*Ll ‘ WWII“H‘HWHII(MINHWIIW I 1 MI OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY MATERIALS: N Place in book return to Hallow.- charge from circulation records THE EFFECT OF AN EXPERIMENTER'S PRESENCE OR ABSENCE DURING ADMINISTRATION OF THE DRAW'A PERSON TEST BY Jeffrey Paul Roach A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1981 C5115 ’/ .. . g // (ff 8/ ABSTRACT THE EFFEIIT OF AN EXPERIMENTER'S PRESENCE OR ABSENCE DURING AUWINISTRATIW OF THE DRAW A PERSCN TEST BY Jeffrey Paul Roach Handler and Reyher (1964) compared figure drawings (man, woman, car) obtained from the same subjects under high and low performance anxiety conditions. Based on their results they con— cluded that manifestations of anxiety in human figure drawings have two sources: a) the anxiety producing characteristics of the laboratory situation, and b) the anxiety producing intrapsychic processes that are activated by drawing the hunan anatomy in an interpersonal context. The current study replicated their work by canparing figure drawings (person, opposite sex person, car) that were obtained when the experimenter was present (high performance anxiety) to figure drawings obtained men the experimenter was absent (low performance anxiety). This study also evaluated the effects of specific experimenters, the sex of these experimen- ters, and the sex of the subjects. The dependent measures were 15 drawing variables, two measures of sympathetic nervous system activation (electrodermal response), and the amomt of time spent on each drawing. The results replicated Handler and Reyher's (1964) findings for the drawing indexes. However, findings for SNS activation and. time spent drawing indicate that differences in the drawings reflect successful avoidance of anxiety by flight fran the situation rather than feelings of anxiety while in the situation. They also indicated that the experience of the sub- jects was quite different, depending on whether the experimenter was present or absent. Vhen the experimenter was present, parti— cipants appeared to be more anxious about how they were impres- sing the experimenter (interpersonal anxiety). Mien the experi- menter was absent, the participants appeared to be more anxious about performing the tasks well (performance anxiety). The pat— tern of results in this study was interpreted via Wicklund's (1975) objective self awareness theory, which adequately explained the findings. This thesis represents the collective contributions of numerous individuals who have dedicated large amounts of time and effort to see its completion. In the initial data collection phase Bob, Louise, Renee, Ann Marie, Ed, and Harold made their contributions as experimenters, while Karen and Jackie served as technicians. The long, tedious, and eye-straining task of coding the GSR data was performed by Anahid, Kathy, Lisa, Julie and Leslie. The frustrating task of rating the drawings was performed by Louise and Bob again, along with Mike, Tim, Carole, and Sue. To all of you, I appreciate your interest in the project and your dedication to performing good work. Special thanks must go to Louise and Bob, who worked for the full duration of the project, and to Anahid and Mike, who made contributions far above the call of duty. I would also like thank Ray Frankman and Helen Benedict, members of my thesis committee. Your kind assistance and generous mder standing is very appreciated. My greatest thanks go to Joe Reyher, my'chairman. Your enthusiastic support has helped sustain me in discouraging times. Your excitement for doing research is contagious, and I have learned a tremendous amomt about doing research as well as what it means to be a researcher from our association. I greatly appreciate all that you have done for me. TABLE OF CCNTENTS Page LIST OF TABLES.......................................iv LIST OF FIGURES......................................Vi LIST OF APPENDICES”................................Vii TEXT A.Introduction....................................l Hypotheses......................................4 B.Method..........................................6 Participants....................................6 Experimenters...................................7 Instruments.....................................7 Procedure.......................................9 C.Results........................................ll Interrater Reliability.........................ll Replication of Handler and Reyher (1964).......12 Hypothesis I...................................13 Hypothesis II..................................14 Other Findings.................................15 DoDixLSSionoooooooooo.00.0.000000000000000000.0016 REFMfi..0.000......O...00......0.0.0.00000000000032 AWICEO0.0.0.0....O0....O0.0...00.0.000000000000035 iii Tabl e (A) UT 10 11 LIST OF TABLES Comparison of Pearson Correlations between Raters on all Drawing Variables for Each Drawing with Interrater Correlations from Previous Studies. Page Specification of the Analysis of Variance Design 27 Including Degrees of Freedom, Error Terms, and Nesting Information. Significant Results for the Main Effect of Presence/ Absence of the Ehtperimenter and for High and Low Performance Anxiety Testing Conditions in the Handler and Reyher(1964) Study. Significant one tailed (l) and tm tailed (2) T-Tests are reported. Mean Squares and F-Ratios of Time Spent for Each Drawing. Mean Squares and F-Ratios for Mean Conductance during Each Drawing. Mean Squares and F—Ratios for Frequency Rate during Each Drawing. Mean Squares and F-Ratios of Shading for Each Drawing. Mean Squares and F-Ratios of Erasures for Each Drawing. Mean Squares and F-Ratios of Delineation Line Absence for Each Drawing. 28 45 46 47 48 49 50 Mean Squares and F-Ratios of Body Simplification 51 for Both Human Figure Drawings. Mean Squares and F-Ratios of Head Simplification 52 for Both Human Figure Drawings. iv 12 13 14 15 16 17 18 20 21 22 Mean Squares and F—Ratios of Vertical Imbalance for Each Drawing. Mean Squares and F-Ratios of Reinforcement for Each Drawing. Mean Squares and F—Ratios of Hair Shading for Both Human Figure Drawings. Mean Squares and F—Ratios of Placement for Fach Drawing. Mean Squares and F—Ratios of Line Discontinuity for Each Drawing. Mean Squares and F-Ratios of Omissions for Each Drawing. Mean Squares and F—Ratios of Transparancies for Each Drawing. Mean Squares and F—Ratios of Body Size for Both Human Figure Drawings. Mean Squares and F—Ratios of Head Size for Both Human Figure Drawings. Mean Squares and F-Ratios of Head :Body Patio for Both Human Figure Drawings. Mean Squares and F—Ratios of Vertical and Horizontal Size and Vertical :Horizontal Ratio for the Car Drawing. S4 55 56 57 58 59 60 61 62 Figure LIST OF FIGURES Page Graphic presentation of the Present/Absent by 29 Sex of Ebrperimenter by Sex of Participant Interaction (AxfbcP) for Mean Conductance levels during A. person figure drawing, B. car figure drawing, and C. opposite sex person figure drawing. Graphic presentation of the Present/Absent by 30 Sex of Experimenter by Sex of Participant Interaction (AxExP) for the number of erasures for the person figure drawing. Graphic presentation of the Present/Absent by 31 Sex of Dcperimenter by Sex of Participant Interaction (AxBxP) for Delineation Line Absences for the person figure drawings. vi Append ix Append ix Append ix Append ix Append ix II III LIST OF APPENDICES vii page 35 45 64 Handler and Reyher (1964) reported that laboratory induced stress (performance anxiety) produced a variety of graphic and stylistic changes on their StbjeCtS drawings of two hunan figures (male,fema1e) and an automobile. They compared figure drawings obtained from the same subjects under a high performance anxiety condition (tested individually in a laboratory while attached to a polygraph) and a low performance anxiety condition (tested in groups in a classroom with implied anonymity). Although all three drawings were affected, a significantly greater number of graphic indexes of anxiety 1 differentiated the testing condi— tions for the human figure drawings than for the automobile dra— wing. They wanted to compare the hunan figures to a figure that was equally difficult to draw 2 ,but less likely to arouse anxie- ty. Since the automobile figure had fewer indexes that differen- tiated conditions, they concluded that manifestations of anxiety had two sources: a) the anxiety producing characteristics of the laboratory situation, and b) the anxiety producing intrapsychic processes that are activated by drawing the human anatomy in an interpersonal context. A later stLrly (Handler and Reyher, 1966) 1 Handler (1967) developed a manual for scoring 2] different gra- phic indexes of anxiety in human figure drawings. Examples are shading, placement, and distortion. 