DIFFERENCES ASSOCIATED WITH CONCEPTUAL TEMPO ON ACADEMIC ACHIEVEMENT, IQ, AND INTERNAL~ EXTERNAL CONTROL MEASURES Dissertation for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY HENRY TAYLOR FRIERSON, JR. 1974 ""“’ rm 71’ I Michigan 5398 Univerei y 4.0" a ‘ t’ WW 7293mm . n“ 1‘ E ’57" “a" This is to certify that the thesis entitled . DIFFERENCES ASSOCIATED WITH CONCEPTUAL TEMPO ON ACADEMIC ACHIEVEMENT, IQ, AND INTERNAL- EXTERNAL CONTROL MEASURES presented by Henry Taylor Frierson, Jr. has been accepted towards fulfillment of the requirements for Ph. D. degree in Educational Psychology Adz/z: Jig/QM “I Major profe sor ( Date November ISI 1921; 0-7 639 ABSTRACT DIFFERENCES ASSOCIATED WITH CONCEPTUAL TEMPO 0N ACADEMIC ACHIEVEMENT, IQ, AND INTERNAL- EXTERNAL CONTROL MEASURES By Henry Taylor Frierson, Jr. Research findings suggest that conceptual tempo is highly associated with academic success. Conceptual tempo refers to the disposition of an individual as related to response time and accuracy in making a choice between two or more difficult alternatives. It has been described as a hypothesis testing process. The majority of past investigations concentrated only on the reflective (slow and accurate) and the impulsive (fast and inaccurate) dimensions of conceptual tempo while ignoring the other two dimensions-efast— accuracy and slow-inaccuracy. One objective of this study was to examine not onlyreflection and impulsivity but fast-accuracy and slow-inaccuracy as well. Socioeconomic status (SES) is another variable that has been reported to be closely related to academic achievement. The purpose of this study was to examine differences amont the four dimensions Henry Taylor Frierson, Jr. of conceptual tempo and also between high and low SES along measures of academic achievement, IQ, and internal-external control. The major hypotheses were (a) conceptual tempo is associated with measures of academic achievement, IQ, and internal-external control to the extent that reflectives and fast accurates will display higher mean scores on all measures and (b) SES is also related to academic achievement, IQ, and internal-external control measures to the degree that high SES §s will demonstrate higher mean scores across all employed measures. Eight dependent variables were employed. Differences in performances on the Metropolitan Achievement Math Test (MAMT), the Minimal Perfor- mance Objective Reading Test (MPORT), an IQ test, teacher ratings, the Intellectual Achievement Responsibility Questionnaire, and the Sense of Control Scale were investigated. The Matching Familiar Figures Test was used to determine the conceptual tempo of subjects and the McGuire-White Scale provided a measure to classify subjects within SES levels. The subjects were 180 fourth grade children from seven different elementary schools in a midwestern city. The results showed that high SES subjects had significantly higher mean scores on each dependent variable, thus indicating that SES was associated with the dependent measures as predicted. Within the two SES groups, conceptual tempo was shown to be strongly asso- ciated with academic achievement only among low SES children. 2 Henry Taylor Frierson, Jr. Although this was only to the extent that fast accurate youngsters were the high performers. There were no significant differences among the other three low SES groups. Overall, however, when disregarding SES reflectives and fast-accurate displayed higher performances on the dependent measures than impulsives and slow-inaccurates. Significant differences were found when reflectives and impulsives were compared on the MPORT and teacher rating, also significant differences were found between reflectives and slow-inaccurates on the MAMT. Fast accurates scored significantly higher on the MPORT, the MART, the MAMT, IQ, and teacher ratings than impulsives, and also performed sig- nificantly higher on the MART, the MAMT, IQ, and teacher ratings than slow-inaccurates. No significant differences between fast-accurate and reflectives nor between slow-inaccurate and impulsives were found. Also no significant differences on the internal-external measures of control were found that indicated any association with conceptual tempo. In relation to academic achievement, the results indicated that of the two components (response time and accuracy) that make up con- ceptual tempo, the most important is accuracy. However, for the low SES child, speed as well as accuracy appears to be the most conducive in relation to school success. The results suggested that if there is continued future attempts to modify conceptual tempo, the concentration Henry Taylor Frierson, Jr. should not be on response time but upon cognitive strategies that will improve the individual's accuracy. DIFFERENCES ASSOCIATED WITH CONCEPTUAL TEMPO ON ACADEMIC ACHIEVEMENT, IQ, AND INTERNAL- ' EXTERNAL CONTROL MEASURES By Henry Taylor Frierson, Jr. A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree Of DOCTOR OF PHILOSOPHY Department of Counseling, Personnel Services, and Educational Psychology I974 © Copyright by Henry Taylor Frierson, Jr. 1974 ii DEDICATION Dedicated to my loving wife Wendolyn and to those who offered me faith and spiritual support. ACKNOWLEDGEMENT Acknowledgement is presented to all of the beautiful children who acted as subjects in this study and also to a good Brother who provided me with the opportunity to collect the data, and finally to those who aided me during the data collection. I would like to especially thank Dr. George Ferre, Dr. Wilbur B. Brookover, and Dr. Maryellen McSweeney. As members of my guidance committee, they demonstrated that their expecta- tions of me were high and would accept nothing less than what their expectations dictated, and at the same time supplying reinforce- ments that enabled the successful completion of this study. I am Offering a special sincere acknowledgement to Dr. Walter G. Hapkiewicz who was not only my academic adviSor and guidance committee chairman but a wonderfully true friend as well. I will always revere the warmth and sincerity that he so willingly shared. His overall contributions to the development of this study was par excellence. I. '- I would also like to extend thanks to Dr. Howard Teitelbaum for his aid in computer programming in addition to his valuable advice and suggestions on methodology. At last I would like to acknowledge my faMily whose strength provided me with a foundation on which I could lean. First of all, /‘ my wife Wendolyn whom I love dearly. She has constantly demonstrated her love for me and has stood by me during my period at Michigan State University. My mother Lorraine Blount, to whom I owe so much it can never be described in words. The same can be truly said of my uncle and aunt, James and FlorenCe Thompson. My father Henry Sr. who wanted so much for his children in all realms of life. My aunt Harriett Groves who Offerred me lOve and encouragement. My grand- mother Florence Frey who has always been proud of her grand-children. My sister Ava, my brothers Clyde,'3ack3 and James, and my cousin Clarissa who by looking up to their "big brother," have made me feel so important and Significant. Finally, my wife's parents, Theodore and Ann Lockhart, who truly see me as their son. TABLE OF CONTENTS LIST OF TABLES ...................... LIST OF FIGURES ...................... LIST OF APPENDICES .................... INTRODUCTION ....................... SECTION I INTRODUCTION ....................... Definition of Cognitive Style ............. A Brief Historical Perspective of Cognitive Style. . . Description of Conceptual Tempo ............. Relationship of Selection Strategies to Conceptual Tempo ....................... REVIEW OF RELATED RESEARCH ................ Conceptual Tempo and School Success .......... Modification of Impulsivity .............. Observing Behavior and Conceptual Tempo ........ The Association of Conceptual Tempo and Information Processing ..................... The Relation of Conceptual Tempo and Problem Solving ...................... vi Page viii ix @000) II II IS 18 19 22 TABLE OF CONTENTS (c0nt'd.) Page Conceptual Tempo and Decision Making .......... 23 Conceptual Tempo and Anxiety .............. 24 Sex Differences in Conceptual Tempo ........... 26 A Rationale For Observing Four Dimensions of Conceptual Tempo ......................... 27 Socioeconomic Status and Conceptual Tempo ........ 29 SUMMARY . ' ................ ~ ......... 32 REFERENCES . . .' ...................... 34 SECTION II JOURNAL ABSTRACT ...................... 39 JOURNAL ARTICLE ...................... 4l Method ......................... 44 Subjects ...................... 44 Instruments and Procedure ........... '. . 45 Design ....................... 50 Results ........................ 5l Discussion ........................ 58 References ...................... .. 64 APPENDIX ........................ 67 vii Table LIST OF TABLES Page The number of $5 per full cell for each category except IQ. Numerals in parentheses represent the number of 55 whose IQ data was collected ........... 51 Cell means for high SES reflective, impulsive, slow- inaccurate, and fast-accurate 55. Standard deviations are in parentheses ............. 53 Cell means for low SES reflective, impulsive, slow- inaccurate, and fast-accurate 85. Standard deviations are in parentheses ............. 53 The means for high and low SES $5 on the eight observation measures. Standard deviations are in parentheses ...................... 54 The means for reflective, impulsive, slow-inaccurate, and fast-accurate $5 on the eight observation measures. Standard deviations are in parentheses. . . 54 viii LIST OF FIGURES Figure Page l. Dimensions of reflection-impulsivity .......... 28 2. Dimensions of four levels of conceptual tempo: Reflection, impulsivity, slow inaccuracy and fast accuracy ..................... 29 LIST OF APPENDICES Appdhdix - Page A. Dissertation ................... 69 B Method ........... . .......... 89 C. Hypotheses .......... I .......... 97 D Overall Report Of Results and Tables ....... 99 E. Graphs of Results ................. llO F. Tests of Hypotheses ................ 113 G. The Matching Familiar Figures Test ........ llS H. Appendix B. From the McGuire-White Scale ..... I45 I. The Intellectual Academic Achievement l49 Responsibility Questionaire (IAR) ........ J. A Sense of Control Scale ............. 15AQ= INTRODUCTION The structure of this dissertation is a departure from the traditional format of five chapters (introduction, procedures, re- sults, summary, and conclusion) that is common at Michigan State University. Instead, it will be in two sections plus the appen- dices. The first section is an extensive overall review of the literature. The objective is to provide the reader with a back- ground pertaining to the central topic of the study. The second section is the study itself which is presented in American Psycho- logical Association journal style. The aim Of the second section is not only to demonstrate research and writing capabilities that would satisfy a major requirement for completing a doctoral pro- gram at Michigan State University, but also to provide the reader a study that can easily be read without expending a great deal of time in the procedure. The third section is the appendices which contain the proposal, a method section, the hypotheses, the tests of hypotheses, graphs, a more extensive results section, and the instruments used in the study. The appendices will serve to pro- vide the reader with added information and materials that are not presented in sections one and two. SECTION I INTRODUCTION Concethal tempo is a cognitive style subset that appears to have widespread educational implications for all individuals.. However, conceptual tempo is just one of several reported subsets of cognitive style. Although conceptual tempo is the focal point, other cognitive style subsets such as "constricted-flexible control," "conceptual versus perceptual-motor dominance," "strong versus weak automatization," analytical field approach, global approach, and selection strategies will also be discussed. The objective is to present at least a brief explanation and description of the other subordinates of cognitive style. Special attention will be paid to the concept of selection strategies which appear to belclosely re- lated to conceptual tempo. An extensive review Of related research in- volving conceptual tempo is presented to emphasize the educational implications of conceptual tempo and to point out the possible re- lationship that it has with socioeconomic status. Also, a discussion of the rationale for the close examination Of two dimensions of conceptual tempo that are largely ignored in studies in the area is presented. Definition of Cognitive Style The contemporary definition Of the cognitive style construct presented by Kagan, Moss, and Sigel (I963) was that cognitive style is a measure of the stability of individual preference in percep- tually organizing and conceptually categorizing the external environ- ment. In the same vein, Hervey (I966) stated that cognitive style is the mode in which an individual organizes his experiences and perceives and understands his environment. At this point the term cognitive style is used widely to refer to a variety of behavioral consistencies in the processing of information (Stanes, I973). A Brief Historical Perspective Of Cognitive Style The construct of cognitive style has its roots in psychoana- lytic theory under the rubric of cognitive control. The concept was first described as the distinctive manner by which individuals face reality (Klein, l95l). In the psychoanalytic theoretical structure, the emphasis was on the relationship between cognitive controls and ego defense mechanisms, and also mediations of perceptual processes along the cognitive control principle of "constricted-flexible control" (Kagan and Kogan, I970). “Constricted-flexible control” refers tO the capability Of responding correctly in a task while facing a field of conflicting cues (Klein, I954). Naming the color word, for example, that is printed by a set of incongruent colored letters, such as the word "blue" printed in green letters. Flexibility ' is exhibited when the individual can successfully perform the task. If the individual, on the other hand, states the color word as being the same as the incongruent colored letters, then this characterizes the constricted mode° I Klein (I958) gave cognitive control the label Of cognitive style. Cognitive Control or style as perceived by Gardner and others (I959) Of ego psychology orientations, was defined as the presence of stabilized ego structures that integrate specific adaptive inten- tions with specific environmental situations. Kagan and Kogan, however, pointed out that the definition provided very little insight in examining individual differences in the functioning Of cognitive controls because of the use Of “intentions“ as a variable. The term "intentions" had a very vague meaning, and therefore it was difficult to employ the ego psychology definition of cognitive style to explain differing modes Of cognitive functioning between individuals. This was especially true when assumptions had to be made about "intentiOns" (for example, in experimental research studies). On an experimental basis, the "intentions" of the subjects should be assumed to be more alike than different. Therefore, one must assume that all the subjects intend_to perform to the best of their ability, unless there is inde- pendent evidence that indicates otherwise. ”Intentions," unfortunately because of its vague meaning,“did not readily lend itself to such an assumption. The restrictions Of the psychoanalytic theoretical 5 framework became evident among researchers in cognitive processes. Those who studied cognitive styles were finally constrained to work out of a conceptual structure that was cognitively rather than psychoanalytically oriented. Broverman (l960a, l960b) stated that although cognitive style is theoretically thought to be a function of common response strengths within particular classes of behavior, it can be viewed as the re- lationship between abilities within individuals. Broverman confirmed the existence Of what he interpreted as two cognitive styles: "Conceptual versus perceptional-motor dominance" and “strong versus weak automatization". "Conceptual versus perceptual-motor dominance," refers to specialization in novel, difficulty or concentration de- manding conceptual behaviors, while at the same time displaying a deficit in skills of novel or difficult perceptual tasks, or vice versa; "strong versus weak automatization," refers to high or low ability in performing simple highly practiced tasks (Broverman, I964). Broverman, Broverman, and Klaiber (I966) described automatized behaviors that are so highly practiCed and overlearned that only a minimum of conscious effort is neCessary for carrying out their execution efficiently. Examples are walking, maintaining one's balance, reading, maintaining perceptual constancies, and talking. When various controls (or later, styles) were recognized by researchers, they were held to be Organized as superordinate structures. Witkin et al (I962) were among the first to look at Specific subsumers Of cognitive styles when they described indivi- duals using an analytical field approach or a global approach. Their work dealt with field dependence-independence. The field independent or analytical field individual tends to experience objects analytically and discrete from their backgrounds. The field de- pendent or global fashion while passively conforming to the influences Of his environment. Witkin and associates indicated that an in- dividual's predisposition toward global or analytical inclinations ' characterized not only his perceptual activity but his problem solving activities as well. Description Of Conceptual Tempo Conceptual tempo, as defined by Kagan, is divided into two dimensions: reflection and impulsivity (or more concretely, Slow accurate). Kagan, Pearson, and Welch (I966) described the reflection impulsivity disposition as: Some children impulsively report the first hypo- thesis that occurs to them, and this response is Often incorrect. The reflective child, on the other hand, delays a long time before report- ing a solution and is correct. The reflective child considers the differential validity of alternative answers, makes fewer errors in read- prose-or