An Investigation of a Simulated Spatial Technique as a Measure of Interpersonal Distancing Behavior

Due to ease of administration, many studies in the area of interpersonal distancing behavior have used simulated spatial measurement t echniques. These have generally incl uded manipulations of inanimate representations of people. They have often been assumed to be equivalent to behavioral measures of interpersonal distancing. The present investigation ' s major objectives were to determine the followings (1) extent to which a simulated spatial measure of interpersonal distancing is related to actual behavior ; (2) whether interpersonal distancing can be understood in terms of a social lear ning model; (J) whether personality variables are related to a simulated spatial measure of interpersonal distancing; and (4) effects of sex differences and body orientation on interpersonal distancing. Sub jects , drawn from undergraduates enrolled in introductory psychology classes , were -administered Rotter's Internality-Externality Scale. According to locus of control scores and sex , four groups were formed a male internals , male externals , female internals , and female externals. Subjects within each of those four groups were randomly assigned to either male or female confederate conditions , yielding the final eight groups (N=lO for each group) . Confederates were chosen so as to match as closely as possible t he age and general appearance of the subject pool. All subjects were administered the Personal Research Form and were individually given two


ABSTRACT
Due to ease of administration, many studies in the area of interpersonal dista ncing behavior have used simulated spatial measurement t echniques. These have generally inc l uded manipulations of inanimate representations of people. They have often been assumed to be equivalent to behavioral measures of interpersonal distancing. The present investigation ' s major objectives were to determine the followings (1) extent to which a simulated spatial measure of interpersonal distancing is related to actual behavior ; (2) whether interpersonal distancing can be understood in terms of a social lear ning model; (J) whether personality variables are related to a simulated spatial measure of interpersonal distancing; and (4) effects of sex differences and body orientation on interpersonal distancing. Sub jects , drawn from undergraduates enrolled in introductory psychology classes , were -administered Rotter's Internality-Externality Scale. According to locus of control scores and sex , four groups were formed a male internals , male externals , female internals , and female externals. Subjects within each of those four groups were randomly assigned to either male or female confederate conditions , yielding the final eight groups (N=lO for each group) . Confederates were chosen so as to match as closely as possible t he age and general appearance of the subject pool. All subjects were administered the Personal Research Form and were individually given two i i experimental tasks : (1) the simulated spatial measure in which they were asked to place pairs of felt figures onto a board; (2) the behavioral measures in which unobtrusive measurements we~e taken of the distance they placed themselves from confederates at side and fr ontal body orientations . For subjects given same sex felt figures , on the simulated spatial measure , confederates in the behavioral measures were also of the same sex as the subjects . For subjects given mixed sex felt figures , the confederates were of the opposite sex , The results indicated that there is no significant relati onship between the simulated and behavioral measures of interpersonal distancing . The findings also provide only partial support for the view that interpersonal distancing may be understood in terms of a social learning model. On the behavioral measure , subjects with an internal locus of control reflected needs related to a desire for i nterpers onal closeness , while subjects with an external locus of control did not , Internals demonstrated closer interpersonal distancing on the simulated measure than externals . However, on the behavioral measure , internals did not maintain closer distance to others than externals, There was only limited evidence supporting the hypothesis that subjects • responses on the simulated measure were related to pe rsonality variables . Distancing at the frontal -body orientation was significantly closer on both the simulated and behavioral measures than for the side -body orientation . On both the simulated and behavioral measures iii there were no statistically significant differences in interpersonal distancing between male and female subjects nor between male and female confederates. Mixed sex felt-figure pairs were placed closer together than same sex pairs. However, there was no significant interaction of sex of subject by sex of confederate on the behavioral measure . Future research and applied implications are discussed.  correlates and measurement, much of the data in this area remains inconsistent and ambiguous. Lett, Clark and Altman (1969) reviewed the major findings in this area and found that "very few relationships are well established" (p. 4).
