damage were i n -cluded i n this group.5 Apparatus The apparatus used i n this experiment, which i s shown i n Figure 4 , consisted of two arrays of red stimulus lights (four i n each array) and two corresponding arrays of black response buttons, (four i n each array). The stimulus lights were each half an inch i n diameter and arranged i n a straight line p a r a l l e l to the table top, with half an inch between each l i g h t . There was a distance of one and a half inches between the two stimulus lig h t arrays, and i n the middle of this space, half an inch lower -22-I - C O N T R O L BOX 2 - T I M E R 3 — CARDBOARD WALL 4 —STIMULUS COVERS 5 — STIMULUS LIGHTS 6— WARNING LIGHT 7 —RESPONSE BUTTONS l ro F I G U R E 4> A P P A R A T U S A N D P O S I T I O N O F S U B J E C T A N D E X P E R I M E N T E R A T A P P A R A T U S . -24-than the stimulus arrays, was situated a yellow warning li g h t , also half an inch i n diameter. The response buttons, which were arranged i n a slight arch, were situated below and forward from the stimulus l i g h t s . The two out-side buttons i n each array were one and three-quarter inches forward from the stimulus l i g h t , while the two inside buttons were one and a half inches forward. The panel on which the response buttons were mounted was one inch below the height of the stimulus l i g h t s . The response buttons were a quarter of an inch i n diameter with half an inch between each button. The arrays were arranged so that there was two and a half inches between arrays and the arrays were seven and a half inches from the front edge of the apparatus. The size and positioning of the response keys was designed to be suitable for the hand size of young children. The apparatus basically followed the standard de-sign of button press apparatus used i n CRT experiments. Three modifications were made from the standard design. As was already mentioned, the dimensions of the response buttons were modified i n order to suit the hand size of young children. Also strips of white tape (1/8) inch wide) were run from each stimulus lig h t to i t s corresponding response button i n order to increase the stimulus-response compatibility. This was done to assist the retarded Ss i n understanding the experimental task. A f i n a l modification was the hinging of metal plates above each of the stimulus - 2 5 -lights, so that when any of the stimulus lights weren't being used for a certain experimental condition they could be covered up by simply dropping the appropriate metal plates over them. This was done to make the experi-mental task less confusing for the young retarded and normal subjects. Position of Subject (S) and Experimenter (E) at Apparatus. The S sat directly i n front of the experimental apparatus (Figure 4) so that when both hands were on the key-board the subject's arms were par a l l e l to each other and perpendicular to the front edge of the apparatus. The height of the chair was adjusted to insure that the forearm of each S was pa r a l l e l with the floor when his hands were placed on the keyboard. Each S was permitted to adjust the distance of the chair from the table to a comfortable position. The experimental apparatus was placed on the table with the front edge one inch from the edge of the table. Six inches to the l e f t of the experimental apparatus a cardboard partition, two feet by two feet, was placed perpendicular to the edge of the tabie and extended two inches over the edge of the table. This partition was placed so that the subject could not view the timer or control apparatus when seated i n front of the experimental appar-atus. To the l e f t of the cardboard partition was the control apparatus and timer which was situated as far as possible (five feet five inches) from the experimental apparatus so -26-th e S would not receive auditory cues from the noise of the e l e c t r i c a l c i r c u i t s . The E sat directly i n front of the timer and control apparatus at a comfortable distance from the table where she could view the S. Experimental Design The present study was a two by four f a c t o r i a l experiment with repeated measures on one factor. Also a Latin square was used i n order to control for a possible learning effect over the four CRT conditions. The experiment involved two different I.Q. levels represented by two groups of Ss. Each S was required to perform a CRT task under four different levels of information. Level or condition one was a test of simple RT with only one stimulus presented. Level two consisted of a choice between two alternatives or one b i t of information. Level three and four respectively consisted of a choice between four and six alternatives. In terms of information theory four alternatives i s equal to two bits of information and six alternatives equals two decimal five eight b i t s of information. Each subject was given 10 t r i a l s on Condition one, 20 t r i a l s on Condition two, 40 on Condition three and 60 on Condition four, for a total of 130 t r i a l s per subject. Since only the right index finger responses were recorded i t was necessary to increase the number of t r i a l s as the number of alternatives were i n -creased over conditions. By increasing the number of t r i a l s i n proportion with the number of alternatives the prob-- 2 7 -a b i l i t y of a right index finger response being required was kept constant over the four conditions. Procedures The experiment began with the S entering the testing room and being seated i n front of the apparatus i n the position previously described. The S was given a suitable