11"! I I Abstract CHILDREN'S REACTIONS TO FRUSTRATION AS REVEALED IN MOTOR PERFORMANCE.AND AGGRESSIVE FANTASY The present study was an attempt to determine certain response effects of frustration in children, namely, whether frustration had a facilitative or inhibitory effect upon motor performance and if any change in these reactions related to aggressive fantasy. Temporal se- quence and locus of frustration, proximity of frustration and the nature of the reSponse class affected were important areas of investi- gation. The reaponse effects were studied in 108 middle class children. The children were divided evenly according to sex and into three age groupings (7% to 8, 9 to 9%, and 10% to 11). An experimental group of seventy-two children was subjected to early trial and extensive frus- tration cOnditions, and a control group of thirty-six children were subjected to late trial frustration conditions. Intentional (though non-arbitrary) interruption on a marble-board task constituted a frus- tration trial. Success conditions involved allowing task completion and giving of rewards. The eXperimental group was subjected to the following trial canditiOns: two success, six frustration, two success, three frustration, and two success. The control group was similar to the above, receiving six success trials instead of six frustration trials. After the six trials in question, both groups were asked to respond to a series of incomplete stories, intended to elicit aggres- sive fantasy. The reaponse variables under consideration were speed of motor performance on the task, force of a gross motor response ime mediately following the task, and, degree of aggressive fantasy following Abstract-2- the series of frustration trials. A.major finding was that, with late trial frustration, and c0ntin- ued prior success, there was a clear-cut facilitative effect on perfor- mance speed, which was not uniformly noted with early trial frustration. This result was interpreted in line with reinforcement and eXpectancy principles. ‘Another maJor result was that frustration effects were evi- dent in consistent inhibition of the extra-task motor reaponse. The in- hibition of this response, together with facilitation of performance speed, indicate that specific effects of frustration depend on the re- sponse class being measured. It was suggested that, at least, certain examples of frustration may activate relevant, instrumental reSponses and at the same time, depress irrelevant, non-instrumental responses. To ascertain the effects of frustration close to the task goal, proximity of frustration, both early and late trial Speeds were com- pared. .A slowing up of motor performance was found near the completion of the task. This effect, however, was independent of frustration con- ditions and appeared to reflect fatigue effects inherent in the task. The effects of frustration upon motor performance were unrelated to aggressive fantasy under the present experimental procedures. CHILDREN’S REACTIONS TOIFRUSTRATION AS REVEALED IN MOTOR PERFORMANCE AND’AGGRESSIVE FANTASY BY Frederick Joseph Behrle A.THESIS Submitted to the School for Advanced Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1959 ACKNOWLEDGMENTS Many peOple deserve Special mention for making this work a reality. As a committee chairman and friend, Dr. 11. Ray Denny has been consistently interested, patient and encouraging. The serious attention given this research by Drs. Howard S. Bartley, Albert I. Rabin, and Charles Hanley has added extensively to it. I wish to extend nnr appreciation to Drs. Frederick Bell and Rae Carlson, who served as Judges in rating the fantasy material. I am particularly indebted to the children and school personnel of the communities of East Lansing, Dewitt, Dimondale, Meridian, Holt and Potterville, Michigan for their willing and indispensable c00peration in this effort. TABLE OF CONTENTS PrOblem Theoretical positions with regard to frustration Frustration terminology Predided reactions to frustration Research with regard to frustration Aggressive responses as they relate to children Pre-test procedure Statement of inquiry Method Sample Apparatus Procedure Results Measures used Inquiries Discussion Summary References Appe ndix : I Incomplete stories II Mean speed, amplitude and fantasy scores Number 1 2 LIST OF FIGURES Page Apparatus 32 Apparatus 33 Mean time scores of 7%-yr.-olds and combined 9- and lO%-yr.-olds for experimental and control groups on all trials hl Mean amplitude scores of experimental and control groups of boys and girls for all trials A3 Mean aggressive scores with respect to age #6 Initial post-frustration performance speed on second half-trials of experimental and control groups with respect to age Corrected cumulative speed.means of experimental ' and control groups with respect to age 52 Corrected means of cumulative amplitude scores with respect to age 5’+ Mean half-trial performance Speeds for trials h—8 with respect to age 57 -PLEASE NOTE: Figure pages are not original COpy,t Light, indistinct type. Filmed as received. UNIVERSITY MICROFIIMS, INC. Number 10 13 1h 15 16 LIST OF TABLES Pge t test of amplitude scores on trial 1 1&1; Analysis of variance on the second half-trial: time scores on the hth and 5th trials #9 Initial amplitude change from trials 2 to 3 19 Analysis of variance of cumulative speed, equated for initial performance 53 Analysis of variance of cumulative amplitude scores, with correction of girls’ scores 55 t test of half-trial time differences of exper- imental and control groups 58 Post-success speed changes from trials 9 to lO 59 Post-success amplitude changes fran trials 8 to 9 60 _t_ test between early and late trial mean changes in learning rate of 75-yr.-olds with initial frustration on _t_ test between early and late trial mean changes in learning rate of 9- and 101 yr.-olds with initial frustration 65 Amplitude change from trials 10 to 11 66 Initial amplitude change fran trials 2 to 3 according to age groupings -67 Cumulative amplitude change from trials 1-2 to 3-8 according to age groupings 68 Analysis of variance of aggression scores with respect to conditions, sex, 86 age classi- fication 71 Analysis of variance of experimental group's aggression scores with respect to age and sex 73 Analysis of variance of control group's ag- gression scores with respect to sex and age 71+ PROBLEM The study of frustration has long been of active interest to those in the behavioral sciences. Though the many efforts in this field have not been rewarding in regard to consistent or general principles governing reactions to frustration, there has been an increasing appreciation of the extreme complexity of the problem. Examination of both the consistency and canplexity of varimls response classes following certain forms of frustration might appear, then, to be a fruitful. approach to the problem. The present study is an attempt to determine the response effects of certain types and amounts of frustration in children. More specifically, the present study is concerned with emining whether certain conditions of frustration have a facilitative or inhibitory effect upon motor reac- tions, and if an change in these reactions relates to aggressive fan- tasy. As a function of the ammmt and temporal sequence of frustration, the following reactions were studied: (a) speed of perfume on a task, itself; (b) strength of gross overt response immediately after the task (amplitude); and, (c) aggressive fantasy following a series of trials on the task. The apparatus used in a recent study (McDomough, 1958) afforded a means of exploring the behavior in (a) and (b). As the McDonough investigation provided, also, a background for the present inquiry, her procedure and results are described briefly. The sample consisted of 100 children of middle class background, evenlydividedaccordingto sex, andgrowedinto five age groupsfrcm 3%- to 9%years of age. The marble-board apparatus used for the -2- task was a modification of that used in the Haner and Brown (1955) ex- periment and will be described in detail in the Method section. In the McDonough study, every child was given at least ten trials to fill the marble-board as quickly as he could. In the first two and last two trials, the child was allowed to complete the task. The intermediate trials were terminated Just before the child filled the board by letting the marbles drOp through the board. These were considered frustration trials. At the end of each trial, the child was instructed to hit a plunger, which was considered the amplitude measure. The recorded amplitude, together with the speed with which the child filled twenty of the 2h-hole board, were considered to be measures of the child's reaction subsequent to frustration (noncompleted) and successful (com- pleted) task conditions. MCDonough predicted a facilitative effect of frustration based on theories suggesting incremental drive preperties of frustration (Amsel, 1958; Spence, 1956; Marx, 1956). However, her results were rather curious in that no such effect was noted in performance speed, while an inhibitory effect was noted in amplitude. The following analyses of her results have led to certain modifications in the experimental design: (a) There was a little change in performance on task speed when frus- tration was introduced in an early trial. This could be accounted for because response patterns to one particular condition (success) were not sufficiently established for disruption of conditions to have any effect (Denny a. Adelman, 1955; Tolman, 1932). The importance of estab- lishment of reSponse patterns was also inherent in.Amsel’s (1958) defin- ition of frustration. He maintained that a prior condition of continuous -3- reward (success) was essential to any drive producing effect of frus- tration. To explore whether later introduction of frustration procedure effects performance, frustration conditions were introduced early in the task experience for one group of children (in replicating the pre- vious research) and. for another group of children, later in the practice period (after having continued successful task experience). (b) McDonough had observed an ihhibitory effect on performance speed close to the completion of the task on frustration trials, which was suggestive of an avoidance reaction close to the point of frustra- tion. Some evidence (Denny a. Adelman, 1955) has indicated that avoidance reactions are set up in temporal and spacial proximity to the frustra- tion goal stimuli. In view of this, the measurement of performance speed close to the point of task completion was thought to be a more sensitive measure of inhibitory or facilitative effects of frustration than the overall trial speed. (c) It was further thought that if no inhibitory or facilitative effect from frustration showed up, it might be revealed in sane other measure. It was thought that aggressive fantasy might be a more sensi- tive measure of frustration. Some evidence has suggested that agges- sive needs of middle class children are revealed to a greater extent in fantasy than in behavioral manifestations (Sanford 33 £1, 19h3). It should be recalled that the sample used in the McDonough study consisted of middle to upper-middle class children. As is often the case with such children, their training may discourage overt motor reactions to frustration, while encouraging verbal and. symbolic means of expression. D. Miller and Swanson (1956) have suggested that middle class expression -14... of needs would be more in terms of indirect, symbolic means than by physical emrts. McNeil's study, cited in the same article, found that middle class children were more expressive with conceptual material than motor reactions; whereas, lower class children showed the Opposite pattern. Bellak (1951+) has also noted that aggressive needs were more likely to be prevented from overt expressions, though they may be strongly manifested in fantasy content. The questions arise, then, will this type of child be more expressive in verbal fantasy than in motor performance subsequent to frustration? It should also be noted that the amplitude measure (from striking the plunger ) may not be an apt measure of aggression, as Haner and Brown (1955) and. McDonough have inferred. The use of aggressive fantasy might be a more appropriate measure . With the introduction of aggressive fantasy as a variable, inter- est in the relationship between motor performance and. aggressive fantasy evolved. According to H. Miller (19M), with continued frustration, aggressive responses are more likely to becane manifest as other responses to the frustration situation becane extinguished. Hence, if certain motor responses becane less daninant, an aggressive response would be expected. If frustration does inhibit certain responses (as in the case of amplitude), it would be expected that as response strength diminished, there would be an increase in sane form of aggression, possibly fantasy. Sane theorists might argue that both motor performance and fantasy reactions to frustration change consistently. Though not specific to fantasy productions as such, certain researchers (Sears, 1937; Zeller, 1951) have indicated that individuals inhibited on frustration tasks -5- also showed inhibitory performance on subsequent measures. From this was inferred a generalized "repressive" effect of frustration. In a broad sense, by assuming a trait theory or theory of generality in re- gard to personality (3f MacKinnon, 19%), one might expect any inhibitory or facilitative effect of frustration to operate consistently upon both the individual’s aggressive and motor performance. Of course, that there should be a direct relation between overt behavior (such as motor performance) and. fantasy has been contested (Lindzey, 1952; Kagan, 1958). Though not in agreement with other findings, Stone (1950), Lesser (1957), and Mussen 8: Naylor (1951;) have presented evidence indicating a positive relationship between individual's overt aggression and aggressive fantasy. In addition to enlarging upon the McDonough design, its replication, at least in part, was a major consideration of the present investigation. We also wanted to understand more of the nature and the stability of the measures used. An extension of the upper range of the age sample was made to ascertain if there was any continuation of develOpmental trends found in her study. Similar to the McDonough research, maturation fac- tors and sex differences were an inherent part of this study. Because of the complex nature of the design, a number of important aspects of the design are enlarged upon in the discussion of the methods section. In smary, this investigation was an effort to learn if facilita- tive or inhibitory effects of frustration exist with regard to certain responses of middle class children. The motor effects were examined with respect to the degree of aggressive fantasy. Variables of age, sex, amount of frustration, temporal sequence of frustration, and proximity of frustration were considered. THEOREI‘ICAL POSITIONS WITH REGARD TO WHICH Frustration terminology Since the term "frustration" has been used to refer to a number of different types of conditionsand events, it is necessary to be aware of the various usages and meanings of the term in psychologr. Levin (1941+) has cautioned that a scientific concept of frustration should not include all the phenanena covered by the popular meaning of the term. Marx (1956) has outlined four different usages of the term. Frustration could be considered as: (1) an independent variable, the complete or partial prevention of attaining a goal; (2) an intervening construct, the hypothetical internal state or condition of the organism that is produced by thwarting or depriving an individual, sanetimes considered irrelevant drive; (3) a dependent variable, which happens as a result of antecedent thwarting conditions, in essence, a response class; and (1+) a phenomenon, a kind of experience of the organism. Besides various usages ascribed to frustration, various authors have attached sanewhat different meanings