in recalling serially learned material, and persists/longer with difficult tasks. The reflective child wants to avoid making an error and and inhibits potentially incorrect hypotheses. The impulsive child seems minimally concerned about mistakes and makes his decisions quickly. Response time or latency and rate of errors are the indices of the reflection-impulsivity dimension. Reflective (slow accurate) individuals have longer response times and lower error rates than those who are classified as impulsive (fast inaccurate). The reflection—impulsivity disposition was found to be very stable--for a psychological attribute--across problem solving tasks with different levels of difficulty. Kagan et al (I964) were the first to report conceptual tempO'S unusual stability over time and generality across tasks. Because of its stability, conceptual tempo is believed to be an important characteristic of an individual's psychological organization (Yando and Kagan, I970). As was pre- viously mentioned, the dimensions Of "fast accurate" and "slow inaccurate" will be incorporated in the study. The disposition Of fast accuracy-slow inaccuracy is assumed to be just as stable as the reflection-impulsivity disposition and also is an equally valid subordinate Of conceptual tempo. Relationship_of Selection Strategies to Conceptual Tempo Closely related to conceptual tempo is Bruner, Goodnow, and Austin's (I956) concept of selection strategies. Conceptual tempo and selection strategies both can serve to indicate how a pupil will perform intellectually in the classroom, for both involve cognitive processes that are school related. Selection strategies are also subsets of cognitive style and appear to be of great potential value in explaining the manner in which reflectives and impulsives process information. Selection strategies describe various ways by which the learning of concepts is attained by controlling the order in which‘ examples and non-examples appear. Bruner and associates described three Objectives of selection strategies: I) to ensure that en- countered instances will contain appropriate information, 2) to make the assimilation of the infOrmation less cognitively strainful and 3) to control the amount of risk in the attainment of a correct solution. Bruner et al depicted four selection strategies. Briefly they are: conservative focusing—-the use of an example of a particu- lar concept as a focal point while systematically changing one attribute of the example at a time in order to discover those attributes which are essential; focus gambling--similar to conserva- tive focusing except more than one attribute is changed during a single period of time; simultaneous scanning--the formulation of several hypotheses about the attributes of a concept followed by a search for appropriate examples; successive scanning—-similar to Simultaneous scanning but only one hypothesis is formulated and tested. Conservative focusing almost always ensures success; it meets all the Objectives of a selection strategy. Focus gambling, on the other hand, produces cognitive strain in that more information must be assimilated and it is a high risk strategy may obtain the correct solution quickly or conversely, be totally incorrect. Simultaneous scanning, also a high risk procedure, produces a great deal of cognitive strain in that more information must be assimilated and it is also a high risk strategy. Simultaneous scanning, also a high risk procedure, produces a great deal of cognitive strain within the individual because of the need to keep track of all the infor- mation in the formulated hypotheses. Successive scanning, a low risk strategy, reduces the liklihood of the encounter of instances embedded with relevant information. FOCUSers have been described as systematic attention deployers and scanners as unsystematic attention deployers (Santostefano and Paley, I964).. Laughlin (I973) discussed a relatively new concept in the area of selection strategies known as tactical strategies. The existence of tactical strategies was actually proposed by Johnson (l97l). However, Laughlin stated that tactical strategies are a set of procedures for testing hypotheses by partitioning the total set of hypotheses into subsets Of defensible or strong hypotheses and indefensible or weak hypotheses. Given that description, tactical strategies are theoretically less efficient than either conservative focusing or focus gambling primarily because the involved processes are more demanding in terms of memory and inference requirements. Selection strategies with high risks Should prove attractive to the impulsive child; the low risk processes , particularly con- servative focusing, should be attractive to the reflective youngster (DeCeddo, I969). Travers (I963) proposed that adjustments in teach- ing be made to coincide with a pupil's selection strategy. Similarly, Kagan stated that adjustments be made by the educational system to meet the cognitive needs Of pupils, especially in the area Of conceptual tempo. IO Although selection strategies and conceptual tempo appear Nery similar, they are conceptually different; Selection strategies are ways in which a concept is learned--by either a focusing or scanning procedure. This occurs thrOUgh a cognitive process of monitoring the sequence in which examples and non-examples of concept attributes appear. Conceptual tempo, on the other hand, refers to Speed and accuracy in making a response to two or more difficult alternatives. Correct response, as associated with conceptual tempo, indicates that a concept has been Iearned,rather than h9w_ it is learned, as is the case of Selection strategies; The two constructs are related in that an individual with a certain concep- tual tempo is inclined to use a selection strategy that is con- gruent with his personality. As an illustration, a reflective individual would not be inclined to use either focus gambling or simultaneous scanning, because of the greater chance for error and the rapid response time inherent in the two strategies. REVIEW OF RELATED RESEARCH Conceptual Tempo and School Success After studying a sample of children over a 2 I/2 year period, Messer (I970a) Observed that youngsters who failed a grade were sig- nificantly more impulsive than those who were promoted. Those young- sters were more impulsive at the start and the end of the 2 l/2 year period. Of interest was Messer's report that the impulsive pupils ~who had experienced failure were comparable in verbal intelligence to their more successful and reflective peers. Therefore, impulsivity appears to be a determinant in school failure. This is in agreement with Kagan's contention that conceptual tempo is related to academic successful. Kagan (I965b), a strong advocate of training impulsive young- sters to be more reflective, found that reading ability was negatively associated with impulsivity. With the use of a word recognition test, Kagan made and confirmed the prediction that reflective children would be more accurate in reporting words than their impulsive counterparts. Of great importance, however, was Kagan's acknowledgement that mastery of the basic reading skills had a greater influence on reading achieve- ment than either reflection or impulsivity. Although Kagan prescribed the training of children who display reading difficulties to be more ll 12 reflective, it appears that the crucial focus should be upon the mastery of the basic components Of reading. Reflection will be of little use unless the basic Skills are mastered. Inductive reasoning, like reading ability, is a necessary skill required for academic achievement; it was also discovered to be negatively correlated with impulsivity (Kagan, Pearson, and Welch, l966a). Probabilistic inferences can be defined as the process of transiting from One proposition or statement that is considered to be true to another whose truth is believed to have followed from that of the former and whose evidence when confirm— ing one truth also confirms, to some extent, the other. Mathe- matics, science, and the discovery method Of instruction all demand the making of inferences for successful learning. Kagan et al. (l966a) reported that impulsive children made more errors in inductive reasoning problems. The errors were attributed to the following reasons: a) impulsives tend to respond quickly in situations where inferences are required, and b) they seem to report the first idea that occurs as in his suggestion of remedy- ing reading problems by training for reflection, Kagan indicated that positive benefits may also be gained in the area of induc- tive reasoning by instructing impulsive youngsters to be reflec- tive. The capacity for recall is another factor that is related to academic success. Kagan (I966) reported that reflective children demonstrated superior recall. However, their superior recall did not I3 appear to be a function of longer response times, rather a persistance in attempts to produce a better cognitive product. This suggests that the reflective children were more highly motivated on reCall tasks. It was observed, in an earlier study, that reflective youngsters dis- play more perserverance on difficult tasks than impulsives with the same verbal ability (Kagan, I965a). Perserverance, particularly when a difficult task is involved, is also another variable that appears to be related to academic success, and it has been shown to be more characteristic for reflective individuals. Attentiveness is also held to be an important component associ- ated with school success. The attention span of a child is looked upon as being an indicator of his capacity to learn the presented subject matter and to perform the related intellectual tasks. Interestingly, the inability to sustain attention was Observed as a characteristic of impulsive children, however, the type of task employed was the more important variable in determining attentiveness in impulsive children (Zelniker, Jeffrey, Ault, and Parsons, I972). Jeffrey et al noted that the ability to sustain attention interacted with the type of task presented to impulsive children. Ault, Crawford, and Jeffrey (I972) observed reflective, im— pulsive, "fast accurate, and "slow inaccurate" pupils. Using in-class teacher ratings, they found that only reflective pupils were perceived as being highly attentive. It was also reported that pupil sex was not a differentiating factor--both reflective girls and boys were rated as highly attentive. One study reported that teachers have a tendency to assign higher ratings to girls who display long response times and to boys who make few errors (Yando and Kagan, I968). Consequently, it appears that teachers are inclined to respond positively to girls who delay and to boys who do not make mistakes. These findings reveal some of the factors behind the higher failure rate for impulsive pupils. The impulsive child, it seems, exhibits certain traits which adversely affect pupil-teacher relations in the present-day school setting, such as quickly blurting out an answer even though the child may be unsure of his/her correctness. This may be due to difficulties in placing effective restraints on tendencies toward action. Kagan et al (I966) reported that there are indications that most impulsives do not appear to be as apprehensive or anxious about making a mistake as does the majority of reflective children. Ault et aI (I972) found that reflective and fast accurate pupils were rated significantly less hyperactive than impulsive and slow inaccurate children by teachers on the Hyperactivity Scale. In an investigation concerning the effects of the drug methylphenidate on the cognitive styles of hyperactive children. Campbell, Douglas, and Morgenstern (l97l) reported that hyperactive children were not only more field dependent, more constricted in their ability to control attention, and slower on measures of automatization, but as expected, also more impulsive than the normal control subjects. Campbell et al (l97l) discovered that methylphenidate had the effect on the conceptual tempo of hyperactive subjects resulting in less impulsive respdnding and an improved ability to inhibit incorrect responses. In-a later ———W '5’" I5 study, Cohen, Weiss, and Minde (I972) found results similar to Campbell et al (I97l) concerning the cognitive styles of hyperactive adolescents. They observed that hyperactive adolescents were inclined to be impulsive rather than reflective. As was demonstrated by the findings of the above studies, re- flective pupils were reported to be more successful in the areas of _ grade promotion, reading ability, inductive reasoning, teacher ratings, and attentiveness. However it was noted that the type of task em- ployed had a bearing on attentiveness as it was related to impulsivity. "Modification of Impulsivity In an earlier effort to train first grade children to be reflective, it was Observed that the only significant effect of train- ing was the increased response latency, and at the same time, only a minimal effect on performance quality was indicated (Kagan, Pearson, and Welch, l966b). The focus of the training was on delaying fast responses rather than improving strategies of increased performance quality. Another investigation indicated that conceptual tempos of teachers, through the concept of modeling, had an effect on pupils. Over a one school year period, children who were placed with ex- perienced reflective teachers became more reflective than those placed with experienced impulsive teachers (Yando and Kagan, I968). I6 Again the difference was Observed in terms of response time. Quality of performance--on the Matching Familiar Figures Task (MFF)--was not affected. . Debus (I970) observed the effects on the conceptual tempo of third grade impulsive children under a control and four treatment conditions: (I) reflective model, (2) impulsive model, (3) change model--the model responded on the MFF in an impulsive manner for the first half of the session and in a reflective mode for the second half, and (4) dual models-—reflective and impulsive models were Observed during the same session. The models displaying impulsive performances were shown to experience no success or reward; the reflective models received reward through praise for being successful. Debus also looked at changes in response latency and error scores over a pretest, and immediate posttest, and a delayed posttest. He re- ported increases in response latencies on the immediate posttest for the experimental treatments under the reflective, change, and dual models. However, the change model condition provided the only significant lasting effect that was shown by the delayed posttest, and this effect was Observed only in female subjects. Another sig- nificant and also highly important finding by Debus was the lack of significant effects on error scores by any of the subjects in the experimental conditions. Again it was demonstrated that an increase in response time has very little, if any, significant relation with quality of performance as measured by the MFF. Denney (I972) also looked at the modeling effect on conceptual tempo, and he employed the Conceptual Style Test as the measure. His | . I7 sample was second grade boys. Using an adult female model displayed on videotape, Denny reported results similar to those of Debus. Re— sponse latencies were affected but error scores remained unchanged. Subjects who Observed the impulsive modeling condition, Shortened their response time; those who observed the reflective modeling condition, increased their reSponse latency. In addition, Denney indicated that the effects Of the conceptual tempo employed in the modeling conditions were generalizable to the subjects' performance on an independent task——the MFF. From a totally different direction, Zelniker et al (I972) re- ported that the modification of tasks produced an effective modifica- tion of impulsive sUbjects' scanning strategy. The new strategy was not only retained but was Observed to have a positive effect on MFF performances, even though the response latencies were not lengthened. Zelniker et al introduced the notion of modifying task variables to improve the effectiveness Of cognitive strategies and problem solving skills. The above studies indicated that cognitive behaviors may be successfully modified, specifically that of response latency, but such modification amy have very little effect on the quality of performance. Contrary to those studies, the Zelniker et al (I969) investigation focused on the change to a more efficacious strategy. They reported that the quality of performance improved with the use Of more efficient cognitive strategies. L¥__ I8 Observing Behavior and Conceptual Tempo In examining observing behavioral patterns of reflective and impulsive fourth grade boys, Siegelman (I969) found that the impulsives ignored two and one half as many alternatives per item as the re- flectives on the MFF. Nelson (I969) supported Siegelmanfs findings when he reported that reflective boys, when performing on the MFF, devoted less time examining the standard figure while allocating more time to the variants than did impulsive boys. Other studies, however, in the area Of observing behavior produced different findings. Drake (I970), for example, found results similar to those of Siegelman (I969) and Nelson (I969), but the results only held true for adults and were just the Opposite for third grade children. Zelniker, Jeffrey, Ault,. and Parsons (1972), on the other hand, failed to find any differences in Observing responses on the MFF between impulsive and reflective third graders. Ault et al (I972) agreed with the basic Observations of the Zelniker et al (I972) stUdy that indicated a lack Of difference in the basic scanning strategies of impulsive and reflec- tive middle elementary grade children on the MFF; all of the subjects relied heavily on the strategy of comparing variants with the standard rather than systematically seeking one feature from each variant at a time. However, in looking at the amount Of visual scanning employed and eye movement scores (the number of eye movements recorded) Of third graders during the performance on the MFF, Zelniker et al (I972) re- ported that reflectives had significantly more eye fixations and had also scanned significantly more variants than impulsives. 