period of time (approximately three to five minutes) to acquaint himself with the equipment. He was then asked to l i s t e n while the instructions (Appendix B) were read to him, interspersed with demonstrations. The instructions were designed to be easily understood by the retarded Ss and were tested and refined during a p i l o t study with two retarded Ss. Following the reading of the instructions the S was told to place his eight fingers (1st, 2nd, 3rd and 4th finger of each hand) on the appropriate response buttons i n the manner described i n the instructions. A l l eight fingers were rested on the response keys for a l l conditions. The S was then given five learning t r i a l s on the condition on which he was going to be tested. It was found from a previous p i l o t study (Appendix A), done by the author, that a minimum of five practice t r i a l s _b" each level was necessary in order for the S to clearly understand the task. On the basis of this information both groups (normals and retardates) were given five practice t r i a l s on each condition. On completion of the learning t r i a l s the experimental t r i a l s were administered. At the beginning of eaeh t r i a l E gave -28-the command "ready" and then presented the warning l i g h t . Two, three or four seconds later a particular stimulus light was presented, depending on the condition being run, and the S responded by pressing the proper response button as quickly as possible. The time of the preparatory i n -terval and the presentation of the stimulus lights were randomly varied within each condition. A table of random numbers was used to determine the order of the stimuli and the preparatory intervals. The i n t e r t r i a l interval was approximately 10 seconds and encompassed the time required by E to record the S's response, to reset the clock and to place the control switch i n position for the next t r i a l . During the experimental t r i a l s the S was observed very closely by E for signs of fatigue or declining interest. If i t was deemed necessary by E the S was given a rest of five minutes. Following the rest period the S was given three practice t r i a l s before the experiment was continued. It was found in a previous p i l o t study (Appendix A) that i t was possible to run both Conditions 1 and 2 i n the same testing session, when they f e l l i n that order, but that 3 and 4 had to be tested separately since they were much longer and the retardates tended to fatigue quite quickly. Analysis of Data The time from when the stimulus l i g h t was presented to when the S pressed the response key was recorded as RT. The recorded value for each condition was the mean of 10 RTs - 2 9 -of the right index finger. Since there were eight Ss i n each group the mean RT for each condition, i n each group, represented the mean of 80 t r i a l s . The mean scores for each S, for each condition and in each group, were analyzed by an analysis of variance test i n order to determine the effect of information on RT, and to compare this effect for the two groups. A trend analysis was performed on the conditions and groups by conditions effects which showed a significant F r a t i o , i n order to see i f a linear trend was occurring and i f there was a difference i n the slope of the trend between groups. The difference i n simple RT (Condition one) between the two groups was tested for significance by both a planned comparison analysis and a t-test. The t-test was used after the results of the planned comparison analysis f a i l e d to achieve significance, since i t was f e l t that the error term used i n the planned comparison analysis wasn't an accurate estimate of variance for comparison of the simple RTs. A further explanation of the reasons for using the t-test i s presented i n Chapter IV. In order to look at the group v a r i a b i l i t y the standard deviation of the S means was computed for each group. This value was then squared to show the group variance of each group. The intra-individual v a r i a b i l i t y was determined by taking the standard deviation of the ten - 3 0 -rigrit index response scores of each S, on each condition. These values were then squared to reveal the i n t r a -individual v a r i a b i l i t y of each S on each condition. CHAPTER IV RESULTS AND DISCUSSION RESULTS Difference Between Group RT Over A l l Pour Conditions It can be seen i n Figure 5 that the retarded group was more detrimentally affected by increases i n information load than the normal group. Figure 5 shows that v/ith each increase i n information (from Condition one through Con-dition four) the difference i n RT between retardates and normals was increased. These results are also presented i n Table 2 which shows the mean RT and variance of each group for each condition. TABLE 2 MEAN RT AND VARIANCE OF RETARDATES AND NORMALS FOR CONDITIONS 1, 2, 3 AND 4 Retardates Normals RT Variance RT Variance Condition 1 0.516 0.0329 0.205 0.0013 2 0.749 0.0247 0.355 0 .0017 3 1.109 0.2367 0.431 0.0012 4 1.196 0.1376 0.463 0 .0027 -31-- 3 2 -F IGURE 5: M E A N R T A S A F U N C T I O N O F I N F O R M A T I O N L O A D - 3 3 -The results shown i n Table 2 revealed a much larger degree of group v a r i a b i l i t y within the retarded group than the normal group. It i s recognized by the investigator that the large difference i n group variances could possibly confound the analysis of the results, since homogeneity of variance must be assumed when using an analysis of variance test. However, i t has been shown (Welch, 1 937 ; David and Johnson, 1 951b ; Box, 1 9 5 2 ; and Horsnell, 1 9 5 3 ) , that i n the commonly occurring case, i n which group sizes are equal, or near equal, the analysis of variance test i s affected surprisingly l i t t l e by unequal variances. Box (1953) recommends, "Since this test i s also known to be very insensitive to non-normality i t would be best to accept the fact that i t can be used safely under most practical conditions." The results shown i n Figure 5 and Table 2 were analyzed by an analysis of variance test, the results of which are shown i n Table 3 . It can be seen that the main effects of Groups, Conditions and Groups by Conditions interaction are sig-nificant at the .01 level of confidence. A trend analysis done on the Conditions effect revealed a significant linear trend (.01 l e v e l ) , and the same analysis on the Groups by Conditions interaction showed a significant linear x linear effect ( .01 l e v e l ) . In order to determine i f the difference i n simple RT of the two groups, which i s shown i n Table 2 , was a -34-TABLE 3 ANALYSIS OP VARIANCE OP RT OVER CONDITIONS 1, 2, 3 AND 4 Source of Variance d.f. M.S. P P Groups 1 44.743 25.89 <.01 Subjects (Groups) 14 1.728 Conditions 3 7.350 28.21 <.01 Linear 1 21.08 81.07 <.01 Quad. 1 0.697 2.68 >.10 Groups x Conditions 3 1.724 6.62 <.01 Linear x Linear 1 4.799 18.45 <.oi Linear x Quad. 1 0.008 0.03 Subjects (Groups) x Con-ditions 42 0.260 Trials 9 0.114 2.33 Groups x Tr i a l s 9 .075 1.53 Subjects (Groups) x Trials 126 0.049 Conditions x Trials 27 0.065 1.41 Groups x Conditions x Tria l s 27 0.053 1.18 Subjects (Groups) x Conditions x Tri a l s 378 0.045 - 3 5 -significant difference a planned comparison was done on the results for Condition 1. Using this analysis, the difference i n simple RT between the normal and retarded group f a i l e d to reach a level of significance. In the planned comparison analysis the mean squares for the Subjects (Groups) by Conditions effect (0.260) was used as the error variance estimate. I t was f e l t that this was not an accurate estimate of error variance when comparing the simple RT of the two groups, since i t included the group variance for the CRT conditions. Table 2 shows that on Conditions three and four the variance -of the retarded group greatly increased while, by comparison, the group variance of the normal group remained the same. It can be seen from Table 2 that the variance estimate i s greatly i n f l a t e d by including the variance of the CRT conditions. Thus for these reasons a simple t-test analysis was used to compare the two simple RT means with the error term used being calculated on the group variance for Condition one, the simple RT condition. With equal Ss and heterogeneity of variance, t was calculated i n the usual manner, but the obtained value of t was evaluated i n terms of the tabled value for half the number of degrees of freedom that would have been available with homogeneity of variance. The t-test analysis revealed a significant difference (.01 level) between the simple RT of the normal and retarded groups. The results of both the planned comparison analysis and the t-test are shown i n Table 4-. ' -36-TABLE 4 T-TEST AND PLANNED COMPARISON ANALYSIS OP DIFFERENCE IN MEAN RT FOR CONDITION 1 t P Planned Comparison 1.22 > .05 t-test 5.18 <.01 Group V a r i a b i l i t y While the group means of the retardates are sig-n i f i c a n t l y different from those of the normals, i t has been found that the variance i n the retarded group, on a l l four conditions, was much greater than that of the normals (Table 2). It was also found that the variance of the normal group remained f a i r l y consistent over the four conditions, while the variance of the retarded group greatly increased on Conditions three and four, the more complex conditions. The variance of both groups has also been shown graphically in Figure 6. It was also discovered that while the retardate group was very variable, there was negligible overlap be-tween the scores of the two groups. A l l but one of the re-tardates scores was higher than the scores of the normal FIGURE 6: G R O U P V A R I A B I L T Y - I N D I V I D U A L M E A N R T S A S A rVJNCTSON C F I N F O R M AT I O N L O A D . -38-group. This over-lapping score occurred on Condition one, the simple RT condition, and can be seen i n Table 5» The observed absence of over-lapping scores reveals that even though the retardates scores are quite variable, their scores, as a group, are d i s t i n c t l y different from those of the normal group, Intra-Individual V a r i a b i l i t y A study of the intra-individual v a r i a b i l i t y of both the retarded and normal group has shown that the retardates were more variable within themselves than the normal SS (Table 5, Figure 7 ) . It was also found that the i n t r a -individual v a r i a b i l i t y of the retarded Ss greatly increased on Conditions three and four while the intra-individual v a r i a b i l i t y of the normal Ss remained r e l a t i v e l y unchanged over a l l four conditions. This trend i s similar to the trend found with the group v a r i a b i l i t y . Table 5 shows the individual variance of each subject on each condition. Figure 7 compares the mean intra-individual v a r i a b i l i t y of the retarded and normal group for each condition. - 3 9 -.5-b C Q < _ _ < D Q > D Z T < _ Z .05 -.01 •• / V / RETARDATES .005-.001., NORMALS 0 I 2 2.58 3 I N F O R M A T I O N (BITS) FIGURE 7: M E A N I N T R A - I N D I V I D U A L V A R I A B I L I T Y A S A F U N C T I O N O F I N F O R M A T I O N L O A D . -40-TABLE 5 INTRA-INDIVIDUAL VARIANCE OP RETARDATES AND NORMALS FOR CONDITIONS 1, 2, 3 AND 4 Conditions Group 1 Retardates Subject 1 0.016 0.108 0 . 