to the term. In sane in- stances, emphasis has been on the thwarting of needs. Rosensweig (19%) has maintained that frustration occurs when a vital need is obstructed in its satisfaction. Maslow and Mittleman (1951) have characterized frustration as taking place when vital needs are blocked (deprivation), constituting a threat to the personality. Symond's (19146) definition suggested that need satisfaction is either blocked, interferred with, or not available. Sargent (191+8) saw frustration as the thwarting of the individual's dominant motives. The same author distinguished frus- -7- tration fran conflict, stating that the former is an objective, environ- mental thwarting, while conflict is the clash of incompatible motives. The distinction between frustration and "stress” has not been made clear. Frustration shall be considered an objective, environmental blocking of a response sequence, without any particular assumptions con- cerning internal states. Stress usually refers to certain states within the organism. Selye (1956) called particular attention to the biologi- cal roots of stress in the following definition: "Stress is the state manifested by a specific syndrome which consists of all the nonspecifi- cally induced changes within a biologic system.” Shakow _e_t_ a}; (1911-5) have emphasized the internal psychological state during stress, noting: ”Stress consists essentially of a threat to the 'ego status’ cf the Subject occasioned by increasing failure in his ability to achieve goals which he is told the general population fairly easily achieves." . Referring to frustration as a blocking of a reSponse sequence springs from a'behavioristic orientation. Dollard 33 9.1. (1939) defined frustration as "that condition which exists when a goal-response suf- fers interference." Child and Waterhouse (1953) also find frustration as a “blocking of a goal-response." Adelman and Maatsch (1955) refered to a *stimulus ccmplex interrupting a response sequence." Ansel and Ward's (1951;) interpretation was quite close to the ex- tinction phenomenon. They defined frustration as a "state which results fran non-reinforcement of an instrumental reSponse which previously was consistently reinforced." Zander (19%) had attempted a rather compre- hens ive definition, emphasizing both individual needs and behavioral -8- components. He has stated that frustration is "that condition which exists when a response toward a goal believed important and attainable by a given person suffers interference, resulting in a change in beha- vior characteristics for that person and situation.“ Frustration will be considered here as an interruption of a response sequence, operationally defined as the prevention of finishing a task near the point of canpletion. Frustration, then, is used as an indepen- dent variable, with reactions subsequent to frustration as dependent variables. In light of the previous discussion, frustration is viewed as a form of environmental manipulation. The purpose of interrupting an activity near the attainment of a goal (completion of a task) is to enable the subject to beccme sufficiently engaged in the task and its canpletion before the behavior sequence is terminated. Essentially, the procedural definition of frustration in this study is that the individual is prevented fran attaining an expected goal. As will becane evident in the methods section, the goal of task canpletion was accen- tuated by a reward procedure. Predicted reactions to frustration A. Aggression The first praninent theory in regards to reactions to frustration was the well-known frustration-aggression hypothesis of Bollard _e_1_:_ _a_l_ (1939). They maintained that frustration always leads to some form of agg'ession. In view of obvious objections to this position, both N. MiJler (19M) and Sears (19%) conceded that it is possible for frus- tration to lead to forms of behavior other than aggression. Initially, Freud (1920) felt that aggression was the ”primordial reaction” to -9- frustration, but later (1922) asserted that agg'ession was a manifesta- tion of the death instinct. Fenichel (1916) suggested that aggression may be a response to frustration; but can also arise spontaneously. Because of the specificity inherent in the original hypothesis of Dollard _e_t_s_._l_, there were a number of violent disagreements to their position. Lewin (19M) asserted that it "would be scientifically meaningless to make the attempt . . . of linking the intensity of frus- tration lawfully with arm specific effect (such as aggression); for one would have to know the type of frustration and the detailed setting in order to make any definitive derivations.” Ichheiser (1950) stated that the original hypothesis neglects the role of misperception, self- perception and social perception. Symonds (1946) has noted anxiety or emotional tension as a cannon reaction to frustration. B. Performance change Sane authors have suggested that frustration leads to a deteriora- tion of performance. Barker §_t_ _a_l_ (l9hl) emphasized the primitivation of performance as one of the outcomes of frustration. Along these lines, T. M. French (191a) suggested that frustration disorganizes goal-seeking into more elementary drives. In making the distinction between frustrated and motivated behavior, Meier (1949) viewed the former as a non-goal oriented, stereotyped, non-adaptive type of response. H. Schaffer (1951+) suggested that with frustration, a form of stress, behavior was governed primarily by sub-cortical rather than cortical processes. Others take note of the increased vigor or restlessness" in per- formance following frustration. Ansel and Ward (19510 have pointed out -10- that frustration provides drive stimulation. Denny and Adelman (1955) also noted the motivational properties of frustration, stating that frustration (nun-reward) has reinforcing prOperties. A more recent attempt to reconcile possibly two different effects of frustration has been suggested by Lawson and Marx (1958). They maintained that frustration may be accompanied by "l. a momentary in- crease in motivation and/or 2. the occurrence of stimuli for responses which interfere with the response involved in a frustrating situation." They cautioned, however, that there were no ways to predict when one or the other effect will predaninate. An example of the two factor theory was that of Child and Waterhouse (1953). They suggested that there are contrasting effects of interference and of increased drive resulting fran frustration. Sane authors have considered roughly two possibilities among frustration responses, namely, along adjustive and non-adjustive lines. Mowrer (1938), Lazarus, Deese and Osler (1952), and Rosensweig (19%) have labeled frustration reactions as: "habit progression” and "habit regression"; goal directed and less goal directed; and "need- persistive” and "ego-defensive” respectively. 0. Reactions'contingent upon the frustrating situation and/ or individual differences A number of theorists feel that the nature of frustration situa- tions and/or the personality or reaction tendencies of the organism should be of primary importance in understanding reactions to frustra- tion. Pastore (1950) asserted that the reasonableness or unreasonableness of the situation, as perceived by the individual, is an important factor '- ‘Iirhi. . J -11- in the post-frustration behavior. StOpol (19510 suggested that certain types of frustration, distraction and failure, yield independent effects upon subsequent performance. Deese gtflgl (1953) felt that there may be some type of interaction between types of frustration and individual differences. Lazarus, Deese and Osler (1952) noted that not all people react uniformly to a particular situation. Billingslea and Bloom (1950) felt strongly that reactions to frustration can best be studied in real life situations. Sargent (19h8) saw the individual's habit patterns, how he inter- preted and.reacted to the situation, as crucial to his'behavior following frustration. Both Lindsey (1950) and Sargent pointed out that the same objective situation.may not‘be experienced.sub3ectively as equally frustrating‘by all individuals. In extensive reviews of frustration studies, both zander (19M) and Lawson & Marx (1958) conclude that most important to the outcome of frustration are the types of frustration and individual reaction patterns. RESEARCH WITH REGARD TO FRUS‘IRATION In reviewing the literature of frustration studies, it becomes apparent that there has been a wide divergence of procedures designed to elicit frustration. Among the methods employed in such experiments have been manipulations, such as arbitrary failure, impossible time limits or solutions, false reports of failure to the subject, delay or denial of reward, verbal derision, withdrawal frail pleasant surroundings, and many others. Along with the wide variety of procedures, theories have been tested on a variety of p0pulations, with no consistent defi- nition of frustration. It is small wonder, then, that the outccme of these investigations has not led to any consistent or coherent body of knowledge. Though we may not have, at this time, an integrated body of inform- ation, research findings can be categorized according to the following reactions to frustration: A. an increase in the level of performance; B. a decrease in the level of performance; C. a change in the. level of performance contingent upon the nature of frustration and/ or indivi- dual differences. Increase in performance In examining various physiological reactions of stable and unstable children, Jost (19111) found an increase of sympathetic reactions fol- lowing frustration. Though there was greater variability in the reac- tions of unstable children, he concluded that physiological measures were less variable, and therefore, better measures of post-frustration reactions than measures of overt behavior. -13- Thompson and Honnicutt (19%) found that children's achievement improved more with either praise or blame than without any external incentives. They pointed out that blame increased the work of the extraverts more than the introverts. In Hurlock's (1952) praise-re- pmf study with children, he found that reproof had an initially facilitative effect upon performance; but that continued reprimanding led to a deterioration in performance. Using a selective learning technique with rats, Amsel and Ward (195% noted the increased drive properties with the introduction of frustrating stimuli. Crespi's (191m) experiment with varying incentives indicated an increased ef- fect when no reward was present, when compared to a small unit of re- ward. Truax and Martin (1957) found all of their adult sample improved in their performance on a simple task following failure, with high anx- ious subjects doing sanewhat better than low anxious subjects. With a canplex task, there were no definitive performance changes. Decrease in performance In the classical study by Barker, Dembo and Levin (191+l), they found a primitivation" in the level of performance of nursery school children following frustration. Along with regression in the con- structiveness of play, they noted more group expression of aggression among the children. In a replication of the above experiment, Block and Martin (1955) also found post-frustration play to be less construc- tive. They observed that "over-controlled" children showed less decre- ment than "under-controlled" children. With induced failure in a card sorting task, McClelland and Apicella (191+?) found that subjects had more difficulty in subsequently -1h- learning the task than those with prior success. Sears (1937) found that failure on a card sorting task related to a decrement in perform- ance on another task of learning nonsense syllables. Zeller's (1951) work lent support to the theory that initial failure in one task seemed to have a generalized decremental effect upon later associated tasks. He found that prior experience of failure in a nonsense syllable task had a decremental effect on learning another form of nonsense syllables. If the subject had a prior experience of success, however, a subsequent failure experience did not have am effect on later performance. Marquart (19118) , in using arbitrary shock and a no solution procedure in a card selection, found that this type of frustration seemed to interfere with the learning of new responses. Postman and Bruner (19%) frustrated a group by presenting a set of pictures too quickly for adequate recognition. They found on sub- sequent trials, in describing another set of pictures, the frustrated group failed to benefit from practice. On the other hand, the control group evidenced learning with successive frustration trials. In continued repititions of drawing a man, children progressively deteriorated in the quality of work, according to Seashore and Bavelas (1912). Resistance to the task was also evidenced by verbalized re- sentments frail the children. When Barker (191+2) confronted children with unpleasant alternatives, they took longer in making decisions and showed greater alternations between choices than when presented with neutral or pleasant choices . Barker suggested the unpleasant situation gave rise to conflicting tendencies within the children. H. Wright (1918) found that an obstruction to a goal can beccme so great that a -15- sudden cessation of effort and/or a withdrawal from the situation can be observed in young children. Change in performance contingent upon nature of frustration situation and/ or individual differences In a number of experiments, emphasis has been given to the frus- trating situation when variations in the quality, degee or nature of frustration are taken into account. Other studies seemed to under- line various individual differences, such as personality factors, or reaction patterns, as crucial to the outccme of frustration. In a few instances, there has been a canbination of the above variables in the design of the study. with increasingly difficult problems, Cowen (1952) found increasing rigidity among his subjects. Also measuring problem solving, Mohsin (19511) discovered that increasing difficulty of the task seemed to have an inhibitory effect upon performance. He noted that there seemed to be a spread of the effect of frustration from the performance on one task to another. By training feeble-minded subjects with success in- terspersed with failure, Grosslight and Child (191w) found that these subjects persisted more in the face of subsequent failure than those who were trained on success only. Osler (1951+) crashed the effects of two types of frustration, fear and failm‘e, upon performance of high school students. She found that failure tended to depress performance on arithmetic problems, whereas, fear yielded no significant change in performance. She concluded that the effect of frustration on intellectual performance was not only a function of the type of stimulus situation, but the type of control -16- used for canparison. Stapol (19510 found that failure seemed to have an incremental effect upon digit symbol performance, while distraction had a decremental effect. He maintained that tolerance for these two types of frustration were independent. Further, Pastore (1952) found that as modifications of an arbitrary set were made in the direction of non-arbitrariness, there was a marked reduction in the number of aggressive responses. Cohen (1955) also found less aggessive responses were made to non-arbitrary stimuli. According to the Hullian goal gradient hypothesis (1938), responses closest to the goal are more strongly learned than those farther fran the goal. Varying the distance 'frcm the goal at which frustration takes place has been attempted by a number of investigators in support of this hypothesis. Lambelt and Solomon (1952) found that rats failing to receive an expected reinforcement close to the goal resisted extinction longer than those frustrated farther fran the goal. They also noted a greater degee of exc itment among the former group, as evidence of a "frustration-produced drive." Adelman and Rosenbaum (1951+) found that