2. Handler (1963) had previously demonstrated that subjects con— sidered these figures to be equally difficult to draw. provided further support for this interpretation since subjects experienced higher levels of sympathetic nervous system activa— tion (GSR frequency and (ER mean conductance) when drawing human figures than when drawing an automobile. The pattern of findings for the graphic indexes in the original stlrly (Handler and Reyher, 1964) suggested an explana— tion as to my the figure drawings differentiated the testing conditions so well. In the high performance anxiety condition, they found significantly less shading, hair shading, erasure, reinforcement, and emphasis lines, and significantly more distor- tion, head simplification, body simplification, detail loss, vertical imbalance, delineation line absence, and line pressure than in the low performance anxiety condition for the human figure drawings. They interpreted this pattern of results as reflecting a ”dimmition in effort, articulation, and detail,” and suggested that their subjects reacted to the anxiety produ- cing characteristics of the task and/or laboratory situation by attempting to finish as quickly as possible and flee the situa- tion. This study serves as a very useful demonstration that projec— tive test performance is affected by the testing situation as well as by the process of performing the task. However, it does not indicate which factors in testing situations alter perfor- mance since the high and low performalce anxiety conditions differed in a number of significant ways. For example, several different dimensions were individmllobserved) versus group (not observed) administration, identified versus anonymous perfor- mance, laboratory versus class room setting, and attached to polygrth versus not attached. It is unclear as to which of these factors, either alone or in combination, produced the results. Of these dimensions, the presence of an observer is likely to be the most significant. A substantial body of literature has documented the effects of an observer's presence on the perfor— mance of a variety of tasks (e.g. Beiman, Israel, and Johnson, 1978; Borkovec, Grayson, and Cooper, 1978; Johnson and Wiese, 1979; Sacco, and Pbkanson, 1978). One study was located that evaluated the effects of the presence of an observer on figure drawing performance. In this study Cassel, Johnson, and Burns (1958) reported that the number of ”interpretable features" and the size of the person figure decreased when the experimenter was present on the House, Tree, Person drawing test. The number of "interpretable features” of the house figure also decreased when the experimenter was present. If the presence of an observer does consistently affect figure drawing performance, then a critical question arises. Are there individual differences among observers in the extent to which their actions or characteristics alter figure drawing performance? Holtzman (1952) reported nonsignificant results for the effects of individual experimenters, their sex (2 male, 2 female), and the sex of the subjects on size, head size, head :body ratio, vertical imbalance, and placement of human figure drawings. Silverstein (1966) had two different experimenters act either ”tough" or ”tender" when they obtained human figure drawings from high and low anxious retarded children (mean age=13.0, mean Stanford-Binet IQ=49.2). He found no dif- ferences on ”qual ity measures" due to the experimenter variables. Starr and Marcuse (1959) reported no differences in the place- ment, size and head:body ratio of human figure drawings obtained one month apart by different experimenters. The purpose of the present investigation was to replicate and extend Handler and Reyher's (1964) study. It was replicated by evaluating the influence of a variable (presence/absence of an experimenter) that is believed to alter performance anxiety and consequently performance on figure drawing tasks. It was exten— ded by evaluating the influences of different experimenters and their sex on performance, and whether such variables differen- tially affected male and female subjects. The effects of these four variables were measured on fifteen graphic indexes of anxie— ty, the amount of time taken to draw each figure, the frequency of (ER skin responses, and (ER mean conductance. Hmtheses A statement of the rationale will precede a formal statement of the hypotheses. If a graphic index is primarily sensitive to externally induced anxiety, then it should significantly diffe- rentiate high and low performance anxiety conditions on all three draings. Out of the fifteen anxiety indexes being considered in this study, Handler and Reyher (1964) reported four that signi- ficantly differentiated the high and low performance anxiety conditions on all three drawings. These were shading, erasure, delineation line absences, and body simplification. Hair shading and head simplification differentiated conditions on both human figure drawings, but ratings were not made on the car drawing in their study for obvious reasons. Since they are conceptually similar to shading and body simplification respectively, it is unlikely that they are primarily sensitive to externally induced anxiety as well. However, if an anxiety index is primarily sensitive to annxiety induced by intrapsychic processes that are activated by drawing the human anatomy in an interpersonal situation, then it should significantly differentiate high and low anxiety conditions on the human figure drawings only. Of the fifteen anxiety indexes considered in the current project, only two fit the pattern in the Handler and Reyher stuiy (1964) . These were reinforcement and vertical imbalance. Several other anxiety indexes were affected in the Handler and Reyher (1964) study, but the results across drawings did not fit either pattern listed above. In most cases only one drawing was affected, but for omissions, the male figure and the automo- bile were both significantly affected. It is unclear what these anxiety indexes actually reflect since the results were not consistent with either pattern. Given this rationale, a formal statement of the hypotheses is in order. Hypothesis I. The presence of an experimenter serves as a source of laboratory induced performance anxiety during figure drawing tasks. Therefore, manifestations of anxiety that are known to be sensitive to external sources will be altered. Specifically, when an experimenter is present there will be less time spent (flight), a higher level of sympathetic nervous system activation (GSR mean conductance and frequency rate), less shadirng and hair shading, fewer erasures, a greater number of delineation line absences, and greater body and head simplifica— tions than when the experimenter is absent on all drawings upon which these measures were taken. Hypothesis II. The presence of an experimenter creates the interpersonal context which results in increased activation of the anxiety producing intrapsychic processes caused by drawing the human anatomy. Therefore, manifestations of anxiety known to be sensitive to anxiety produced by activated intrapsychic pro- cesses will be altered. Specifically, there will be increased reinforcement and vertical imbalance in the human figure draw- ings, but not in the car figure drawings. METHOD Participants One hundred and five undergraduates from introductory psychology courses participated in this experiment. Nine of these participants were excluded and randomly replaced during the data collection phase. Seven of them drew incomplete figures prior to procedural changes made to compensate for this problem. Two had resistance levels above 500,000 ohms which led to inaccurate recording of the skin responses by the polygraph that was used. The remaining ninety—six [articipants (48 males, 48 females) represented the desired number that was needed to complete the balanced design. Ebcperimenters The experimenters were six undergraduate advanced psychology majors (3 female, 3 male). They ranged in age between 20 and 23, with the exception of one woman who was 33. years old. All were caucasian, but were basically dissimilar in physical appearance. None of the experimenters recognized the participants that they tested. Instruments Handler Draw a Person Rating Scales. Handler (1967) developed scales for rating human figure drawings on twenty graphic indexes of anxiety. The following indexes were selected for use in the present study: shading, erasure, delineation line absence, body simplification, head simplification, vertical imbalance, reinforcement, hair shading, placement, line discon- tinuity, omission, size, head size, head:body ratio, and trans— parancy. Due to time and technical limitations the following indexes were not used: emphasis lines, light line-heavy line, line pressure increase, distortion, and detail loss. Handler (1967) did not describe prcedures for rating the automobile in his manual. Therefore, analogous scales had to be developed which closely approximated the procedures used to rate the human figures (see Appendix I). Problems arose when this translation was attempted for the size variables and body simplification. Only the height of the human figure was measured. Since a large amount of variance appeared to be present for both the height and the length of the automobile drawings, both dimen- sions were measured. These variables and their ratio (verti- cal:horizontal) were analyzed in the present stuiy. Attempts at devising a reliable scale of body simplification for the car figure consistantly failed. After approximately ten hours of alterirng the scale and training raters, agreement between raters still failed to rise above chance levels. Therefore, this variable was not rated for the automobile figure drawing. For each of the fifteen indexes, one pair out of eight potential raters was trained using Handler's manual. Before rating experimental data, each pair of raters had to achieve a level of 80% agreement on an independent set of 40 drawings. Once this critereon was achieved, this pair rated the entire set of experimental drawings for that variable. The dependent variables used in the analysis were computed by averaging the two ratings on each drawing for each variable. Grass (Model :5) Six Chanel Polygraph. Skin resistance was continuously recorded on paper tape by the polygraph. Mean skin conductance was derived by averaging the reciprocals of resistance levels measured at twenty secornd intervals. The first reading was taken ten seconds after the participant began each task. Transforming the skin resistance level to its reciprocal, skin conductance, normalizes the heavily skewed resistance distribution. This variable provides a global measure of sympathetic nervous system arousal for each task. ‘0 The frequency of skin responses was measured by counting the number of pen deflections that were greater than the distance representing 1000 ohms for each task period. This sum was then divided by the amount of time taken to complete that task to obtain responses per second. This variable measures the reactivi- ty of the participant's sympathetic nervous system to the task. The time for each task was calculated by measuring the distance between marks on the paper tape made by the participant pressing the event marker button. This distance was easily converted into seconds since the paper tape progressed at a set rate (2.5 centimeters per second). Procedure Each participant was greeted and ushered into an eight foot square room. This room contained two chairs, a square table in one corner, an extra recliner in the opposite corner, and a small stand which