132 0 .551 2 0.46? 0 . 1 07 0.643 . 0 .389 3 0.003 0 . 0 15 0.145 0.090 4 0.041 0 . 0 15 0 .110 0.162 5 0.009 0 . 012 0 .009 0 .021 6 0.005 0.008 0.006 0 . 002 7 0 .001 0.005 0 . 007 0 .011 8 0 . 012 0 . 007 0.004 0 . 015 Group II Normals 1 0.002 0.008 0.004 0.000 2 0.003 0.003 0.000 0.003 3 0.001 0.004 0.011 0.001 4 0.001 0.002 0.000 0.003 5 0.001 0.000 0.007 0.004 6 0.000 0.004 0.000 0.010 7 0.004 0.002 0.002 0.002 8 0.001 0.000 0.003 0.002 DISCUSSION Simple RT, Retardates and Normals The f i r s t hypothesis of the present study, that retardates, as a group, have a slower than normal RT, was supported since the RT of the retarded Ss was found to be significantly slower than the RT of the normal Ss. These results are supported by the findings of a large number of investigators (Berkson, 1960a, 1960b, 1960c; Berkson and Baumeister, 1967; Baumeister and Kellas, 1968; Jones and Benton, 1968; and Jones and Hinkle, 1 970 . Table 6 compares the simple RT of retardates and normals i n the present study with the results of two other studies. TABLE 6 COMPARATIVE ANALYSIS OP SIMPLE RT OP RETARDATES AND NORMALS IN THE PRESENT STUDY WITH THE RESULTS OF TWO OTHER STUDIES Present Study Baumeister & Kellas, 1968 Jones & Hinkle, 1970 Retardates Mean Variance 0.516 0.0329 0.311 0.0077 0.67 0.0676 Normals Mean Variance 0.205 0.0013 0.158 0.0013 0.34 0.0081 -400--41-Jones and Hinkle suggested as an explanation of slower simple RTs i n retarded Ss that, "perhaps even the simplest task to a retardate, presents i t s e l f as a choice situation. This i s suggested by the fact that the simple RT for the retard-ates was equal to the CRT for normals." The findings of the present study are contradictory to the theory offered by Jones and Hinkle. On the basis of the present results (Table 2) i t i s apparent that the slov/er simple RT of the retarded S cannot be wholly ex-plained by the theory that retardates deal with a simple RT task as a CRT situation, since i t i s observed that the simple RT of the retarded group i s greater than the CRT of the normal group, even on the six choice conditions i n the present study. I f the simple RT task does present i t s e l f as a CRT situation to the retarded S there must also be some other com-plication occurring i n the central mechanisms of the retarded Ss, which causes the simple RT of the retardates to be greater than the CRT of the normals. It has been hypothesized, i n the present study, that the slower simple RT of retardates i s caused by a slower than normal information processing rate i n the decision mechanism of retarded Ss. It has become well established that, as a group, the mentally retarded have a slower simple RT than normals, however researchers have not yet discovered which mechanism, or combination of mechanisms underlie t h i s evidenced slow-ness. Many theories have been offered as an explanation of this difference betv/een normals and retardates. Berkson , -42-(1960c) f e l t that some aspect of the response i n i t i a t i o n or execution, rather than the sensory or choice components of RT, may he effected i n retardates. Baumeister, Hawkins and Kellas (1965a) looked at general arousal l e v e l , and Holden ( 1965) suggested a pre-stimulus arousal defect. Annett ( 1957 ) f e l t that slower simple RTs i n retardates were the result of slower information transmission by re-tardates. If the decision mechanism i s at least one of the mechanisms involved i n the decreased performance of retardates then i t might be expected that increasing the information load i n a CRT experiment would be accompanied by a continually increasing d e f i c i t i n performance when compared to the performance of normal Ss. The following section examines this p o s s i b i l i t y . CRT, Retardates and Normals The second hypothesis of this study, which stated that retarded Ss process information i n their decision mechanisms at a slower rate than normal, was strongly supported by the results of this study. The analysis of variance table shown i n Table 3 indicates that the main effects, Groups, Conditions and Groups by Conditions inter-action were a l l significant at the .01 le v e l . The s i g n i f i -cant Groups effect indicates that the retardates' mean RT for a l l conditions was significantly different than the normals' mean RT for a l l conditions. The conditions effect confirms that there was a significant difference i n the mean RT recorded for each condition, for both groups. A - 4 3 -trend analysis done on the Conditions effect revealed a significant linear trend, indicating that the change which occurred for both groups over conditions was l i n e a r . This can be seen i n the graph presented i n Figure 5. A linear trend was expected for the normal group since this has already been shown by Hicks ( 1952 ) and Hyman ( 1 9 5 3 ) . The finding of a significant l i n e a r trend for the re-tarded group i s supported by the findings of Annett ( 1 9 5 7 ) » Berkson (1960c), Hawkins, Baumeister, Koenigsknecht and Kellas (1965), D. Jones and Benton (1968), and J. Jones and Hinkle ( 1 9 7 0 ) . The Groups by Conditions interaction, which was also found to be significant, confirms that the two groups reacted differently to the Conditions or increasing task complexity. A trend analysis done on the Groups by Con-ditions interaction indicated that the change occurring was a significant (.01 level) linea r change. This change can be seen i n Figure 5 which shows that the retarded group reacted much more