the response of adults frustrated closer to the goal took longer to extin- guish than responses of subjects frustrated farther fran the goal. Children frustrated near the completion of a task showed more intense physical reaction than those frustrated near the beginning of a task, according to Haner and Brown (1955). They concluded that frustratim produced near the goal showed a greater aggressive reaction than when frustration occured farther from the goal. Adelman and Maetsch (1955) attempted various means of extinguishing learning habits in rats. Their results suggested that reactions to extinction were a function -17- of the type of response encouraged during frustration, and the way this reaction interacted with the original learning. Where responses were incompatible with the original learning, the original habit extinguished quickly, suggesting an interference effect of frustration. If the re- sponses were compatible, the original learning extinguished.more slowly, suggesting a facilitative effect of frustration. waterhouse and Child (1953) also concluded that frustration would decrease the quality of the performance to the extent that frustration evoked inter- fering responses. They found that adults reporting a high degree of task interference showed a small decrement in performance under failure when compared to successful conditions. Those subjects reporting low interference showed an increment during failure as compared to success- ful performance conditions. Studies involving high and low anxiety youps have becane somewhat popular in the last decade. Deese 93 a}; (1953) found that following frustration, the non-anxious subjects are more variable and.poorer in their perfOrmance. ,Anxious subjects improved.somewhat in their per- formance. Using sanewhat similar learning conditions as the above study, Lazarus, Deese and Hamilton (1951;) found no difference in the performance of high and low anxious groups,‘both showing a decrement in performance following frustration. Lucus (1952) found that, with increasing degrees of failure, fewer number of errors occured with the low anxiety group, and greater number of errors occured with the high anxious subjects. This author suggested that there is an interaction between anxiety and the amount of failure. Sarason at g (1952) found that two types of threatening instructions (tasksrelevant and non-task- -18- relevant) had Opposite effects on high and low anxiety groups. Adams (19140), using varying degrees of failure on stable and neurotic groups, found no difference between the two groups in efficiency on manual performance or in their susceptibility to varying degrees of frustration. Bennett and J orden (1%) compared the picture-frustration scores of secure and insecure groups, finding the former to be more im- punative, while the latter group was more extrapunative. On the other hand, Rosenswieg and Sarason's (191+2) results suggested that impunative reactions to the Picture-Frustration Test were intimately related to such personality traits as preference for repressive defense measures and high suggestability. Exta-punative reactions appeared to be associ- ated with nonrepressive defense measures and low suggestiblity. Zander (19%) found that children who were rated inferior on per- sonality measures displayed regressive and] or inattentive behavior following frustration, whereas, those children with superior ratings showed more aggressive and/ or attentive reactions. He found that considerable individual variation in post-frustration behavior was evident. He suggested that these differences in reaction my be the result of individualized habits of meeting frustration and the degrees to which the personality of each subject was threatened by the same situation. The author also found that boys showed more non-adjustive forms of behavior, while girls showed more c00peration following frus- tration. Marquis (l9h3) also noted the wide variations in individual reactions to a’frustration among infants. Lazarus and Eriksen (1952) found an increase in variability of performance in a high school group under failure when cmpared to a youp under success conditions. They -19- also found that those students with a high grade point average improved under failure conditions, while students with a lower average did more PM”. WIVEREPONSESASTEEYRELATETOCW Probably one of the most basic concerns of a culture is how to deal with and modify the aggressive responses of its offspring. 0n the other hand, a clear understanding of how and when in the child's de- ve10pment aggressive reactions cane under some form of control is n0t in the offing. Research has suggested a few sign posts, but much re- mains in the realm of theory. Researchers have labeled as "aggession" any one of a number of reactions ranging fran intense physical violence to fantasy. As enlarged upon in the following section, aggession shall be defined here as "an expression or action whose goal is injury to an organism.” Cultural considerations In the acculturation process, Child (19510 maintains that there are twokinds of effects in the socialization of aggressive behavior: those that tend to strengthen the tendency to be aggressive, and those that lead to control or inhibition of this tendency. Such facilitation or inhibition of aggressive tendencies have been related to class status and cultural variations. Davis (1952) noted that middleclass adolescents were punished for physical aggression, while lower class children were frequently rewarded in this area. The same author (19111) noted that the middle class, in general, most strongly controls areas of behavior such as aggression. In a cross cultural study, whiting and Child (1953) concluded that the severity of inhibiting socialization of aggession in any particular culture was related to anxiety about aggression in that culture. With a can- parison of middle class and lower class nursery school groups, Brody (1955) observed more verbal responses and sane forms of verbal aggres- -21.. sion in the former group. On the other hand, in another study by Sears, MacCoby and Levin (1957), there were no differences in aggression as related to socioeconanic status. Parental punishment Sane experimenters have assumed that punativeness of the parent was a form of frustration to the child; and, in line with the frustration- awession hypothesis, would precipitate some fonn of aggession on the part of the child. Research, however, attempting to relate punativeness or restrictiveness of the parent to aggression in their respective off- spring has not led to any consistent or linear type of relationship (Levin and Sears, 1956; Sears, MacCoby, a Levin, 1957; Hollenberg a Sperry, 1951; Sears, Whiting, Nowlis, a Sears, 1953). Page (191+l) offered an interesting alternative to the above hypothesis, concluding, "The occurence of aggression in childen apparently may be said to pre- suppose laxity or absence of control over the child more than . . . the presence of frustrating conditions . . ." DevelOpmental . There is also sane meagre information about the course and modifi- cations of aggressive responses as the child becanes older. Waters 21.; a_l' (1957) found that overt aggressive responses seem to increase fran two to four years of age (33 Sears, 1951). Halpern's (1953) analysis of children's Rorschach responses suggested a prevalence of aggressive responses, without anxiety indices, at about six years; relatively mature records, with less manifestation of aggression, at about eight years; and tendencies towards exaggerated control (inhibition), at about ten years. Ames _e_t_ gr (1952) described a somewhat different pat- -22- tern of Rorschach responses. They suggested suppression of aggressive tendencies by the seven-year-olds, whereas, the eight- and nine-year- olds revealed considerable aggressivity toward the cards. With children at about ten years and over, Sanford gt 31_ (194-3) came to a conclusion similar to Halpern's interpretation, concerning responses to the TAT. He found that there was considerable concern with guilt and guilt re- duction in the stories of these children. Rosensweig (19L8) found in responses to his Pioture Frustration Test a progressive decrease in extrapunative responses, and an increasing tendency for intr0punative and impunative responses to become evident with increasing age (fran four to ten years). Sex differences Almost without exception, child studies have noted greater overt and fantasy aggression in boys than girls (Levin & Sears, 1956; Sears, Whiting, Nowlis, & Sears, 1953; Sanford 933 211, 1913). Sears, MacCoby, and Levin (1957) found no sex differences in aggressive fantasy. Post-frustration aggressive fantasy In measuring responses to TAT pictures, both Bellak (1919) and Shakow gt _a_l_ (1916) found an increase in aggressive themes following frustration. Grandall (1951) found no difference between frustrated and nonfrustrated subjects on formal aspects of TAT stories, but noted greater punishment expectancy in the frustrated group. Lindzey and Riecken (1951) deinonstrated that, subsequent to social frustration, TAT stories reflected more of an increase in the number of aggressive acts carried out by "heroes" than those carried out by "other" figures. Rodnick and Klebanoff (19%) found that poorly adjusted subjects' TAT stories were shorter and contained more aggressive elements following frustration -23- while the adjusted subjects gave longer stories, yet did not show an increase in aggression following frustration. ‘According to Yarrow (l9h8), children exhibited.an increase in certain ferns of aggression in pro- Jective play after frustration. PRE-TEST PRWEIIIRE Measurement of aggessive fantasy Certain considerations led us to use responses to incanplete stories as a means of measuring aggressive fantasy. Standard projec- tive techniques were not thought suitable for our purposes, as no one of these instruments had a sufficient number of stimuli that tend to elicit aggressin responses. The decision to use incomplete stories was based on Korner's (19119) success in using this method to tap young children’s aggressive fantasy. As previous studies did not contain enough incanplete stories that would meet the present research design, a number of original stories were devised. It was considered important to try out a series of these stories on a small sample of children of a similar pOpulation and age range as those of the investigation. A total of ten incanplete stories was used in the pre-test procedure (Appendix I). Three of these stories (nos. 2, 6 and 9) were taken in modified form fran Korner's (1949) study with 1p, 5- and 6-year-olds. She found that these stories elici- ted the highest number of aggressive responses among her repetoire of stories. Two of the devised stories (nos. 1 and h) were similar to those used in the Andersons' (1951+) research. In devising the stories, the following criteria were used: (1) All stories present children in sanewhat frustrating situations which are likely to provoke full responses of an agyessive nature. (2) Questions regarding the canpletion of the story center upon the charac- ter who experiences sane form of frustration. (3) Story content in- -25- volves not too ordinary, but realistic occurrences to counteract the stereotyping of answers. (h) The content of each story is assumed to have sanewhat the same appeal for both sexes through the age range of 7’00 11. (5) The content is believed to be comprehensible to a large majority, if not all, the children. In addition, two neutral stories, whose content was of a more pos- itive nature, were used as filler items. The purpose of the filler stories was to make it more difficult for the child to 'catch on” to the experimental intent of the stories. . ' ‘ Criteria of aggressive fantasy Massive fantasy, as used here, is a particular verbal reapcnse to theincanplete stories. A response which contains arm expression or action whose goal is injury to an organism is considered aggressive. This general definition is essentially that of Dollard at 9:1. (1939). Such responses involving assault, destruction, anger, resistance, etc., could rightfully be included under such a definition. It is noted that sane authors, as Murray (1938) and Jackson (1951;), have included a. cannon connotation relating to initiative, daninance, assertion under aggression. This meaning of the m is excluded fun the present defi- nition. Interpertation of aggressive fantasy was in terms of its presence and intensity. Direction of aggression or punishment themes with regard to aggression was not considered in this investigation. It was assumed that certain forms of aggression were more intense than others. Overt acts were considered more intense, and naninal expression to passive forms of agg‘ession were evaluated as less intense. According to the -26- following schema, Judges were to tabulate the intensity of aggression in each story response: Tally Nature of aggression 1.5 Active -physical attach intentional destruction 1.0 Nominal expression -expression of anger statements of revenge 0.5 Passive and inferred -non-canpliance accidental destruction This means of tabulation was adapted fran Korner's (19119) ratings of incanplete stories. Admittedly, the method of tallying aggressive responses is arbitrary, but use of a particularza ting procedure was thought essential for consistency. Raters were especially instructed to score manifest content with a. minimum of interpretation. There was no limit placed upon the number of aggressive tallies in each story, provided such tallies related to sanewhat discrete happenings. Selection of experimental stories The sample used for the pre-testing of the incanplete stories can- prised eleven college faculty children (6 girls and 5 boys). An attempt was made to select chfi ren at the oldest and the youngest agerange of the final sample. The order of presentation of the stories was reversed at randan. The 3; related each of the stories to the individual child, and wrote down verbatim his response. The children were instructed to be free in the use of imagination, much as in the standard administration of the TAT and Symonds' (1919) Picture Test. The responses to each story were then coded for agyessive fantasy -27.. by two raters with graduate training in psycholog. Then, the scores of the two raters were averaged. In selecting final stories, the following requirements were considered: the story should (a) elicit a sufficient number of aggressive responses; (b) have a reasonable amount of variability among respondents in aggressive fantasy; and (e) have responses which can be reliably coded with regard to the degree of agyessive fantasy. It was found that the mean ratings of aggressive fantasy for each child’s story ranged fran .18 (story no. 5) to 1.18 (story no. 3), with a mean of .55 for all stories. According to the above criteria, stories nos. 6, 7, 8 and 10 were satisfactory and so, were selected for use in the experiment mm. The mean ratings of these four stories were .lb2. Inter-rater reliability of the scores on the four stories was found to be .87h. The stories selected and a filler story (no. 9a) were used in the main investigation in the following order: nos. 6, 93, 7, 8 and 100 STATEMENT OF INQUIRY In the survey of frustration theory and research, there did not appear to be a substantial or consistent body of knowledge from.which definitive hypotheses could be made in relation to this investigation. Therefore, instead of employing hypotheses, it seemed wise to summar- ize the major areas of concern as points of inquiry. As a minimum, the following points came under consideration: 1: a. ‘With the initial introduction of frustration, is there an in- hibitory1 or facilitative effect upon motor performance (speed and amplitude) as compared with no introduction of frustration? b. Does a series of frustration trials have an inhibitory or facilitative effect upon motor performance when compared with a series of non-frustration trials? 2: Are inhibitory or facilitative effects following frustration most pronounced near the point of frustration? 