19 It is interesting to note the dissimilar findings on the Observing behaviors of reflective and impulsive youngsters. One reason why dissimilar findings were evident may be because precise instruments that measure the observing behaviors of individuals in relationship to their conceptual tempo have not yet been developed. Since selection strategies often incorporate observing behavior, and if selection strategies and conceptual tempo are related, it also appears that the lack Of differences between impulsives and re- flectives would be evident in the area of observation strategies, but as was reported that was not the case. The Association of Conceptual Tempo and Information Processing Nuessle (I972) reported that reflective subjects were more efficient at processing information. He posited that the reflective cognitive style facilitated information processing because it provided for more effective retrieval and recoding of information. Nuessle suggested that reflection-impulsivity differences are strongly related to developmental differences in information processing. He observed that the reflection periods of Older subjects (9th graders) were longer than those younger (5th graders). The older subjects were significantly more proficient at processing information. Nuessle found, however, that when reflective younger subjects and impulsive older ones were specifically studied, they did not differ signifi- cantly on information processing. The lack Of developmental 20 differences between younger reflective and older impulsive sjbjects implies that the reflection-impulsivity dimension is stable: re- flective subjects maintain their more efficacious information 'pnocessing capabilities through periods of development, and im- pulsives continue to lag. In a study on problem solving strategies, Ault (I973) also found that younger (mean age 6.7 years) reflective children were more reflective than older (mean age 10.9 years) impulsive youngsters on the MFF. McKinney (I973) observed that reflective children were more likely to use a conservative focusing strategy, and impulsives tended to employ a random trail and error procedure when processing information. McKinney also noted that only a small proportion of his subjects employed a scanning strategy; however, the impulsives were observed to use the scanning strategy more frequently. "McKinney implied that the impulsive child may benefit more, if there is a desire to modify his impulsive tempo, by being instructed to use conceptual strategies more efficiently, instead of requiring him to delay his response time. McKinney indicated that the random or hypothesis scanning strategy employed by impulsive children is an inefficient conceptual strategy that increases one's memory load Of the amount of information that is retained, by the same token, forcing impulsives to delay their responses may actually impair their performances on tasks where the memory load is great. McKinney (I973) also questioned the wisdom of those who believe that the conceptual tempo of an impulsive child should be modified to be more reflective in the area of response latency. He cited that 2I performance quality appeared to be ignored by proponents Of the delaying of response times. McKinney posited that modifying an impulsive child's response time is not a solution, but a more efficacious mode of processing information may be the answer for a more effective and productive modification. In other words, through effective training in conceptual strategies, the ime pulsive child's response Iatence may not increase, which should actually be of little or no concern, but his quality of perfor- ‘mance across problems should increase, and that should be the major Objective. In an investigation Of information processing and the modification of impulsivity in middle and lower class boys, Hieder (I97l) reportedthat task strategy instruction produced a sig— nificant decrease in errors for both groups on the MFF and sentence construction. Task strategy instruction was compared with the effects of forced delay of response latency, increased motivation through the Offering of rewards for correctiveness, and a control condition in which the subjects were instructed on the MFF in the standard manner suggested by Kagan. Task strategy of instruction was found to be the superior method. Hieder's findings reinforce the position that the learning of effective cognitive strategies is more effective in decreasing errors and improving performance quality than just delaying response times. 22 The Relation of Conceptual Tempo and Problem Solving Yando and Kagan (I970) stated that a child's conceptual tempo was a highly stable feature of his problem solving behavior; also, conceptual tempo is a good predictor of errors. In other words, an individual who is disposed toward impulsivity tends to produce more errors than one who is reflective. It was reported that under memory support conditions, im- pulsive children solved significantly more problems than under the no-memory support condition; however, only a slight, nonsignificant incrbase was observed for reflectives under memory support conditions (McKinney and Banerjee, I973). McKinney and Banerjee (I973) noted that both reflectives and impulsives solved the same number of problems when memory aids such as a constant source of feedback of a correct response in a problem solving task was provided. They also observed that memory aids during concept attainment were significantly more beneficial for impulsive children. Ault (I973) suggested that from evidence presented in other reports that emphasized training techniques in scanning strategies that has significantly decreased errors, the focal point should not be on response speed but on the problem strategy employed. She Ob— served that impulsive children exhibited less mature cognitive strategies than those classified as reflective. Siegelmann (I969) was one of the first to present the argument that Ault postulated. She proposed that the focus should be on attempts to train algorithms or attention deployment strategies in an effort to modify the negative in.» r 23 effects of impulsivity. Furthermore, Siegelman stated that although a long response time may be necessary, it is not a sufficient con- dition for reflective (slow and accurate) responding. One may observe an example of this when various subjects on the MFF display long response latencies but also a high rate of errors. In order to be classified as reflective, One must not only display a long response latency but a low error rate as well. The tenor of individuals such as Ault, McKinney, and Siegelman is contrary to the earlier solutions for impulsivity advanced by Kagan and his associates. The newer theme is an emphasis on the cog- nitive strategies employed by impulsives in an attempt not to increase response times, but to decrease the amount Of errors across problem solving tasks through more effective cognitive strategy training methods. Conceptual Tempo and Decision Making In a recentustudy that looked at differences in reflective and impulsive children in the area of decision making, Mann (I973) reported that the reflection-impulsivity dimension was assOciated with caution- haste in decision making. Not surprisingly, Mann found that reflective subjects took significantly more time in making decision on ego. involving tasks. However, no substantial differences were reported between reflectives and impulsives in the content and quality Of the decisions Observed. 24 Upon studying the risk factor involved in decision making, Kopfstein (I973) reported that conceptual tempo was not related to risk taking behavior. He stated that such findings were totally unexpected, for the hypothesis was that impulsives-would be more inclined toward taking risks in the decision making process. The findings of Mann (I973) did not show a deficiency in the impulsive style in terms of decision making quality, even though the impulsives made decisions in a shorter time frame. It appears that the greatest consideration should be given to the quality of decision making rather than the time required to make a response as having serious implfications concerning cognitive strategies in problem solving. Kopfstein indicated that caution was just as prevalent in impulsives, contrary to previous assumptions, as in reflectives. Both Kopfstein (I973) and Mann (I973) added further strength to the position that strategy, not the speed Of response, should be the primary focus in the teaching of individuals. Conceptual Tempo and Anxiety In earlier reports concerning conceptual tempo, it was strongly held that the impulsive child displayed more anxiety because of his own expectations of failure, and therefore, he answered quickly in order to reduce tension; on the other hand, the child who does not expect to fail is less anxious about his ability and so tends to 25 exhibit more reflection (Kagan, Rosman, Day, Albert, and Phillips, I964). Later a gradual shift toward viewing the reflective in- . dividual as being more anxious about failure became more prevalent. Kagan (I966) reported that the belief that reflective children display more inclination toward anxiety received only marginal support in his study of the generality and dynamics of conceptual tempo. . Kagan suggested a curvilinear relationship that demonstrated that if a child is more anxious about possible errors than quick success, he will be reflective, but if his anxiety over gaining quick success is greater than committing errors, then he will be impulsive. Kagan (I966) further postulated, under the assumption that impulsives are less successful, that an individual may become so accustomed to failure that he no longer protects himself and as a result is less anxious about the prospects of failure. When anxiety was induced, longer decision times were observed for both impulsive and reflective children, and also fewer errors for impulsives who displayed increased response latencies (Messer, l970b). Such findings provided support for the position that anxiety over intellectual performance is an antecedent of a reflective conceptual tempo. Presently, there is strong evidence that suggests that the reflective individual is more anxious over making a mistake and as a result is more cautious; but for the impulsive person, there is mini- mal anxiety over a potential error (Kagan and Kogan, I970). From previous observations concerning Conceptual tempo and anxiety, Kagan 26 and Kogan deduced that in being cautious and highly concerned with committing an error, the reflective individual is consequently going to take longer in responding or repOrting an answer in his attempt to avoid errors; on the other hand, impulsives are only minimally Iconcerned with errors and as a result are more likely to respond quickly in tasks that involve the choOsing of the correct answer from similar and difficult alternatives. Sex Differences in Conceptual Tempo Earlier studies provided indications that no significant sex differences were found to be associated with conceptual tempo (Kagan l965a and I956b). A recent study by Kopfstein (I973) also indicated that there was no Significant differences between fourth grade boys and girls on conceptual tempo. An earlier investigation by Ault et al (I972) also reported no significant sex differences on th MFF test. Lewis, Rausch, Goldberg, and Dodd (I968) also reported the lack of significant sex differences on conceptual tempo. However, they observed that the rate of errors were more strongly related to IQ in boys. The reports of no sex differences were of interest primarily because the belief that boys would be more inclined toward impulsivity, especially in view of the contention that impulsives were more aggress— ive, less inhibited, and less anxious about success (Kagan and Kogan, I970). The socialization of girls is normally thought to be away from gee—e: ,, _ 27 those characteristics mentioned, while boys would have less social conditioning Of young boys and girls have little effect on creating sex differences in conceptual tempo. A Rationale For Observing Four Dimensions of ConceptuEl Tempo Kagan and others who studied conceptual tempo, categorized subjects along the reflectioneimpulSivity dimension on the criteria of response latency and rate of errors. The median response latency and error-rate of the Observed saMple was always incorporated in the classification of conceptual tempo. A subject whose response time was above the median and whose error rate was below, was classified as reflective. The subject whose response latency fell below the median and whose error rate was above, was categorized as impulsive. The reflection—impulsivity dimension model of conceptual tempo is depicted in Figure I. There was always a fairly large group of individuals--approximately one-third--who fit neither the defined reflective nor impulsive modes. upon reviewing studies of conceptual tempo, it was apparent that those subjects who did not fit into dichotomous reflection-impulsivity dimension were ignored. 28 Median Error Rate + /\ ‘ [Reflective +\_ Median Response Latency \ » _/ , . Impulsive Fig. I.--Dimensions of Reflection-Impulsivity \/ Shulman, Loupe, and Piper (I968), in a study Of aduls sub- jects, reported that conceptual tempo and inquiry competence were related. Reflectives generally proved to more effective inquiries ' than impulsives. It should be adknowledged. however, that Shulman and assotiates found that the error rate component was significantly more important in predicting inquiry behavior than the amount of reflection time. In essence, an individual may fall in either rcategory of reflectivity nor impulsivity because of an Observed tendenCy toward a low error rate coupled with a short reflection time, but because of his low error rate, the individual is nevertheless an effective inquirer. On the other hand, the converse can be observed in an individual with a long reflection time as well as a high error rate. That particular individual's ability as an effective inquirer should be lower. Such Observations indicate a need to look at individuals who do not fit into Kagan's reflective-impulsive dichotomy. Fig. 2 is an illustration of the conceptual tempo model which in- corporates the "fast accurate" and "Slow inaccurate" dimensions. Shulman et al noted that by focusing upon only reflectivity and 29 Median Error Rate + .. Reflective slow Inaccurate+ + \ Median Response / \ _ /' Latency Fast Accurate Impulsive Fig. 2.--Dimensions of Four Levels of Conceptual Tempo: Reflection, Impulsivity, Slow Inaccuracy and Fast Accuracy impulsivity and throwing out those subjects who do not fit either made, one may be consistently confounding reflection and accuracy. Variables such as reading ability or inductive reasoning that correlate consistently with reflection time, may actually highly correlate with the aspect of reflection which involves one's capa- bility for discrimination accuracy. Socioeconomic Status and Conceptual Tempo Socioeconomic status (SES) is another factor that is by and large accepted as an important component that is associated with a child's academic success (Battle and Rotter, I963). SES is actually overwhelmingly acknowledged as having considerable influence on whether or not a child succeeds in school, and the evidence has been 3O abundant (Havighurst and Janke, I944; Havighurst and Breese, I947; Pierce-Jones, I959; Wilson, I963). Kagan (I966) posited that high SES children tend to be more reflective than those from a low SES background. He stated,=”Investigators working With 'culturally deprived' children believetthat one reason for their poor intellectual performance is their impulsive orientations.” In a study of middle and lower class boys (ages 9 to l2), SChwebel (I966) reported differences and also effects Of impulsivity‘bn the performance Of verbal tasks: the middle claSs subjeCts performed higher on the verbal tasks, and the lower class subjects had higher inclinations toward impulsivity. Schwebel stated that social class differences in verbal performance were attributable to language«speech competence and to reflection-impulsivity. It appears that the lower SES child has the factors Of low verbal performance (viS-a-vis the academic environment) and his predisposition toward impulsivity as impingements working against his achieving school related success. This, in all probability, is a cumulative process--the further the low SES child advances in school, the more stifling or debilitating his school experiences. According to Rist (I970), a child is labeled by the time he leaves the first grade as to whether or not he is to experience overall success or failure in the classroom. The primary basis for his label- ling is the child's appearance (essentially whether or not he/she is clean, neat, and somewhat fashionably dressed as perceived by signif- icant others such as teachers and school administrators). Since 3I appearance is highly related to SES, it is assumed that a child who has a more "desired" appearance is more likely to come from a family with a higher SES than a child who is depicted as having a less "desired" appearance. A child perceived as having a high SES familiar background, enjoys a higher probability of experiencing a more receptive and posi- tive attitude (conscious or unCOnscious) than those viewed as being less economically fortunate. Moreover, a greater amount of benign attention from teachers and other school related personnel might be granted to those seen as being of a higher SES as opposed to those classified as economically disadvantaged or of a lower SES. Such occurrences; which are related to teacher-expectations-of—pupiIs, are also closely associated with academic and other school oriented success. Pupils who are either reflective or quick ("fast accurate“) and of highSES backgrounds may be viewed more favorably by their teachers and as a consequence will experience more academic success than im- pulsive youngsters of low SES backgrounds. Clustered between these groups might be the children who are reflective and low SES, slow ("Slow inaccurate") and high SES, qUick and low SES, impulsive and high SES, and slow and low SES in that respective order. If such aforementioned differences are observed, the impli- cations then suggest that a child's conceptual tempo along with his SES background predicts whether or not he is more prone to experience success, failure, or even mediocrity in the schools by the time he has completed the early elementary grades. A further implication is ‘3 at..