contained the paraphanalia needed to clean and prepare the electrodes. The room was brightly lit, and the temperature was maintained at 72 degrees. The participant was seated at the table facing the wall. A. clipboard which contained the experimental packet was secured to the table directly in front of the participant. A wall on the participant's right contained a hole through which the electrode and event marker wires passed. The experimenter's chair was on the participant's left which allowed the experimenter to face the participant when he or she was present. Once the participant was seated, the experimenter read the 10 instructions, answered questions about the procedure, and had the subject sign an informed consent form. These instructions informed the participants that they could expect to perform several drawing tasks and respond to two brief questionaires while their "electrodermal activity" was recorded by the “machine in the other room." They were asked to follow the written instructions in the packet carefully, to use a fresh pencil for each drawing task, to hold the hand with electrodes attached as still as possible, and to avoid speaking once they had started. Half of the participants learned that the experimenter would be present, but would not speak. The other half were told that the experimenter would wait in the other rooum until they had comple— ted the tasks. The experimenter then attached the electrodes to the first and second fingers of the participant's non-prefered hand. The polygraph was calibrated, and the habituation period was imposed. Once habituation started, the present experimenter remained silent and the absent experimenter left the room. The participant recieved the signal to begin (a knock on the door by a techni— cian) following stability of the resistance level for at least two minutes. After hearing the signal the participant opened the packet which contained six pages. The participant was asked by written instructions at the top of each page to a) complete a 23 item modified trait anxiety questionaire, b) draw six geometric figuures (a warm up task), c) Draw a Person, 6) Draw a Car, 11 e) Draw a Person of the Opposite Sex and f) complete a modified 24 item state anxiety questionaire. These tasks appeared in the packet in the order presented above with the exception of the placement of the state anxiety questionaire. It was systematical- ly placed behind one of the three figure drawings. A note at the bottom of each page reminded the participants to press the event marker button. Due to time and technical limitations, the state and trait anxiety questionaire data will not be analyzed. RESULTS Interrater reliability was estimated by computing the Pearson product-moment correlations between raters on each varia— ble for each drawing. These correlations were very similar to those reported by Handler (1967) and Attkinson et al (1974) (see table 1). The reliability estimates ranged between .67 and .97 and had a median of .90. Insert Table 1 about here A four factor analysis of variance design was used to evalu- ate the influences of the independent variables and their inter- actions on dependent measures (see table 2). All of the factors were completely crossed with the exception of the individual experimenteru) variable which was nested in the sex of experi- menter variable (a specific experimenter cannot be male and female). The individual experimenter variable was treated as a 12 random factor since it sampled a population of potential experi- menters and is a non-replicable factor in future research. The other three variables were treated as fixed factors. F-tests of main and interaction effects for the presence/absence of the experimenter(A), sex of experimenter (E), and sex of partici- pant(P) were initially based upon error terms with only four degrees of freedom. Power was increased for these tests by using a pooled error term and thereby increasing degrees of freedom. Appendix II explains the pooling procedure in detail. Insert Table 2 about here All of the F-tests for the main effects of presence/absence of the experimenter were converted to t-tests. One tailed t-tests were performed for all of the graphic indexes which had previous- ly differentiated high and low performance anxiety cornditions in the Handler and Reyher study (1964) . The t—values were considered to be significant only if they were in the same direction as the Handler and Reyher results. Two tailed t-tests were performed for the rest of the variables i.e. those variables which were either not included or did not differentiate testing conditions in the Handler and Reyher (1964) study. Replication o_f Handler and Reyher (1964) . The results of thi study replicated Handler and Reyher's (1964) finding that human figure drawings are affected by the external situation and by the internal processes that are activated by drawing the human 13 anatomy in an interpersonal context. Table 3 summarizes the findings for the presence/absence of the experimenter and compares them to the Handler and Reyher (1964) results. The presence/ absence of the experimenter significantly affected six drawing indexes on the person drawing and seven drawing indexes on the opposite sex person drawing. Only two of the drawing indexes for the car figure were affected. Chly the graphic indexes specified by either hypothesis I or Hypothesis II differentiated the experimenter present from the experimenter absent conditions. Hypothesis I. The results provided strong support for Hypothesis I which predicted directional effects due to presence/ absence of the experimenter on all three drawings for all of the dependent variables that were assumed to be primarily sensitive to externally induced anxiety. With the exception of head simpl i— fication for the opposite sex person drawing and delineation line absence for the car drawing, the six graphic indexes specified by Hypothesis I were significantly affected by the presence/ absence of the experimenter on all drawings for which these indexes were rated. All effects were in the expected direction. In addition to the main effects for presence/absence of the experimenter, four of these graphic indexes were qualified by significant interactions which included presence/ absence as one of the factors for particular figure drawings. These were erasure, delineation line absence, and head simplification on the person drawing and shading on the car drawing. A description of the 14 interactions for erasure and delineation line absence will not be provided here since they are both described in a later section. A significant present/absent by sex of.participant (AxP) interac- tion occured for head simplification on the person drawing. Head simplification was greater for female participants when an experimenter was present, but about the same for female partici- pants when the experimenter was absent or for male participants regardless of whether the experimenter was present or absent. Shading on the car figure had a significant IxAxP interaction. Since shading on the car figure is of little interest, it will not be described here. Neither GSR measure was significantly affected by the presence/absence of the experimenter during any of the drawings, but the amount of time spent did significantly differentiate the present from the absent conditions on all three drawirugs. when the experimenter was present, participants typically spent 40 to 60 seconds less than when the experimenter was absent. Hypothesis II. The results provided partial support for Hypothesis II which predicted directional. effects due to pre- sence/absence of the experimenter on hunan figure drawings only for variables that were assumed to be primarily sensitive to intrapsychic processes activated by drawing the human anatomy in an interpersonal context. Both vertical imbalance and reinforce- ment were were significantly affected by presence/absence of the experimenter for the opposite sex person figure drawing. Howe- ver, neither of these indexes was significantly affected on the 15 person or car drawings. Although delineation line absence was predicted to differen- tiate present and absent conditions on all three drawings , it did so only for the human figures. Consequently, it might also be considered to be affected primarily by the anxiety producing intrapsychic processes that were activated by the drawing task. However, the rating procedures used to rate the car figure may not have been similar enough to the procedures used to rate the human figures to draw that conclusion. None of the other varia— bles differentiated the present and absent conditions for the human figures only. Other Findings Twenty-three F-tests for the other main effects (E, P, and I) arnd interactions (AxF, AxP, BxP, AxExP, IxA., IxP, IxAxP) were significant. Five of these have important implications for the present sturiy, and the remaining 18 are of secondary interest. Since these latter findings are so numerous, they will not be presented here. The interested reader will find a description of these findings in Appendix IV. The five significant findings refered to above were all significant AxExP interactions for the mean conductance levels during all three drawings and for erasure and delineation line absense on the person drawing. These findings are especially germane because they reflect the effects of the interpersonal situation on the participant's anxiety level. Figure 1 demons- trates the remarkable consistancy of this interaction across the 16 three drawirng tasks for mean conductance. Durirng all three drawings, when the experimenter was present the participants experienced higher levels of sympathetic nervous system (SNS) acticvation (anxiety) when tested by opposite sex experimenters than when tested by same sex experimenters. However, when the experimenter was absent, participants experienced higher 80S activation when tested by same sex experimenters than when tested by opposite sex experimenters. Figures 2 and 3 demonstrate that the AxExP interaction for erasure and for delineation line absence were strikingly similar to the Ame interactions for mean conductance. The increased elevation in the present condition for both