3: Does the introduction of success after frustration have an inhibi- ting or facilitating effect on motor performance? h: .After a series of trials, does changing from success to frustration have a facilitating or inhibiting effect on motor performance, and how do such effects compare with early trial frustration? 58 ,Are inhibitory or facilitative effects of frustration on amplitude of extrartask reSponses dependent upon the age of the child? Wbition is defined as a decrement from expected or preceding performance, while facilitation is defined as an increment over expec- ted or preceding performance. -29- 6: Do children showing changes in motor performance under frustration, diplay corresponding changes in their aggressive fantasy? 7: Is there greater aggressive fantasy among children subjected to frustration than those subjected to success? 8: a. Is there a difference in boys' and girls' aggressive fantasy following frustration? b. Is there a difference in boys' and girls' aggressive fantasy following success? L'EI'HOD Sample The test sample was made up of 108 grammar school children selec- ted fran eight schools in central Michigan. To obtain a representative sample of middle class children, those children whose parents were in white collar occupations were selected. At randan, seventy-two of these children were chosen as the experimental group and thirty-six as the control group. As will be described later, the former youp were primarily subjected to frustration and the latter group, primari- ly success. Since the study was concerned with reactions to frustra- tion rather than to success, larger sampling was considered among the experimental subjects. The subjects were divided according to the following age levels and conditions: 21-33; :Elperimental- -Control- boys girls boys girls 7%- - 8 12 12 6 6 9 - 9% 12 12 6 6 10%,- - 11 12 12 6 6 The age groups of 7%t08and9to9-é—were the same age groupings used in the McDonough study for the purpose of replication. Further- more, the selection of these groups was based on the fact that prior experimentation showed that children were more consistently cooperative to the procedures used and the reaction measures most stable with this age group. Extension of the age range by including an older group of children was an attempt to trace developmentally performance reactions of the McDonough study, as well as age changes in aggessive fantasy. ‘I‘IIII'I -31- By separating the boys and girls in the sample, the design took into account possible sex differences. Apparatus The same marble-board apparatus used in the McDonough study (adap- ted fran Bauer and Brown, 1955) served as the experimental task for the present investigation. Haner and Brown have pointed out the suit- ability of this type of apparatus in frustration procedures. The task and apparatus were considered a "game," which would engage the interest and cooperation of children. The taskalso provided for a continuous reSponse, with a clear starting point and goal. Further, the task enabled E to thwart _S_ at any point in the task, without allowing _S_ to realize the arbitrary nature of the thwarting. The apparatus is shown in Figures 1 and 2. Canon to many chil- dren's toys, the apparatus was colored in bright yellow and blue, which was intended to increase the interest and game element of the task. The base box was 12:20:25 in front, rising to a height of h-3/h" in the back. The top box was perforated with four rows of six holes, ' slightly larger than %" in diameter. The child was required to place different colored marbles in these holes. Inside the top box was a sliding frame attached to a handle which extended to the rear of the apparatus . When the frame handle was pulled by the g, the marbles in the holes fell into the box. The frame was then returned to its origi- nal position by Q for the next trial. The floor of the marble box was slanted in a manner which caused the marbles to roll toward the right front corner where they entered a tunnel leading to a small tray, just large enough to contain a single marble. Only when §_ picked up a C \ ~. I ('0'!) 8.? E: (U \\\ ’/-‘—“- ‘N V ‘ ‘11}va . _( 1, , u . 914; ll! Jul («33335351381 l..\ \ . - _ M _ m min . . “ vi. .‘ llL ..44 . .\ 0| 1! .1..- ll. \ll‘llli) _. i; illizvl [1:11.88 )Iul («2L1 5|“! . \ [it o 1‘. -3h. marble did another becane available by rolling into the tray. In the center of the base box and L3/h“ fran the marble box was located a wooden plunger, 2-3/h" in diameter. Atop the wooden plunger was 111,—" of foam rubber covered with linen. The foam rubber provided cushioning when striking the plunger, in order to eliminate any noxious effects of contact. Beneath the plunger, there was a steel spring which provided resistance when the plunger was hit, and also recoiled the plunger to the initial position. On a lever principle, depression of the plunger activated a recording device. The recording mechanism consisted of two wooden columns, lit—3;” high, placed on either side of the hole fran which the plunger lever emerged. Resting on the end of the lever was a wooden block, grooved on both sides so that it could ride up and down between the two columns. Each time S hit the plunger, the block was propelled upward by the push of the plunger. It was necessary for the purpose of recording the height of the block that it remain elevated after the S's blow. This was accanplished by creating tension between the two columns and the wooden block by means of a pulley-like arrangement of amread in series with a turnbuckle screw and a light weight spring, as seen in Fig. 2. A shield attached to the appwatus obstructed _S_‘s view of the recording device and the marble-relasing mechanism. Measurement of the force of the blow upon the plunger was read directly fran a scale (measured in 10th of inches) on the face of one of the columns. A marker on the wooden block (riding between the two columns) indicated the force of the blow at the maximum height obtained. This apparatus was designed to measure the force of the §‘s initial im- -35- pact on the plunger rather than prolonged physical pressure, as initial impact was believed to be a better indicator of immediate reaction to the experimental conditions. A Important to the investigation was determining whether the ampli- tude deflections represented a linear measurement. This was attempted by ascertaining if proportional increases in known amounts of force upon the plunger would result in prOportional increases in the marker deflection. This was accanplished by drOpping various standard weights fran a constant height onto the plunger and recording the marker de- flections. Standard weights of .05, .l, .2, and .5 kgm. were dropped five times fran a constant distance above the plunger. The means of the resulting deflections fran these weights was found to be directly preportional to the amount of weight drOpped. It was concluded that the amplitude deflections were essentially a linear measure of the force applied to the plunger. Procedm'e Children were seen individually in a vacant schoolroan, after being told they were going to play a game with E. During the examina- tion, the child sat in front of the apparatus which was placed on a table of suitable height. The E sat behind the appwatus, manipulating the wooden frame, timing the trials and recording the marker elevation. All the children were given the following instruct ions : "Look at this board with the holes in it. The idea is to take one of these marbles (E pointing to marble container), put it in a hole, then take another marble, put it in a. hole. . . until you get the whole board filled with marbles . Every time you get the whole board filled with marbles, you get a little prize that you can take hane with you. Any -36. time you don't get the board filled fast enough, within the time allowed, all the marbles will fall through the holes. Every time the marbles drop through the holes, you hit this (3.3 pointing to the plunger).' g gestured with his fist over the plunger as if to hit and then told g to hit it once before starting, to see how it worked. _S_ then added, "Whenever themarbles drop through the holes, whether you've filled the holes or did not finish filling the holes, you hit the plunger.” The task essentially required the child to fill a 2u-ho1e marble- bosrd as quickly as he could, then hit the plunger. As hitting the plunger was not a part of the response sequence of filling the board, it was considered an extra-task response. Since reward was contingent upon task canpletion, this incentive was used to exacerbate the goal of canpletion. On successful trials, the marbles were released after filling the last hole and a toy was put in an enveloPe, which he “was told he could take hane after school. Toy rewards were similar to those which Bijou (1955) found had a universal appeal: for children. Toy charms, as era- ser figurines, whistles, cars, alligator snappers, were given to boys and girls alike; while rings and fans were given exclusively to girls and army trucks and soldiers were givenexclusiuely to boys. On frustration or unsuccessful trials, the marbles were released before the canpletion of the task. On these trials, the E after the child had filled either the 20th, 21st, 22nd or 23rd hole, said, "Time is up," ostensibly looking at a stop watch upon releasingth e marbles. Calling attention to the time was intended to offset any perception of arbitrariness of failure. -37- Every child was given fifteen trials on the marble board, as well as asked to respond to five incomplete stories selected from.the pre- testing procedure at the end of the eigth trial. The experimental or frustration group was given two initial successful trials, six unsuc- cessful trials, two successful trials, three unsuccessful trials, and two successful trials. The same procedure was followed for the control group, receiving six successful trials instead of six unsuccessful trials. Successful trials were considered those in which the child filled the board and was awarded a prize. Frustration trials were considered those in which the marbles drop before filling the board and no reward was given. The time between trials averaged 30 to 35 secs. Therefore, practice periods were essentialhy distributed. Below is schematized the procedural order of conditions for the experimental and control groups: Trial number Experimental Control I and 2 success success 3 to 8 frustration success - incomplete stories incomplete stories 9 to 11 success success 11 to 13 frustration frustration 1h to 15 success success With the experimental group, the initial success was for purposes of engaging the cOOperation of the Child. In order to ascertain initial, as well as cumulative effects of frustration, six separate failure tri- als were then presented. Incomplete stories (as fantasyindicators) were introduced after these trials because it was assumed.that the most -38. marked effect of frustration would be after a long series of unsuccess- ful trials. The stories were administered in the same manner as des- cribed in the pre-test procedure. ,Following these stories, all the‘gg were given two successful trials. These trial conditions were of use in ascertaining the effects of success that had been preceded by frus- tration. Up to this point, order of conditions and number of success trials were identical to the McDonough study, with the exception of introducing a fantasy'measure. In addition, three frustration and two concluding successful trials were given. These additional frustration trials made comparison.possible'between post-frustration reactions of children who underwent previous frustration and reactions of children who had only prior success. The basic purpose ofthe control youp was for direct canparison of reactions obtained under the experimental procedure. 'With this method, the effects of early and late introduction of frustration could be analyzed. Further, it was noted that McDonough' s examination of reactions to differential trial conditions was restricted to se- quential trial canparbons. The present design afforded a control group means of comparison, with variable conditions of frustration and success. The control group, also, offered an opportunity of studying performance on the apparatus with a series of continually reinforced (nonrfrustration) trials. RESULTS Measures used There are three kinds of measures used in this study: performance speed, amplitude and aggressive fantasy. The mean scores of each mea- sure for the experimental, age and sex groupings are presented in Appen- dix II. Performance speed Performance speed was the only measure of motor performance on the task itself. Performance Speed was defined as the time to canplete a trial or half-trial. A mlete trial consisted of placing 2O marbles in the board; the first half-trial, of placing marbles l to 10; and the second half-trial of placing marbles 11 to 20. As dis- cussed under "Problem," performance Speed was separated into first half- and second-half-trial times since the second half-trial time was considered a more sensitive measure of inhibitory or facilitative ef- fects. The rationale for considering the sensitivity of this measure was: (1) this time represented task performance close to the point of subsequent frustration (learned effects) and, (2) work (fatigue-like )1 effects associated with the second half-trial meant performance was farther fran a ceiling, hence allowing any facilitative effects to be- cme manifest. W refers to the mean performance Speed of trials 1} to 8, the trials which followed the frustration procedure and preceded administration of the incanplete stories. A measure of reliability was obtained by comparing individual per- formance speeds between trials where experimental conditions were 1 pg Inquiry 2 in ”Results" section. 4+0- constant over a series of trials. During and prior to the 7th and 8th trials, conditions were constant for both the experimental and control groups. The correlation between performance speeds on these trials is .792. This result shows a high degree of consistency in the speed measure. The mean performance speeds of ithe experimental and control groups of boys and girls appear in Fig. 3. The 9- and lO%-—year groups were canbined for graphical clarity. Although learning was clearly evident, there seemed to be no appreciable sex differences, both in level of performance and rate of learning the marble-board task. How- ever, as will becane evident later, faster performance speeds were associated with increasing age; Amplitude Amplitude, also a motor response, measured the force of initial impact upon the plunger, subsequent to each trial. As noted in the "Procedure," amplitude was designated as an extra-task response. Cilnmlative E. litude refers to the mean scores of trials 3 to 8. These trials were selected since frustration procedures directly pre- ceded plunger manipulation during this trial sequence and were followed by the incanplete stories. The amplitude measure was taken fran the recorded deflection of the marker attached to the plunger. Due to the construction of the apparatus, it was necessary to make daily correc- tions of individual amplitude scores over the testing period. On each' testing day, a .19 kgn. weight was dropped five times through a 2h"-1ong tube onto the plunger. The mean of the resulting mrker deflections for each day was recorded. (These means varied fran 1.9" to 2.3”, with Frinfii Mean time States of 7f»yr.-cide and canbined 9- and iOpvyr.-tlds for eXprrimeuial All tsutrci urrurs for all trials Experimental PK" ‘ --x Cmil‘tfl Ii?- ----- s 33 f‘ o n-m wro a 0 fl H-u- Powwow» DH. 00mm i ’ l f 8 5‘ _fl”_ mfi~___ ‘5 \f I". 20 > .'< 5" r 1 ' 7f d \ t ! 1 i ‘ 7 5. \ i g ‘ i 3 3 I 1 : ; I 2 't I i " ( g g 3 € 2 -f i ' 9 l . . g ' A' ‘ I L 11 ,_-_~,_ L_..,-_...._--._......ll_-.. P—- ’3 _— Iflh— H 12 13 14 :5 (a: a La.» ~‘ N 5.