- 32 that children ere channeled into academic or intellectual categories-- based upon SES and conceptual tempO--as early as age seven or eight. SUMMARY In summary, it was emphasized that conceptual tempo was one of several cognitive style subsets and it is closely related to selec- tion strategies, another cognitive style subset, which describes the learning of concepts by eithera focusing or scanning procedure. Con- ceptual tempo was defined as a hypothesis testing procedure that refers to speed and accuracy in making a response to two or more diffi- cult alternatives. Two dimensions Of conceptual tempo initially des- cribed were reflection and impulsivity. Reflective individuals were depicted as those who demonstrated a slow response time but a low error rate, and impulsives were illustrated as individuals with short response times and high error Scores. Various investigations of conceptual tempo reported that re- flective pupils were more successful in the areas of grade promotion, reading ability, inductive reasoning, teacher ratings, attentive, and they were also more anxious than their impulsive counterparts. Modifi- cation of conceptual tempo appears to have shifted away from the stress on altering response latency for impulsives to an emphasis on devel- Oping more efficient cognitive strategies. 33 A rationale was presented for examining the fast-accurate and slow-inaccurate dimensions of conceptual tempo. It was posited that the possible confounding of reflection time and accuracy exists because of the exclusion of those individuals who do not fit the reflection-impulsivity mode. _Also, accuracy rather than reflection time might be the component more highly correlated with variables such as reading and inductive reasoning. Socioeconomic status (SES) was also reported to be related to conceptual tempo to the extent that reflectives were more likely to be from high SES backgrounds and impulsives from lower socioeconomic levels. It was suggested that a child's SES and conceptual tempo had considerable effects on school achievement. REFERENCES Ault, R.L., Crawford, D.E., Jeffrey, W.E. Visual scanning strategies of reflective, impulsive, fast-accurate, and slow-inaccurate children on the Matching Familiar Figures Test. Child Development, I972, 43, l4l2—l4l7. Ault, R.L. Problem-solving strategies of reflective, impulsive, fast- accurate, and slow inaccurate children. Child Development, I973, 44, 259-266. Bartel, N.R. Locus of control and achievement in middle and lower- class children. Child Development, l97l, 42) l099-IIO7. Battle, E.S. and Rotter, J.B. Children feelings of personal control as related to social class and ethnic groups. Journal pf Personality, I963, 34, 482—490. Brookover, W.B., Gigliotti, R.J., Henderson, R.D., and Schneider, J.M. Elementary school social environment and school achievement. Research report published by the College of Urban Development, Michigan State University, East Lansing, Michigan, I973. Broverman, D.M., Dimensions of cognitive style. Journal pf Person- ality, I960a, 28, l67—l85. Broverman, D.M., Cognitive style and intra-individual variations in abilities. Journal pf Personality, l960b, 28, 240-256. Broverman, D.M., Generality and behavioral correlates of cognitive styles. Journal pf_Consulting Psycholpgy, I964, 28) 478-500. I Broverman, D.M., Broverman, I.K., and Klaiber, E.L. Ability to automatize and automatization cognitive style: a validation 5 study. Perceptual gpg_Motor Skills,l966, 23, 4l9-437. Bruner, J.S., Goodnow, J.J., and Austin, G.A. 4_Study pf_Thinking. New York: John Wiley & Sons, Inc , 956. Campbell, 5.8., Douglas, V.I., and Morgenstern, G. Cognitive styles In hyperactive children and the effect of Methylphenidate. Journal pf_Child Psychology gpg_Psychiatry, l97l, 423 55-67. 34 35 Cohen, N.J., Weiss, C., and Minde, C., Cognitive styles in adolescents previously diagnosed as hyperactive. Journal of Child Psychology and Psychiatry, I972, 13, 203-209. Coleman, J.S., Campbell, E.G., Hobson, C.J., McPartland, J., Mood, A.M., Weinfield, F.D., and York, R.L. Eguality pf_Educational Opportunity, Washington, D.C.: U.S. Government Office, I966. Crandall, V.C., Katkovsky, W. and Crandall, V.J. Children's beliefs in their own control Of reinforcement in intellectual—academic achievement Situations. Child Development, I965, 33, 9l-IO9. Debus, R.L., Effects of brief observation of model behavior on conceptual tempo of impulsive children. Developmental Psyghology, I970, g, 22-32. DeCecco, J.P. The Psychology gf_Learning and Instruction: Educational PS cholo . Englegood Cliffs, New Jersey: Prentice-Hall, Inc., I9UU. ~ - Denney, D.R. Modeling effects upon conceptual style and cognitive tempo. Child Development, I972, 43, l05-ll9. Drake, D.M. Perceptual correlates of impulsive and reflective behavior. Developmental Psycholpgy, I970, g, 202-2I4. Eska, B., and Black, K.N. Conceptual tempo in young grade-school children. Child Development, I971, 43, 505—5I6. Gardner, R W., Holzman, P.S., Klein, G.S., Linton, H.B., and Spence, D.P. Cognitive control; a study of individual consistencies in cognitive behavior. Psychological Issues,l959, 1, No.4. Havighurst, R.J. and Janke, L.L. Relations between ability and social status in a mid-western community. III. Primary mental abilities. Journal of Educational Psychology, I947, 33, 24l—247. . Hervey, S.D. Cognitive style and preferences in school tasks. Un- published Ph.D. dissertation, I966, Michigan State University. Hieder, E.R. Information processing and the modification of an "impulsive conceptual tempo.” Child Development, I97l, 42, 1276-l28l. Johnson, C.S. Objective identification of strategy on a selection task. Journal of_Experimental Psychology Monograph, I97I, 99, I67-I96 36 Kagan, J. Impulsive and reflective children: Significance of con- ceptual tempo. In J.D. Krumboltz (Ed.), Learnin and the_ Educational Process. Chicago: Rand-McNaIIy,-I965a, l33-I6l. Kagan J. Reflection-impulsivity and reading ability in primary grade children. Childeevelopment, l933b, 33, 609-628. Kagan, J. Reflection-impulsivity: The generality and dynamics of conceptual tempo. Journal of Abnormal Psychology, I966, 21, I7-24° Kagan, J., Rosman, B.L., Day, 0-, Albert, J., and Phillip, W. In- formation processing the child: significance of analytic and reflective attitudes. Psychological Monograph, I964, Z3, Kagan, J 5., Moss, H.A., and Sigel, I.E. Psychological significance of st Ies of conceptualization. In Wright, J.C. Kagan, J.S. (Eds. Basic Cognitive Processes in Children. Monographs of; the Societ for Research in-Child Development, I96 , __, T27 Serial No. 86), 73-7124.— ‘ Kagan, J. andKagon, N. Individual variation in cognitive processes. In P.H. Mussen (Ed.), Carmichael's Manual gf_Child Psychology, (3rd ed.), New York: Wiley, I970, l273-1365. Kagan, J., Pearson, L., and Welch, L. Conceptual impulsivity and ‘ inductive reasoning. Child Development,l966a, 32, 583-594. Kagan, J., Pearson, L., and Welch, L. Modifiability of an impulsive tempo. Journal pf_Educational Psychology. l966b, 32, 350-365. Klein, G.S., Need and regularion.. In M.R. Jones (Ed.), Nebraska S m osium on Motivation. Lincoln: University of Nebraska Press, I9543'224-274 Klein, G.S. Cognitive control and motivation. In G. Lindzey (Ed.), Assessment pf_Human Motives. New York: Rinehart, I958, 87-Il8. Kopfstein, D. Risk-taking behavior and cognitive style. Child Development, I973, 44, l90-l92. Laughlin, P.R. Selectien strategies in concept attainment. In R.L. Solso Ed. Contemporary Issues ip‘Cognitive Psychology. The Loyola Symposium. Washington, D.C., 973, 277-3 I. Lewis, M., Rausch, M., Goldber, S., and Dodd, C. Error response time and IQ: sex differences in cognitive style of preschool children. Perceptual and Motor Skills, I968, g3, 563-58. 37 Mann, L. Differences between reflective and impulsive children in tempo and quality of decision making. Child Development, I973 44, 274-279. McKinney, J.D. Problem solving strategies in impulsive and reflective second graders. Developmental Psychology, I973, 3, I45. McKinney, J.D. and Banerjee, C. Concept attainment by reflective and impulsive children. Unpublished manuscript, Frank Porter Graham Child DeveIOpment Center, I973. Messer, S. Reflection-impulsivity: Stability and school failure. gournal pf_Educational Psychology, l970a, 31, 487-490. Messer, S. The effect of anxiety over intellectual performance on reflection-impulsivity in children. Child Development, l970b, 41, 723-735. Mosteller, F. and Moynihan, D.P. (Eds.). Qp_E ualit of_Educational Opportunity. New York: Random House, Inc. 972. Nelson, T.F. The effects Of learning in attention deployment on observing behavior in reflective and impulsive children. Unpublished Doctoral Dissertaion, University of Minnesota, I968. Neussle, W. Reflectivity as an influence on focusing behavior of children. Journal of_Experimental Child Psyohology, I972, 14, 265-276. Pierce-Jones, J. Social Stratification and academic achievement. In A.J. Passow (Ed.), Education in Depressed Areas, N.Y. Teachers College Press, Columbia University, I963. Rist, R.E. Student social class and teacher expectations: The self-fulfilling prophecy in ghetto education. Harvard Educa- tional Review, I970, 43, 4ll-45l. Santostefano, 8.6., and Paleys.E.. DeveIOpment of cognitive controls 3 in children. Child Development, I964, 33, 939-949. Schwebel, A.I. Effects of impulsivity on performance of verbal tasks in middle and lower-class children. American Journal of Orthopsychiatry, I966, 33, I3—2l. Shulman, L.S., Loupe, M.J., and Piper, R.M. Studies of_the Inqpimy Process, Michigan State University, East Lansing, Michigan: Educational Publicational Services, College of Education, I968. 38 Siegelman, E. Reflective and impulsive observing behavior. Child Development, I969, 43, l203-12l2. Stanes, D. Analytic reSponses to conceptual style test as a function of instructions. Child Development, I973, 44, 389-39I. Travers, R.M. Essentials of_Learning. New York: The MacMillan C0,: I963. - Yando, R.M. and Kagan, J. The effect Of teacher tempo on the child. Child Development, I968, 33, 27-34. Yando, R.M. and Kagan, J. The effect Of task complexity on reflection- impulsivity. Cognitive Psychology, I970, 1, l92-200. Wilson, A.B. Social stratification and academic achievement. In A.H. Passow (Ed.), Education 1p_Depressed Areas. N.Y. Teachers College Press, Columbia, University,7l963. Witkin, H.A., Dyk, R.B., Faterson, H.R., Goodenough, D.R., and Karp, D.A. Psychological Differentiation. New York: Wiley, I962. Zelniker, T., Jeffrey, W.E., Ault, R., and Parsons, J. Analysis and modification of search strategies of impulsive and reflective children on the Matching Familiar Figures test. Child DeveIOpment, I972, 43, 32l-335. SECTION II JOURNAL ABSTRACT DIFFERENCES ASSOCIATED WITH CONCEPTUAL TEMPO AND SOCIOECONOMIC STATUS 0N ACADEMIC ACHIEVEMENT, IQ, AND INTERNAL—EXTERNAL MEASURES Henry T. Frierson, Jr. Michigan State University One hundred eighty fourth grade 55 were examined to determine differences on academic achievement, IQ, and internal—external con- trol measures in relationship to conceptual tempo and socioeconomic status (SES). Each S was categorized as either reflective, impul— sive, slow-inaccurate, or fast-accurate. Also, each S was classified as high or low SES. High SES 55 demonstrated higher mean scores on each dependent variable. This indicated that SES was associated with the dependent variables as predicted. Within the two SES groups, conceptual tempo demonstrated a strong relationship with academic achievement only among low SES children, and for low SES youngsters the fast-accurate dimension had the strongest association with school success. For the high SES child the major factor for school success was his high socioeconomic level. Overall, when SES is ignored, reflectives and fast—accurates had higher performances on the depen— dent variables, thus suggesting that accuracy not response time is the more important component in conceptual tempo in relation to 39 40 school success. The results also indicated that the measures of internal-external control were not significantly associated with conceptual tempo. JOURNAL ARTICLE DIFFERENCES ASSOCIATED WITH CONCEPTUAL TEMPO AND SOCIOECONOMIC STATUS ON ACADEMIC ACHIEVEMENT, IQ , AND INTERNAL -EXTERNAL MEASURES Conceptual tempo is reported to be highly associated with a child's academic success or failure (Kagan l965a). "As measured by the Matching Familiar Figures (MFF) test (developed by Kagan, l965). conceptual tempo is diVided into tWo dimensions; reflection and im- pulsivity. The MFF is composed of a series of six pictures of familiar objects of which only one is identical to a presented Istandard. 55 who display long response times (latencies) and low error scores on the MFF are classified as reflective, and those who demonstrate short response times and high error scores are cate— gorized as impulsives. Kagan (l965b) reported that conceptual tempo manifests pervasive generality across varied task situations and importantly, it is linked to some fundamental aspects of an individual's personality structure. The reflective disposition is generally believed to be the one most closely related to academic success. Messer (1970) observed that youngsters who failed a grade were significantly more impulsive, although all 55 were comparable in verbal intelligence. Kagan (1965b) reported that reading ability was negatively associated with impulsivity. Inductive reasoning, perceived 41 42 as being a requisite for academic success, was also discovered to be negatively correlated with impulsivity (Kagan, Pearson, & Welch, 1966). The inability to sustain attention was observed as a characteristic of impulsive children (Zelniker, Jeffrey, Ault, & Parsons, l972). Another investigation reported that teachers tended to assign higher ratings to girls who displayed lOng response times and to boys who made few errors (Yando & Kagan, l968). On the basis of this evidence it appears that teachers are inclined to respond positively to girls who delay and to boys who do not make mistakes. Finally, it was reported that most impulsive individuals do not appear to be as appre- hensive or anxious about making a mistake as do those who are reflec- tive (Kagan et al., I966). While there is considerable research on the reflection- impulsivity dimensions, there has been very little published research involving the two groups that do not fit the reflection-impulsivity dichotomy. Those 85 who were characterized by short response latencies and low error scores (fast-accurates) or long response latencies and high error scores (slow-inaccurates), as demonstrated on the MFF, were usually excluded from the reSearCh sample. The excluded 55 make up two groups who usually constitute approximately a third of the popu- lation when conceptual tempo is studied. Shulman, Loupe, & Piper (I958) contended that the two "extreme" groups possessed qualities that were certainly worth investigating. They argued that by focusing only upon reflectivity and impulsivity, researchers may be consistently con- founding reflection and accuracy. They also noted that variables 43 such as reading ability or inductive reasoning that correlate con- sistently with reflection time, may actually be related to the aspect of reflection that actually involves the capability for discrimination accuracy. Such a hypothesis cannot be validly examined by excluding the two "extreme" groups. Among the first studies that examined the two "extreme" groups were those by Ault, Crawford, & Jeffrey (1972) and Ault (l9la). In the Ault et al (1972) investigation, it was reported that of the four groups (reflective, impulsive, slow-in— accurate, and fast-accurate), only reflective pupils were perceived by their teachers as being highly attentive. They also fOUnd that reflective and fast-accurate pupils were rated significantly less hyperactive than impulsive and slow-inaccurate youngsters by teachers on the Hyperactivity Scale. Ault (l973) reported that reflective and fast-accurate Ss employed more mature cognitive strategies than impulsive and slow-inaccurate Ss in problem solving tasks. A major aim of this study was to further examine the two "extreme“ groups with respect to the relative importance of accuracy and response latency. Socioeconomic status (SES) is also a factor that is associated with a child's academic success (Battle & Rotter, l963). According to Risf (1970), a child is labeled by the time he leaves the first grade as to whether or not he is to experience overall success or failure in the classroom. Rist postulated that the criteria for such labeling are strongly related to SES; those pupils who are positively perceived by significant others such as teachers and school administrators would more likely be of a higher SES than those not as positively viewed. 44 Specifically, the purpose of the present study was to examine differences among reflective, impulsive, fast-accurate, and slow- inaccurate fourth grade children who were also either in the high or low SES category. The major hypotheses were (a) conceptual tempo is associated with measures of academic achievement, school related intelligence. and internal-external control to the extent that reflective Ss and fast-accurate 55 will demonstrate higher mean scores on all measures, and (b) SES is also related to academic achievement, school related intelligence, and internal-external control measures to the degree that high SES 55 will produce higher mean scores across all dependent measures employed in the study. Eight dependent variables were investigated: scores on the Intellectual Achievement Responsibility Questionaire (IAR), the Sense of Control Scale, teacherratings, the standardized Metropolitan Achievement Reading Test (MART),-the standardized Metropolitan Achievement Math Test (MAMT), the criterion—referenced Minimal Perform- ance Objective Math Test (MPOMT) also from the MEAP, and IQ. Method Subjects The Ss were I80 fourth grade children (99 males and 81 females) from seven different elementary schools in a city of approximately 200,000 in the midwest. The data were collected during the midpoint 45' of the school year. Two schools were located in low socioeconomic regions and the remaining five were situated in middle to high socio- economic areas. Ss were categorized by the criteria of reflection- impulsivity on the basis of the MFF test. SES classification of the areas where the schools Were lOcated were made through the use of the Michigan Educational Assessment PrOgram's index of SES. Each S was then individually assigned to high SES or low SES categories on the basis of the McGuire-White Scale.) Therefore, the actual sample was specifically high SES youngsterS‘from high SES schools and low SES youngsters from low SES schools who were either reflective, impulsive, slow-inaccurate, or fast-accurate. Instruments and Procedure MFF, The MFF was made up of l2 items containing pictures of familiar objects. For each item there were six variants arranged in two rows of three pictures on the bottom card of two attached 8 1/2“ X ll” cards. 