negatively to the increase i n i n -formation than the normal group. These results support the major hypothesis of the present study, that retardates process information at a slower rate than normals. In terms of Welford's model, which i s shown i n Figure 2 , the results indicate that the slower than normal simple RT common to the retarded population i s p a r t i a l l y or wholly due to some deficiency i n the decision mechanism. The de-cision mechanism, according to Welford, i s responsible for -44-calling-out from the effector mechanism the appropriate response to match a particular information signal received from the perceptual mechanism. It encompasses two com-ponent processes; response selection and ret r i e v a l of motor programs from motor memory, which oversees the desired motor execution. The results of the present study suggest a deficiency i n one or both of these component processes of the decision mechanism. Since the present CRT experiment was designed to minimize the effects of the per-ceptual and effector mechanisms (Hyman , 1 953 ) i t i s f e l t by the investigator that the present study indicates that the decision mechanism i s almost entirely responsible for the delay i n RT of retardates. If the perceptual or effector mechanisms are at a l l responsible for causing the delay i n retardates i n the present study, i t i s f e l t that the effect of these mechanisms would be very minimal. However, i t i s realized by the investigator that other studies using a com-plex visual display or requiring a complex motor response might find that the slow simple RT responses of retardates are due to either the perceptual or effector mechanisms. The present study confirms that the decision mechanism of retarded Ss i s p a r t i a l l y responsible for the slower than normal simple RTs found i n retardates. Further studies must be done to determine the degree of responsibility of the decision mechanism. It i s f e l t , by the investigator, that further studies may reveal that the degree of involvement of any one of the three mechanisms varies betv/een Ss. V/ith testing -45-i t may be possible to determine i n which, of the three mechanisms a retarded person i s most limited. Consequently educators would then know which of the mechanisms to con-centrate their efforts towards when teaching a particular student. For example, i f a retarded child was found to be limited mainly by his effector mechanism he could be assigned remedial motor tasks. •Similar to the results of the present study, Annett (1957 ) found an interaction between intelligence and i n -formation load. Annett tested three levels of retarded Ss (high, medium and low grade), and found that for a l l three groups of Ss the regression of time on information was linear, and the slopes of each l i n e were s i g n i f i c a n t l y different from each other. These results are shown graphic-a l l y i n Figure 8. Annett concluded from his study that low and mid grade retardates have a lower information capacity than normals (although he did not test normals on the experimental task he used i n his study). Because of weaknesses i n the design of Annett's experiment, the results must be viewed with some skeptism. It i s f e l t by the investigator that Annett's peg-board assembly task f a i l e d to minimize the roles of the perceptual and effector mechanisms. Also, by recording Reach time Annett was actually measuring movement time (MT) rather than RT and thereby involving the effector mechanism. Annett's study has implicated the decision mechanism as the cause of the delay, however his experimental design -46-INFQRMATTON (BITS) FIGURE 8^ M E A N T I M E S A N D R E G R E S S I O N O F T I M E O N I N F O R M A T I O N ( A N N E T T 1957^ • -47-f a i l e d to isolate the decision mechanism. The present study was designed to eliminate as much as possible the roles of the perceptual and effector mechanisms and see i f the relationship found i n Annett's study s t i l l remained. The results of the present study revealed a significant interaction between Intelligence Groups and Conditions (information load) confirming that the decision mechanism i s p a r t i a l l y responsible for the delay i n RT i n retardates. A recent study by Berkson (1960c) offers a theory which i s contradictory to that advanced by the investigator. Berkson f a i l e d to find a significant interaction betv/een Intelligence and RT Task (information load) but did find a significant interaction betv/een Intelligence and Response Task (task requiring a complex motor response). He con-cluded that I.Q. i s related to functions involved i n the i n i t i a t i o n or performance of a response., D I t i s f e l t by the investigator that Berkson f a i l e d to find a significant I.Q. by RT task interaction because of the design of his experimental task. Berkson required his Ss to suppress a button with their right index finger. When a stimulus l i g h t came on they were to release the suppressed button and turn the li g h t off by pressing the appropriate response button. RT was measured as the elapse i n time betv/een the presentation of the stimulus l i g h t and the release of the suppressed button. It i s f e l t by the i n -vestigator that i t i s possible the Ss released the suppressed button and then made their choice of which button to press. -48-With the excitable nature observed i n the retarded children i n the present study and their strong desire to do the task as quickly as possible, as they were instructed, i t i s f e l t possible that the experimenter did not obtain true CRT