- (I. .0 L- ( \ Cir- -hg- a grand mean of 2.2" and prObable error of I .Oh8".) The daily mean was then compared to a constant mean deflection of 2.2". The differ- ence between the daily mean and the constant mean deflection was ap- prOpriately added to or subtracted from the amplitude scores Obtained on that particular day. Thus, amplitude scores were corrected daily for variations in the apparatus. A.measure of reliability was Obtained'by comparing individual scores on those trials where conditions were constant over a series of trials. Comparison of scores from the 6th to the 7th and from the 7th to the 8th trials suited this purpose. The correlation between the amplitude scores on the 6th and 7th trials is .782, and between the 7th and 8th trials is .753. This finding indicates a high degree of consistency in the amplitude measure. In Fig. h, appears the mean amplitude of the experimental and control groups of boys and girls during the experimental trials. From the initial trial on, it is evident that the boys hit the plunger harder than girls. This difference appears to be consistent for all age groups1 throughout the performance period. To ascertain the sex differences in the plunger scores, aqt_test was computed on the amp- litude scores of the first trial (Table 1). 1 median test of first trial amplitude scores with respect to age and a similar test with respect to sex show no significant age differen es x? of 2.89, n.s. for df 2) and significant sex differh ences (X‘3 of .118, p 6 .05 for 511; 17. F I (SLR E 4 uh3. Mean amplitude scores of eXperlmental and control groans of QQCflMHU3> MCDHOOU') boys and girls for all trials E - Boys F—~-——w Girls X —~- g C - BOYS 7;. - _. _ _. .__ _45 500 r- Cir153--“—‘ 1‘1 1"] 9'» 4. 5 t I, . , I ‘. x -' w ' 1‘ . Q , N I ' V [Z ' t} I .~ G 1 4o 0 E I ‘ m ‘I . '\‘ 1 [P f ‘K + X i / I) I :L ' I ' ‘ \ l.— / M 3. 5 ! , J, | "I k f r‘ 7 q f a >‘ ‘& "‘ 3 30 O i f "r g X ‘ 1 i 17 i x C'. 't 2 . 5 g - ‘ , )7 g \\ m ’ ~ _ l ( l “\ PK Q 1" ‘sq._ . ‘ / 9.\ ./ 1 . i ? o 0 ; Ox \\ [A / \\ 1'3 ., {3/ / w\- .' \‘ 4 ~~¥ 4 1,} wk. ,. 17“," 1.5 i 1 o O (I , T , b . , .5 j y i O L__i..-- _L .L . L A 1 1 L .~______.__;___'.__H,___ 1 2 3 4 5 6 7 8 9 10 11 12 13 M 15 Trials Table l t test of amplitude scores on trial 1 Group Boys Girls Mean 3.372 2.32:2 Variance 6.13 2 .71 n 51+ 51+ t = 2.5h p (.02 for 2-tail test, df lo6 An _1: of 2.26 (p ( .01 for _d_f_ 53 a. 53) for the variance ratios indicates non-hanogeneity of variance. However, probability values of _t_ are not substantially influenced by variance heterogeneity in the case of the 2-tail test. Hence, the resulting E can be considered significant, in- dicating a reliable difference between boys' and girls' initial ampli- tude scores. Aggressive fantasy Responses to the incanplete stories were rated for aggressive fantasy in the same manner as responses to the pre-test stories. Two judges, with advanced degrees in psychology, independently rated the aggressive fantasy responses to the experimental stories. As the even- tual fantasy measure was to be based on the individual scores totaled over four stories, the inter-rater reliability was computed fran the total of the judges' ratings on these stories. An 5 of .891 indicates that the measure of aggressive fantasy was adequately rated by the judges. The average of the judges' ratings was used as the measure of aggressive fantasy. The mean aggressive scores of the experimental and control groups -h5- at each age level are illustrated in Figs. 5a, b and c. As will became evident later, analysis of the entire sample reveals that age groups differed significanly in their aggressive fantasy. The results, how- ever, are not clear. For, though both experimental and control groups showed an increase in aggressive fantasy fran the 7%- to 9-year levels, the control group seemed to show a continued rise, while the experi- mental group declined in aggressive fantasy fran the 9- to lOfi-year levels. It is reasonable to conclude that there was an increase in aggressive fantasy fran the 7%- to 9-year groups (3 I 3.03, p< .01 for if; 70). On the other hand, any conclusions regarding the differences in agg‘essive fantasy between the two older age levels would seem inappr0priate . -h6- W.I-in 9sta.m m>om. maufiu douucou on a mad 0 «a x ,r ,\ x“. x x .3. .// t I. \ X, X ’1‘ ynllx 93m x... a 2:0 Hmucoeflquxw ADV a manual .-—.—.--.. ”—I--h- - ‘7 m'-“an-. . in” o . N - -- \h .\ \ r . x I _x u , , x z x . \ x i x wllln Houwcoo w ¥-nux Hmucm5«umaxw wmm ow woodman cu“: mouoom ceammoummm can: :3 ”-4 V) ‘1 ’31 u‘ {350“ ‘1' -h7- Inquiries In taking up each inquiry, the results pertaining to motor per- formance were assayed in the following order: first, performance speed; then, amplitude. Inquiry l: _a_._ With initial introduction of frustration, is there an inhibitory or facilitative effect upon motor performance as canpared with no introduction of frustration? Performance speed The initial effect of frustration upon performance speed was ob- tained by canparison of the experimental and control group’s perform- ance speed on trials 1» and 5. As a sensitive measure of performnce speed, the second half-trial time was used. At each age level, the groups were found to have the same means on the previous trial (3rd) with respect to this measure. The mean scores on the LLth and 5th trials for the experimental and control groups according to age levels are presented in Fig. 6. Fig. 6 suggests that the introduction of frustration had a differential effect upon the age levels (facilita- tive for the youngest and inhibitory for the older). Thus, a 2x3 analysis of variance of conditions and age was made to test for inter- actions between age and conditions, as well as for main effects (Table 2). Hartley's test indicates variance hanogeneity with an Emax of 2.79 (N.S. for 93 23 and 6 groups). Significant age difference shows that faster performance was associated with increasing age. Although no significant difference was found between conditions, a significant interaction between age and conditions indicates that the presence or absence of initial frustration did have a differential effect on the various age groups of the sample with respect to speed of performance. fiizELIrb {trim eagle-aw 4 . . . o . ~ . $15. ... [Iv.>bc\ -h8- initial pirinfrU3traiion perftrmauce Speed on «scvnd hoifrltialt of experimental art-j .t‘.t“r‘.1r<'l qualms n39h rpgyurt {C 809 ’ i 3 7"” ‘.r.. 1‘ ‘r‘ Airs.» «f. 1d ,' ', f A 4 I: 8. new". ( v52 ~e ‘\ 'L“ 'e e \ .. ‘\ 4 ‘I '\ \ I 9' " {'1 L+.f-*ridl 1 \ "N 117.0 \\)<,- in ii“ 5.1 u..- ._. 4p. \ou—JF— 10% Age grOUps OOU4Q‘VHM '~ -— Table 2 Analysis of variance of the second half-trial time scores on the lIrth 86 5th trials Source of variation SS df MS F P Between conditions .56 l .56 - Age 579.85 2 289.92 31.04 .01 cm 67.35 2 33.68 3.61 .05 within 952. 87 102 Total 1600.63 107 Amplitude To ascertain the immediate effect of frustration on the amplitude of reSponse, the initial amplitude change (amplitude on trial 2 less amplitude on trial 3) of the experimental group was compared with the control group. Table 3 reveals that a large preportion of the experi- mental group either decreased, or had similar amplitude of response; while the reverse trend was true of the control groups. Table 31 Initial amplitude change frcm trials 2 to 3 Group 7 Increase _ Not increase _ Totals Experimental 1 20 I 52 l 72 Control ‘ 22 l 11+ I 36 Totals 12 66 x2 - 9.88 p < .01 Note - - No change in 8 experimental and 3 control S3. I Since the amplitude scores as a rule did not lend themselves to parametric analysis, it was necessary to use nonparametrics. Through- out this study, parametrics were used whenever possible (_e_.g.- with hanogeneity of variance). -50.. The resulting £2 of 9.88 is significant at the 1% level, indicating that the difference of amplitude of the two groups appeared to be the result of the different experimental conditions. This suggests that the initial introduction of frustration had an inhibitory effect upon the amplitude or vigorousness of response. In smnnary, it was found that the initial introduction of frustra- tion has differential effects upon speed of performance (task response) among different age groups, and had a general inhibitory effect upon the vigor of an extra-task reSponse. -51- Inquiry l: 2 Does a series of frustration trials have an inhibitory or facilitative effect upon motor performance when compared with a series of non-frustration trials? Performance speed I In order to canpare the cumulative speed of the experimental groups with their controls, it was necessary to take into account initial per- formance differences at certain age levels. Both the 9- and lO§-year- old girls of the experimental group were slower than their controls on pre-frustration trial 3. The remainig groups were at equal levels of initial performance. The equating of all groups with their controls was accauplished by anitting two girls of each of the older experimental groups who were the slowest on trial 3, as well as the fastest girl from each of the older control groups, removing six in all fran the sample. Figure 7 shows the mean ctmlulative speed, as corrected above for initial performance, of the experimental and control groups at each age level. Observation of the graph suggests that there was an inhibi- tory effect of frustration on the youngest group and a facilitative effect on the two older groups. In order to obtain a strong test of interaction between age and experimental conditions, the 9- and 10%;- year-old groups were combined in an analysis of variance. Table h rep- resents a 2x2 analysis of variance of cumulative Speed with respect to age and conditions. The variances of the groups are hcmogeneous, using Hartley's test, with a nonsignificant gm of 1.97 (g; 1+3 and 1+ groups). Analysis reveals a highly significant differnce between age groups, but no significant difference between conditions and no significant inter- action between conditions and age. An analysis of ctmnilative speed, using the second half-trials time means of the experimental and control 'E, £r.7.rw‘sfl4 -52- FIGURE 7 Ccrrected cunmlative speed means of experimental and control grOUpS with reapect to age Contr. a :73 I Mean EXPCI'. x time of trials 26 4-8 in sec. ‘;5 d” f 24 . . E) \ \ '\ . \ 23 '*‘ ’X : ”s i :22 ~4- .5 75 9 105 Age Groups -53- groups on the three age groups was also attempted. The results are es- sentially similar to the complete trial time analysis: a highly signi- ficant difference in cumulative speed scores between ages, a nonsigiifi- cant difference between conditions, and nonsignificant interaction be- tween conditions and age. Table h Analysis of variance of cumulative speed, equated for initial performance Source of variation SS df MS F P Between conditions ' 58.88 1 58.88 - Age 9,978.37 1 9,978.37 50.92 .01 CxA 163.65 1 163.65 2.36 n.s. Within 19,205.83 98 195.97 Total 29,706.73 101 Amplitude As described in the section on measurement, the boys' initial trial amplitude scores are significantly Izher than the girls'. The mean dif- ference in amplitude between the sex groups was 1.03. In order to can- bine the female and male groups with respect to amplitude, this mean difference was added to each girl's amplitude scores. Fran these cor- rected scores, the mean amplitude of the control and experimental groups was found for trials 3 to 8 (mmmlative amplitude). The cor- rected scores of these two groups with respect to age are presented in Fig. 8. A 2x3 analysis of variance of these scores was then attemp- ted with regard to conditions and age classification (table 5). «L‘ r'.’ -. . A v . ._ cf)'- 1: v“—" p (P y,,. 1““ , .i J .'(‘. u . ‘ .‘f ' 7 \ . .“F‘?” v\,'. 4",. Ag. . 9.}. — . W-' I -~‘ 5‘ ‘5 it- v. J. \- ~. - _ ""3. " ‘2 (1 x . . ' ' ‘ -_...- K A ,I f a- \ 1"\ i . I . I r . ’ - l ’s ,— k 1' , t3 _ ,- i. ' . l I i . t ' c 4 x . . L. “7“ j_ _ _ -5h- Table 51 Analysis of variance of cumulative amplitude scores, with correction of girls' scores Source of variation SS df MS F P Between conditions 51m.99 l 5M.99 5.1+2 .05 Age 63 .8? 2 31.9h - CxA 6ll.17 2 305.59 3.09., *- n.s. Within 19, 21+7 .21; 102 100.117 Total 11,167.57 107 A Hartley test reveals a nonsignificant Fmax of 2.72 (if; of 23 and 6 groups), indicating hanogeneity of variance. The results of this analy- sis show a significant difference in amplitude between experimental and control groups, but no significant difference with reSpect to age or age-condition interaction. The foregoing indicates the amplitude response was inhibited overa series of frustration trials. Wx groups were canbined to mks use of the largest possible sample. It should be noted that a comparable analysis of sex and con- ditions was not applicable in view of non-hanogeneity of variance (gm of n.06, p (.01 for 9.15. of 17 and h groups). -56- Inquiry 2: Are inhibitory or facilitative effects following frustration most pronounced near the point of frustration? In Fig. 9a are shown the means of the experimental group’s speeds of performance on the first halves and the second halves of trials h to 8, broken down for each age group. Similar scores for the control group appear in Fig. 9b. Trials h to 8 were selected for this compari- son'because they encompass a difference in experimental conditions over a period of time. To test for the presence of intraetrial inhibitory or facilitative effects, a difference score was obtained: for each individual, the mean time on the first halves of trials h to 8 were subtracted.from.the mean time on the second halves. In the experimental group, the mean difference score was 8.80 seconds, and its standard error, .676 sec. The I; of 13.00 is significant at the 1% level (9; 71), indicating that the experimental group was slower on the second halves of trials h to 8. For the controls, the mean difference was 9.06 seconds; its standard error, .8h3 sec. The 2,0f 10.75 is significant at the 1% level (93.35): again indicating slower preformance in the later halves of these trials. While'both slowed in the second halves of trials h to 8, inhibition may have affected the groups differently. For this reason, difference scores of the two samples were compared in ant test. The data are shown in Table 6. The variances are homogeneous. There is no signifi- cant difference in intra-trial inhibition of performance speed between the experimental and control groups. From these analyses, it appears that the inhibitory effects occur- ring near the completion of the task were not influenced by the frustra- tion and reward conditions. Slower speed on the second half of the museum mm< mom a an mom a me a l a 11} a r e _ L ‘ q a q 4 1 q \V \1 If 1r- its LI. t lull _ _ ozoam Monacoo Any asoam Hmucwemuoaxm “av W i o----.; cuisines 0. a ado; pom was 0» voodoo“ can: miv mange» new museum oucmEaOVuoa Ammuwimao: coo: . e _ manual o.oe o.ae o.aa o.ma o.ea o.na own use». Manna imam: -58. Table 6 t_test of halfbtrial time differences of experimental and control groups Group Experimental Control Mean 8.80 9.06 variance 32.7 2S.h n 72 36 t 3 .2h n.s. for df 106 trial appears to be a temporary form.of inhibition, as trial periods were distributed. -59- Inquiry 3: Does the introduction of success after frustration have an inhibiting or facilitating effect on motor performance? As success was introduced after frustration during the 9th trial for the experimental group and during the lhth trial for the control group, an analysis of motor performance was considered during each of these periods. A. Success with extensive history of frustration Performance speed If success following a long history of frustration either inhibi- ted or facilitated performance speed, we'd expect to find differences in going iron trials 9 to 10. In Table 7 appears a tabular canparison of changes in performance speed for the experimental and control groups during these trials. Table 7 Post-success speed changes fran trials 9 to 10 Group Increase 7 Not increase Totals Experimental 39 I 33 . 