drawing indexes was reflected by the significant main effect for A on both variables. DISCUSSICN Overall, Handler and Reyher's (1964) findings that human figure drawings are affected by two sources of anxiety were replicated by the results in the present stuly. As was antici- pated the presence of an experimenter resulted in a number of graphic and stylistic changes in the figure drawings of the participants. Almost all of the drawing indexes that were hypothesized to be primarily sensitive to external sources of anxiety (Hypothesis I) were affected for each drawing (13 out of 15). However, findings were less clear for the two indexes that were hypothesized to be sensitive to internal sources of anxiety which are activated by the interpersonal task of drawing the human anatomy. Since both indexes were significantly affected by 17 the presence/absence of an experimenter on the opposite sex figure rejection of Hypothesis II is not warented at this time. The pattern of results for the graphic indexes arnd time variables confirmed Handler and Reyher's (1964) speculation that partici— pants react to the anxiety producing characteristics of the figure drawing situation by finishirng as quickly as possible and leaving (flight). The participants who were observed spent approximately fourty to sixty seconds less on each drawing than participants who were not observed. Likewise, the differences on the graphic indexes reflected less ”effort, articulation, and detail" in the drawings of observed participants, Since the observed participants spend less time drawing, the drawing indexes probably reflect less opportunity to improve on and add detail to the drawing. The findings for the GSR mean conductance indicated that observed participants did not flee the situation because of feel irngs of high anxiety. None of the mean conductance measures had a significant main effect for the presence/absence of the experimenter. If flight from the situation was motivated by the desire to reduce high anxiety, then sympathetic nervous system activation should have been consistantly higher across tasks for observed participants. Although these GSR variables were not directly affected by the presence of an experimenter they were affected by fascinating interactions. Results for mean conduc— tance indicated that when participants were observed, they were more anxious in the presence of opposite sex experimenters. 18 However, when participants were not observed, they were more anxious when tested by same sex experimenters. The implications of these interactions are elucidated when considered from the framework of Duval and Wicklund's (1972, Wicklund, 1975) theory of objective self awareness. According to this theory, concious attention is directed either entirely toward the self or toward external events in the environment. Particular stimuli, such as a mirror or a tape recording of the participants voice, focus a person's attention on some dimension of self i.e. results in objective self awareness. All other stimuli distract attention away from self and toward the environ- ment. when attention is directed toward some dimension of self, the result is self evaluation of the discrepancy between attain- ment and aspiration on that dimension. If this discrepancy is positive (typically due to a recent success experience), then a person should experience positive affect and will seek out situa- tions that stimulate objective self awareness (approach response). However, if the discrepancy is negative, an indivi— dual will experience negative affect and will actively attempt to avoid stimuli which result in objective self awareness. In situa— tions where the discrepancy is negative and objective self aware— ness is inescapable, the person will attempt discrepancy reduc- tion. The degree of affect that is experienced is considered to be a joint function of the degree of the discrepancy and the proportion of attention focused on that discrepancy during a time interval . In this study, objective self awareness was probably induced by the procedure of recording Galvanic Skin Resistance. Most of the participants, if not all, were aware that some aspect of their internal state was being monitered since electrodes were attached by finger clamps to their non—prefered hand. If objec- tive self awareness was induced in both observed and non-observed participants, then why did these groups differ on a number of variables? Wicklund (1975) argued that the presence of an audience results in objective self awareness when the person realizes that the attention of the audience is focused on some feature of self. Consequently, observed participants may have simply reacted to two stimuli (polygraph arnd observer) that induce objective self awareness rather than just one (polygraph). However, the pattern of the interactions for GSR mean conductance suggest a slightly different interpretation ; The presence of an observer appeared to alter the dimension upon which self evalua- tion occurs. A brief description of the situation of this experi— ment will help to make this point. The participants were probably representative of eighteen to twenty-two year old undergraduates who attend introductory psychology courses at a large state university. In the laboratory situation they were met by complete strangers, who were approxi— mately the same age or slightly older. They were asked to perform a task that should result in a large negative discrepancy between actual and ideal performance for most participants. Thus partici- pants were faced with the situation of trying to create a 20 favorable impression in the eyes of a stranger while performing a task that they probably do not do well. When participants were observed, their sympathetic nervous system activation was higher in the presence of opposite sex experimenters than in the presence of same sex experimenters. This suggests that the dimension of evaluation of their objective self awareness was their appeal or acceptability to the experi- menter. This was inferred because in this period of their life most college students have major concerns about their sexual appeal to and ability to develop relationships with members of the opposite sex. Thus anxiety was greater when opposite sex experimenters were present because participants were more anxious about being accepted by a strange member of the opposite sex. Since there was little opportunity for the participant to speak, they could only impress the observer with their physical appea— rance and their performance on the drawing task. Since self evaluation on both dimensions was likely to result in a negative discrepancy, the participants needed to avoid negative affect. In addition to focusing awareness upon their acceptability to the experimenter, the presence of an observer also effectively prevented the participant from avoiding self awareness while in the situation, Consequently, the participants had to finish quickly and leave because of their inability to successfully distract their attention away from self. when participants were not observed, they were more anxious when they had same sex experimenters than when they had opposite 21 sex experimenters. This suggest that objective self awareness was focused primarily on their ability to perform the drawing task rather than on their acceptability or appeal to the experimenter. This inference is based on the assumption that college students at this age tend to be more competative with members of the same sex. Thus participants were more anxious because they feared the critisizm of the same sex experimenter more than the critisizm of the opposite sex experimenter. Since an observer was not present, participants were able successfully to focus attention outward by attending to the task i.e. the actual mechanics of drawing the figures absorbed their attention and prevented focus on self. Thus they spent more tiume drawing, because they could avoid objective self awareness and because they wanted to improve their performance and thereby reduce the discrepancy between actual and ideal performance. Some of the most useful findings in this study occured for the eraSure and delineation line absence indexes. Both of these were qualified by significant AbeuP interactions on at least one drawing variable that strongly resembled the AxExP interactions for mean conductance described above. However, they also had significant main effects for the presence/absence of the experi- menter. This pattern of findings indicates that both erasure and delineation line absence on the person drawing have two different determinants. Both appear to be sensitive to the participants anxiety that is induced by the interpersonal context and to the amount of time spent in the situation (avoidant defense). 22 The results of this study have strong implications for future research on the validity of the Draw a Person test (DAP) as a personality assessment instrument. Major literature reviews have consistently reported mixed or negative findings for studies that attempted to document the relationship between particular aspects of personality and drawing variables (Swanson, 1957, 1968; Roback, 1968; Harndler and Reyher, 1965). The current findirngs provide further confirmation for Handler and Reyher's (1965) suggestion that inconsistancies in the DAP research results were the product of a lack of understanding about and control over situational effects on drawing performance. The major implication for future research on the DAP is that the relationships between particular aspects of personality and drawirng variables may be obscured if testirng conditions result in participants finishing the drawings quickly and fleeing the situation. when participants rush to complete the drawing tasks, they leave impoverished drawings that tend to have little varia- tion between participants on a number of indexes that are believed to be indicative of aspects of personality. when testing conditions are arranged that encourage participants to invest time and effort in the drawing task, variation between participants on these drawing variables ternds to increase. The greater degree of variation between participants should reduce restriction of range problems in correlation studies and thereby increase the likelihood that significant relationships may be found . 