scores for the retarded Ss. . A study by Hawkins, Baumeister, Koenigsknecht and Kellas ( 1 965 ) also f a i l e d to find a significant interaction between Intelligence and Task Complexity. This study pur-ported to improve on what the experimenters f e l t were the weaknesses of Berkson's study, by using a response which did not vary with task complexity and which was r e l a t i v e l y free of complex motor elements. It i s f e l t by the investig-ator that, while improving on Berkson's study the design of the study by Hawkins et a l . was s t i l l weak since task com-plexity was only increased from one to two alternatives. Hawkins et a l . also recognized this as a weakness of their study since they mentioned i n the discussion that a task involving a larger number of alternatives may have produced the predicted interaction.' Group V a r i a b i l i t y The results of this study (Table 2) revealed a high degree of variance in the retarded population as compared to the normal population. This large degree of group v a r i a b i l i t y noted i n the present study has also been found i n other studies dealing with the retarded population., Berkson and Baumeister ( 1967 ) noted that i n addition to -49-being slower than normals i n a RT task, retardates are more variable both between and within Ss. Baumeister and Kellas (1968) also noted that the RT responses of retarded Ss are much more variable than those of normal Ss. Many suggestions such as, individual differences i n arousal levels, individual differences i n receptability to encouragement, and individual differences i n noise levels have been offered to explain the large degree of group v a r i a b i l i t y i n RT responses of retardates. However, the results of the present study seem to indicate a relationship between v a r i a b i l i t y and informational load. Table 2 and Figure 6 clearly show a large increase i n the group v a r i -a b i l i t y of retardates on Conditions three and four, the more complex conditions. The group v a r i a b i l i t y of the normal Ss, by comparison, i s consistent over a l l four con-ditions. These results support the theory that the RTs of retarded Ss are influenced more by increases i n information load than normal subjects and consequently lend indirect support to the second hypothesis of this study. Intra-individual V a r i a b i l i t y The results (Table 5) of the present study showed that the retarded Ss were more variable within themselves than the normal Ss. This finding i s supported by both Berkson and Baumeister (196?) and Baumeister and Kellas (1968). Baumeister and Berkson f e l t that the high i n t r a -individual v a r i a b i l i t y of the retarded Ss was an indication - 5 0 -they were not working at their maximum efficiency l e v e l . Further they suggested that the true RT of retarded Ss may be near that of normals, but that they are simply less e f f i c i e n t at maintaining this optimum response l e v e l . Baumeister and Kellas (1968) f e l t that the high degree of intra-individual v a r i a b i l i t y i n retarded Ss was due to variable attentional or arousal fluctuations of retarded Ss. Figure 7 and Table 5 reveal that the intra-individual v a r i a b i l i t y of the retarded Ss of the present study greatly increased on the more complex conditions (three and four). The intra-individual v a r i a b i l i t y of the normal Ss, by comparison, was consistent over a l l four conditions. These results appear to indicate a relationship between i n t e l -ligence and intra-individual v a r i a b i l i t y . Thus i t would appear that the present study supports the finding that retarded Ss tend to be more variable than normal Ss both as a group and within themselves, and adds the fact that v a r i a b i l i t y of performance, for the retarded Ss increases as they are faced with increasing information demands. The exact cause of this increase in v a r i a b i l i t y i s not known. CHAPTER V SUMMARY AND CONCLUSIONS The purpose of this study was to determine whether the delay i n RT of retarded Ss v/as the result of retard-ates processing information i n their D.M. at a slower rate than normals. A CRT experiment, v/ith four different levels of information load, was used to examine the information processing rate of retarded and normal Ss. Eight, right-handed boys between 10 and 12 years of age, with average intelligence (I.Q. 95-105), and eight, right-handed boys between 10 and 12 years of age, with below average intelligence (I.Q. 40-54), were tested on a l l four conditions of the CRT task. An analysis of variance test was performed on the data. It was found on the basis of the test analysis that both the simple and choice RT of retarded Ss was si g n i f i c a n t l y slower than the simple and choice RT of the normal Ss. A significant Conditions effect revealed that there v/as s i g n i f i -cant differences i n the mean Rts recorded over a l l four con-ditions. A trend analysis done on the Conditions effect showed that the change which occurred, for both groups, over Con-ditions v/as linear. A significant Intelligence Groups by Con-ditions interaction further revealed that the two i n t e l l i -gence groups reacted differently to the four conditions, or - 5 1 -- 5 2 -increasing task complexity. A trend analysis done on the Groups by Conditions interaction indicated that the change occurring was a significant (.01 level) linea r change. These results indicated that, i n terms of the task used, the delay i n RT of retarded Ss i s the result of deficiencies i n the D.M. of retarded Ss. I t was also found that