72 Control 18 l 18 36 Totals 57 51 x2 a .0112 n.s. Note - - No change in h experimental 8c 1 control _S_s. The resulting 32 is not significant. Therefore, when conditions change fran (extensive) frustration to success, performance speed does not ap- pear to be reliably different fran performance Speed under continued success 0 -60- Amplitude Inhibitory or facilitative effects of success upon amplitude would be anticipated by amplitude changes frcm trials 8 to 9. Tabulation of changes in amplitude during this period for the experimental and control groups is presented in Table 8. Table 8 Post-success amplitude changes from trials 8 to 9 Group _ Increase , Not increase _ Totals Experimental l 1+2 I jO 1 72 Control I 16 l 20 | 36 Totals 58 50 X2 I 1.111 n.s. Note - - No change in 8 experimental 8c 5 control S8. The resulting x2 is not significant, indicating that with the introduc- tion of success after (extensive) frustration, there was no reliable change in amplitude scores frcm those given success all along. B. Success with extensive history of success Performance speed If success following frustration and an extended period of success either inhibited or facilitated performance speed, we'd expect to find differences in going frcm trial 11+ to 15 among the controls. To test such effect, difference scores were obtained by subtracting each indi- vidual's performance speed on trial 15 from his speed on trial 1’4. The mean difference score was .823 seconds; its standard error, .367 sec. The 3 of 2.30 is significant at the 5% level (g; 35), indicating perform- ance speed on trial 15 was reliably faster. -61- _Amplitude Changes in amplitude from.trials 13 to lh'would reveal facilita- tion or inhibition effects of success after frustration among the con- trols. Individual changes in amplitude during these trials were re- corded in a sign test analysis. .A.5 of 2.33 is significant at the 5% level, indicating a greater than chance number of subjects increased in amplitude after success. Hence, amplitude scores were reliably facilitated. In summary, the introduction of success subsequent to frustration had no demonstrable effect on.motor performance after a considerable degree of frustration, but after a long history of success (and a limited.amount of frustration), had a facilitative effect on motor performance. -62- Inquiry h: After a series of trials, does changing from success to frustration have a facilitating or inhibiting effect on motor performance and how do such effects canpare with early trial frustration? Both the experimental and control groups had changed frcm success to frustration conditions between the 10th and 11th trials, with the former group having been subjected to prior furstration and the latter having no history of experimental frustration. Performance speed A. Late trial frustration Any inhibitory or facilitative effects of frustration upon per- formance speed would be expected in speed changes subsequent to the 11th trial. To measure individual changes in performance speed from success to frustration conditions, the mean timel on trials 12 and 13 were subtracted fran the mean of trials 10 and 11. A positive differ- ence was interpreted as a facilitative effect on performance Speed. For the experimental group, the mean time difference (trials 11 and 12 less 10 and 11) was I .053 seconds; its standard error, .379 sec. The resulting t of .139 is not significant (5111-71). Thus, in changing from success to frustration among children with a prior his- tory of frustration, there was no demmatrable CW in performance speed. For the control group, the mean time difference was I 1.708 seconds: its standard error, .516 sec. The resultant j_t_ of 3.31 is highly significant at the 1% level (9; 35). Since there is a reliable improvement in Speed scores, the introduction of failure after a. I Throughout this inquiry, time chosen was that of the second half-trial in view of the sensitivity of the measure. -63- sizeable history of success had a facilitative effect upon performance speed. However, when frustration was re-introduced after a series of frustration trials, and a.minimum of success, there seemed to be no noticeable effect on performance speed. B. Early and late trial frustration The results of Inquiry la imply that early trial frustration had an immediate facilitative effect upon performance Speed of the 7%-year-olds in contrast with an inhibiting effect on the older group. In this con- nection, concern arose whether the initial effects of early trial frustration were Similar to late trial frustration for these age groupings. For example, among the 7%-year-olds, is the improvement in performance speed noted in early trial frustration the same magni- tude as for late trial frustration, independent of the history of frustration? .Any comparison of early and later frustration effects in- volves comparing performance Speed during early practice periods and late practice periods where different learning effects may be present. In order to cancel out differential learning effects, performance Just prbr to frustration was compared with performance Just after frustra- tion. Thus, it was possible to determine approximately the degree to which.performance varied with frustration without contamination from learning effects. Such perb rmance changes found in early trial frus- tration were then compared with those found in later trial frustration. The change in performance in the early trials was obtained by Sub- tracting the average performance of the two trials immediately following frustration (trials 3 to h and h to 5) from the performance on the pre- frustration trial (2 to 3). Change in performance during the late -61;- trials was obtained by using similar canputations in comparing post- frustration period (trials 11 to 12 and 12 to 13) with the pre-frustra- tion period (trial 10 to 11). A positive score implied inhibition of performance; a negative score was indicative of a. facilitative change (acceleration). Admittedly, this method might not be the most direct means of comparing early and late trial frustration effects 3 however, it appeared to be the best means available with the present design. With the 7%"3’881-0163, a t test was made between the mean changes in early trial period and changes found in the late trial period (Table 9). A _t of 1.16 suggests that a facilitative effect with late furstra- tion trials is not significantly greater than with early trial frustra- tion among Tit-year-olds. Table 9 t test between early and late trial mean changes in learning rate of 7%yr.-olds with initial frustration Group Experimental Control Early trial change Late trial change Mean -.h33 -l'.667 Variance 5.239 7 10.928 n 2h 12 t I 1.16 n.s. for if 31+ Similarly, with the canbined 9- and lOé-year-old groups, early and late trial changes were canpared in a _t_ test (Table 10). . . .- l. o . I" .3 . s. 3! 1.: . . . . I . "n.\_ ‘ ‘I . n ‘ n -_ . u . ........ ............ . ...... .....vv .- .. 0 le~ .h ' 0 do , -- - ..v-*~- ~- —~ .........-.. . .... ... ..... . .... -. 7‘ H V . . Table 10 t test between early and late trial mean changes in learning rate of 9- and lO%-yr.-olds with initial frustration Group Experimental Control Early trial change Late trial change Mean I .363 - .533 Variance 2.272 2.60’+ n 118 21; t 3 2.27 p ( .05 for if 70 The _t_ is significant, indicating a reliable improvement in performance frcm early to late trial frustration periods among the older children. Thus, when the older children had a more extensive history of success, frustration resulted in facilitation of performance Speed, as was evi- dent in the 73-year-old group, both for early and late frustration. Amplitude Facilitative or inhibitory effects of later period frustration upon amplitude would be evident in changes in amplitude scores frcm the 10th to 11th trials. To test the frustration effects in the experimental group, the changes in amplitude scores between these trials were tabulated. A Sign test reveals a reliable decrease in amplitude scores between these trials (_z_ 8 2.95, p (.01). The same procedure was followed for the controls. There is also a reliable decrease in the amplitude scores of the control group which was also frustrated at this point (2.. " 250’ p (.02). -66- Although both groups Show a decrement in amplitude, frustration may have affected the groups differently. Therefore, the change in amplitude of the experimental group was compared with the change of the controls in a g2 analysis (Table 11). Memll Amplitude change from trials 10 to 11 Group Increase Not increase Totals Experimental 18 gfik 72 Control 7 22, 36 Totals 25 83 xfiyates I .021 n.s. Note - - No change in 10 experimental and 2 control §5' The resulting analysis indicates inhibitory effects upon amplitude are not significantly different between the experimental and control groups. The foregoing computations reveal inhibitory effects on the ampli- tude of the extrartask response when conditions changed from.success to frustration; however, the inhibitory effects appear to be relatively in- dependent of prior history of frustration or success. -67- Inquiry 5: Are inhibitory or facilitative effects of frustration on the amplitude of extra-task responses dependent upon the age of the child? .As the MCDonough investigation suggested inhibitory effects of frustration on amplitude increased from.the 7%- to 9-year-old groups, this inquiry is directed at examination of the amplitude reaction among the same age children and older. Similar to the procedure of the Mo- Donough study, sequential trial comparisons of pre- and post-frustra- tion behavior were considered. Initial frustration effect To test for possible age variations with initial introduction of frustration, amplitude scores on trial 2 were compared with amplitude scores on trial 3, for each age group (7%, 9, and 10%) of the experi- mental sample. The Sg'who increased and those who decreased or had the same amplitude scores during these trials were tabulated and broken down into age groupings for a.§? analysis. The data appear in Table 12. Table 12 Initial amplitude change from.trials 2 to 3 according to age groupings Age A: 9 10% . Totals Increase 5 8 I 7 J 20 Not increase 419 16 117 J 52 Totals ah ah at X? 3 .h3 n.s. (df 2) There is no significant relationship between age groups and (facilitas -68- tive or inhibitory) changes in amplitude with initial frustration. Cumulative frustration effects To test age as a variable under a series of frustation trials, the mean of the §s' amplitude scores on trhl s l and 2 was compared with the mean scores over trials 3 to 8 for each age grouping. Those §s who in- creased.and those who decreased or had the same mean amplitude scores during the trials in question were tabulated at each age level in a 2&2 analysis. The data are presented in Table 13. Table 13 Cumulative amplitude change fran trials 1-2 to 3-8 according to age groupings Age 7% J 10% Totals Increase j 12 8 . 29 Not increase 15 12 16 1&3 Totals 21+ 21+ 2h 32 = 1.50 n.s. (or 2) f“. There is no significant difference between the age levels regarding (facilitative or inhibitory) changes in amplitude with cumulative frus- tration. Fran these analyses, it appears that facilitative or inhibitory effects of frustration, initial or cumulative, on amplitude were not influenced by the age variables under consideration. E‘rrIrfwflomemgslx .. fluba..u.lxl..., 1H]. . .. . . . em. \ 1% w . a -69- Inquiry 6: Do children showing changes in motor performance under frus- tration, display corresponding changes in their aggressive fantasy? The first step in analyzing this inquiry was a comparison of the degree of aggressive fantasy with performance measures of amplitude, and speed among the experimental _S_s M frustration trials. The clmlula- tive performance Speed of each individual was correlated with aggressive fantasy scores, resulting in an 3 of .01. Using the cumulative ampli- tude scores, the correlation of these measures is .01. Thus, there ap- pears to be little relationship between aggressive fantasy and inhibiting or facilitating effects on motor performance during frustration. Aggessive fantasy was also related to inter-trial amplitude changes with frustration. Amplitude change scores were obtained by subtracting the mean of the individual's cmmdative ampltiude score (trials 3 to 8) from the mean of pre-frusta tion trials 1 and 2: the higher the score, the geater the inhibition. These scores were then correlated with aggressive fantasy. An 5 of -.221 (n.s. for d_f_’ 70) was obtained, suggesting only a mild negative relation between indivi- dual inhibition in amplitude and intasity of aggressive fantasy. To further explore the relationship between aggressive fantasy and motor performance, a comparison of these variable during success conditions was attempted. First, the control group's motor perform- ance was computed for successful trials 3 to 8, the resulting scores essentially measuring cumulative Speed and cumulative amplitude. These scores were then correlated with aggressive fantasy. An _r_ of -.208 (n.s. for _d_f 3%) between cumulative speed and aggressive fantasy was obtained ; and between cumulative amplitude and aggressive fantasy, an r of .213. Though not significant, there is a slight tendency for -70- faster perfOrmance speed and greater amplitude to be associated with greater aggressive fantasy among successful children. In conclusion, no demonstrable relationship was found between in- hibition or facilitation of motor performance of chidren subjected to frustration and their aggressive fantasy under the present experimental procedures. -71- Inquiry 7: Is there greater aggressive fantasy among children subjec- ted to frustration than those subjected to success? A canparison of the mean aggression scores of the experimental and control groups (Fig. 5a) reveals a somewhat complex picture. There does appear to be an interaction effect between conditions and age (younger experimental children showing greater aggressive fantasy, whereas, the oldest group showing less than their controls). However, the data were not suitable for direct analysis of such interaction. 0n the other hand, a 2x2x3 analysis of variance with conditions-sex- age classification was possible (Table 1h). Table 11% Analysis of variance of aggression scores with reSpect to conditions, sex & age classification Source of variation SS df MS F P Between conditions 1.90 l 1.90 1.21 n.s. Sex 3.08 1 3.08 1.97 n. s. Age 111.51 2 7.26 11.63 .05 5x0 .oh 1 .011 .. M 3.52 2 1.76 1.12 n.s. AxS 1.02 2 .51 - AxSxC .52 h .13 - Within 1117.30 91 i 1.567 Total . 171.89 107 variances are homogeneous (n.s. Emax of 8.3h,‘d£.1l, with 12 groups). .As the main conditions effect is not significant, aggressive fantasy between the experimental and control groups did not differ reliably. ”its? »..._.~.,.. I! \ (tortiwuw‘ -72- Interaction between conditions and age is also not significant. There- fore, we could not conclude the presence or absence of frustration had a differential effect on the various age groups with respect to aggres- sive fantasy. In general, our results did not Show substantially greater aggres- sive fantasy in children having frustration from.those having success. None the less, differences in aggressive fantasy between various age groups were evident. -73- Inquiry 8: aIs there a difference in boys' and girls‘ aggressive fan- - tasy following frustration? Observation of the mean aggression scores of the experimental groups of boys and girls with respect to age (Fig. 5b) suggests a greater degree of aggressive fantasy among the boys. A 2x3 analysis of variance (n.s. F of 3.91;, if; 11 with 6 groups) of aggression with —max respect to age and sex classification was canputed to test sex differ- ences (Table 15). Table .15 Analysis of variance of eXperimental group's aggression scores with respect to sex 8c age Source of variation SS df MS F P Between sex 1.76 1 1.76 1.09 n.s. Age 12.00 2 6.00 3.71 .05 SxA 1.39 2 .70 - Within 106.78 66 1. 62 Total 121.93 71 No significant difference between male and female children was found, yet aggression scores seemed to vary significantly between age groups. Although boys tended to Show more aggression than girls, there were no conclusive sex differences in aggressive fantasy among children under frustration chnditions. 4+.