0n the top card wasa picture of the standard. Only one variant per item was identical With the standard. The objective was to choose the one variant that was identical to the standard. A stop- watch was used to record the latenCy to the first response for each item. Previous studies indicated that the mean cut-off for categor- izing reflective and impulsive fourth grade 55 was approximately l0 secs. for the mean response latency and l0 total errors for the error score (Kagan, I966; Kopfstein, I973). Those values were used in this 46 study for classifying reflectives (l0 secs. or above and 9 errors or below), impulsives (below l0 secs. and l0 errors or above), fast-accurates (below l0 secs. and 9 errors or below), and slow- inaccurates (10 secs. or above and 10 errors or above). On the MFF test, the mean response latency and error score was l5.84 secs. and 5.7l for reflectives, 7.05 secs. and l3.4l for impulsives, 14.85 secs. and l2.67 for slow-inaccurates, and 7.82 secs. and 6.87 for fast-accurates. The MFF test was administered according to the prescribed standard procedure. For all 12 items of the MFF test, each S was asked to select the one figure from six variants that was identical to the standard. Each S was also asked to vocalize as well as point, with his finger, to the picture of his choice. The latency of each S's first response to each variant was recorded in half seconds. The stopwatch was kept from the S's view. Errors for each item were also recorded. A maximum of six errors per item was allowed. Whenever six errors were recorded, the S was shown the correct figure and E then proceeded to the next item. A score sheet for errors was also kept out of the S's view. McGuire-White. The McGuire-White Scale (Kennedy, 1969) was used to identify and categorize high and low SES $5. This was done by examining Ss' parental educational attainment and occupational status. Those 55 whose parents were at the upper level of the scale were classified as high SES, and those whose parents were at the lower level were designated as low SES. For instance, high SES parents were in the top two levels (l and 2) for occupational status which included 47 professionals, high status businesSmen, large scale proprietors, and high level white collar positions, also the tOp two levels (I and 2) for educational attainment which included completion of a graduate school, a professional school, or a four year college or university. Low SES 35' parents, on the Other hand, were on the bottom two levels (6 and 7) for occupational status and the bottom three levels (5,6, and 7) for educational attainment. The various levels on the scale were used as values in assigning SES Classification 35. The values from the levels for occupational status and educational attainment were added together thus getting a tOtal value that served as an index of SES. For example, values of 2 to‘4 indicated high SES and values of ll to l4 indicated low SES. 4 IAR, The Intellectual Achievement Responsibility Questionaire (IAR), developed by Crandall, Katkovsky, & Crandall (I965), was de- signed to measure the degree of control and individual perceives regarding his own reinforcement responsibility in academic situations. In essence, the IAR assesses whether or not an individual perceives himself as his own controlling agent of reinforcements and therefore views himself as responsible for successes as well as failures. The test-retest correlation for the IAR, after a 2-month interval between the administration of the first and second tests, was .69. The follow- ing are two examples of positive and negative event items from the 34 forced choice item IAR scale. 48 If a teacher passes you to the next grade, would it probably be a. because she liked you, or I+ b. because of the work you did? When you have trouble understanding something in school, is it usually a. because the teacher didn't explain it clearly, or I- A. because the teacher didn't listen carefully. Sense of Control. The Sense of Control Scale, developed by Brookover, Gigliotti, Henderson, & Schneider (l973), is based upon the work of Coleman and others (l966). Brookover and his associates described the scale as a measure of a child's feeling of personal efficacy over his environment in relationship to his school perfor- mance. The Sense of Control Scale is a five item Likert—type mul— tiple choice scale. Hoyt's Analysis of Variance reliability co- efficient for the scale was .65. The two internal-external measures, the IAR and the Sense of Control Scale, were administered to $5 in groups. A tape recorded voice of an adult female reading the complete IAR questionnaire and Sense of Control Scale was provided for the 55 so that they might follow along at the recorded pace if desired. The primary objective for the recording, however, was to control for the possibility of any reading difficulties that might have been encountered by the Ss. 49 Teacher Ratings. Teacher ratings were gathered from the teachers of each S. The teacher ratings served as indices of classroom success as perceived by the teaCher. E transformed the ratings, which were in evaluation terms, into numerical values (I to 4). An example of the evaluation terms used by the teachers were "poor," "unsatis- factory," and "needs improvement." Those terms had values of l, as compared to "excellent" and “OutStanding" which had values of 4. The values of 2 and 3 were generally interpreted as fair or average or good or above average respectively. The ratings were a cumulative evaluation of each S's academic achievement and social habits. MARI, Scores from the standardized Metropolitan Achievement Reading Test (MART) were Compiled for each S. The MART served as one of the measures of the 55' academic achievement. MAMT, Scores from the standardized Metropolitan Achievement Math Test (MAMT) were also gathered. The MAMT was another measure of the 35' academic achievement. I EEENEL. Scores from the Minimal Performance Objective Reading Test (MPORT) were obtained as another measure of academic achievement. ‘The MPORT was a criterion-referenced test that contained 23 "minimal performance objective" items. MPOMT, Scores from the Minimal Performance Objective Math Test (MPOMT) were also compiled. Like the MPORT, the MPOMT was a criterion-referenced instrument that served as a measure of academic achievement. The MPOMT contained 35 ”minimal performance objective" items. 50 . Design The design was a 2 X 4 analysis with eight dependent variables. Because of the number of dependent variables, multivariate analysis of variance (MANOVA) was the statistical technique employed. A total of 37 Ss were eliminated for various reasons such as missing data and absenteeism, All observations except 10 scores were obtained for each 3. However, because of the necessary computer programming constraints for MANOVA, it was necessary to supply sub- stitute IQ values for any missing IQ values for any missing IQ data encountered. Since each S was categorized on the criteria of concep- tual tempo and SES, a group mean of each dependent variable could be calculated. The cell mean for obtained IQ scores was used as the value for those 55 (42) who had missing.IQ data. This procedure, hOwever, does not alter the group mean differences on 10 scores, al- though there will be a reduction of within group variability, espe- cially for the low SES category which had the largest number of missing IQ Data. Table l illustrates the number of $5 in each full cell as opposed to the cells containing missing IQ data. 51 Table 1.--The Number of $5 Per Full Cell for Each Category Except IQ. Numerals in Parentheses Represent the Number of Ss Whose IQ Data was Collected. High SES Low SES Reflective 29 (27) 26 (16) Impulsive 35 (32) 36 (23) Slow-Inaccurate 9 ( 8) 15 ( 7) Fast-Accurate 17 (16) 13 (10) Total 90 (83) 90 (56) Results The test for interaction (SES X conceptual tempo) indicated no significant interaction (multivariate F = 1.25, d: = 3/172,_p<.20). The cell means are presented in Tables 2 and 3 for high and low SES Ss classified as reflective, impulsive, slow-inaccurate, or fasta accurate. Although there was no significant interaction, of interest was whether or not any significant contrasts existed within SES groups among reflective, impulsive, slow-inaccurate, and fast-accurate $5 on any of the eight dependent variables. Through the use of the Scheffe' post hoc comparison test, it was shown that differences between low SES fast—accurate and slow-inaccurate $5 on MPOMT scores and teacher ratings were significant (p§.05 for MPOMT and £5.01 for 52 teacher ratings). Also, significant teacher ratings differences (25.05) were found between low SESifast-accurate and impulsive 35. All differences favored fast-accurate 55. For high SES $5, the post hoc procedure showed that reflective 55 had significantly (95.05) higher teacher ratings than impulsive $5. The hypothesis that significant differences on the eight dependent variables were related to SES was supported (multivariate F = 21.65, gf_= 1/172, p§.0001). When univariate F tests were applied, differences on all eight observations resulted in probability values ranging from<.001 to<.0007. The SES cell means for each dependent variable, as shown in Table 4, indicated that the high‘SES 55 performed significantly better on all eight measures. 53 AN_.mV Fm.e_ A_o.mV No.mN AoN.V me n Ago NPV NN.No_ ANo.FV e“ m ANm.PV mm.m Amm.NV N~.NN ANe. AV Ne : ee<-emee ANN.NV Nm.e_ AFF.SV _F.eN A_N.V mo.N Amm.m V MN.om Aem. V eo.N Ame. V ~¢.N Amo.NV AN.o. Amm.~V Nm.e eeeeH-opm Amm.NV mm.m_ Amp.mV mm._N Amo.V NN.N AFN.N V om.mw Ape. V No.m Ame. V me.N Ame.NV em._N Ao_.mV Nm.m e>em_=eeV AmN.NV me.m_ Aom.mV om._N ANN. V «m. N Aom.~ V oo.Nm Amo._V em m Amo._V Pm.N nem.mV Pm.mN ANN.AV NN.N e>eeee_eem um Nma ucmvcmum .mm mumcauumm_:QEH .m>wuumpmom mum 304 com mcmw: _Pmu11.m m_nmh Aam.NV .N.m_ ANA.NV mo.eN Rom. V co. N A_A.mVV Nm.mpp Amo._V mA.e Amm. _V co. m “Nm.mV mm.om ANN.NV mm.e_ ee<-umee Amm.NV he.ep A__. «V PF. eN ANN. V NN. m Ae_._VV ee.o_P ANm._V mm.e ANN. NV me. e Ace.NV oo.Nm Ace.AV NN.o_ QueeH-epm A_N.NV me.m_ ANN. NV NN. NN Ana. V me. N ANN._NV me.m__ ANO._V oe.e AON._V me.e Amm.eV oe.mN ANN.NV eN.eF e>_mV=ee~ Amm.NV No.9, ANN. NV “a. eN Ace. V mo. m ANN.o_V eN.m_V ANo.FV 0N.e Ace. _V me. e ANN.mV e_._m ANN.mV SN.N_ e>weee_eem um Nmo ugmucmum .mm mumczoo< «men use .Oumczouch-Zo_m .o>_m_:aEH .m>wuumpmmm mum saw: so» mcmmz F—ouuu.w m_nmh 54 o N AN_.NFV NN.NNV ANN._V NN.e V e.N A_N.N V NP.NN ANN._V NN.N V N N ANN.NPV NN.NN ANN. V NN.N ANV.FV N N ANN.N V NN.¢NF ANN._V No.4 V ANN.NV NN.NN ANN.NV NN.e_ ee<-NNNN apN.NV o_.NN ANN.NV NN.NN A_N. Ne e NN.N ANN.NV eN.NN NN.NV NN.N_ eeNeH-e_N NN N NN N V N ANN.NV NN.N_ ANN.NV FN._N ANN.V N ANN.NV NN.NF ANe.NV _N.NN ANN.V N ANN.NV eN.e_ ANN.NV NN.NN ANN.V N ANN.NV NN.NN ANN.NV No.op NSNN_=aeN AN_.NV NN.NN ANN.NV NN.N. N>NNNN_NNN um .xmo ucmucoum .chzmmwz cowum>cmmno ucmwm mzp co mm wpmczuu_m~:aEV .m>wpowpwmm cow mcmmz mgpnu.m m_nmh ANN.NV eN.N_ ANN.NV NN._N ANN.V Ne.N AVN.N V _N.NN ANN. V NV.N ANN. V NN.N ANe.NV NN.NN ANN.NV NN.N NNN NOS Amo.mv Fm.m— Amm.mv om.mm Amm.V mm.~ Avm.npv om.vp— nom._V wm.¢ Amo._v No.¢ Amm.mV Fm.om Awk.ov mm.m_ mum car: um m<~ mp OH Pzmo vcmucmum .mmczmmmz cowpm>gwmno acmwm asp :0 mm mmm 304 ucm now: co» Ncmmz m:p--.¢ m_nmh 55 Conceptual tempo cell means on each dependent variable are shown in Table 5. The hypothesis that conceptual tempo was associated with the eight observation variables at the traditional levels of significance was not supported (multivariate F = 1.42, df_= 3/172, 95.09). Although traditional levels of significance were not attained, such findings as above should not be disregarded. The interpretations and inferences at such a level, however, should be carried out with caution. one way of proceeding cautiously is to insure control by using the~BonferrOni Inequality Index. The purpose of this index is to hold the experimentwise error rate at a conservative level of confidence such as .05. In using the Bonferroni Inequality Index, the level of significance for each univariate F was maintained at .05 by dividing the number of dependent variables into .05, thereby yielding a confidence level of .006 per variable. Of the univariate F tests, as displayed in Table 6, only teacher ratings were significant at the .006 level. 56 Table 6.--Univariate Analysis of Variance for Each Dependent Variable as Associated with Conceptual Tempo. Variable Mean Square Between Univariate _p Less Than 10 584.88 3.28 .02 . MART 4.85 2.52 .06 MAMT - 3.79 3.44 .02 MPORT 146.50 2.96 .03 PMOMT 93.31 1.77 .15 TR 3.57 5.40 .002 IAR 26.46 ' 1.88 .13 SC 4.55 .64 .59 57 Because of differences found among 55 on the factor of conceptual tempo, a conservative post hoc comparison technique using the least square estimates of effects and the standard error of least squares was applied to examine contrasts among reflective, impulsive, slow-inaccurate, and fast—accurate $5 on the ' dependent variables. That post hoc procedure is an estimate of the group mean differences by using the standarderror comparisons were not made on the variables of IAR and sense of control in view of the small univariate F ratios (F~= 1.88, 25.14 and F = .64, p§.59 respectively). .The post hoc analysis on the remaining six variables showed that reflective 35 had higher MPORT scores (95.05) and higher teacher ratings (25.01) than impulsive Ss. Reflective 55 also had higher MAMT scores (25.05) than slow-inaccurate Ss. Fast-accurate 85, it was found, had higher MPOMT scores (95.05), MART scores (95.05), MAMT scores (p§.05), IQ scores (95.01, and teacher ratings (35.001) than impulsive Ss. Fast-accurate 85 also compiled higher MART scores (95.05), MAMT scores (95.01), 00 scpres (95.05), and teacher ratings (p§.05) than slow-inaccurate $5. No significant differences on any measures were found betWeen reflectives and fast-accurates nor between impulsives and slow-inaccurates. 58 Discussion . Since previous studies suggested that low SES children were more inclined toward impulsivity than high SES youngsters (Kagan, 1966; Schwebel, 1966), it was predicted that children characterized as both low SES and impulsive would have the lowest scores on measures of academic achievement. No SES X conceptual tempo interaction, however, was found in the present study. Furthermore, Kagan (1966) did not actually test his assumption that impulsivity is related to low SES youngsters, and Schwebel (1966) based his conclusions upon scores from verbal tasks which high SES pupils traditionally perform better on than those from low SES backgrounds. Eska and Black (1971) also stated that fast response times were indicative of low SES children. In the present study, however, 55 were placed in the four conceptual tempo categories on apriori criteria instead of employing the 55' median response latency and error scores for classification purposes. aIt was observed that there were 49 fast responding low SES Ss and 51 fast responding high SES 55 out of the total sample of 180 which was divided exactly in half on the basis of SES. Such results were counter to those who contended that short response latency was a disposition more common to low SES children. The data supported the hypothesis that SES would be strongly associated with the dependent variables to the extent that high SES 55 would exhibit higher mean scores across all dependent measures. The findings of strong associations for SES with academic achievement and the greater school related success of the high SES 35 were consistent 59 with earlier studies in the area (Havighurst and Janke, 1944; Havighurst and Breese, 1947; Pierce-Jones, 1959; Wilson, 1963). High SES 55 also displayed higher feelings of control over their school environment as shown on the IAR and Sense of Control Scale. This . appears to coincide with the greater success in academic achieve- ment that was demonstrated by the high SES $5. The indications are that if one group experiences more sChOol related success than another, then it follows that they should also have higher feelings of control over their school environment. I Another view of conceptual tempo was gained by including fast- accurate and slow-inaccurate Ss in the sample. The overall contention resulting from previous research was that reflective pupils, for the most part, demonstrated higher levels of academic achievement (Kagan, l965a, 1965b, and 1966; Kagan, Pearson, & Welch, l966a and 1966b; Schwebel, 1966; Yando & Kagan, 1968; Messer, 1970). However, the results of the present study indicated that the relective 85 were not the consistent high performers. The high SES slow-inaccurate 85 achieved highest mean teacher ratings and MPOMT scores, and the high SES fast-accurates displayed the highest mean scores on the MART, IQ, and equalled the mean score Of the high SES reflective 55' on the MAMT. For low SES $5, the fast-accurates were the consistent high performers within that group on each of the dependent measures. One of the more surprising findings was in the area of teacher ratings. Although the high SES slow-inaccurate Ss compiled the highest mean teacher rating for the entire sample, the opposite was observed 60 for the low SES slow-inaccurate Ss who compiled the lowest mean teacher rating. The differences found between high SES slow- inaccurates and low SES slow-inaccurates might be explained in terms of teacher eXpectations (Rist, 1970). The teacher may perceive the high SES slow-inaccurate youngster as trying his hardest . to succeed and hence rewards him for his efforts. The high SES-slow- inaccurate Child may also be perceived as non-threatening, for he does not display traits of high activity that is attributed to im- ‘pulsive children, nor does'he”quitkly provide the correct answer to a question or problem as might a fast-accurate child. He might appear to be more dependent upon the teacher than either fast-accurate or reflective children, and this might also work in his favor in a relationship with the teacher. 