both the group and i n t r a -individual v a r i a b i l i t y of the RT responses of the retarded Ss was much greater than the group and intra-individual v a r i a b i l i t y of the RT responses of normal Ss. On Conditions three and four, the more complex conditions,there was a large increase i n both the group and intra-individual v a r i a b i l i t y of the RT responses of the retarded Ss. By comparison the group and intra-individual v a r i a b i l i t y of the RT responses of the normal SS was consistent over a l l four conditions. These results provided additional support for the theory that retarded Ss are more severely affected by increases i n information load than normal Ss. The conclusions of this experiment were: 1. That the simple RT of retarded Ss i s slower than the simple RT of normal Ss. 2 . That retardates process information i n t h e i r D.M. at a slower rate than normals, and that this con-tributes to the delay i n RT of retarded Ss. 3. That the RT responses of retarded Ss are more variable, both as a group and within individuals, than the RT responses of normal Ss. - 5 3 -4. That both the group and intra-individual v a r i a b i l i t y of the RT responses of retarded Ss increases with increases i n information, while by comparison, the group and intra-individual v a r i a b i l i t y of RT responses of normal Ss i s unaffected by increases i n information. Recommendations 1. That further studies testing RT of retarded subjects use a gross motor response, similar to that used i n Austin's study (1969), since retarded subjects appear to have d i f f i c u l t y with fine motor movements. 2. Further individual study of the perceptual decision or effector mechanisms of retarded Ss, i n comparison with normal Ss, i s recommended i n order to determine the unique characteristics of the retarded population i n relation to these mechanisms. 3. Further study of the perceptual decision and effector mechanisms of retarded Ss i s recommended i n order to determine to what degree each mechanism contributes to the delay i n RT of retarded Ss. REFERENCES Annett, J. The information capacity of young mental de-fectives i n an assembly task. Journal of Mental Science, 1 0 3 : 6 2 1 - 6 3 1 , 1 957 . ; Austin, P. L. Gross motor performance and motor learning of educable mentally retarded boys.Unpublished Study, University of•Alberta, 1969 . Baumeister, A., Urquhart, D., Beedle, R., Smith, T. Reaction times of normals and retardates under different stimulus intensity changes. American Journal of Mental Deficiency, 6 9 : 1 2 6 - 1 3 0 , 1 964 . Baumeister, A., Hawkins, W., Kellas, G. The interactive effects of stimulus intensity and intelligence upon reaction time. American Journal of Mental Deficiency, 6 9 : 5 2 6 - 5 3 0 , 1 9 6 5 . Baumeister, A., Dugas, J., Erdberg, P. Effects of warning signal intensity, reaction signal intensity, preparatory interval, and temporal uncertainty on reaction times of mental defectives. Psychological Record, 1 7 ( 4 ) : 5 0 3 - 5 0 7 , 1 9 6 7 . Baumeister, A., Kellas, G. Distribution of reaction times of retardates and normals. American Journal of Mental Deficiency, 7 2 ( 5 ) : 7 1 5 - 7 1 8 , 1968 . Baumeister, A., Berry, P., Forehand, R. Effects of secondary cues on rote verbal learning of retardates and normal children. Journal of Comparative and Physio-l o g i c a l Psychology,, 6 9 : 2 7 3 - 2 8 0 , 1969 . Berkson, G. An analysis of reaction time i n normal and mentally deficient young men - a duration threshold experiment. Journal of Mental Deficiency, 4 : 5 1 - 5 8 , I 9 6 0 ,a. Berkson, G. An analysis of reaction time i n normal and mentally deficient young men, - variation of com-plexity i n reaction time tasks. Journal of Mental Deficiency, 4 : 5 9 - 6 7 , 1960,b. Berkson, G. An analysis of reaction time i n normal and mentally deficient young men, ~ variation of stimulus and response complexity. Journal of Mental De-ficiency, 4 : 6 9 - 7 7 , I 9 6 0 , cl : - 5 5 -Berkson, G., Baumeister, A. Reaction time v a r i a b i l i t y of mental defectives and normals. American Journal of Mental Deficiency, 7 2 ( 2 ) : 262=266, 1 967 . Blackburn, Benton, A. Simple and choice reaction time and cerebral disease. Confinia Neurologica, 1 5 : 3 27 - 336 , 1 955 . Box, G. Non-normality and tests on variances. Biometrica, 4 0 : 3 1 8 - 3 3 5 , 1953 . Brinley, Botwick. Preparation.time and choice i n relation to age differences. Journal of Gerontology, 1 4 : 226 - 228 , 1959 . Costa, L.D. Visual reaction time of patients with cerebral disease as a function of length and constancy of preparatory interval. Journal of Perceptual Motor S k i l l s , 1 4 : 3 9 1 - 3 9 7 , 1 9 6 ? : Czudner, G., Rourke, B. Simple reaction time i n "brain-damaged and normal children under regular and irregular preparatory interval conditions. Journal of Perceptual Motor S k i l l s , 3 1 : 7 6 7 - 7 7 3 , 1 970 . Deich, R. Reading time and error rates for normal and re-tarded readers. Journal of Perceptual Motor S k i l l s , 3 2 : 6 89 - 690 , 1 9 7 1 . Dingman, H., Silverstein, A. Intelligence, motor dis-a b i l i t i e s and reaction time i n the mentally retarded. Journal of Perceptual Motor S k i l l s , 1 9 : 791 - 794 , 1964. E l l i s , N.? Sloan, W. Relationship between intelligence and simple reaction time i n mental defectives. Journal of Perceptual Motor S k i l l s , 7 : 6 5 - 6 7 , 1 957 . E l l i s , N. International review of research i n mental re- tardation. New York: Academic Press, 2 : 2 9 - 5 6 , T~5~; F i t t s , P., Posner, M. Human Performance. Belmont, C a l i -fornia: Brooks/Cole Publishing Co., 1 