- Inquiry 8: _b_ Is there a difference in boys' and girls' aggressive fan- tasy following success? Observation of the mean aggreSsion scores of the control groups of boys and girls with respect to age (Fig. 5c) again intimates greater aggressive fantasy in boys. A 2x3 analysis of variance of the control group's aggression scores with respect to sex and age was attempted (Table 16). Table 16 Analysis of variance of control group's aggression scores with respect to sex 8: age Source of variation SS df MS F P Between sex 1.36 1 1.36 1.01 n.s. Age 6.03 2 3.02 2.21; n.s. SxA .15 2 .08 - Within 1+0 Q2 30 1 . 35 Total 118.06 35 Variances are homogeneous (n.s. Emax of 8.23, if; 5, with6groups). No significant difference was found between sexes, indicating no definitive sex differences in aggressive fantasy among children having Just success conditions. Further, as was found in the aforementioned 2x2x3 analysis of variance, no reliable sex differences in aggressive fantasy exists over the entire sample (Table 11;). DISCUSSION The present investigation was designed to examine response' pat- terns of’middle class children subjected to task frustration. Specific frustration effects were evident in consistent inhbition of the ampli- tude of the extrartask response and in either increased or decreased performance speed, depending upon the locus and sequence of frustration and the age of the child. Little or no frustration effect was evident in aggressive fantasy. Changes in the experimental conditions (frustra- tion and success) and the sex of the child were also related to the amplitude of the extra-task response. An important finding of this study seemed to be that the effect of frustration on performance Speed was dependent upon whether frustae tion was introduced early or late in training (i.e. after two successes or after ten successes). There was a clear-cut facilitative effect following frustration after ten success, including the older children (9 and 10%~years), who, if anything, were inhibited after only two suc- cesses. Apparently, it is necessary to establish habit patterns to the task or expectations of success (via continued reinforcement) before a change to frustration conditions yields a facilitating effect on task performance. That under late trial frustration an incremental effect was found on task performance appears to be in accord with research sug- gesting energizing effects of frustration (Mer, 1956; Spence, 1956; Amsel, 1958). On the other hand, as will be pointed out later, the assumption of a generalized drive effect of frustration, as energizing all response classes (Spence, 1956), was not supported by the data. -76- The importance of expectancy patterns was also evident in.Abe1's (1936) study. In an investigation of sensori-motor learning of children, she found that the absence or presence of rewards, from.the initial trial on, had no differential effect on the rate of learning. However, when a new incentive was introduced after a series of learning trials, bet- ter learning was evident. In comparison.with present findings, it ap- pears then, that whenever a new condition (reward in the Abel study, frustration in the present study) is introduced for the first time 232?. 2 E33123. 212 trials, a facilitative effect is present in children's perb rmance. From.the above findings and interpretations, the effec- tiveness of late trial frustration appears to have important implica- tions for procedures in future researth. .A provocative finding was that frustration had an inhibiting ef- fect on the amplitude of the extra-task response. The same findings were Obtained in the forerunner of the present study by MbDonough (1958), using similar procedures, the same apparatus, and children the same age and younger. Thus, the effect appears stable. This inhibitory ef- fect led to the examination of the nature of the extraetask response. Temporally, the extra-task response is not within, but outside, the start and end phases of the instrumental response sequence of placing the marbles. Further, this response is only minimally related to such instrumental behavior via instructions. Therefore, one might conclude that the energizing effects of frustration occur only to the instrumen- tal response sequence. .Although frustration had an energizing effect on the extra-task response in the Haner and Brown (1955) study, it is possible to infer that their plunger response was more integral to the -77- instrumental chain or, at least, had a more instrumental quality. In their study, pushing the plunger turned off a (possibly noxious)‘buz- set, which had signalled the dropping of the marbles. Stating the above more simply, one could say that in the present study, the"ufineces- sary" plunger response is partially'by-passed, after failure on the pre- ceding trial, in the child's eagerness to fill the board on the next chance. This particular finding and interpretation are clearly contra- dictory to Spence's (1956) position in maintaining that all response classes are activated by frustration. To check on the above, one could use the following research design: Instead of having the plunger re- sponse outside the instrumental response sequence, embed it directly into the reSponse chain, by having § hit the plunger each time a row of six marbles is completed. Under these conditions, it is likely that amplitude would.be facilitated by frustration. The amplitude measure also showed that boys hit the plunger harder than girls for all age groups studied. .Although methany (l9h1) and.Meredith (1935), who measured manual strength directly, have poin- ted out greater physical strength in.boys, this would appear to be too Simple an explanation. An indication that amplitude is not a direct measure of physical strength is that, with increasing age, there was no noticeable increase in amplitude. Older children, due to maturation, are stronger (Methany, 19111; Jokl & Cluver, 19111) and could hit harder if strength were an important variable. The fact that physical expres- sion is probably discouraged for girls more aptly accounts for the less vigorous amplitude reSponse among the girls. It was found that the initial effects of frustration facilitated -78- performance Speed in younger children (7%; years) and inhibited perform- ance speed of the older children (9 and 10%- years); this differential effect was not as apparent after a series of frustration trials. Here an adequate explanation is not obvious. The older children may be more aware of the ramifications of not succeeding than younger children are and, thereby, may have become discouraged by their performance. Some evidence that failure may be more meaningful for the older child was revealed in Sanford's (19116) research on the recall of completed and uncompleted tasks. He found that younger children (under 10) tended to remember successful tats, whereas there was an increasing tendency for older children to remember failures. Further, the effects found in the present study were most noticeable immediately after the introduction of frustration, reflecting the momentary effects of frustration (Lawson & Marx, 1958). It appears that some form of adaptations takes place after a series of frustration trials. If inhibitory or facilitative effects of frustration upon motor reSponseS are related to the intensity of aggressive fantasy, it was not evident with the procedures used in this study. It may be that frustration procedures were too mild, or as discussed earlier, were introduced too early in the task experience to have any noticeable ef- fect upon aggressive fantasy. Since it was observed in the success group that the speed of performance and vigor of the amplitude response were minimally and positively associated with the intensity of aggres- sive fantasy, one might explore further how individual response patterns or motivational states may influence the above response measures. Although aggressive fantasy was to some extent more pronounced among -79- children under frustration conditions than among those under success conditions as found by Bellak (1951+), Shakow 33 _a_l_ (19115), Yarrow (191+8), failure to find sizeable differences between these groups may also be related to the experimental procedure or possibly to the nature of the measure itself. Although an increase in aggressive fantasy was found from the 7% to 9 year groups, the implications of this result are not clear. The finding seems to agree with.AmeS gt El (1952), in their account of in- creasing aggression to Rorschach stimuli from the 7 to 9 year level. On the other hand, Helpern (1953) noted the Opposite pattern in study- ing Rorschach reSponses and Rosensweig (1988) found a decrease in ex- trapunative responses during these ages.’ It seems feasible that the above findings are Obscured by different methods of measuring aggressive fantasy. It also is possible that different forms of aggression, as suggested in Rosensweig's research, do not show similar changes with increasing age. SUMI-EARY The present study was an attempt to determine certain response effects of frustration in children, namely, whether frustration had a facilitative or inhibitory effect upon motor performance and if any change in these reactions related to aggressive fantasy. Temporal se- quence and locus of frustration, proximity of frustration, and the na- ture of the response class affected were important areas of investiga- tion. The response effects were studied in 108 middle class children. The children were divided evenly according to sex and into three age groupings (7%- to 8, 9 to 951,-, and 10;: to 11). An experimental group of seventy-two children was subjected to early trial and extensive frustra- tion conditions, and a control group of thirty-six children were sub- jected to late trial frustration conditions. Intentional (though non- arbitrary) interruption on a marble-board task constituted a frustra- tion trial. Success conditions involved allowing tad: canpletion and the giving of rewards. The experimbntal group was subjected to the fol- lowing trial conditions: two success, six frustration, two success, three frustration, and two success. The control group was similar to the above, receiving six success trials instead of six frustration trials. After the Six trials in question, both groups were asked to reSpond to a series of incomplete stories, intended to elicit aggressive fantasy. The response variables under consideration were Speed of motor performance on the task, force of a gross motor reSponse imnediately following the task, and, degree of aggressive fantasy following the series of frustra- -81.. tion trials. A.major finding was that, with late-trial frustration, and contin- ued prior success, there was a clear-cut facilitative effect on per- formance speed, which was not uniformly noted with early trial frustra- tion. This result was interpreted in line with reinforcement and ex- pectancy principles. Another major result was that frustration effects were evident in consistent inhibition of the extra-task response. The inhibition of this response, together with facilitation of performance Speed, indicate that specific effects of frustration depend on the re- sponse class being measured. It was suggested that, at leaSt, certain examples of frustration may activate relevant, instrumental responses and at the same time depress irrelevant, non-instrumental responses. To ascertain the effects of frustration close to the task goal, proximity of frustration,‘both early and late trial speeds were come pared. .A slowing up of motor performance was found near the completion of the task. This effect, however, was independent of frustration con- ditions and appeared to reflect fatigue effects inherent in the task. The effects of frustration upon motor performance were unrelated to aggressive fantasy under the present experimental procedures. WINES Abel, L. B. The effects of shift in motivation upon learning of a sensori-motor task. Arch. Psychol., N.Y., 1936, 29, No. 205. Adams, C. R. Individlfi. differences in behavior resulting from experi- mentally induced frustration. J. P_sycho1., l9’+0, 10, 157-176. Adelman, H. 8: Maatsch, J. Resistance to extinction as a function of the type of reSponse elicited by frustration. J. exp. Psychol., 1955: 50: 61‘65- Adelman, H. & Rosenbaum, G. Extinction of instrumental behavior as a function of frustration at various distances from the goal. J. exp. Psychol., 1951!», ’47, #294132. Ames, L. B., Learned, J., Metraux, R., 8: Walker, R. Child Rorschach responses: Develgpmental trends fran 2 to 10 years. New York: P. G. Hoeber, 1952. Amsel, A. The role of frustrative non-reward in noncontinuous reward situations. Psychol. Bull., 1958, 55, 102-119. Amsel, A. 8: Ward, J. S. Motivational preperties of frustration: II Frustration drive stimulus and frustration reduction in selective learning. J. cg. Psychol., 19515 113, 37-h7. Anderson, H. 8. 3: Anderson, Gladys L. Children's perception of social conflict situations: A study of adolescent children in Germany. Amer. J. OrthOpsychiat., 1953+, 2h, 216-257. Barker, R. G. An experimental study ofthe resolution of conflict by children. In McNemar, Q. 8: Merrill, M.A. Studies in personality. New York: McGraw-Hill, 1982. Barker, R. G., Dembo, Tamara, 8t Lewin, K. Frustration and regression: An experiment with young children. Univer. Ia. Stud. Child We1f., 19171, 18’ NO. 1. Bellak, L. An experimental investigation of projection. _P_e_:ychol. Bull. , 1942: 39, M39-490- Bellak, L. The Thematic Amerception Test andthef Children' s Appfircep- tion Test in clinical use. New York: Grune & Stratton, 195 . Bennett, C. M. 8: Jordan, T. E. Security-insecurity and the direction of aggressive reSponses to frustration. J. clin. Psychol., 1958, 115166-157. . 1‘.» ..L'. a “-45.4.IItIII-5'AII’kwo‘Jiug \;||\'n' -83- Bijou, S. .A systematic approach to an experimental analysis of young children. Child Develpm., 1955, 26, 161-168. Billingslea, F. Y. 8: Bloom, H. The canparative effect of frustration and success on goal-directed behavior in the classroom. J. abnorm. soc. Psychol., 1950, 1+5, 510-515. Block, Jean 8: Martin, B. Predidion of behavior of children under frus- tration. J. abnorm. soc. Psychol., 1955, 51, 281-285. Brody, Margaret. Patterns of aggression in nursery school. Child Develpm., 1955. 26, 3-11. Child, I. L. Socialization. In G. Lindzey (Ed.), Handbook of Social payphology, Vol. II. Cambridge, Mass.:Addison-Wesley, 195h, pp. 655-692. Child, I. L. 8c Waterhouse, I. K. Frustration and the quality of perform- ance: II A theoretical statement. Psychol. Rev., 1953, 60, 127-139. Cohen, A.H. Social norms, arbitrariness and status of the agent of frus- tration in the frustration-aggression hypothesis. J. abnorm. soc. Psychol., 1955, 51, 222-226. Cowen, E. L. The influence of varying degrees of pyschological stress on problem-solving efficiency. J. abnorm. soc. Psychol., 1952, 1+7, 512‘1590 Crandall, V. J. Induced frustration and punishment-reward expectancy in Thematic Apperception Stories. J. consult. Psychol., 1951, 15, too-1101+. Crespi, L. P. Amount of reinforcement and level of performance. Psychol. Rev., 19th, 51. 3h1-357. Davis, A. American status system and the socialization of the child. Amer. soc. Rev., 19ul, 6, 3u5-35h. Davis, A. Socialization and the adolescent perfonality. In Swanson, Newcomb & Henley (Eds.) Readings in Social Psychology. (2nd ed.) New York: Holt, 1952,. pp. 520-531. Deese, J., Lazarus, R. 8., & Keenan, J. Anxiety, anxiety reduction and stress in learning. J. exp. ngchol., 1953, #6, 55-60. Denny, M. R. a Adelman, H. Elicitation theory: I An analysis of two typical learning situations. Psychol. Rev. , 1955, 62, 290-296. Dollard, J. Doob, L. W., Mile r, N. E., Mowrer, 0. R., 8c Sears, R. R. Frustration and aggression. New Haven:l939. Fenichel, 0. The psychoanalytic theory of neurosis. New York: Norton, 19h5. n 9 . . . c v e D O . a ' c . O o O . n c — . o . v . I H . . a o ._ .in.l .. .