23 The results of this study also have major implications for the use of the [AP in other research settings. As a research instrument, the DAP is valuable because a variety of variables (Handler's graphic indexes of anxiety) are easily measured, highly reliable, and highly sensitive to particular types of situational manipulations. As a research methodology, the Draw a Person test procedure has high potential, because it effectively creates a situation that heightens anxiety about performaing adequately. Thus it is a particularly useful procedure for studying the effects of a variety of interpersonal and situational variables on performance. The clinical implications of this study are numerous. First of all, the DAP can be used to assess a client's ability to cope with an anxiety producing situation. The clinician could obtain drawings under conditions that are likely to provoke a flight response and under conditions that encourage the investment of time and effort in the drawings. Traditional personality inter- pretations may be made from the latter set. Comparisons of the two sets of drawings on the anxiety indexes that are sensitive to situational manipulations and on the amount of time taken to complete all drawings should provide valuable information about the client's ability to cope with anxiety producing situations. when drawings from both sets show effort arnd adequate detail, the diagnostic conclusions might be either excessive concerns about performance, or a general ability to cope with situationally produced anxiety. when the drawings obtained in the high anxiety 24 situation are impoverished and the ones obtained in the low anxiety situation show greater effort arnd detail, then the clini- cian might conclude that the client has difficulty in coping with an interpersonal situation that made him or her anxious. Further- more, the client's response was to avoid the situation by leavirng. when drawings are impoverished in both situations, the implications might be either strorng resistance to being assessed or excessive anxiety induced by the process of drawing the human anatomy. Comparisons of drawings obtained when the assessor is present and when the assessor is absent might provide valuable information about the client's ability to focus on self when in the presence of a therapist. The ability of the client to tolerate objective self awareness in the Draw a Person test situation while being observed might serve as an effective predictor of whether that client would be able to tolerate the strorng anxiety inducing process that occurs from the intensive self focus in insight oriented psychotherapies. A final implication is that this study clearly demonstrates that the assessor should always consider the influences of the cornditions of testirng when clinically interpreting the DAP. Failure to consider the effects of the interpersonal situation on projective test performance can easily result in confused arnd inappropriate interpretations. The methodology of comparing figure drawings obtained in different situations and of comparing human figure drawings to the relatively neutral car figure 25 drawing should facilitate the process of determining whether the enurce of variation in human figure drawings is situational or intrapsychic . 26 Table 1 Cbmparison of Pearson correlations Between Raters on All Drawing 'Variables for Each Drawing wdth Interrater Correlations from Previous Studies. Previous Studies Present Study ._A_ _B_ _C_ variable Name All Ad] Adl Per OSPer Car Shading .90 .93 .83 .88 .82 .85 Erasures .91 .87 .93 .91 .69 .78 Delin. Line Abs .89 .80 .90 .97 .96 .81 Body Simp. .71 .73 .76 .83 .81 — Head Simp. .88 .80 .86 .89 .82 —- Vert. Imballance .91 .70 .90 .82 .77 .77 Reinforcement .76 .94 .77 .94 .88 .84 Hair Shading .87 .87 .90 .95 .90 - Placement 1.00 - - .94 .90 .95 Line Discont. .84 .97 .67 .91 .91 .78 Omission .90 .92 .87 .94 .94 .97 Transparency .74 .97 .97 .83 .66 .81 Body Size .94 -— -- .97 .94 -- Head Size .99 - -- .96 .95 -— Head:Body Ratio .95 -— -- .87 .90 -- vert. Size —— - -- -- -— .93 Horz. Size - - —- —- —- .96 A= Attkisson, waidler, Jeffrey, and Lambert (1974) 8: Handler (1963) C= Handler and Reyher (1964) 27 Table 2 Specification of the Analysis of Variance Design Including Degrees of Freedom, Error Terms, and Nesting Information. Source M g Nestig Error Term Sex of Buperimenter E l I Presence/Absence of F A l IxA Pres/Abs by Sex of E AxE l IxA. Sex of Participant P l IxP Sex of P by Sex of B PxF. l IxP Sex of P by Pres/Abs PxA 1 IxAxP Sex of P by Pres/Abs by Sex of E PxAxE 1 IxAxP Individual mperimenter I 4 E S Ind E by Sex of P IxP 4 E S Ind E by Pres/Abs IxA 4 B S Ind E by Pres/Abs by U) Sex of P IxAxP 4 E U) Subjects 72 IthchP none 28 Table 3 Significant Results for the Main Effect of Presence/Absence of the Duperimenter in the Present Study and for High and Low Performance Anxiety Testing Conditions in Handler and Reyher (1964) Study. Significant one tailed (l) and two tailed (2) T-tests are reported. Handler and Reyher Present Study Variable Name Male Female Auto Per OSper Car Time N/A N/A N/A 2 2 2 Mean Conductance N/A N/A N/A Frequency Rate N/A N/A N/A Shading 2 2 2 l l 1 Erasures 2 2 2 l 1 l Delin. Line Abs. 2@ 2 2 1 1 Body Simp. 2 2 2 l 1 N/A Head Simp. 2 2 N/A 1 N/A Vert. Imballance 2 2 l Reinforcement 2 28 1 Hair Shading 2 29 N/A 1 1 N/A Placement 2 Line Disc. 2 mission 2 2 Transparancy Body Size 2* N/A Head Size N/A N/A Head:Body Ratio N/A N/A Vertical Size(Car) N/A N/A N/A N/A N/A Horizontal Size N/A N/A N/A N/A N/A Vert:Horz Ratio N/A N/A N/A N/A N/A N713 Not applicable for this variable since a test was not performed. 6 Significant at p<.10; All other Handler and Reyher (1964) results were significant at p<.05. * Significantly smaller in high performance anxiety condition when tested in that situation first. However, they were significantly larger when tested in the low performance anxiety condition first. 29 A. .8 . Male E 1. Female F 1.1 MJale P Female P Present B. .8? .fi- Male E 1.0- 1.1L (u- 1.2- Female B Male p Feinale p Present C. .8» .9- ‘- Male E 1,n¢ 1.1:- uL Female E Kale P Female P Present .84- ‘- Female E .9"- 1.0- 1.1- Male E Méle P Female P Absent .8 ' Female E 1.0 1 O 1 Male F 1.2 Male P Fe4ma1e P Absent . Female E 1. 1.1 Male F Tale P Feinale P Absent Figure 1. Graphic presentation of the Present/Absent by Sex of EXperimenter by Sex of Participant Interaction (AxEXP) for the mean conductance level of subjects during the A. person figure drawing, B. car figure drawing, and C. opposite sex person figure drawing. 30 1.00”t 1.00 .75? .75 no Male E 0501- Male E .50 025". Female E .25 Male P Feinale P Kale P Fehale P Present Absent Figure 2. Graphic presentation of the Present/Absent by Sex of Experimenter by Sex of Participant interaction (AxeP) for the number of erasures for the person figure drawings. 2.0— 2. 1.5- \ Female F. l. - Male E 1.0~ 1. >< Female E .5- .. Male E .L 4. i . Male P Female P Male P Female P Present Absent Figure 3. Graphic presentation of the Present/Absent by Sex of Duperimenter by Sex of Participant interaction (AxExP) for Delineation Line Absences on person figure drawings. REFERENCES REFERENCES Attkisson, C.C., Waidler, V.J., Jeffrey, P.M., and Lambert, W.E. Interrater reliability of the Handler Draw a Person Scoring. Perceptual and Motor Skills, 1974, 38, 567-573. Beiman, 1., Israel, 8., and Johnson, S. A. During trainirng and posttraining effects of live and taped extended progressive relaxation, self-relaxation, and electromyogram biofeedback. Journal g Consulting and Clinical Psychology, 1978, g, 314-321. Borkovec, T.D., Grayson, J. B., and Cooper K. M. Treatment of general tension: Subjective and physiological effects of progressive relaxation. Journal o_f Consulting and Clinical Cassel, R.H., Johnson, A., and Burns, W.H. Examiner ego—defense and the H—T—P test. Journal of Clinical Psychology, 1958, 14, Handler, L. Anxiety Indexes in the Draw a Person Test: A Scoring Manual. Journal o_f Projective Techniques and Personality Assessment, 1967, 31, 46-57. Handler, L. and Reyher, J. The effects of stress on the Draw a Person test. Journal o_f Consulting Psychology, 1964, _2_8_, 259-264. Handler, L. and Reyher, J. Figure drawing indexes: a review of the literature. Journal of Projective Techniques and Personalitl Assessment, 1965, _2_9, 305-313. 32 (A) w Handler, L. arnd Reyher, J. The relationship between GSR and anxiety indexes in projective drawings. Journal o_f Consulting Psychology, 1966, _3_0_, 60-67. Holtzman, w. H. The examiner as a variable in the draw a person test. Journal 9_f Consulting Psychology, 1952, 16, 145-148. Johnson, L. S. and Wiese K. S. Live versus tape recorded assessments of hypnotic responsiveness in pain—control patients. International Journal 93 Clinical and Experimental Hmosis, 1979, _2_7_, 74—84. Roback, H. B. Human figure drawings: their utility in the clinical psychologist's armamentarium for personality assessment. Psychological Bulletin, 1968, _'_7_(_l_, 1-19. Sacco, W. P. and Hokanson, J. E. EXpectations of success and anagram performance of depressives in a public and private setting. Journal 3: Abnormal Psychology, 1978, _81, 122-130. Silverstein, A. B. Anxiety and the quality of human figure drawing. American Journal o_f Mental Deficiency, 1966, :72, 607-608. Starr R. arnd Marcuse, A. Reliability in the Draw a Person test. Journal o_f Projective Techniques, 1959, __2_3_, 83-86. Swenson, C. H. Jr. Empirical evaluation of human figure drawings. Psychological Bulletin, 1957, 51, 431-466. Swenson, C. H. Jr. Empirical evaluation of human figure drawings. Psychological Bulletin, 1968, 10, 1-19. Wicklund, R. A. A theory of objective self awareness. In L. Berkowitz (Ed.) Advances 12 Buperimental Social Psychology 34 (Vol. 8), New York, Academic Press, 1975. APPENDIX I 35 Appendix I contains the scoring criteria for the twelve variables that were used in rating the car figure drawing. An attempt was made to construct scoring criteria which would approximate Handler‘s (1967) procedures for rating human figure drawirngs as closely as possible. The