One major factor contributing to the inconsistent findings in this area, according to Evans and Howard (1973), is · the absence of an accurate, reliable measure. While actual behavioral measures would seem to be most desirable, the indicates that two objects belong together he has employed some schema or plan. If these objects are people or people symbols, the schema employed may be considered, by definition, a social schema" (Kuethe, 1962a, p. Jl There was also a significant tendency to place a child figure closer to a female than a male figure. From the consiste. nt grouping found, Kuethe concluded that there are common social schemas (e.g. people . belonging together). He believes that responses differing from these common social schemas are indicative of "disturbances in normal social· thinking" (p. JS).

Various investigators have used the felt-figure technique
and variants of it as a measure of psychological distance, which has generally been defined as the desire for or perception of interpersonal closeness· (e.g. Weinstein, 1965r Fischer, 19671 Tolor & Orange, 1969'>. To a more limited extent, this measurement technique has also been used as a measure of interpersonal distanc.ing behavior, with the assumption of it being equivalent to actual interpersonal distancing (e.g. Little, 19651 Meisel & Guardo, 19691 Pedersen, 197Jc).
The commonly used term of personal space was first suggested by Sommer (1959) for the phenomenon ·of interpersonal distancing. He defined personal space as "an area with invisible boundaries surrounding a person's body into which intruders may not come" (1969, p. 26). Another often used term is that . of proxemics coined by Hall (1966) for the stµdy of man's use of space. For the purposes of this study the generic term of interpersonal distancing behavior will be used and viewed as encompassing the more specific terms of proxemics, personal space and social schemas.
The main focus of the present investigation will includes (1) determination of the extent .to which a simulated spatial technique, which is essentially a refinement of Kuethe's measure, is directly comparable to actual interpersonal distancing behaviors (2)  investigations have provided at least indirect evidence that 3 simulated measurement techniques and "real life" behavior are in some manner related. Kuethe and Weingartner (1964) report that homosexual males when asked to place male and female figures on a felt field often did not show the normal schemata of man-woman pairings. Homosexual males also tended to place two male figures closer together than man-woman figure pairs. Guardo (1969) (1968) 4 completed a second experiment. Actual distance subject maintained between himself and an experimenter who had previously been described as epileptic was recorded. Results were similar to that reported on the figure placement task. Subjects maintained greater distance from the experimenter described as having epilepsy than to the "normal" experimenter. Levinger and Gunnar (1967) have reported that students who closely paired figural representations of themselves and that of a professor tended to sit in front of the classroom while those who placed greater distance between the representations tended to sit to the rear of the classroom. Tolor (1975) found that the distances subjects place between "self" symbols and symbols representing such social stimuli as policemen, boss, mother, friend, etc. were reported by t he subjects as corresponding to their actual behavior. Thus, in this study the subjects• use of representational space was at least perceived by them as being equivalent to actual behavior.
There have been a s eries of studies which have more directly compared the relationsh~p between simulated and actual measures of interpersonal distancing. Unfortunately, the results are far from conclusive due to differences irt experimental conditions and measures used. In a study with college students, Little (1965) found a very high correlation (,!:=.??) between the students• placement of photographs of male and female silhouettes in various settings, and their placement of actresses recreating scenes from the silhouette task. This is the highest r eported correlati on in the liter ature between a simul ated and behavioral measure of interpersonal distancing.
However, this appears to be an artifact of the measures used in t he study. The behavioral measure is so similar to Little's simulated/pr ojective measure that the subjects may have in essence been performing the same task. The only diff erence between the two tasks seems to be the use of live actresses instead of silhouettes. Gottheil, Corey, and Paredes (1968) photographed through a two-way mirror a seated subject and experimenter. The subject •s physical distance (nose to nose) from the experimenter was measured from the photographs. Subjects were asked to place cylindrical magnets representing "father", "mother", "best male friend", "God", and "interviewer" in relation to a magnet representing "self" according to how c l ose the subject felt towards them. The physical distance from the experimenter and the distance between the "self" and "experimenter" magnet were moderately correlated. again rank ordered their preferred interaction distances .