0n the other hand, the low SES slow-inaccurate child may be perceived in terms of just being ”slow" i.e. a slow learner in most instances, and therefore, is evaluated accordingly. The low SES fasteaccurate child, as Opposed to his slow- inaccurate counter part, is probably perceived as the ”beacon of light" in a disadvantaged setting who is going to be successful and overcome the possible detrimental effects of a low socioeconomic back- ground, and therefore he attains high teacher evaluations that far surpass the other 55 within the low SES group. The results suggested that of the two components that comprise conceptual tempo, response time and accuracy, that accuracy is the (Imore important factor in relation to academic achievement. In earlier studies, the primary suggestion was to modify the response latency of impulsive children by increasing their response latency, therefore 61 = making them reflective (Kagan, Pearson & Welch, 1966b; Yando & Kagan, 1968). However, the present study is in accord with more recent, research which shows that slowing down response speed has little or no effect on performance (Zelniker, Jeffrey, Ault & Parsons, 1972). In order to further examine the importance of accuracy, a Scheffe' post hoc comparison procedure was used to contrast accurate Ss (reflectives and fast-accurates) with inaccurate Ss (impulsives and slow-inaccurates). A comparison was then made between fast 55 (impulsives and fast-accurat8%9*and slow Ss (reflective and slow- inaccurate). Significant differences (95.05) were attained on every dependent measure except the MART when accurate and inaccurate 33 were contrasted, but no significant differences were found when fast and slow 85 were compared. The results from the post hoc procedure suggested that accuracy was by far the more significant factor and that response latency was of little importance when related to measures of academic achievement. The data indicated that overall, the high SES child met with relative success regardless of his speed in responding or accuracy as determined by the MFF test. The single exception was teacher ratings, which favored accurate $5. The important factor for the high SES child was his high socioeconomic level, not the accuracy or response latency components of conceptual tempo. For the low SES child, on the other hand, the accuracy component of conceptual tempo appeared to be highly related to academic achievement. Therefore, it seems that not only is it to the low SES child's advantage to be accurate, but it would be best if he was fast as well. Of the high SES group, the only 85 who 62 demonstrated consistently 10wer performances were the impulsives, as predicted, and in that respect a slower response-time may be related to academic achievement in a high SES setting. Nevertheless, it should be noted that only on teacher ratings were there significant differences within the hi.gh SES group. Differences associated with conceptual tempo on the two internal-external control measures were not significant. The results suggested that an individual's conceptual tempo has little or no effect on how he perceives the control4he has over his school environment. However, the data indicated that low SES fast-accurates displayed IAR scores 6X7: 23.92) that were in the same range (X'= 22.77 to'X = 24.97) as the high SES Ss. Such results obtained from low SES fast-accurates might be expected. Since they have the highest attained standards across all measures for the low SES group, it was not surprising that, the low SES fast-accurates would demonstrate relatively high feelings of control over their school environment. In future investigations of conceptual tempo, prior to the administration of instruments, standards should be established such as the MFF test, that serve toCategorize the conceptual tempo of subjects. Efforts should be made to set up norms for conceptual tempo dimensions at various developmental stages that would act as guide- lines of conceptual tempo regardless of age. This should provide con- sistency established for classifying conceptual tempo, situations may and have probably already occurred where one study classifies certain 55 as impulsive, whereupon another study categorizes the same 55 as reflective. This may occur because the median response latency and 63 error scores (which in all liklihood varies across samples) for each specific sample serves as the cut-off point in designating conceptual tempo dimensions. With such occurrences, valid replications would be difficult to accomplish. For further research, past studies that had been carried out with only reflective and impulsive subjects in the sample should be replicated but with the addition of fast-accurates and slow-inaccurates as subjects. Moreover, it is suggested that all future research on conceptual tempo include fasteaccurate and slow- inaccurate subjects in the research sample. Examining and determining the cognitive strategies used by subjects in the four conceptual tempo groups in relation to problem solving and information processing should also prove valuable. Stringent efforts should be made to reveal whether or not various modes of instruction when paired with an in- dividual's conceptual tempo, do in fact have any beneficial effects on the individual's learning capabilities. Finally, if attempts to modify conceptual tempo are continued, the primary focus should be upon improving the accuracy component as Opposed to increasing response latency. REFERENCES Ault, R.L., Crawford, D.E., Jeffrey, W.D. Visual scanning stra- tegies of reflective, impulsive, fast-accurate, slow- inaccurate children on the Matching Familiar Figures Test. Child Development, 1972, 43, 1412-1417. Ault, R.L. Problem-solving strategies of reflective, impulsive, fast-accurate, and slow-inaccurate children. Child Develop: ment, 1973, 44, 259-266. Battle, E.S., and Rotter, J.B. Children feelings of personal control as related to social class and ethnic groups. Journal of_Personality, 1963, 31, 482-490. Brookover, W.B., Gigliotti, R.J., Henderson, R.D., and Schneider, J.M. Elementary school social environment and school achievement. Research report published by the College of Urban Development, Michigan State University, East Lansing, Michigan, 1973. Coleman, J.S., Campbell, E.G., Hobson, C.J., McPartland, J., Mood, A.M., Weinfield, F.D., and York, R.L. Equality of Educational Opportunity, Washington, D.C.: U.S. Government Ovvite, 1966. Crandall, V.C., Katkovsky, W. and Crandall, V.J. Children's beliefs in their own control of reinforcement in intellectual-academic achievement situations. Child Development, 1965, 33, 91-109. Eska, B., and Black, K.N. Conceptual tempo in young grade-school children. Journal of_Educational Psychology, 1944, 33, 357-368. Havighurst, R.J. and Janke, L.L.w Relations between ability and social status in a mid—western community. Ten-year old children. Journal of_Educationa1 Psychology, 1944, 33, 357-368. Havighurst, R.J. and Breese, F.H. Relation between ability and social status in a mid-western community. III. Primary mental abilities. Journal of Educational Psychology, 1947, 33, 241-247. 64 65 Heider, E.R. Information processing and the modification of an ”impulsive conceptual tempo.” Child Development, 1971, 42, 1276-1281. Kagan, J. Impulsive and reflective children: Significance of ‘ conceptual tempo. In J.D. Krumboltz (Ed ), Learnin and The Educational Process. Chicago: Rand-McNally, T965a :1 33-161. ' Kagan, J. Reflection-impulsivity and reading ability in primary grade children. Chilngevelopment, l965b,3§, 609-628. Kagan, J. Reflection-impulSivity: The generality and dynamics of conceptual tempo. Journal oj;Abnormal Psychology, 1966, 71, 17-24. Kagan, J., Pearson, L., and Welch, L. Conceptual impulsivity and inductive reasoning. Child Development, l966a, 32, 583-594. Kagan, J., Pearson, L., and Welch, L. Modifiability of an impulsive tempo. Journal oj3Educationa1 Psychology, l966b, 32, 350-365. Kennedy, W.A, A Follow-up Normative Study of Negro Intelligence and Achievement. Monogra hs of the for Research in Child Development, I969, 34,171IE7' ‘77 Kopfstein, D. Risk—taking behavior and cognitive style. Child Development, 1973, 44, 190-192. ‘ McKinney, J.D. Problem solving strategies in impulsive and reflective second grades. Developmental Psychology, 1973, 3, 145. Messer, S. Reflection-impulsivity: Stability and school failure. Journal of_Experimental Child Psychology,l972, 14, 265—276. Pierce-Jones, J. Socio—economic status and adolescents' interests. Psychological Reports, 1959, 5, 683. Rist, R.E. Student social class and teacher expectations: The self— fulfilling prophecy in ghetto education. Harvard Educational Review, 1970, 43, 411—451. Schwebel, A.I. Effects of impulsiv1ty on performance of verbal tasks in middle and lower-class children. American Journal of Orthopsychiatpy, 1966, 36, 13-21. 66 Shulman, L.S., Loupe, M.J., and Piper, R.M. Studies olehe Inquiry Process, Michigan State University, East Lansing, MiChigan: Educational Publicational Services, College of Education, 1968. Yando, R.M. and Kagan, J. The effect of teacher tempo on the child. Child Development, 1968, 33, 27-34. Wilson, A.B. Social stratifiCation and academic achievement. In A.H. Passow (Ed.), Education lp_Depressed Areas. N.Y.: Teachers College Press, Columbia University, 1963. Zelniker, T., Jeffrey, W E., Ault, R., and Parsons J. Analysis and modification of search strategies of impulsive and reflective children on the Matching Familiar Figures Test. Child Development, 1972, 43, 321—335 APPENDICES APPENDIX A Ph.D. DISSERTATION PROPOSAL Henry T. Frierson, Jr. College of education Department of Educational Psychology Committee Members: Dr. Walter G. Hapkiewicz, Chairman Dr. Maryellen McSweeney Dr. Wilbur B. Brookover Dr. George Ferree 67 68 Introduction A child's cognitive style is most important in affecting his learning and his adaptations to varying methods of instructions as well as instructors. For instance, when a pupil's cognitive style is congruent with the instructions he receives and also the cognitive style of his teachers, the chances for academic success are enhanced. On the other hand, when the converse is evident, the child's chances for academic success are minimized. Depending on the circum— stances, whether or not a child's cognitive style is congruent or incongruent with that of his teachers, such a situation may serve to help a child achieve a high academic self concept, or relegate his academic self-concept along with his total selfeesteem to the lowest eve . A cognitive style such as conceptual tempo--the major focus of this study--and selection strategies can serve to indicate how a pupil will perform intellectually in the classroom. Both are con- cerned with cognitive processes and products that are school related.' Conceptual tempo refers to response speed and accuracy in tasks such as those that involve problem solving. Selection strategies, on the other hand, refer to the way in which concepts are learned. For example, an individual either scans or focuses on various attributes that enable him to either accept or reject a presented concept that is to be learned or one that is a false representation of the ”correct” concept. Conceptual tempo and selection strategies are related. As an illustration, an individual who uses a particular selection strategy will be predisposed toward a conceptual tempo disposition that matches the selection strategy.. For instance, if an individual employs a conservative focusing strategy (systematically examining attributes of a concept in order to discover those that are essential), his conceptual tempo is likely to be of a reflective nature (long response time but few errors). Conservative focusing ensures success in learning a concept; reflectivity enhances the probability of choosing the correct concept from similar alternatives. The purpose of this study is to look at conceptual tempo--a subset of cognitive styles--and socioeconomic status in identifying differences in IQ reading achievement, arithmetic achievement,teacher ratings, intellectual achievement responsibility, and sense of control in middle grade elementary school children. Definition of Cognitive Style The contemporary defjn.itaon of the cognitive style construct presented by Kagan, Moss, and Sigel (1963), was defined as the stability of individual preference in perceptually organizing and conceptually categorizing the external environment. In the same vein, Hervey (1966) stated that cognitive style is the mode in which an individual organizes his experiences and perceives and understands his environment. At this point the term cognitive style is used widely to refer to a variety of behavioral consistencies in the processing of information (Stanes, 1973). 69 A Brief Historical Perspective of Cognitive Styles The construct of cognitive styles has its roots in psycho- analytic theory under the rubric of cognitive control. In the psychoanalytic theoretical structure, the emphasis was on the re- lationship between cognitive control and ego defense mechanisms, and also mediations of perceptual processes along the cognitive control principle of ”constricted-flexible control” (Kagan and Kogan, 1970). "Constricted-flexible control” refers to the capability of responding correctly in a task while facing a field of conflicting cues (Klein, 1954). Naming the color word, for example, that is printed by a set of incongruent colored letters, such as the word ”blue“ printed in green. Flexibility is exhibited when the individual can successfully perform the task. If the individual, on the other hand, states the color word as being the same as the incongruent colored letters, then this characterizes the constricted mode, Klein (1958) gave cognitive controls the label of cognitive styles. Cognitive control or style as perceived by Gardner and others (1959) of ego psychology orientations, was defined as the presence of stabilized ego structures that integrate specific adaptive intentions with specific environmental situations. Kagan and Kogan, however, pointed out that the definition provided very little insight in ex- amining individual differences in the functioning of cognitive controls because of the use of “intentions” as a variable. The term ”inten- tions” had a very vague meaning, and therefore it was difficult to employ the ego psychology definition of cognitive style to explain differing modes of cognitive functioning between individuals. This was especially true when assumptions had to be about ”intentions“ (for example, in experimental research studies). On an experimental basis, the “intentions” of the subjects should be assumed to be more alike than different. Therefore, one mus1 assume that all the subjects intend to perform to the best of their ability, unless there is independent evidence that indicates otherwise “Intentions,” unfortunately because of its vague meaning, dld not readily lend itself to such an assumption. The restrictions of the psychoanalytic theoretical framework became evident among researchers in cognitive processes. Those who studied cognitive styles were finally constrained to work out of a conceptual structure that was cognitively rather than psychoanalytically oriented. When various controls (or later, styles) were recognized by researchers, they were held to be organized as superordinate structures. Witkin et a1 (1962) were among the first to look at specific subsumers of cognitive styles when they described individuals using an analytical field approach or a global approach- Their work dealt with field dependence-independence. The field independent or analytical field individual tends to experience objects analytically and discrete from their backgrounds. The field dependent or global individual, on the other hand, is inclined to perceive his environ— ment in a global fashion while passively conforming to the influences of his environment. Witkin and associates indicated that an individual's predisposition toward globai or analytical inclinations 70 characterized not only his perceptual activity but his problem solving activities as well.- Problem » ' The construct of conceptual tempo, a subsumer of cognitive style, is the major concern and primary focus of this study. Through a multivariate analysis approach, the study is designed to examine differences in IQ, reading achievement, arithmetic achievement, teacher ratings, intellectual achievement responsibility and sense of control in middle grade elementary pupils. Kagan (l965a) posited that conceptual tempo plays a very important role in a child's academic success or failure. In most instances a child is initially responsible for the types of school experiences he undergoes only to the extent that certain inherent interacting dispositions serve to affect his tendency toward "motivational zeal'I or a display of apathy whenever he is to express products of intellectual quality. Of the interacting dispositions, conceptual tempo is one of the most important. Kagan reported that conceptual tempo points out a child's predispoSition to "reflect” upon the quality of a cognitive product or, in contrast, to make an "impulsive" and unconsidered response. Kagan (1956b) also reported that Conceptual tempo manifests per- vasive generality across varied task situations and, importantly, it is linked to some fundamental aspects of a child's personality structure. 