major problem in the construction of analogous scales occurred for human figure drawings that require separation of the body into parts. Handler (1967) divides the body into several areas, 1) head (including facial features), 2) neck, 3) one or both hands, 4) one or both feet, 5) one or both legs, 6) one or both arms and, 7) trunk. Followirng several attempts, a system was developed for dividing the car into five distinctive areas: 1) The area forward of a vertical line drawn tangentiallly to the front edge of the front tire. (typically includes the front bumper, head lights, grill, and part of the hood), 2) The area backward of the vertical line drawn tangentially to the back edge of the back tire (typicallly includes the back bumper, tail pipe, tail lights, and part of the trunk) 3) The area above a horizontal line that is drawn between the point where the wind shield meets the hood, and the point where the back window meets the trunk (typically includes all windows arnd roof) 4) one or both tires, and 5) the remainder of the car (corresponds to the trunk of the human figure). Division of any one of these five areas into two lesser areas did not add additional precision to the measurement process. hiring the training for erasures index, the raters kept track of each incident of finding two erasures in one car area. It was quite 36 rare to find two erasures in one area, although it was not rare to find more than one erasure on a drawing. Combination of two areas into one did seem to result in a loss of precision in the measure though. Other problems in the development of analogous irndexes will be discussed when the scales are presented. I. Snading Shading within any essential car area is scored. A design or emblem or any consistent pattern of lines, (e.g. cross hatching) is scored as shading. Enuhaust smoke is also scored as shadirng. Score 0 when there is no shading. Score 1 when there is shading on any one car area. Score 2 when there is shading on any two car areas. Score 3 when there is shading on more than two car areas. II. Erasure An erasure on any car area is scored. Score 0 when there are no erasures. Score 1 when there is erasure on any single car areas. Score 2 when there is erasure on any two car areas. Score 3 when there is erasure on three or more car areas. III. Delineation Line Absence This index refers to the absence of lines on the car that divide it into various parts. The extreme case is a car figure that is merely an outline, or a shell. The drawing was scored for the presence of 1) at least one vertical line that delineates a window, 2) at least two vertical lines that delineate a car 37 door, and 3) lines that connect the tires to the car. Other possible delineation lines were judged to have occurred too infrequently to be of use in differentiating performance. Score 0 if all three criteria are met. Score 1 if one of the criteria is not met. Score 2 if two of the criteria are not met. Score 3 if all of the criteria are not met. IV. Vertical Imbalance The angle formed by the bottom edge of the protractor (which was parallel to the horizontal edge of the paper) and a line which passed along the bottom side of the car was measured. The number of degrees that deviated from the 0 degree mark was counted. Score 0 if figure has less than a 2 degree deviation. Score 1 if the figure has a deviation between 2 1/2 and 8 1/2 degrees. Score 2 if the figure has a deviation between 9 and 17 degree Score 3 if the figure has a deviation greater than 17 degrees V. Reinforcement Reinforcement consists of retracing of lines (lines that have been redrawn, or gone over). This does not include shading. Reinforcement is often confused with sketchiness of a line. Some subjects habitually draw using a sketchy line arnd therefore if most of the drawing is sketchy, reinforcement should not be scored. Also, lines that have been erased and redrawn with a single line should not be scored as reinforcement. 38 Score 0 if less than a quarter of the figure is reinforced. Score 1 if approximately a quarter of the figure is reinforce Score 2 if approximately half of the figure is reinforced. Score 3 if approximately three-quarters or more of the figure is reinforced. VI. Placement The drawing receives points depending on where the figure lies in relation to the vertical and horizontal axis. For the vertical axis, the car figure receives 0 points if the vertical axis passes entirely between both tires, 1 point if it touches either tire, 2 points if it falls outside of the tires but still passes through the car, and 3 points if it does not intersect the car. For the horizontal axis, the car figure receives 0 points if it passes between line A (the line that separates the windows and roof from the car body) and line B (the line that runs along the bottom of the car); 1 point if it passes above line A or below line B while still touching part of the car, and 2 points if it does not intersect with the car at any point. These points are summed for the two axes. Score 0 for 0 points. Score 1 for 1 point. Score 2 for 2 points. Score 3 for 3 or more points. VII. Line Discontinuity Line Discontinuity refers to the frequency of broken lines used in the drawing, and to the spaces left between various body 39 parts. an careful inspection these body parts may show them to be unconnected A line discontinuity is scored if it is possible to go from the outside of the car body wall to the inside of the ‘ car body wall without crossirng a body line. If the drawing is done with a sketchy line it is difficult to determine whether line discontinuity is to be scored. Line discontinuity should not be scored if, despite the sketchiness, it is impossible to go ffom the outside of the car body wall to the inside without crossing a body line. Score 0 if there are no more than three line discontinuities in a drawing Score 1 if four or five line discontinuities are present. Score 2 if six, seven, or eight line discontinuities are present. Score 3 if nine or more line discontinuities are present. VIII. Omission Score if there is an omission of any essential car area or when the figure is placed so that one or more essential car body areas has been cut off by the edge of the paper (paper—chopping). Score 0 if there are no omissions. Score 1 if one car body part is omitted. Score 2 if two car body parts arenomitted. Score 3 if three or more car body parts are omitted. IX. Transparancies Transparency is scored when part of a car body area which ordinarily is not visible due to the structure of the automobile 40 is visible in the drawing (e.g., ability to see the tires through the fender). Score 0 if there are no transparencies in the drawing. Score 1 if one transparency is present. Score 2 if two transparencies are present. Score 3 if three or more transparencies are present. X. Horizontal Size This index was not directly related to the human figure drawing index for body size since the car figure is typically wider than it is tall. New'cutoff points had to be set in order to reduce it to a four point scale; These points were selected on the basis of two criterea: 20% of the participants should be in each of the 0 and 3 point groups and 30% of the participants should be in the 1 and 2 point groups, and 2) this distribution should be symetrical about the mean of the distribution. Horizon— tal size was measured by placing an axis line along the bottom of the car. Perpendicular lines were extended to the farthest points in the left and right of the car. The distance between these perpendicular line was measured to the nearest 1/16th of an inch. Score 0 if the horizontal distance is between 4 and 5/8 and 5 and 5/16 inches. Score 1 if the horizontal distance is between 5 and 3/8 and 6 and 3/8 inches. Score 1 if the horizontal distance is between 3 and 9/16 and 4 and 11/16 inches. Score 2 if the horizontal distance is between 6 and 7/16 and 41 7 and 1/8 inches. Score 2 if the horizontal distance is between 2 and 1/2 and 3 and 1/2 inches. Score 3 if the horizontal distance is greater than 7 and 3/16 inches. Score 3 if the horizontal distance is less than 2 and 7/16 inches. XI. Vertical Size Vertical size is measured by placing a line tangential to the bottom of both tires. A perpendicular axis was constusted at the midpoint of this line. A perpendicular line was then drawnn between the axis and the highest point on the roof of the car. The distance between the two parallel lines along the axis was then measured. Score 0 if the vertical distance is between 1 and 3/4 and 2 and 1/8 inches. Score 1 if the vertical distance is between 2 and 3/16 and 2 and 7/16 inches. Score 1 if the vertical distance is between 1 and 1/2 and l and 11/16 inches. Score 2 if the vertical distance is between 2 and 1/2 and 3 and 1/8 inches. Score 2 if the vertical distance is between 1 and 1/16 and l and 7/16 inches. Score 3 if the vertical distance is greater than 4 and 1/8 inches. 42 Score 3 if the vertical distance is less than 1 inch. XII. Vertical:Horizontal Ratio The vertica1:horizontal ratio is measured by dividing the vertical distance by the horizontal distance. This variable was not recoded into four categories like the others, but left in its raw form. APPENDIX I I 43 In this design each of the mean squares which involved the individual experimenter variable (hereafter known as I) served as error terms for the F—tests of the main and interaction effects of sex of experimenter (E) , presence/absence of the experimenter (A), and sex of the participant «3). Specifically, the error term for E was I, for A and AxE it was IxA, for P and ExP it was IxP, and fbr AxP and AxExP it was IxAxP; The mean squares for participants (residual variance) served as the error term for all F tests which involved the individual experimenter variable (I, Ixa, IxP, IxAxP). This design resulted in limited power for F-tests of A, E, P and their interactions since mean squares involving I were all based on only four degrees of freedom/ Power for these Fatests was increased by pooling appropriate error tenms and thereby increasing the degrees of freedom. For each analysis the pooled error term was formed by adding the sums of squares for participants (residual variance) and the sums of squares of all effects involving I that were not significant at p<.25. The pooled mean square was then produced by dividing this sum by the sum of the degrees of freedom for each included term. New F—ratios were then computed by dividing the mean squares of the A, E, and P main effects and interactions by the pooled mean square if the F-tests of these effects had been previously based on an error term that was included in the pooling. If one of these Fatests was based on an error term that was not included in the pooling (the F-test for that error term was significant at p<.25), then that F—test was not recalculated using the Pooled 44 term. In this case the orriginal denominator was larger than the pooled error term and therefore~provided a more conservative test. APPENDIX I I I 45 Table 4 Mean Squares and F-Ratios of Time went for Each Drawing PERSCN DRAWING OS PER SCN DRAWING CAR DRAWING SOURCE MS( 5)a F-RATIO MS( 5) F-RATIO MS( 5) F-RATIO E 0.000 .000 0.037 NS 0.020 NS A 1.030 7.12*** 1.588 7.61*** 0.526 6.21*** AXE 0.017 NS 0.072 NS 0.014 NS P 0.462 NS 0.963 4.61* 0.506 NS PXE 0.006 NS 0.079 NS 0.02'.7 NS PXA 0.053 NS 0.039 NS 0.246 NS PXAXE 0.148 NS 0.027 NS 0.089 NS I 0.100 NS 0.201 NS 0.064 NS IxP 0.036 NS 0.107 NS 0.101 NS IxA 0 . 