(J) In vivo participation . Subjects were asked to approach the experimenter from a distance of 15 feet until they were at a distance comfortable for a conversation. Pedersen (1973b) used a simulated measure in which subjects were instructed to place a profile representing another person "as close as is comfortable for you in most situations" to a profile representing "self." The profiles were drawn facing left, front, right, and one with a top view, Two behavioral measures were taken. The awareness measure involved subject approaching, being approached and both subject and experimenter approaching each other as close as possible without feeling uncomfortable. In the "unawareness" behavioral measure subjects were asked to bring two chairs to the center of a room facing each other and to be seated . ~easureme nt was taken of the distance from the front of one chair to the other. In the situation where the simulated task orientation was most similar t o the awareness behavioral measure , the measures were moderately correlated, The simulated measure was not pre dictive of t he "unawareness" behavioral measure.
In contrast to t he above findings, Dosey and ~eisels (1969) and Love and Aiello (1976) found no signi f icant relationship between simulated and behavioral measures, Dosey and Meisels (1969) compared three measures of personal space in an investigation exploring the effect of stress on pers onal space. The measure included: (1) distance subject maintaine d from another subject when asked to approach him or her; (2) distance subject placed a silhouette. representing hims elf, in relation to a printed silhouette of the opposite sex; (J) choice of either a near or far seat from the experimenter. intimacy of conversation, etc., also have an equilibrium point. · As the frequency of these behaviors increases, affiliation needs are also increasingly satisfied. However, after a certain point an increase in these behaviors results in anxiety. They suggest that an equilibrium develops for "intimacy" which is a joint function of eye-contact, physical proximity, intimacy of topic, amount of smiling, etc. They deduced that if one of the aspects of intimacy changes, then in order to maintain the equilibrium point, one or more of the other components will move in the reverse direction. If this is not possible and equilibrium cannot be restored, then avoidance forces will predominate in the situation of too much 12 intimaeys in the situation of not enough intimacy, a person will :feel a lack of affiliative satisfaction.
From this theoretical formu~ation Argyle and Dean derived 12 empirical deductions. They cite Exline•s (196J) study as confirming one of· these deductions. Exline found that eye contact increased when the intimacy level of a conversation decreased. Argyle and Dean (1965) also reported results that support their deduction that interpersonal distance be~een individuals will be greater with eye contact than without and that with less interpersonal distance, eye contact will decrease. Goldberg, Kiesler and Collins (1969) attempted to . replicate Argyle and Dean's (1965) study. They reported that subjects spent less time gazing at the interv~ewer•s eyes when seated at 2 1/2 feet than when seated 6 feet from them. This finding supports Argyle and Dean's theory. However, Goldberg et. al (1969) also state. that their finding can only be viewed as a demonstration that inte~action distance influences behavior.
·rt is not a test of any theoretical interpretation of the data and it does not enable one to sort among various theoretical -ideas .which could account for the data in terms of underlying variables. Argyle and Dean's theory of "intimacy", it should be noted, accounts for these data only by redescribing them in terms of the construct of "intimacy." The theory is not a testable statement of under- These factors of situational context and history of reinforcement are integral aspects of social learning t heory.
There are four classes of variables in social learning theorya behaviors, expectancies, reinforcements and psychological situations. In its most basic form, the general formula for behavior is that the potential for a behavior to occur in any specific psychological situation is a function of the expectancy that the behavior will lead to a particular reinforcement in that situation and the value of that reinforcement (Rotter, 1975, p. 57).
Expectancies are determined by reinforcement histories.
Positive reinforcements in a situation would strengthen the expectancy that positive reinforcements will again occur in 14 that particular situation. This specific expectancy generalizes A major outgrowth of social learning theory is the measurement of generalized expectancies. This has been referred to as internal versus external control of reinforcement or locus of control. Rotter (1966) defines ·this concept as followsa When a reinforcement is perceived by the subject as following some action of his own, but not being entirely contingent upon his action, then, in our culture it is typically perceived · as the result of luck, chance, fate, as under the control of powerful others, or as unpr~dietable because of the great complexity of forces surrounding him. When the event is interpreted in this way by an individual, we have labeled this a belief in external control. If the person perceives that the event is contingent upon his own behavior or his own relatively permanent characteristics, we have termed this a belief in internal control (p. 1).