'Thus, conceptual tempo appears to be an integral part of a child's personality. If such is the case, then implications can be made concerning the teaching of individuals in a manner that is clearly related to their conceptual tempo. It should be explicitly noted that Kagan, the initiator of conceptual tempo research, identified and was only concerned with two dimensions of conceptual tempo--ref1ection and impulsivity. Since a large number of individuals do not fall in either of the two dimensions, the present study is structured to incorporate four dimensions of conceptual tempo instead of the usual two dimensions. The two new dimensions, "fast accurate," and "slow inaccurate," should be as much an influencing factor on the aforementioned dependent variables as reflection and impulsivity. In order to maintain a proper perspective concerning the four dimensions, one should view the reflection- impulsivity dimension as "slow accurate" and "fast inaccurate” dispositiOns, even though the terms reflective and impulsive will be used for traits indicating slow accuracy and fast inaccuracy. Description of Conceptual Tempo Conceptual tempo, as defined by Kagan, is divided into two dimensions: reflection and impulsivity ("slow acCurate“ and "fast inaccurate"). Kagan, Pearson, and Welch (1966) described the re- flection-impulsivity disposition as: Some children impulsively report the first hypothesis that occurs to them, and this response is often incorrect. The reflec- tive child, on the other hand, delays a 71 long time before reporting a solution and is correct. The reflective child considers the differential validity of alternative answers, makes fewer errors in reading prose or in recalling serially learned material, and persists longer with difficult tasks. The reflec- tive child wants to avoid making an error and inhibits potentially incorrect hy- potheses. The impulsive child seems mini- mally concerned about mistakes and makes his decisions quickly. Response time or latency and rate of errors are the indices of the reflection--impulsivity dimension. Reflective ("slow accurate") individuals have longer response times and lower error rates than those who are classified as impulsive ("fast inaccurate"). The reflection-impulsivity disposition was found to be very stable across problem solving tasks with different levels of diffie culty. Therefore, conceptual tempo is believed to be an important characteristic of an individual's psychological organization (Yando and Kagan, 1970). As was previously mentioned, the dimensions of "fast accurate” and l'slow inaccurate" will be incorporated in the study. I'Fast accurate" and "slow inaccurate" will be incorporated in the study. "Fast accurate” individuals have long response times and high error rates. The disposition of fast accuracy-slow inaccuracy is assumed to be just as stable as the reflection- impulsivity disposition and also equally valid subordinates of conceptual tempo. Relationship of Selection Strategies to Conceptual Tempo Closel' rElated to conceptuél tempo is Bruner, Goodnow, and Austin's (1956 conceptualization of selection strategies. Selection strategies are also subsets of cognitive style and appear to be of great value in explaining the manner in which reflectives and impulsives ' process information. ~ Selection strategies describe various ways by which the learning of concepts is attained by controlling the order that examples and non-examples appear. Bruner and associates described three ob- jectives of selection strategies: (1) to ensure that encountered instances will contain appropriate information, (2) to make the assimilation of the information less cognitively straining, and (3) to control the amount of risk in the attainment of a correct solution. Bruner et a1 depicted four-selection strategies. Briefly they are: conservative focusing--the use of an example of a particular concept as a focal point while systematically changing one attribute of the example at a time in order to discover those attributes which are essential; focus gambling--similar to conservative focusing except more than one attribute is changed during a single period of time; simultaneous scanning--the formulation of several hypotheses about the attributes of a concept followed by a search for appropriate examples; successive scanning—-similar to simultaneous scanning but only one hypothesis is formulated and tested. Conservative focusing 72 almost ensures success; it meets all the objectives of a selection strategy. Focus gambling, on the other hand, produces cognitive strain in that more information must be assimilated and it is a high risk strategy--one may obtain the correct solution quickly or conversely, be totally incorrect. Simultaneous scanning, also a high risk procedure, produces a great deal of cognitive strain in that more information must be assimilated and it is a high risk strategy-- one may obtain the correct solution quickly or conversely, be totally incorrect. Simultaneous scanning, also a high risk pro- cedure, produces a great deal of cognitive strain within the individ- ual because of the need to keep track of all the information in the formulated hypotheses. Successive scanning, a low risk strategy, reduces the likilihood of the encounter of instances embedded with relevant information. Focusers have been described as systematic attention deployers and scanners as unsystematic attention deployers (Santostefano and Paley, 1964)., Selection strategies with high risk should prove attractive to the impulsive child; the low risk processes, part1cularly conserv- ative focusing, should be attractive to the reflective youngster. (DeCeddo, 1969). Travers (1963) proposed that adjustments in teach- ing made to coincide with a pupil's selection strategy. Similarly, Kagan stated that adjustments be made by the educational system to meet the cognitive needs of pupils. Although selection strategies and conceptual tempo appear very similar, they are conceptually different. Selection strategies are ways in which a concept is learned. lhis OCCurs through a cognitive process of monitoring the sequence in which examples and non—examples of concept attributes appear. Conceptual tempo, on the other hand, refers to Speed and accuracy in making a response to two or more difficult alternatives. Correct response, as associated with conceptual tempo, indicates that a concept has been learned, rather than hpw_it is learned as is the case for selection strategies. The two constructs are related in that an individual with a certain conceptual tempo is inclined to use a selection strategy that is congruent with his personality. ~As an illustration, a reflective individual would not be inclined to use either focus gambling or simultaneous scanning, because of the greater chance for error and the rapid response time inherent in the two strategies. Review of Related Research Messer (1970) observed that children who failed a grade were significantly more impulsive, although they were comparable in verbal intelligence to their peers who had not failed. This was in agree- ment with Kagan's contention that conceptual tempo is related to academic success. Earlier, Kagan (l965b) found that reading ability was negatively associated with impulsivity. Impulsivity was also discovered to be negatively correlated with inductive reasoning (Kagan, Pearson, and Welch, 1966). Both reading ability and in- ductive reasoning are perceived as being necessary skills needed for academic achievement. 73 It was reported that teachers have a tendency to assign higher ratings to girls who display long response times and to boys who make few errors (Yando and Kagan, 1968). Consequently, it appears that teachers are inclined to respond positively to girls who delay and to boys who do not make mistakes. These findings reveal some of the factors behind the higher failure rate for impulsive pupils. It. seems that the impulsive child exhibits certain traits which adversely affect pupil—teacher relations in the present-day school setting, such as quickly blurting out an answer even though the child may be unsure of its correctness. This may be due to difficulties in placing effective restraints on the tendencies toward action. Kagan et al (1966) reported that there are indications that most impulsives do not appear to be as apprehensive or anxious about making a mistake as does the majority of reflective children Ault, Crawford, and Jeffrey (1972) observed reflective, im- pulsive, ”fast accurate,” and IIslow inaccurate“ pUpils. Using in-class teacher ratings, they found that only reflective pupils were perceived as being highly attentive. It was also reported that pupil sex was not a differentiating factor--both reflective girls and boys were rated as highly attentive. Siegelman (1969) found that impulsive boys ignored two and one- half times as many alternatives per item on Kagan's Matching Familiar Figures task. She also noted that reflectives took significantly longer looks at the alternatives In a related study, Mann (1973) reported that the reflection-impulsivity dimension was associated with caution—haste in higher ratings to girls who display long response times and to boys who make few errors (Yando and Kagan. 1968). Con- sequently, it appears that teachers are inclined to respond positively to girls who delay and to boys who do not make mistakes. These findings reveal some of the factors behind the higher failure rate for impulsive pupils. It seems that the impulsive child exhibits certain traits which adversely affect pupil-teacher relations in the present—day school setting, such as quickly blurting out an answer even though the child may be unsure of its correctness. This may be due to difficulties in placing effective restraints on the tendencies toward action. Kagan et a1 (1966) reported that there are indications that most impulsives do not appear to be as apprehensive or anxious about making a mistake as does the majority of reflective children. Ault, Crawford, and Jeffrey (1972) observed reflective, impulsive, “fast accurate,” and “slow inaccurate” pupils. Using in- class teacher ratings, they found that only reflective pupils were perceived as being highly attentive. It was also reported that pupil sex was not a differentiating factor—-both reflective girls and boys were rated as highly attentive. Siegelman (1969) found that impulsive boys ignored two and one- half times as many alternatives per item on Kagan’s Matching Familiar Figures Task. She also noted that reflectives took significantly longer looks at the alternatives. In a related study, Mann (1973) reported that the reflection-impulsivity dimension was associated with caution—haste in decision making in children. 74 Not surprisingly, he found that reflective subjects took significantly more time in making decisions on ego—involving tasks. Nuessle (1972) reported that reflective subjects were more efficient at processing information. He posited that the reflective cognitive style facilitated information processing because it provided for more effective retrieval and recoding of information. Nuessle suggested that reflection- -impulsivity differences are strongly related to developmental differences in information processing. He *observed that the reflection periods of older subjects (9th graders) “were longer than those younger (5th graders). The older subjects were significantly more proficient at processing information. This ‘ observation would appear to suggest that the reflective dimension in- creases with age and differences between individuals decreases. NuesSle fOund, however, that when reflective younger subjects and im- pulsive older ones were specifically studied, they did not differ significantly on information processing. The lack of developmental differences between younger“refléCtive and older impulsive subjects implies that the reflection- impulsivity dimension is stable: reflective subjects maintain their more efficacious information processing capabilities through periods of develOpment, while impulsives continue to lag. In a study on problem solving strategies, ' A ult (1973) also found that younger reflective children were more " reflective than older impulsive youngsters on the Matching Familiar Figures Task. Rationale for Observing Four Dimensions of Conceptual Tempo Kagan and others Who- studied conceptuaT'tempo, categorized subjects along the reflection-impulsivity dimension on the criteria of response latency and rate of errors. The median response latency and error-rate of the observed sample was always incorporated in the classification of conseptual tempo. A subject whose response time is above the median and whose error rate is below, is classified as reflective. The subject whose response latency falls below the median and whose error rate is above,.is categorized as impulsive. The reflection-impulSivity dimension model of conceptual tempo is depicted in fig. 1. There is always Medjen Error Rate +.- . _ A Reflective \: Median Response Latenqy- /. Im ulsive v " ./ \ Fig. 1 75 a fairly large group of individuals--approximately one-third--who fit neither the defined reflective nor impulsive modes. Upon reviewing studies of conceptual tempo, it is apparent that those subjects who did not fit into the dichotomous reflection-impulsivity dimension were ignored. - , Shulman, Loupe, and Piper (1968), in a study of adult subjects, reported that conceptual tempo and inquiry competence were related. Reflectives generally proved to be more effective-inquirers than im- ‘pulsives. It should be acknowledged, however, that Shulman and associ- found that the error rate_component was significantly more important in predicting inquiry behavior than the amount of reflection time. In essence, and individual may fall in either category of reflectivity nor impulsivity because of an observed tendency toward a low error rate coupled with a short reflection time, but because of his low error rate, the individual-is nevertheless and effective inquirer. 0n the other hand, the converse can be observed in an individual with a long reflection time as we11faS‘a high error rate. That particular individual's ability as an effective inquirer should be lower. ‘Such observations indicate a need to look at individuals who do not fit into Kagan's reflective-impulsive dichotomy. Fig. 2 is an illustration of the conceptual tempo model which incorporates the "fast accurate" and "slow inaccurate" dimensions. Shulman et al Median Error Rate + A\ Reflective Slow Inaccurate+ / \ Median Response \ /. Latency Fast Accurate' Impulsive \d/ Fig. 2 noted that by focusing-Upon only reflectivity and impulsivity, throwing out those subjects who do not fit either mode, one may be consistently confounding reflection and accuracy. Variables such as reading ability or inductive reasoning that correlate consistently with reflection time, may actually highly correlate with the aspect of reflection which involves one's capability for discrimination accuracy. Relationship Between Socioeconomic Status and Conceptual Tempo * Socioeconomic status SES is another factor that is by and large accepted as an important determinant in a child' 5 academic success (Battle and Rotter, 1963). Kagan (1966) posited that high SES children tend to be more reflective than those from a low SES background: "investigators 76 working with 'culturally deprived' children believe that one reason for their poor intellectual performance is their impulsive orien- tations". Ina study of middle and lower class boys (ages 9 to 12), Schwebel (1966) reported differences and also effects of impulsivity on the performance of verbal tasks. the middle class subjects performed higher on the verbal tasks, and the lower class subjects had higher inclinations toward impulsivity. Schwebel stated that social class differences in verbal performance were attributable to language- -speech competence and to reflection- -impulsivity. It appears that the lower SES child has the factors of low verbal performance (vis- -a- vis the academic environment) and his predisposition toward impulsivity as impingements working against his achieving school related success. This,-in all probability, is a cumulative process--the further the low SES child advances in school, the more stifling or debilitating his school experiences. According to Rist (1970), a child is labeled by the time he leaves the first grade as to whether or not he is to experience overall success or failure in the classroom. The primary basis for his labelling is the child's appearance (essentially whether or not he/she is clean, neat, and somewhat fashionably dressed as perceived by significant others such as teachers and school administrators). Since appearance is highly related to SES, it is assumed that a child who has a more "desired" appearance is more likely to come from a family with a higher SES than a child who is depicted as having a less "desired" appearance. A child perceived as having a high SES familial background, enjoys a higher probability of experiencing a more receptive and positive attitude (conscious or unconscious) than those viewed as being less economically fortunate. Moreover, a greater amount of benign attention from teachers and other school related personnel might be granted to those seen as being of a higher SES as opposed to those classified as economically disadvantaged of of a lower SES. Such occurrences; which are related to teacher- -expectations- of pupils, are also closely associated with academic and other school oriented success. v f Pupils who are either reflective or quick (“fast accurate“) and of high SES backgrounds may be viewed more favorably by their teachers and as a consequence will experience more academic success than impulsive youngsters of low SES backgrounds. Clustered between these roups might be the children who are reflective and low SES, slow ("slow inaccUrate“) and high SES, quick and low SES, impusive and high SES, and-slow and low SES in that respective order. If such aforementioned differences are observed, the implications then suggest that a child's conceptual tempo along with his SES background predicts whether or not he is more prone to experience success, failure, or even mediocrity in the schools by the time he has completed the early elementary grades. A further implication is that children are channeled into academic or in- tellectual categories--based upon SES and conceptual tempo--as early as age seven or eight. 77 METHOD Purpose One objective of this study is to examine existing differences between impulsive, "fast accurate," reflective and "slow inaccurate middle grade elementary school children on six criteria variables:. (1) IQ, (2) arithmetic achievement, (3) reading achievement, (4) teacher rating, (5) intellectual achievement responsibility, and (6) sense of control. Another objective is to examine the inter- actions of conceptual tempo and SES with the six dependent variables. , - In the design of the study, conceptual tempo and SES are the independent variables.» Arithmetic achievement test scores, reading achievement test scores, IQ scores, teacher ratings intellectual achievement responsibility, and sense of control are the dependent variables.f Differences within the dependent variables associated with conceptual tempo and SES will be examined, as well as the interactions that the two independent variables may have with the six observations. The design of the study is illustrated in Fig. 3. Multivariate analysis will be the statistical analysis technique employed in the study. 