089 NS 0. 109 NS 0 . 050 NS IxAxP 0.137 NS 0.450 2.04# 0.051 NS RESIDUAL 0.157 - 0.220 - 0.089 - a lean square was multiplied by ten raised to the power indicated by the number in the parentheses. * Significant at p=.05. ** Significant at p=.025 *** Significant at p=.01 @ F—test is significantly small at p=.95 # Significant at p=.25, thus not included in the pooling procedure. x F-test is not based on the pooled error term and dbl,4. 46 Table 5 Mean Squares and F-Ratios for Mean Conductance during Fach Drawing PERSCN DRAWING OS PERSCN DRAWING CAR DRAWING SOURCE MS {-7) a F-RATIO MS {-7) F-RATIO MS (-7) F-RATIO E 3.079 NS 0.128 NS 0.386 NS A 1.387 X NS 0.517 NS 0.161 NS AXE 0.060 NS 5.349 NS 1.344 NS P 0.613 NS 0.001 .000 0.446 NS PXE 0.094 NS 0.257 NS 0.285 NS PXA 0.185 NS 2.327 NS 0.827 NS PXAXF 6.917 5.53** 12.358 5.03* 11.124 7.24** I 1.344 NS 3.200 NS 3.285 NS IxP 1.587 NS 2.971 NS 1.294 NS IxA 2.609 1.99# 3.034 NS 1.080 NS IxAxP 0.935 NS 2.703 NS 1.847 NS RESIDUAL 1.307 - 2.376 - 1.637 -— a Mean square was multiplied by ten raised to the power indicated by the number in the parentheses. * Significant at p=.05. ** Significant at p=.025 *** Significant at p=.01 @ F-test is significantly small at p=.95 # Significant at p=.25, thus not included in the pooling procedure. x F-test is not based on the pooled error term and df=1,4. 47 Table 6 Mean Squares and F-Ratios for Frequency Rate during Each Drawing PERSON DRAWING OS PERSCN DRAWING CAR DRAWING SOURCE MS {-2 ) a F-RA'I‘IO MS {-2) F-RATIO MS (-2) F-RATIO E 0.111 NS 0.011 NS 0.065 NS A 0.876 x NS 0.628 NS 0.013 NS AxE 0.613 x NS 0.008 NS 0.065 NS P 1.540 NS 0.000 .008 0.011 NS PxF 0.236 NS 0.434 NS 1.337 NS PxA 0.203 NS 0.002 NS 0.229 NS PXAXE 0. 081 NS 0 . 081 NS 0. 161 NS I 0.328 NS 0.579 2.78* 0.114 NS IxA 1.651 2.78* 0.335 1.61# 0.378 NS IxP 0.321 NS 0.335 NS 0.378 NS IxAxP 0.195 NS 0.121 NS 0.221 NS RESIDUAL 0.594 - 0.209 - 0.372 - a Mean square was multiplied by ten raised to the power indicated by the number in the parentheses. * Significant at p=.05. ** Significant at p=.025 *** Significant at p=.01 @ F—test is significantly small at p=.95 # Significant at p=.25, thus not included in the pooling procedure. x F—test is not based on the pooled error term and df=l,4. 48 Table 7 Mean Squares and F—Ratios of Shading for Each Drawing PERSON DRAMUNG OS PERSON DRAWING CAR DRAWING SOURCE MS( 0)a F-RATIO MS( 0) F-RATIO MS( 0) F-RATIO E 0.940 NS 1.898 NS 0.065 NS A 4.815 8.49*** 4.815 5.28** 6.773 4.28** AXE 0.023 NS 0.128 NS 1.628 NS P 0.940 NS 3.961 4.34* 1.378 NS PXE 0.128 NS 0.211 NS 0.023 NS PXA 0.065 NS 0.003 NS 0.753 NS AXEXP 0.003 NS 2.836 NS 0.211 NS I 0.362 NS 1.021 NS 1.346 NS IXG 0.221 NS 0.312 NS 3.950 2.48% IXA 0.747 NS 0.229 NS 1.576 NS IXAXP 0.253 NS 0.490 NS 4.263 2.67* RESIDUAL 0.605 - 1.001 - 1.596 -— a Mean square was multiplied by ten raised to the power indicated by the number in the parentheses. * Significant at p=.05. ** Significant at p=.025 *** Significant at p=.01 @ F-test is significantly small at p=.95 # Significant at p=.25, thus not included in the pooling procedure. x F-test is not based on the pooled error term and df=l,4. 49 Table 8 Mean Squares and F-Ratios of Erasures for Each Drawing PERSON DRAWING OS PERSON DRAWING CAR DRAWING SOURCE MS( 0)a F-RATIO MS( 0) F-RATIO MS( 0) F-RATIO E 0.585 NS 0.167 NS 1.148 NS A 9.065 9.50*** 0.010 8.82*** 6.253 5.25** AXE 0.065 NS 0.667 NS 0.219 NS P 0.586 NS 4.167 NS 2.190 NS PXE 0.023 NS 0.094 NS 1.378 NS PXA 0.128 NS 0.667 NS 0.753 NS PxAXE 4.378 4.59* 0.260 NS 2.503 NS I 0.224 NS 0.146 .12@ 1.591 NS IxP 0.297 NS 1.130 NS 0.581 NS IxA 0.229 NS 0.292 NS 0.763 NS IxAxP 0.448 NS 0.923 NS 0.565 NS RESIDUAL 1.098 - 1.248 -- 1.261 -— a Mean square was multiplied by ten raised to the power indicated by the number in the parentheses. * Significant at p=.05. ** Significant at p=.025 *** Significant at p=.01 @ F-test is significantly small at p=.95 # Significant at p=.25, thus not included in the pooling procedure. x F-test is not based on the pooled error term and df=1,4. 50 Table 9 Mean Squares and F-Ratios of Delineation Line Absence for Each Drawing PERSON DRAWING OS PERSON DRAWING CAR DRAWING SOURCE MS! 0)a F-RATIO MS( 0) F—RATIO MS( 0) F-RATIO E 0.586 NS 2.042 X NS 1.628 NS A 9.690 7.22*** 3.375 3.60* 0.023 X NS AXE 2.503 NS 2.667 NS 0.003 X .008 P 0.128 NS 0.375 NS 5.273 12.70*** PXE 0.065 NS 1.500 NS 0.023 NS PXA 0.128 NS 0.167 NS 0.440 NS PXAXE 5.753 4.29* 0.667 NS 1.148 NS I 1.219 NS 1.505 1.56# 0.237 NS IXP 0.948 NS 0.844 NS 0.305 NS IXA 1.104 NS 0.302 NS 0.700 1.60# IXAXP 0.416 NS 1.682 1.74# 0.310 NS RESIDUAL 1.434 - 0.965 -- 0.436 - a Mean square was multiplied by ten raised to the power indicated by the number in the parentheses. * Significant at p=.05. ** Significant at p=.025 *** Significant at p=.01 8 F—test is significantly small at p=.95 # Significant at p=.25, thus not included in the pooling procedure. x F-test is not based on the pooled error term and df--l,4. 51 Table 10 Mean Squares and F-Ratios of Body Simplification for Both Human Figure Drawings. PERSON DRAWING OS PERSON! DRAWING CAR DRAWING SOURCE MS( 0)a F-RATIO MS( 0) F-RATIO MS( 0) F-RATIO E 0.211 NS 0.510 NS -— - A 3.190 8.27*** 5.042 11.28*** - - AXE 0.128 NS 1.042 NS - - P 0.065 NS 0.010 NS - - PXF 0.315 NS 0.844 NS - - PXA 0.128 NS 0.042 NS - -- PXAXE 0.234 NS 0.167 NS - -— I 0.286 NS 0.307 NS - - IXP 0.151 NS 0.521 NS - -- IXA 0.120 NS 0.120 NS - - IXAXP 0.036 .080 0.448 NS - - RESIDUAL 0.438 - 0.469 -- -- -- a Mean square was multiplied by ten raised to the by the number in the parentheses. * Significant at p=.05. ** Significant at p=.025 *** Significant at p=.01 @ F-test is significantly small at p=.95 # Significant at p=.25, thus not included in the pooling procedure. indicated x F-test is not based on the pooled error term and df=1,4. 52 Table 11 Mean Squares and F-Ratios of Head Simplification for Both Human Figure Drawings. PERSGJ DRAWING OS PERSON! DRAWING CAR DRAWING SOURCE MS ( 0 ) a F-RATIO MS( 0) F-RATIO MS( 0) F-RATIO E 0. 094 X NS 0. 094 NS -- -- A 2.667 X 4.79* 1.041 NS — ._ AXE 0. 167 X NS 0. 260 NS — -- P 3. 375 X NS 0. 667 NS — .. PXE 0. 375 X NS 0. 260 NS — -- PXA 1.760 4.46* 0.167 NS — - PXAXE 0. 260 NS 0. 094 NS — -- I 0.802 2.02# 0.365 NS — -- IXP 0.594 1.49% 0.385 NS — - IXA 0.557 1.40# 0. 119 NS — - IXAXP 0. 338 NS 0. 130 NS — -- RESIDUAL 0. 398 -- 0 . 441 — —- .. a Mean square was multiplied by ten raised to the power indicated by the number in the parentheses. * Significant at p=.05. ** Significant at p=.025 *** Significant at p=.01 @ F-test is significantly small at p=.95 # Significant at p=.25, thus not included in the pooling procedure. x F—test is not based on the pooled error term and df=1,4. 53 Table 12 Mean Squares and FBRatios of'vertical Imballance for Each Drawing PERSCN DRAWING OS PERSCN DRAWING CAR DRAWING SOURCE MS( 0)a F-RATIO MS( 0) F-RA'I‘IO MS( 0) F—RATIO E 0.000 .008 0.042 x NS 0.327 NS A 0.094 NS 1.760 2.86* 1.260 NS AxF 0.167 NS 0.375 NS 1.402 NS P 0.167 x NS 0.042 NS 3.010 4.07* PxE 0.844 x NS 1.760 NS 0.135 NS PxA 0.167 NS 2.042 NS 1.084 NS PxAxF 1.760 NS 0.010 NS 0.327 NS I 0.490 NS 0.943 1.50# 0.782 NS IxP 1.678 2.41# 0.651 NS 0.916 NS IxA 0.083 .128 0.239 NS 0.725 NS IxAxP 0.167 NS 0.729 NS 1.549 2.139 RESIDUAL 0.696 -- 0.627 -- 0.728 - a Mean square was multiplied by ten raised to the power indicated by the number in the parentheses. * Significant at p=.05. ** Significant at.ps.025 *** Significant at p=.01 @ F-test is significantly small at p=.95 # Significant at p=.25, thus not included in the pooling procedure. x F—test is not based on the pooled error term and dfal,4. 54 Table 13 Mean Squares and F-Ratios of Reinforcement for Each Drawing PERSON DRAWING OS PERSON DRAWING CAR DRAWING SOURCE MS( 0)a F-RATIO MS! 0) F-RATIO MS( 0) F-RATIO E 0.010 NS 1.042 NS 0.042 X NS A 1.260 X NS 4.594 5.16** 1.500 NS AXE 0.010 X NS 0.510 NS 0.375 NS P 0.260 NS 0.094 NS 0.167 NS PXF 0.010 NS 0.094 NS 0.042 NS PXA 0.844 NS 0.167 NS 0.000 .008 PXAXE 0.010 NS 0.667 NS 0.167 NS I 0.721 NS 1.060 NS 1.198 1.69# IXP 0.581 NS 0.383 NS 0.010 .028 IXA 0.940 1.49# 0.779 NS 0.359 NS IXAXP 0.747 NS 0.300 NS 0.349 NS RESIDUAL 0.630 - 0.948 - 0.708 - a Mean square was multiplied by ten raised to the power indicated by the number in the parentheses. * Significant at p=.05. ** Significant at p=.025 *** Significant at pr.Ol 8 F-test is significantly small at p=.95 # Significant at p=.25, thus not included in the pooling procedure. x F-test is not based on the pooled error term and df=1,4. 55 Table 14 Mean Squares and F-Ratios of Hair Shading for Both Human Figure Drawings. PERSON DRAWING OS PERSON DRAWING CAR DRAWING SOURCE MS( 0)a F-RATIO MS( 0) F-RATIO MS( 0) F-RATIO E 0.128 NS 0.375 X NS - - A 4.378 5.21* 3.010 4.22* - -- AXE 2.503 NS 0.167 NS - - P 0.023 NS 0.375 NS - - PXE 0.128 NS 0.094 NS -— - PXA 1.378 NS 0.000 .008 - PXAXE 1.148 NS 0.844 NS - - I 1.130 NS 1.206 1.71# - - IXP 1.536 1.85# 1.117 1.57# - - IXA 1.010 NS 0.518 NS -— -— IXAXP 0.490 NS 0.930 NS - - RESIDUAL 0.832 —- 0.712 - -- - a Mean square was multiplied by ten raised to the power indicated by the number in the parentheses. * Significant at $.05. ** Significant at $.025 *** Significant at $.01 8 F-test is significantly small at $.95 # Significant at $.25, thus not included in the pooling procedure. x F-test is not based on the pooled error term and df=1,4. 