969 . Gallagher, J. Comparison of retardates and normals on subjective organization i n short term memory. Amer-ican Journal of Mental Deficiency, 7 3 : 6 6 1 - 6 6 5 , i w : : Groden, G. Mental a b i l i t y , reaction time, perceptual motor and motor a b i l i t i e s i n handicapped children. Journal of Perceptual Motor S k i l l s , 2 8 : 2 7 - 3 0 , 1 969 . - 5 6 -Hawkins, W., Baumeister, A., Koenigsknecht, R., Kellas, G . Simple and disjunctive reaction times of normals and retardates. American Journal of Mental De-ficiency, 6 9 : 536 - 539 , 1965 . Hick, W. Gn the rate of gain of information. Quarterly Journal of Experimental Psychology, 4 : 1 1 - 2 6 , 1 9 5 2 . Holden, E. Reaction time during unimodal and trimodal stimulation i n educable retardates. American Journal of Mental Deficiency, 9 : 183^190, 1965 . Hyman, R. Stimulus information as a determinant of re-action time. Journal of Experimental Psychology, 4 5 : 1 8 8 - 1 9 6 , Jones, D., Benton, A. Reaction time and mental age i n normal and retarded children. American Journal of Mental Deficiency, 7 3 ( 1 ) : 1 4 3 - 1 4 7 , 1968 . Jones, J., Hinkle, B. Var i a b i l i t y i n reaction time of normal and educable mentally retarded children. Paper presented at the meetings of the Southeastern Psychological Association, 1 970 . Kellas, G . , Baumeister, A. Effects of warning signal duration on the reaction time of retardates. American Journal of Mental Deficiency, 7 2 ( 5 ) : 6 6 8 - 6 7 3 , 1968 . Spitz, H. Information Transmission i n an absolute judgement task with feedback, using normal and retarded subjects. Journal of Comparative and Physiological Psychology, 4 : 85-92, 1967. Spitz, H. Effects of stimulus information reduction on search time of retarded adolescents and normal children. Journal of Experimental Psychology, 8 2 : 4 8 2 - 4 8 7 , 1969 . Stevens, H., Heber, R. Mental retardation: A review of research. Chicago: University of Chicago Press, 1964. Welford, A. Performance, biological mechanisms and age: A theoretical sketch. Behavior, Aging and the Nervous System, edited by Welford, A. and Birren, J. Springfield, I l l i n o i s : C. C. Thomas, 1 9 65 . Welford, A. Single-channel operation i n the brain. Acta Psychologiea, 2 7 : 5 - 22 , 1967 . Welford, A. Fundamentals of S k i l l , London: Methuen, 1 968 . - 5 7 -APPENDIX A RESULTS PROM PILOT STUDY - 5 8 -RESULTS OP TWO SUBJECTS FOR FOUR LEVELS OF INFORMATION LOAD INFORMATION SUBJECT SUBJECT A (Male 17 yrs) • Bits 1 Bits 2 Bits 2.58 : .203 .359 .973 . 684 .4-32 . 375 .533 . 734 . 615 .397 .574- .876 .132 . 424 .601 .711 .14-2 .350 .407 .691 .159 .299 .4-31 .753 .213 .400 .722 .893 .181 .3^7 .581 .751 .191 .560 .373 .901 .198 .301 .614 .962 SUBJECT B (Female 11 yrs) .212 .391 .233 . 444 .241 . 350 .217 .372 .199 .451 .209 . 477 .202 .516 . 194 .497 .221 .481 .213 .518 .995 .606 .726 .406 .714 .573 . 584 .358 .534 .648 .736 .390 .538 .376 .521 .366 .600 .550 .443 .308 -59-MEAN RESULTS OF TWO SUBJECTS FOR FOUR LEVELS OF INFORMATION LOAD INFORMATION 0 BITS 1 BIT 2 BITS 2.58 BITS SUBJECTS SUBJECT A .246 .381 .580 795.6 SUBJECT B .214 .450 .641 558 - 6 0 -APPENDIX B INSTRUCTIONS USED IN PILOT STUDY AND STUDY -61-INSTRUCTIONS - (subject's name) , this i s a game to see how fast you can press the buttons with your fingers. In order to play properly you must put one of your fingers on each of the black buttons, (demonstrate). Make sure you curl your other fingers out of the way, (demonstrate). Try pressing each button one at a time, with the finger that i s covering i t . (Practise 3 or 4 times, or u n t i l subject does five consecutive t r i a l s without a mistake). Now, when I say "ready" make sure that your fingers are on the buttons, and watch the yellow l i g h t , (E. points to warning l i g h t ) . Just after I say "ready" the yellow l i g h t w i l l come on, (demonstrate). As soon as that l i g h t goes off one of the red lights w i l l come on, (demonstrate). - You must try to make the red l i g h t go off as fast as you can by pressing the button that i s joined to i t by the tape (3 or 4 t r i a l s ) . For the f i r s t game I am going to cover a l l but two, (1, 4 or 6) of the red l i g h t s . This means that as soon as the yellow l i g h t goes out either this l i g h t (point) or this l i g h t (point) w i l l come on. You must decide which li g h t has come on and which button to press to make i t go off. Remember you must turn i t off as fast as you can. (5 practice t r i a l s on each condition to be tested). -62-APPENDIX C INBIYIBUAL SCORE SHEETS - 6 3 -NAME: AGE: CONDITION A CONDITION B Foreperiod Results Stimulus Results 2 1. 1 w 1. 4 2. 2 2. 3 3 . 2 3 . 4 4. 2 4. 3 5. 2 5. 4 6. 1 6. 4 7 . 1 7. 4 8. 1 8. 1 9 . 1 9 . 3 10. 2 10. 1 2 11. 3 1 12. 2 2 1 3 . 4 2 14. 1 1 1 5 . 4 1 16. 1 1 1 7 . 1 2 18. 4 1 1 9 . 2 2 20. -64-NAME; AGE: CONDITION C CONDITION D Fore- Stimuli Results Stimuli Results Stimuli Results Stimuli Results period 4 2 1. 4 3 1 . 1 1. 1 31. 4 3 2. 1 3 2 . 2 2. 4 3 2 . 4 2 3. 3 33. 6 3. 5 33. 3 4 4. 2 34. 3 4. 1 34. 4 3 5. 1 35. 3 5. 5 35. 2 4 6. 2 36. 1 6. 3 36. 3 1 7. 3 37. 3 7. 4 37. 2 4 8. 4 38. 2 8. 5 38. 2 4 9. 3 39. 3 9. 2 39. 4 3 10. 1 40. 3 10. 6 40. 2 4 11. 4 11. 1 41. 2 3 12. 1 12. 6 42. 2 4 13. 4 13. 1 43. 3 4 14. 5 14. 5 44. 2 1 1 5 . 3 1 5 . 6 45. 2 1 16. 3 16. 4 46. 2 3 1 7 . 6 1 7 . 3 47. 2 1 18. 2 18. 2 48. 4 2 1 9 . 6 1 9 . 4 49. 3 1 20. 1 20. 5 50. 4 2 21. 5 21. 4 51. 2 3 22. 2 22. 6 52 . 2 2 2 3 . 5 2 3 . 2 53. 2 2 24. 2 24. 6 54. 2 2 2 5 . 1 2 5 . 6 55. 3 2 26. 1 26. 2 56. 4 1 2 7 . 5 2 7 . 3 57. 3 4 28. 4 28. 4 58. 2 1 29. 2 29. 5 59. 3 3 30. 4 30. 6 60.