-...xlucdafl£v .l... . -8h- French, '1‘. M. An analysis of the goal concept based upon study of reac- tions to frustration. Psychoanal. Rev., 19h1, 28, 61-71. Freud, S. A general introduction to psychoanalygis. New York: Borri 8c Liveright, 1920. Freud, S. Beyond the pleasure principle. London: Inter. Psychoanal. Pr., 1922. Grosslight, J. H. 8: Child, I. L. Persistence of a function of previous experience of failure followed by success. Amer. J. Psychol., 191+7, 70, 378-387. Halpern, Florence. A clinical approach to children's Rorschachs. New York; Grune 8c Stratton, 1953. Haner, C. F. 8: Brown, Patricia A. ’Clarification of the instigation to action concept in the frustration-aggression hypothesis. g. abnorm. soc. Psychol., 1955, 51, zen-206. Hollenberg, Eleanor 8: Sperry, Margaret. Some antecedents of aggression and effects of frustration on doll play. Pers., 1951, l, 32-1L5. Hull, C. L. The goal gradient hypothesis applied to some ”field force" problems in the behavior of young children. Psychol. M, 1938, #5, 271-299. Hurlock, E. An evaluation of certain incentives used in school work. J. educ. Psychol., 1925, 16, 115-159. Ichheiser, G. Frustration and aggression or frustration and defense: .A counter hypothesis. J. gen. Psychol., 1950, #3, 125-2191 Jackson, Lydia. Aggression and its interpretation. London: Methuen 8: COO, 1951+. Jokl, E. 3: Cluver, E. I. Physical fitness. J. Amer. Med. Ass., 19%, 116, 2383-2389. Kagan, J. Socialization of aggression and the perception of parents in fantasy. Child. Developm., 1958, 29, 311-320. Korner, Anneliese F. Sane aspects of hostility of young children. New York: Grune 8c Stratton, 19119. Lambert, W. W. 8: Solunon, R. L. Extinction of a running response as a function of distance of block point frm the goal. J. egg. pysiol. Psychol., 1952, 1&5, 269-279. Lawson, R. 8:. Marx, M. H. Frustration: Theory and experiment. Genet. Psycholo Mansr.. 1958. 57, 393-56“. -35- Lazarus, R. 8., Deese, J. 8: Hamilton, R. Anxiety and stress in learning: The role of intraserial learning. _J. exp. Psychol., 1951+,“ 1+7, 111-11%. Lazarus, R. S., Deese, J. 8: Osler, Sonia J. F. The effect of psycholo- gical stress upon performance. Psgghol. Bull. , 1952, 1+9, 293-317. Lazarus, R. S. 8: Eriksen, C. W. Effects of failuree stress upon skilled performance. J. efl. Psychol., 1952, 1&3, 100-105. Lesser, G. S. The relationship between overt and fantasy aggression as a function of maternal response to aggression. J. abnorm. soc. Psychol., 1957, 55, 218-221. Levin, H. 8: Sears, R. R. Identification with parents as a determinant of doll play aggression. Child Develjm" 1956, 27, 135-153. Lewin, K. Constructs in psycholow and psychological ecolog. Authority and frustration. Univ. Ia. Stud. Child. We1f., 191m, 20, 1-30. Lindzey, G. Frustration tolerance, frustration susceptibility and overt disturbance. Psychiat., 1950, 13, 205-211. Lindzey, G. Thematic Apperception Test: Interpretive assumptions and related empirical evidence. Psychol. Bull. , 1952, 1+9, 1-25. Lindzey, G. 8: Riecken, H. W. Inducing frustration in adult subjects. J. consult. IISChOlc, 1951, 15, 18-23. Lucus, J. D. The interactive effects of anxiety, failure and intraserial duplications. American J. Psychol., 1952, 65, 59-66. MacKinnon, D. W. The structure of personality. In J. McV. Huht. Per- sonality and the behavior disorderg, New York: Ronald Pr. , 1911117 Ch. 1. Maier, N. R. F. Frustration. New York:McGraw-Hi11, 191+9. Marquart, Dorothy I. The pattern of punishment and its relation to ab- normal fixation in adult human subjects. J. gen. Psychol., 19118, 39, 107-1% . Marquis, Dorothy P. A study of frustration in new born infants. ,1. eg. Psychol., 19h3, 32, 123-138. Marx, M. H. Some relations between frustration and drive. In Jones, M. (Ed. ), Nebraska smosimn on motivation. Lincoln: Univer. Nebr. Pr. , 1956. Maslow, A. H. 8: Mittlemann, B. Principles of abnormal psycholog. (2nd ed.) New York: Help er, 1951. McClelland, D. C. 8s Apicella, F. S. Reminiscense following experimen- tally induced failure. J. e333. Psychol., 1917, 37, 159-169. -86.. McDonough, Leah B. A developmental study of motivation and. reactions to frustration. Unpublished doctoral dissertation, Michigan State Univer., 1958. Metheny, E. The present status of strength testing for children of ele- mentary school and pre-school age. Res. Quart” 191+1, 12, 115-130. Meredith, H. V. The rhythm of physical growth. Univer. Ia. Child. Helf., 1935: 11: NO- 3- 1vii11er, N. E. The frustration-aggression hypothesis. Psychol. Rev., 19h1. #8. 337-352. Miller, D. R. 8c Swanson, G. E. Study of conflict. In Jones, M. (Ed.) Nebraska Symposium on motivation. Lincoln: Univer. Nebr. Pr. , 1956. Mohsin, S. M. Effect of frustration on problem-solving behavior. ,_T_. abnorm. soc. Psychol., 195%, 1&9, 152-155. Mowrer, O. H. Frustration as an experimental problem: II Sane research implications of the frustration concept as related to social and educational problems. Charact. 8c Pers., 1938, 7, 129-135. Murray, H. A. mlorations in personality. New York: Oxford Pr., 1938. Mussen, P. H. 8: Naylor, H. K. The relationship between overt and fan- tasy aggression. J. abnorm. soc. Psychol., 1951+, 1+9, 235-21+0v Osler, Sonia F. Intellectual performance as a function of two types of psycholOgical stress. J. ea. Psychol., 1951+, #7, 115-121. Page, R. M. Aggression and withdrawal in relation to possible frustra- tion factors in the lives of children. Unpublished doctoral dis- sertation, Northwestern Univer., 191+1. Pastors, N. A neglected factor in the frustration-aggression hypothesis: A comment. J. Psychol., 1950, 29, 271-280. Pastors, N. The role of arbitrariness in the frustration-aggression hypothesis. J. abnorm. soc. Psychol., 1952, 1+7, 728-731. Postman L. 8: Bruner, J. S. Perception under stress. P hol. Rev. , 1938. 55, ssh-323. L— Rodnick, E. H. & Klebanoff, S. G. Projective reactions to induced frus- tration as a measure of social adjustment. Psychol. Bull. , l9lt2, 39. #89. Rosensweig, S. An outline of frustration theory. In J. MCV. Hunt (Ed.) Personality and the behavior disorders. New York: Ronald Pr. , 19M», Ch. 11. -87- Rosensweig, S. The children's form of the Rosensweig Picture-Frustra- tion 81311”. J. szChOIQ, 19,48, 26, lhl-l9lo Rosensweig, S. 8: Sarason, S. An experimental study of the triadic-hypo- thesis: Reactions to frustration, ego defense and hypnotizability. Charact. & Pers., 19h2, 11, 1-19. Sanford, R. N. Age as a factor in recall of irlerrupted tasks. Psychol. Rev., 19h6, 53, 23u-2ho. Sanford, R. N., Adkins, Margaret M., Miller, R. B. 8: Cobb, Elizabeth. Physique, personality and scholarship. Monog'. soc. res. child. mvelEO’ 192+3’ 8’ NO. 1. Sarason, S. B., Mandler, G. 8: Graighill, P. G. The effect of differen- tial instructions on anxiety and learning. J. abnorm. Soc. Psychol., 1952, 1&7. 561-565- Sargent, S. S. Reaction to frustration- A critique and hypothesis. PgChOIO Rev., 1948, 55, 108-llll-o Schaffer, H. R. Behavior under stress: a neurophysiological hypothesis. Psychol. Rev., 195%, 61, 323-333. Sears, R. R. Initiations of the repression sequence by experimental failure. J. 8520 PEiYChOlo, 1937, 20, 570-5800 Sears, R. R. Non-aggressive reactions to frustration. Psychol. Rev., 19111, 1+8, 3h3-3u6. Sears, R. R. Symposium on genetic pachology:III Effects of frustration and anxiety on fantasy aggression. Amer. J. Brtho;gsychiat., 1951, 21’ h98-5050 Sears, R. R. , MacCoby, Eleanor E. 8: Levin, H. Patterns of child rearing. Evanston: Raw, Peterson 8: Co. , 1957. Sears, R. R., Whiting, J. W. M., Nowlis, v. & Sears Pauline, Some child-rearing antecedents of aggression and dependency in young children. Genet. Psychol. Monqgn 191+3, 1+7, l35-23’+. Seashore, H. G. 8: Bavelas, A. A study of frustration in children. ,1. genet. Psychol., l9h2, 61, 279-31h. Selye, H. The stress of life. New York: McGraw-Hill, 1956. Shakow, D., Rodnick, E. H. & Lebeaux, T. A psychological study of a schizophrenic: exemplification of a rushed. J. abnorm. soc. Psychol., 19h5, ho, 15h-17h. Spence, K. W. Behavior theory and conditioning. New Haven: Yale Univer. Pr., 1956. o o . D Q I o o I I a _ O o a C Q A . . a O . o p i . c c . O I . a . nlfilib‘lllll.’ ill; lltnbdiutI-IJI! 0- .unu. .o .0 It’s- . ‘-‘. -88- StOpol, M. S. Rorschach performance in relation to two types of stress. J. consult. Psychol., 1951+, 18, 11-15. Stone, A. A. The effect of sanctioned overt aggression in total instiga- tion to aggression. Unpublished honors thesis, Harvard Univer., 1950. Symondsi,‘6 P. M. The dynamics of human adiustment. New York:Amer. Book Co., 19 . Symonds, P. M. Adolescent fantagy. New York:Columbia Univer. Pr. , 1919. Thompson, G. G. 8: Honnicutt, C. W. The effect of repeated praise or lhme on the work achievement of :htroverts and extraverts. J. educ. Psychol. , 19th. 257-266. Tolman, E. C. mosive behavior in animals and men. New York:Appleton- Century-Crofts, Inc., 1932. Truax, C. B. 8: Martin, B. The immediate and delayed effect of failure as a function of task complexity and personalization of failure. J. abnorm. soc. PsEholq 1957, 55, 16-20. Walters, J., Peace, Doris 8c Dahms, Lucille. Affectional and aggressive behavior ofpreschool children. Child Develpm., 1957,28, 15-26. Waterhouse, I. K. 8: Child, I. L. Frustration and the quality of perform- antezlll An experimental study. J. Pers., 1953, 21, 298-311. Whiting, J. W. M. 8: Child, I. L. Child training and personality: A cross-cultural study. New Haven:Ya1e Univer. Pr., 1953. Wright, H. F. The emct of barriers upon strength of motivation. In R .G. Barker et al (Eds.), Child behavior and development. New York: McGraw-Hill, 191.73. Ch. 20. Yarrow, L. J. The effect of antecedent frustration on projective play. P_sychol. Monogr., 191+8, 62, No. 293. Zander, A. F. A study of experimental frustration. Psychol. Monogr., 191m, 56, No. 256. Zeller, A. F. An experimental analogue of repression: II The effect of individual failure and success on memory measured by relearning. J. exp. Psychol., 1951, 112, 32-38. l. 2. 3. S. 7. APPENDIX I Incomplete stories Jim and.Bill (Joan and.Dorothy) were at a picnic by a lake. All of a sudden, Jhm (Joan) took Bill's (Dorothy's) ball and threw it into the lake. What does Bill (Dorotrnr) feel like doing when his (her) ball is thrown into the lake? A.boy (girl) likes to go and play with his (her) friends in the street. His (her) mother says, "Put on your rubbers (boots) be- fore you go out. It is wet outside." It is too much bother for the boy (girl) to put on rubbers (boots). as (she) does not like to do it at all. What does the boy (girl) feel like when his (her) mother tells him (her) he (she) must put them on? A.bcy (girl) was having a good time putting together a puzzle. Just as he (she) was about to finish it, another child steps on it, and messes the puzzle up. What does the boy (girl) feel like doing when his (her) puzzle was ruined? Some money was missing from the teacher's desk. The teacher asked a‘boy (girl) about the missing money. The teacher then told him (her) that she believed that he (she) took the money. But the boy (girl) knows that he (she) did not take the money. What does the boy (girl) rai like doing about the teacher saying he (she) did? Larry and Joe (Joyce and Sandy) had permission to go together to the zoo. They had looked forward to going for a long time. On the day that they were supposed to go, Larry (Joyce) became sick and Joe (Sandy) could not go alone. What does HOe (Sandy) feel like doing about not going to the zoo? A.boy (girl) had a lot «of fun making a mess in the living room. Mother tells him (her) to clean up before father comes home. But he (she) had so much fun making a mess. What does he (she) feel like doing when told to clean up? A.boy (girl) was reading a story in a comic book. Just as he (she) was at the end of the story, he (she) found that the last page was torn out of the book. What does the boy (girl) feel like doing when he (she) discovers that the page is missing? .A'boy (girl) became angry with his (her) friend. His (her) mother and father think that the friend is right and scold him (her) for being so mean to his (her) friend. What does the boy (girl) feel like doing when his (her) mother and father do that? -90- 9. One day, a'boy (girl) did something naughty. His (her) friend saw him (her) and told so that he (she) was punished. How does the boy (girl) feel about his (her) friend for tattling on him (her)? 10. A.boy (girl) was told that he (she) could not play with his (her) new h00p in the house. One rainy day, he (she) wanted to play very much'with the hoop, but his (her) mother would not let him.(her) as long as he (she) was in the house. What does the boy (girl) feel like doing nOW? Filler stories 1a. A boy (girl) went out to a party with a friend and received a lot of prsents. As he (she) was looking over his (her) belongings, he (she) found that he (she) had a lot more than his (her) friend. What does he (she) feel like doing now? 9a. A boy (girl) was getting a special present for doing something for his parents. He (she) wondered what he (she) was going to get. When he (She) got the present, what does he (she) feel like doing? -91 - «.o« 0.0« 1.00 0.0« 0.0« 0.0« 0.«« 0.0« 0.3 0.0« 0.0« 0.«« o.¢« 0.0« 0.0« «.3 0.0« 0.0« 0.0« 0.0« 0.: o.~« 0.H« 0.0« 0.0« «.«« 0.0« 0.0« 0.0« 0.0« 0.0« 0.0« e.«« «.«« 0.~« H.¢« 0.0« v.«« 0.0« 0.0« 0.«« «.«« 0.0« 0.0« 0.0« t.«« 0.0« 0.«« 0.«« 0.«« «.n« v.0« A.v« 0.0« 0.0« 0.0« 0.0« ~.0« 0.0« v.0« 0.0« «.~« 123 0..—« 0..~« h.«« 0.«« 0.«« 0.0« 0.«« 0.0« 0.3 0.0« ¢.h« 0.0« «.0« ¢.0« c.«« 0.0« 0.«« 0.«« «.0« 0.0« 0.0« 0.0« 0.0« 0.0« 0.0« 0.0« 0.0« 0 .«« c.«« 0.0« 0.«« 0.0« 0.0« 0.0« o.¢« 0.0« 0.0« «.t« 0.0« 0..0« 0.0« ¢.0« 0.0« 0.0« 0.0« 0.0« 0.0« o.¢« 0.0« h.¢« 0.0« 0.0« 0.0« «.0« «.~.« 0.0« 0.00 0.0« 0.0« 0.0« «.t« 0.0« 0.0« 0.0« v.0« 0.0« 0.0« 0.0« 0.0« 0.0« 02—0 0.00 0.0« «.0« 0.0« 0.0« 0.0« «.0« 0.0« 1.3 0.0« 0.0« 0.0« 0.0« 0.0« «.00 0.00 0.0« 0.0« hag 0.0« 0.0« 0.0« 0.0« 0.0« 0.0« 0.0« 0.0« 0.0« 0.0« 0.0« 1.00 0.0« 0.0« 0.0« H.0« 0.0« 0.0« .0.0« 0.0« 0.0« «.5 V.h« ~.0« 0.0« 0.00 0.3 0.— 0." 0A «A .3 o." 0 0 b 0 0 V 0 « a soda." 7... :3 can... 102.. con-sucros— nal. 2 ME 9».- 383 .5:8 n.s.—.3 ca“ 0 at." 2.5- 0.33 .55 3.15 oh). 383 ha :1. 'y b! if; ’l’l’-’l‘ th‘ / J 1 o 0 0 j , . J / ., . / o J J. J ‘ J a o x o . . .. . J J J ,p . O C , / J I J G ‘ O J J H: J J J J U 0 I. I; J ‘/ I, J ,x o o J . A. I I. 1 z o o J I. J 1 . . J . p o J x J ) x . L. c . r J u I .J z. o O 1 H .) \J ) J J O . o I"-.D.[ . H _ . j / yo .1 \J O J _\ . . I J .I, U 1 ./ 1 J ) \ a . j J « O I J . 4 , .- 5, J I/ J .. , J N- A . o x/ J . . IN“ 1 I ‘i 00.0 00.« 00.« «0.0 00.« «0 .0 0«.0 00.0 0«.0 00.0 00.0 00.0 0«.0 0«.0 00.0 00.« 00.« 00.0 00.0 00.0 00.« 00.0 0«.« 00.0 00.0 00.0 00.0 0«.« 00.0 00.0 00.0 00.0 00 .0 0«.0 00.0 00 .0 0«.0 00.0 00.0 00.0 00.0 «0.0 00.0 8.0 00.0 00.« 00.« 00.« 00.« 00.« 0«.0 00.« 00.« 00.« 00.« 00 .0 00.« 00.« 00.« 00.0 «0.0 00.0 0«.0 00.« «0.« «0.0 0«.0 00.0 00.0 00.0 «0.0 00.0. 00.0 00.0 00.0 00.« 00.« 00.0 00.« 00.0 00.« 00.« «0.0 8.0 00.« 0«.0 00.« 00.« 00.« 00.« 00.0 «0.0 00.« 00.« 00.« 00.0 «0.0 «0.« 00.« 00.« «0.« 00.« 00.« 00.0 00.0 «0.« 00.« 00.« 00.« 00.0 00.« 00.« 00.0 00.0 «0.« 00.« 00.« 00.« 00.0 00.« «0.0 00.0 00.0 8.0 00.« «0.0 00.0 00.0 «0.0 00.0 8.0 00.0 00.0 00 .0 «0.0 «0.« 00.« 00.0 00.0 00.0 00.0 00.« 00.0 8.« 00.« 00.« 00.« 00.0 «0.0 «0.0 00.0 00.0 00.« 00.« 00.« 00 .0 «0.0 00.« 00.« 0«.0 00.0 00 .« 00.0 00.0 00.« 00.0 00.0 00.0 00.« 00.0 00.0 00.0 0«.0 00.0 0«.0 00.0 00.0 «0.0 «0.0 «0.0 00 00 00 «0 00 00 0 0 0 0 0 0 0 « 0 0.0.0.0. .943. .1500}. 3.! 3 fig .0.- .030 .550 .0.- .050 a?” ogd g.- .0000 35.300 .0800 0:“ .n» 0 .050 .350 .0.- .030 on!” . ..o.. ... . - fl _ ,0 o .r c; 1 J . o . l / o A. .L - \ . w . f,. . I , I A g y. ... , 1 .. e c ., m .o . ..W . ., . o J. . . o. q. 9/ 1 z u a u, 1 A . ./ 0. ,. ‘ 1"“ J . 0 p > .O 4 y 1 . . o /. , o o ,0 I , 0 0 v 7 .u‘ r .L .0 I U. 1 . . 0. ,o z 4 I / ,o 0 . a J I. . o J I 5 ._ . ob. , . 2 J. ., , . o . . ; .. ., .x ./ o . . ) a / F o ,. u . .r. m . I, .r 0 . ; J z , o .o . iii-152110,...” A .Iw .17.?)qu ntkl.’ . ... fl. \ 1’! . y 'l’!‘ {1. -93.. mm II In»! aux-«av. fantasy «or» Ag. 7} ’ 10% Expert-um group 6131. o“ 1.” 1.15 Icy: 1.10 2.00 1.03 Cum]. group 6131‘ .58 1000‘ 10“ '0’. .7, 105° 1.,2 .- 'nuw‘l ’ ....y ‘5? 4".II.. xv... O..." . ... 5,; i MICIIWHWIW iinIllEIITIW WW!“ 3 1193 03083 0167