78 Fogpcou mo mmcom - .o.m mar—envwcogmom pcwsw>mwgo< szpomFFmch - mwvso< owpmaspws< - .< .gpw;< psoEm>mwco< mcwvmwm - .< .mvm mmpwgsuoacH so” mw>wm_:aEH mmpwgzoo< amok mm>wpumemm .u.m mand teacher expectations: The self-fulfilling prophecy in ghetto education. Harvard Educational Review, 1970, 49_, 411-451. Santostefano, 8.6., and Paley, E. DevelOpment of cognitive controls in children. Child Develppment, 1964, 35, 939-949. Schwebel, A.I. Effects of impulsivity on performance of verbal tasks in middle and lower-class children. American Journal prOrthopsychiatry, 1966, 3Q, 13-21. Shulman, L.S., Loupe, M.J., and Piper, R.M. Studies pf [pe_ Inquiry Process, Michigan State University, East Lansing, Michigan: Educational Publicational Services, College of Education, 1968. Siegelman, E. Reflective and impulsive observing behavior. Child Development, 1969, 49, 1203—1212. Stanes, D. Analytic responses to conceptual style test as a function of instructions. Child Develgpment, 1973, 44, 389-391. Travers, R.M. Essentials pf Learning. New York: The MacMillan Co., 1963. Yando, R.M. and Kagan, J. The effect of teacher tempo on the child. Child Development, 1968, 39, 27-34. Yando, R.M. and Kagan, J. The effect of task complexity on reflection-impulsivity. Cognitive Psychology, 1970, 1, 192-200. Witkin, H.A., Dyk, R.B., Faterson, H.R., Goodenough, D.R., and Karp, S.A. Psychological Differentiation. New York: Wiley, 1962. APPENDIX B METHOD Purpose - .~ One objective of this study is to examine existing differences between impulsive, "fast accurate," reflective and "slow inaccurate" middle grade elementary school children on eight criteria variables: (1) IQ, (2) standardized reading achievement test scores, (3) crite- referenced reading achievement test scores, (4) standardized arithmetic achievement test scores, (5) criterion- referenced arith- metic achievement test scores, (6) teacher ratings, (7) intellectual ,achievement responsibility, and (8) sense of control. Another objective is to examine the interactions of conceptual tempo and SES with the eight dependent variables. In the design of the study, conceptual tempo and SES are the independent variables. Both reading achievement test scores, both arithmetic achievement test scores, IQ scores, teacher ratings, intellectual achievement responsibility, and sense of control are the dependent variables. Differences within the dependent variables associated with conceptual tempo and SES will be examined as well as the interactions that the two independent variables may have with the eight observations. The design of the study is illustrated in Figure 3. Multivariate analysis will be the statistical technique employed in the study. 87 88 .m .mvc —ogp:ou to mmcmmuuum zywpwawmcoammm pam5m>mwso< szuumpymchuummwcu< ovumsgpwc< umocmgmmmmucowgmp_guuump<mwgo< ovum2;uwg< umeugmvcmumuump<w_go< mcwummm umucmcwcmmucowgmumgu-umpmwso< mcwumwm,mmNWuLmucmumluwhwm—=aEH .mwamgaoo< owed mw>wuow—wmm um m_m_=aeV Amm.~V oe.m_ Aom.mV om.P~ Aew.V em.~ Aom.~ V oo.~m Amo._V s Amo._V _m.N Aem.mV .m.m~ ANQ.NV mm.w a>_uoa_com um moo ungcmpm .mm mumcauume:QEH .m>wuuw—$wm mum :04 Low mama: _Fmoln.m mpnmh Amm.~V _~.mp ANN.mV mo.e~ Vom.V co.m “_N.m_V Nw.m__ Amo.PV ms Amm._V co.m Ama.mV mm.om Amm.NV mm.o_ oo<-pmee Amm.~V Ne.¢_ A__.¢V __.¢N Amm.V NN.m AeV.__V e¢.op_ ANm._V mm AN~._V me.¢ Ace.NV oo.~m hoe.~V N~.m_ coacV-opm A_m.~V m¢.m_ Am~.mV “K.NN Amm.V m¢.N Amm.PNV me.m__ Awo._V as Ao~.FV m¢.¢ Amm.mV oo.m~ A¢N.NV mN.¢_ o>me=QsV Amm.~V No.8, A-.mV N¢.¢N Aom.V ao.m Amm.o_V ¢N.m__ Ako._V on Ame.PV me.e A-.mV e_.Fm ANN.mV o~.5_ m>_oum_cox um mmo ugmucmum .mm mumgzoo< amen use .mpmgsoomcH-Zo_m .m>wmpaa5_ .m>wpoopmmm mum saw: so; memo: _pmuuu.m «pack 100 Table 4.--Sample Correlation Matrix for the Eight Observation Variables. MPORT MPOMT MART MAMT IQ IAR SC TR MART 1.00 MAMT .66 1.00 MART .65 .42 1.00 MAMT .63 .49 .72 1.00 IQ .38 .30 .48 .47 1.00 IAR .27 .22 .36 .31 .17 1.00 SC .29 .26 .40 .41 .28 .28 1.00 TR .58 .50 .53 .55 .36 .25 .40 1.00 The model to be tested for fit is y = u + a1 + Bj + eij; where: y = observation variables p = population a1 = socioeconomic status Bj = conceptual tempo eij error term The claim of the model is that there will be no interaction. As shown by tables 5 and 6, the estimated means and residuals in the form of t statistics respectively, the vast majority of the t scores were not significant. Thus the model seemed appropriate. The test for interaction yielded a multivariate F of 1.2447 (df = 3/172, p_< .20) that indicated no significant interaction. Thus the previous decision concerning the absence of interaction in the test of the fit of the model (y = u + a1 + Bj + e instead of y = p + a1 + Bj +y + e) is confirmed. The hypothesis of a significant difference related to socio- economic status was supported by the multivariate F of 21.6510 (df = 1/172, p_< .0001). Hummel and $1190 (1971) stated that a test of MANOVA followed by univariate ANOVA is a suggested technique when dependent measures are correlated. They reasoned that such an approach results in a conservative experimentwise error rate (the probability that at least one comparison will be significant when in fact the null 101 Table 5.--Estimated Cell Means for All Eight Variables MPORT MPOMT MART MAMT IQ TR IAR SC Hi SES Ref. 17.37 31.00 4.65 4.74 115.09 3.01 24.25 15.74 Hi SES Imp. 14.15 29.54 4.41 4.39 111.07 2.58 23.29 15.52 » Hi SES SI 15.41 28.75 4.32 4.15 112.45 2.72 22.92 14.98 Hi SES FA 17.72 32.60 5.18 4.91 119.87 3.20 24.84 15.93 Lo SES Ref. 9.32 23.46 2.90 3.36 92.17 2.62 22.30 13.83 Lo SES Imp. 6.10 22.00 2.67 3.02 88.15 2.19 21.35 13.61 Lo SES SI 7.36 21.21 2.58 2.78 89.53 2.33 20.98 13.08 Lo SES FA 9.67 25.06 3.44 3.54 96.94 2.81 22.90 14.02 Table 6.--Residuals in the Form of t Statistics MPORT MPOMT MART MAMT IQ TR IAR SC Hi SES Ref. .52 .18 -.Ol .20 .11 .87 1.81 1.17 Hi SES Imp. .20 .09 .12 .04 1.89 -1.21 -1.49 -.35 Hi SES SI .47 1.81 .50 .88 -.61 2.49* 1.29 -.48 Hi SES FA -1.34 —1.96* -.59 -1.10 -2.08* -1.71 -1.43 -.56 Lo SES Ref. -.52 -.18 .01 -.20 -.11 -.87 -1.81 -l.l7 Lo SES Imp. -.20 -.O9 —.12 -.04 -1.89 1.21 1.49 .35 Lo SES SI -.47 —1.81 -.50 -.88 .61 —2.49* -1.29 .48 Lo SES FA 1.34 -l.96* .59 1.10 2.08* 1.71 1.43 .56 *p<.05 Note.--The residuals in the form of t statistics tested whether or not the observed mean value differed significantly from the estimated mean. 102 hypothesis is true for an investigation with more than one companion) regardless of the number of variables and the proportion of variance that are in common. The use of the univariate F in the present study followed the method suggested by Hummel and Sligo (1971). As illus- trated in Table 7, the univariate F for each observation was signifi- cant; thus indicating strong associations with SES on each measure. Table 7.--Univariate ANOVA Related to SES Variable Mean Square Between Univariate F p Less Than IQ 24546.69 137.6414 .0001 'MATR 144.54 74.9657 .OOOl MATM ' 91.16 82.8259 .0001 MARI 3050.45 61.6797 .0001 MAMT 2722.22 51.5394 .0001 TR 8.02 12.1385 .0007 IAR 192.20 13.6513 .0003 SC 174.05 . 24.3830 .OOOl The level for testing each univariate F will be in accordance to the Bonferroni Inequality index. The purpose is to hold the experi- mentwise error rate at .05 when the Finn Computer Program (1970) is employed, because the Finn program is designed to compute independent F tests. Therefore, the level of for hypothesis testing was a = .05/8 = .006. The value 8 represents the number of dependent variables. The observed combined means for SES, as shown in Table 8, indicated that the high SES Ss performed better on all eight mea- sures. All differences were significant as demonstrated by the univariate F tests. 1(13 A_N.NV o_.mp AmN.eV oo.¢N AFN.V mo m Am_.¢_V om.mo_ Aom._V Nm.¢ s Amm.oV NN.NN ANN.NV NN.e_ ou<-Nm~L AON.NV NN.NF “No.3V FN._N ANN.V N¢.N AFN.N V N_.Nm Amo._V mN.m AFN.~V m .m ANN.mV so.¢N mN.NV Nm.o_ coacV-oPm AmN.NV mm.¢_ Amq.mV _N.NN AmN.V an N Aoo.m_V me.mm flew. V oN.m m Amm.NV NN.mN Aem.mV No.0. w>me=QsV VON.NV ¢N.¢_ Asm.mV NN.NN ANN.V mm N ANN.N V om.¢o_ Amo._V mo.¢ N Am_.NV ¢¢.NN ANo.oV om.m_ a>_uum_cm¢ um .mma ccmvcmum .mwczmmmz :owum>cmmao ucmwm mcu co mm mumczuowm—=QEH .m>_uow_+mm Lo» mama: mcH-1.m m_nmh Amo.NV em.m_ Aoo.NV NN.FN ANN.V ms.N A_N.N V _m.om flew. V m_.m ANN. V NN.N Ame.NV NN.NN ANO.NV 0N.N mum :04 Amo.NV .m.m_ ANN.NV ON.NN AmN.V mN.N ANN.N_V ON.¢__ VON._V Nm.¢ ANo.VV No.s ANo.mV _m.om ANN.QV mm.m_ mum ;m_z um moo vcmucmum .mmcsmmmz :o_pm>cmmno ucmwm ms“ :0 mm mum :04 use zm_: com mcmwz m:p--.¢ mpnmp 104 Table 8.--The Observed Combined Means For SES ...................... 3, —r MPORT MPOMT MART MAMT) IQ TR IAR so High SES 15.99 30.51 4.62 4.58 114.20 2.85. 23.86 15.61 Low SES 7.76 22.73: 2.83' 3.15 90.81 2.43 -21.79 13.64 The hypothesis that conceptual tempo was-associated with the eight observation variables was not supported at the .05 level of significance by the multivariate F of 1.4172 (df + 3/172, p <.09 but p >.05). Of the univariate F tests, as displayed in Table 9, only teacher rating is significant at the .006 level (the level of significance when BOnferroni's Inequality index is employed). However, for the univariate F tests per se, IQ, MART, MATM, along with TR were variables with p values of less than .05. Table 9.--Univariate ANOVA Related To Conceptual Tempo Variable Mean Square Between Univariate F p Less Than IQ 584.88 3.2796 .022 MATR 4.85 2.5167 .060 MATM 3.79 3.4453 .018 MART 146.50 2.9623 .034 MAMT 93.31 1.7666 .155 TR 3.57 5.4003 .002 IAR 26.46 1.8794 .135 SC 4.55 .6368 .592 The observed combined means for conceptual tempo are shown in Table 10. 105 Table lO.--The observed Combined Mean For Conceptual Tempo MPORT MPOMT MART MAMT 10 TR IAR sc Reflective 13.56 27.44 3.83 4.09 104.30 2.83 23.33 14.84 Impulsive, 10.07 25.72 3.53 3.70 99.45 2.38 .22.31 .4.55 Slow Inaccurate 10.37 24.04 '3.23 3.29 98.12 2.48 21.71 13.79 Fast Accurate 14.23 29.33 4.42 4.32 109.90 3.03. 24.00 25.10 Because of the differences observed in the factor of conceptual tempo, a post hoc comparison using the least squares estimates of effects and the standard error of least squares estimates was applied. This was done in order to examine specific contrasts. Comparisons made on the variables of IAR and sense of control were not presented in view of the small univariate F ratios (F = 1.8794, p <.l3 and F = .6368, p <.59 respectively). it was found that no significant: contrasts exiSted in those two areas. Tables 11 and 12 show the conceptual tempo comparisons along the academic related measures. 106 Table 11.-—Contrasts of Conceptual Tempo Through the Application of Least Squares Estimates of Effects (LSE) and the Standard Error of Least Squares Estimates (SE of LSE). . MPORT MPOMT MART compar‘sons SE of LSE LSEE SE of LSE LSEE SE of LSE LSEE Ref. - Imp. 1.253757 3.2169? 1.306006 1.4593 .249528 .2403 Ref. - SI 1.727913 1.9627 1.785679 2.2465 .341176 .3307 Ref. - FA 1.596708 -.3525 1.650087 -1 5999 .315269 .5292 Imp. - SI 1.665151 ‘-1.2543’ 1.720819 .7827 .328783 .0905 Imp. - FA 1.533348 -3.5695* 1.584609 -3 0592 .302759 .7694* SI - FA 1.936517 -2 3152 2.001257 -3 8464 .382364 .8599* Table 12.--Contrasts of Conceptual Tempo Through the Application of Least Squares Estimates of Effects (LSEE) and the Standard Error of Least Squares Estimates (SE of LSE). Comparisons MAMT IQ TR SE of LSE LSEE SE of LSE LSEE SE of LSE LSEE Ref. - Imp. .188531 .3430 2.399799 4.0172 .146089 .4336** Ref. - SI .257775 .5866* 3.281202 2.6390 .199745 .2888 Ref. - FA .238201 .1787' 3.032051 4.7758 .184578 .1907 Imp. - SI .248412 .2436 3.162021 -l.3782 .192490 - 1449 Imp. 4 FA .228749 -.5217* 2.911735 -8.7930** .177253 —.6243*** SI - FA .288895 -.7653** 3.677328 -7 4148* .223859“ -.4795* For Tables 11 and 12: *p<.05 ** p < .01 *** p < .001 107 Through the application of the post hoc procedure it was found that reflective 55 had higher Michigan Assessment Reading Test scores (p_<.05) and higher teacher ratings (p_<.01) than impulsive Ss. Reflective 85 also had higher Metropolitan Achieve- ment Math Test scores (p_<.05) than slow inaccurate 85. Fast accurate Ss had higher Michigan Assessment Math Test scores (p.< .05), Metropolitan Achievement Reading Test scores (9 <.05), Metropolitan Achievement Math Test scores (p_<.05), 10 scores (B_< .01), and teacher ratings (9 <.001) than impulsive 55. Fast accurate Ss also compiled higher Metropolitan Achievement Reading Test scores (p_<.05), Metropolitan Achievement Math Test scores (p_<.01), IQ scores (p_<.05), and teacher ratings (p_<.05) than slow inaccurate $5. No significant differences on any mea- sures were found between reflective and fast-accurates or between impulsives and slow—inaccurates. 1 = Reflective 2 = Impulsive 3 = Slow-Inaccurate 4 = Fast-Accurate 120 90 85 80 108 19 17 15 13 11 109 26 25 21+ 23 22 21 20 19 18 17 16 15. 14 13 12 11 110 / Viz-”K \ \ \ \\/ 2 3 '-+ / 2 3 H APPENDIX F TESTS OF HYPOTHESES The level of significance for testing all hypotheses will be at the .05 probability level. General Hypothesis 1: Both reflective and fast-inaccurate subjects on the eight dependent variables: scores on standardized math achievement tests, standardized reading achievement tests, criterion-referenced math achievement tests, criterion-referenced reading achievement tests, teacher ratings, IQ, intellectual achievement responsibility, and sense of control. 1. Both reflective and fast accurate subjects will have higher standardized math achievement test scores. The hypothesis was accepted for fast accurates over impulsives (p_<.05) and slow-inaccurates (p_<.01), and also for reflectives over slow-inaccurates but no significant difference was found between reflectives-and impulsives. 2. Both reflective and fast accurate subjects will have higher standardized reading achievement test scores. The hypothesis was accepted for fast-accurates over impulsives (p_<.05) and slow-inaccurates (p_<.05), but no significant differences were found between reflectives and impulsives nor reflectives and slow-inaccurates. 3. Both reflective and fast-accurate subjects will have higher criterion-referenced math achievement test scores. The hypothesis was rejected. No significant differences were found. 4. Both reflective and fast-accurate subjects will have higher criterion-referenced reading achievement test scores. ' The hypothesis was accepted for fast-accurates over impulsives ( <.05) and reflectives over impulsives (p_<.05), but no significant d1fferences were found between reflectives and slow-inaccurates nor between fast-accurates and Slow—in80curates. 111 112 5. Both reflective and fast-accurate subjects will have higher teacher ratings. The hypothesis was acce ted fast- accurates over impulsives (p_<. 001) and slow- inaccurates lp_<. 05), and also for reflectives over impulsives (p_<. 01) but no significant difference was found between reflectives and slow- inaccurates. 6. Both reflective and fast- accurate subjects will have higher IQ scores. The hypothesis was accepted for fast-accurates over impulsives (p_<.01) and slow-inaccurates (p_<.05), but no significant differences were found between reflectives and impulsives nor reflectives and Slow- inaccurates. 7. Both reflective and fast- accurate subjects will have higher intellectual achievement responsibility scores. - The hypothesis was rejected. No significant differences were found. ~ 8. Both reflective and fast-accurate subjects will have higher sense of control scores. . The hypothesis was rejected. No Significant differences were found. . ’ General Hypothesis 2: Reflective and fast-accurate subjects will not significantly differ on the eight dependent variables. The hypothesis was accepted. No significant differences were found on any of the dependent variables when reflectives and fast— _ accurates were contrasted. General Hypothesis 3: Impulsive and slow-inaccurate subjects will significantly differ on the eight dependent variables. The hypothesis was rejected. No significant differences on any of the dependent variables were found when impulsives and slow- inaccurates were contrasted. General Hypothesis 4: High socioeconomic status and low socioeconomic status subjects will significantly differ on the eight dependent variables. , The hypothesis was accepted for all dependent variables Gmultivariate F= 21. 65, df_= l/172, p s. 0001). General Hypothesis 5: Conceptual tempo and socioeconomic status will significantly interact on the eight dependent variables. , ) The hypothesis was rejected (multivariate F = 1.25, gf_= 3/172, ' p <.20 APPENDIX G MATCHING FAMILIAR FIGURES Answer Sheet Set 1 - F Notez. First two items are practice Item 1. house .g. . . l 2. scissors. . . 6 3. phone . . . . 3 4. bear ..... 1 5. tree ..... 2 6. leaf ..... 6 7. cat ..... 3 8. dress 5 9. giraffe . . . 4 10. lamp ..... 5 11. boat ..... 2 12. cowboy. . . . 4 Kagan 9/29/65 113 114 DIRECTIONS FOR MATCHING FAMILIAR FIGURES "I am going to show you a picture of something you know and then some pictures that look like it. You will have to point to the picture on this bottom page (point) that is just-like the one on this top page (point). Let'sdo some for practice." E shows practice items and helps the child to find the correct answer. "Now we are going to do some that are a little bit harder.' You will see a picture on top and six pictures on the bottom. Find theone that is juSt like the one on top and point to it." E will record latency to first response to the half-second, total number of errors for each item and the order in which the errors are made. If S is correct, E will praise. If wr0ng, E will say, I'No, that is not the right one. Find the one that is just. like this one (point)."' Concinue to code responses (not times) until child makes a maximum of six errors or gets the item correct. If incorrect, E will show the right answer. It is necessary to have a stand to place the test booklet on so thatboth the stimulus and the alternatives are clearly viSible to the S at the same time. The tWopages should be practically at right angles to one another. Note: It is desirable to enclose each page in clear plastic in order to keep the pages clean. L"'1 W111 115 H6 \w 118 x , \Q \K \\\\. !_\ s w (1 E” 0 ONO 0 o o 00 o, 0000 00".) 09 O 121 123 o- .-- .Hm.—.—~_ .1 125 127 131 132 133 134 ~ r- — » . -..__ WWW, 00000 135 .- I . LI. .55.!!! il‘kflll‘, .ILIIIIIIIII'II.‘ 'l 5| .Inol’ 5E.l.;i’hl I'I-IIII E‘lglu.‘ o) (.‘l {A p .1" ,I E .[I’i’xk. I. In 136 U/ 137 f1); ..) I 3 n 7; . I o. 2. fig‘liti Ia‘ar.....