56 Table 15 Mean Squares and F-Ratios of Placement fer Each Drawing PERSON DRAWING OS PERSON DRAWING CAR DRAWING SOURCE MS( 0)a E-RATIO MS( 0) F-RATIO MS( 0) F-RATIO E 4.815 5.92** 0.010 NS 1.148 NS A 1.148 NS 1.500 NS 1.148 NS AXE 0.065 NS 0.094 NS 0.315 NS P 0.440 NS 0.667 NS 0.065 NS PXE 0.065 NS 1.260 NS 1.898 NS PXA 0.440 NS 0.042 NS 0.023 NS PXAXE 0.023 NS 0.094 NS 0.586 NS I 0.385 NS 0.260 NS 1.349 NS IXP 1.104 NS 0.104 .138 0.474 NS IXA 0.927 NS 0.156 NS 0.333 NS IXAXP 1.833 2.25# 0.302 NS 0.281 NS RESIDUAL 0.815 -- 0.797 -- 1.095 - a Haan square was multiplied by ten raised to the power indicated by the number in the parentheses. * Significant at $.05. ** Significant at p=.025 *** Significant at $.01 8 F-test is significantly small at p=.95 # Significant at p=.25, thus not included in the pooling procedure. x F—test is not based on the pooled error term and df=1,4. 57 Table 16 Mean Squares and F-Ratios of Line Discontinuity for Each Drawing PERSCN DRAWING OS PERSON! DRAWING CAR DRAWING SOURCE MS( 0)a F-RATIO MS( 0) F—RATIO MS( 0) F-RATIO E 4.594 x NS 0.510 x NS 0.510 x NS A 7.594 x NS 0.510 x NS 0.260 x NS AxF 0.510 x NS 0.010 x NS 0.260 x NS P 0.260 NS 3.010 NS 3.010 x NS PxF 1.260 NS 0.010 NS 0.094 x NS PxA 0.094 NS 0.094 NS 0.010 NS PxAxF 0.094 NS 2.344 NS 0.094 NS I 3.271 2.08# 2.301 1.84 1.823 2.42# IxP 1.167 NS 3.073 2.45# 2.177 2.89* IxA 3.146 2.004 2.010 1.60# 1.135 1.51% IxAxP 0.625 NS 1.156 NS 0.615 NS RESIDUAL 1.573 - 1.254 - 0.753 - a Mean square was multiplied by ten raised to the power indicated by the number in the parentheses. * Significant at $.05. ** Significant at $.025 *** Significant at $.01 @ F-test is significantly small at $.95 # Significant at p=.25, thus not included in the pooling procedure. x F-test is not based on the pooled error term and df=1,4. 58 Table 17 Mean Squares and F-Ratios of Omissions for Each Drawing PERSON DRAWING OS PERSON DRAWING CAR DRAWING SOURCE MS( 0)a F-RATIO MS( 0) F-RATIO MS( 0) F-RATIO E 1.260 x NS 1.042 x NS 1.760 NS A 4.167 x NS 2.042 NS 0.375 NS AxF 0.260 x NS 0.375 NS 0.042 NS P 2.344 x NS 0.375 NS 0.667 NS PxE 2.042 x NS 0.667 NS 0.167 NS PxA 0.844 NS 0.167 NS 0.094 NS PxAxE 0.375 NS 0.000 .008 1.260 NS I 1.982 1.64# 1.776 1.55# 2.318 NS IxP 2.294 1.90# 2.849 2.49# 1.893 2.73* IxA 2.253 1.87# 1.115 NS 0.467 NS IxAxP 0.836 NS 0.802 NS 0.263 NS RESIDUAL 1.206 -- 1.146 - 0.693 - a wean square was multiplied by ten raised to the power indicated by the number in the parentheses. * Significant at p=.05. ** Significant at $.025 *** Significant at p=.01 8 F-test is significantly small at $.95 # Significant at.p=.25, thus not included in the pooling procedure. x F-test is not based on the pooled error term and df=1,4. Table 18 Mean Squares and F-Ratios of Transparancies for Each Drawing PERSON DRAWING OS PERSON DRAWING CAR DRAWING SOURCE MS( 0)a F-RATIO MS( 0) F-RATIO MS( 0) F-RATIO E 1.898 4.76* 0.094 NS 0.167 NS A 0.440 NS 0.260 NS 0.094 NS AXE 0.441 NS 0.375 NS 0.510 NS P 0.441 NS 1.042 NS 0.260 NS PXE 0.211 NS 0.094 NS 0.010 NS PXA 0.003 NS 0.094 NS 0.375 NS PXAXE 0.003 NS 0.042 NS 0.042 NS I 0.521 NS 0.255 NS 0.174 NS IXP 0.505 NS 0.333 NS 0.628 NS IXA 0.010 .028 0.255 NS 0.388 NS IXAXP 0.245 NS 0.271 NS 1.560 2.58* RESIDUAL 0.416 - 0.319 - 0.604 NS a Mean square was multiplied by ten raised to the power indicated by the number in the parentheses. * Significant at $.05. ** Significant at $.025 *** Significant at $.01 8 F-test is significantly small at $.95 # Significant at $.25, thus not included in the pooling procedure. x F-test is not based on the pooled error term and dfi1,4. 60 Table 19 Mean Squares and FBRatios of Body Size for Both Human Figure Drawings. PERSCN DRAWING OS PERSQQ DRAWING CAR MWING SOURCE MS ( 0) a F—RATIO MS ( 0) E-RATIO MS ( 0) F-RATIO E 1.500 NS 0.128 NS -— - A 0.260 NS 0.023 NS - - AXE 0.167 NS 1.378 NS —- -— P 0.167 NS 0.586 NS —— - PXE 0.094 NS 0.753 NS - - PXA 0.667 NS 0.023 —- -- PXAXE 1 . 260 NS 1 . 898 NS — -- I 0.349 NS 0.120 .148 - - IXP 1.161 NS 0.521 NS —— -- IXA 0.401 NS 0.193 NS -— - IXAXP 0.589 NS 0.094 .116 - -- RESIDUAL 1.026 -— 0.858 - -- —- a wean square was multiplied by ten raised to the power indicated by the number in the parentheses. * Significant at p=.05. ** Significant at p=.025 *** Significant at p=.01 8 F-test is significantly small at p=.95 # Significant at p=.25, thus not included in the pooling procedure. x F—test is not based on the pooled error term and dfél,4. 61 Table 20 Mean Squares and F—Ratios of Head Size for Both Human Figure Drawings. PERSQJ DRAWING OS PERSCN DRAWING CAR DRAWING SOURCE MS ( 0) a F-RATIO MS ( 0) F-RATIO MS ( 0) F-RATIO E 3. 010 NS 2. 667 NS — -- A 0. 844 NS 2. 041 NS -— —- AXE 0. 260 NS 1. 042 NS -—- -- P 0.844 NS 0.375 X NS — - PXE 0.260 NS 0.042 X NS — - PXA 0. 094 NS 1 . 500 X NS — —- PXAXE 1 . 760 NS 0 . 667 X NS — - I 0. 698 NS 0. 396 NS - - IXP 1 . 302 NS 2. 521 NS — -- IXA 1 .615 NS 0.042 NS — -— IXAXP 3.552 2.85* 3.083 NS - -- RESIDUAL 1. 247 - 1 . 361 — —— -- a Mean square was multiplied by ten raised to the power indicated by the number in the parentheses. * Significant at $.05. ** Significant at $.025 *** Significant at $.01 8 F—test is significantly small at $.95 # Significant at $.25, thus not included in the pooling procedure. x F—test is not based on the pooled error term and df=1,4. 62 Table 21 Mean Squares and F-Ratios of Head:Body Ratio for Both Human Figure Drawings. PERSON DRAWING OS PERSON DRAWING CAR DRAWING SOURCE MS( 0)a F-RATIO MS( 0) F-RATIO MS( 0) F-RATIO E 0.094 NS 2.042 NS —— —— A 0.844 NS 0.667 NS -— - AXE 0.010 NS 1.042 NS -— -— P 6.510 7.80*** 2.042 NS —- -- PXE 1.260 NS 0.167 NS - - PXA 5.510 NS 3.375 NS -— -— PXAXE 3.010 NS 0.000 .008 - - I 0.208 NS 0.427 NS - - IXP 0.292 NS 0.260 NS - - IXA 0.833 NS 1.010 NS - - IXAXP 1.292 1.44# 2.844 3.15** -— —- RESIDUAL 0.899 -- 0.903 - -— -- a Mean square was multiplied by ten raised to the by the number in the parentheses. * Significant at $.05. ** Significant at $.025 *** Significant at $.01 8 F-test is significantly small at $.95 # Significant at $.25, thus not included in the pooling procedure. power indicated x F—test is not based on the pooled error term and df=1,4. Table 22 Mean Squares and F—Ratios of Vertical and Horizontal Size and Vertical :Horizontal Ratio for the Car Drawing PERSCN DRAWING OS PERSON DRAWING CAR DRAWING SOURCE MS( 0)a F-RATIO MS( 0) F-RATIO MS( 0) E-RATIO E 3.010 NS 2.667 NS 0.007 X NS A 0.844 NS 2.042 N022 NS AXE 0.260 NS 1.042 NS 0.021 NS P 0.844 NS 0.375 X NS 0.125 6.12** PXE 0.260 NS 0.042 X NS 0.046 NS PXA 0.094 NS 1.500 X NS 0.030 NS PXAXE 1.760 NS 0.667 X NS 0.000 NS I 0.698 NS 0.396 NS 0.044 1.98# IXP 1.302 NS 2.521 1.85# 0.004 .178 IXA 1.615 NS 0.042 .030 0.008 NS IXAXP 3.552 2.85* 3.083 2.26# 0.009 NS RESIDUAL 1.256 - 1.361 - 0.022 NS a Mean square was multiplied by ten raised to the power indicated by the number in the parentheses. * Significant at $.05. ** Significant at $.025 *** Significant at $.01 @ F-test is significantly small at $.95 # Significant at $.25, thus not included in the pooling procedure. x F-test is not based on the pooled error term and df=1,4. APPENDIX IV 64 Appendix IV contains brief descriptions of the eighteen sig- nificant E-Tests that were not considered in the main text. Significant Effects Due 51 the Sex 9f the anerimenter. Male experimenters resulted in more, off-centered placement and a greater number of transparancies for the person figure drawings. Significant Effects due £3 the Individual Experimenter. Significant differences occured between individual experimen- ters in the GSR frequency rates of the participants that they tested during the opposite sex person figure drawing task. Significant Participant Sex Differences. Female and Male participants differed on a number of dependent variables. Female participants spent less time than males in drawing the opposite sex person and the car. They also used more shading on the opposite sex person figure, and had fewer delineation line absenses and a larger vertical:horizontal ratio for the car drawing. Finally, female participants had a larger head:body ratio for the person figure, and less vertical imbalance for the car figure. Significant Interactions which Included Presence/Absence. A significant presence/absence by sex of participant (AxP) interaction occured for head simplification on the person drawing. Head simplification was greater for female participants when an experimenter was present, but about the same for female participants when the experimenter was absent and for the male participants regardless of whether the experimenter was present 65 or absent. Significant individual experimenter by present/absent (IxA) interactions occured for CSR frequency rate on the person drawing. One experimenter resulted in a higher rate when he was absent, while four experi- menters resulted in a higher rate when they were present. Frequency rate did not essentially vary across present and absent conditions for the remaining experimenter. Significant individual experimenter by present/absent by sex of participant (IxAxP) interactions occured for the head size of the person figure for head:body ratio on the opposite sex person figure, and for shading and transparancies on the car figure. Since these interactions include the effects of unique experimenters (a non-repl icable factor) and since they are very complex, they will not be described here. Other Significant Interactions (nnot including A as a factor) Significant individual experimenter by sex of participant interactions (IxP) occured for line discontinuity and omission on the car figure. These interactions are also very complex and of little interest since they only occured for the car figure; they will not be described either. MICHIGAN STATE UNIV. LIBRARIES ill1|"ml”Hm”WWWW"VWIWII‘HIWHI 31293107221255