From these basic assumptions Duke and Nowicki (1972) have postulated that locus of control has a mediational role in interpersonal distancing behavior. Since externals generally perceive a lack of personal control in novel situations, then they should prefer to maintain greater distance from others than internals. Duke and Nowicki's citing of previous research (Tolor, Brannigan & Murphy, 19701 Tolor and Jalowiec, 1968) as supporting their view is quite misleading. In Tolor, Brannigan and Murphy's (1970) study. using a simulated measure, it was in fact reported that for males there was no relationship between inte~personal distance and locus of control orientation. For females, internals placed themselves closer to a stimulus figure representing "sister" than externals. But this finding is contrary to Duke and Nowicki's hypothesis that generalized expectancy would not be a factor in determi. ning interpersonal distance with stimuli with whom there is past experience as there obviously is with a "sister". Even more astounding is that Tolor and Jalowiec (1968) report a non-significant relationship between body boundary (personal space) and internal versus external expectancy.
Other investigations using simulated measures have provided only partial support for the hypothesized role of locus of control in interpersonal distancing behavior. Brannigan and Tolor (1971) used a simulated measure consisting of seven concepts (sister, stranger, fath~r, brother, neighbor, best friend, mother) and presented two concepts paired at a time in all possible combinations. Each pairing was printed on a single sheet of paper. The subjects were instructed to place an "X" in one of five spaces between the two concepts, depending on the degree of closeness felt. For females there was a significant relationship between distance placed between self and parental distance and internality. The relationship for males approached significance.
Duke and Mullens (197J) report that locus of control scores were significantly correlated with interpersonal distance (using a simulated measure) for a schizophrenic and non-schizophrenic patient group, but not for a normal group. Tolor (1975) reported that locus of control was not related to interpersonal distance for females. However, males placement of "self" symbols in relation to "mother" figures on a sheet of paper was significantly related to externality.
Externally-oriented men ·placed themselves farther from "mother" than male internals.

Personality Correlates
Results from investigations relating interpersonal distance to personality variables have been ambiguous in their findings. Lett, Clark and Altman (1969) concluded that it has been well established that personality abnormality is associated with large interpersonal distancing behavior. Horowitz, Duff and Stratton (1964) and Duke and Mullens (197J) both report that schizop:t:irenics maintain greater distance from Qthers than do non-schizophrenic groups. Using Kuethe's social schema technique, Weinstein (1965) found that emotionally-disturbed boys placed human figures further apart than geometric figures, significantly more often than did normal boys. Fisher (1967) reported that normal boys arranged human figures more closely together than did disturbed boys. Similarly, Gerber (1973) found that disturbed boys placed dolls representing themselves and "mother" at greater distances than normal boys. 18 Yet there are also studies which did .not find this relationship between psychopathology and interpersonal distance. For example, Duhamel and Jarmon (1971) found no difference between emotionally-disturbed boys and their siblings in the placement of "mother" and "son" figures. Tolor (1968) and . Tolor, Warren and Weinick (1971) report no difference in distance placement of figures for normal and disturbed c~ildren.
Gerber and Kaswan (1971) report that youngsters with learning difficulties did not place dolls representing family members at greater distances than their "normal" siblings. Tolor (1971), in contrast to Duke and Mullens (197J) and Horowitz, Duff, and Stratton•s (1964) findings, report that schizophrenics placed human figures closer together than normal adults.
Investigations of specific personality dimensions have also reported contradictory findings. Leipold (196J) found that introverted and anxious undergraduates sat further away from the. experimenter than extroverted undergraduates with a lower anxiety level. Males high on the Heterosexual Scale, from the Edwards Personal Schedule, were found by Harnett, Bailey and Gibson (1970) to allow females to approach closer than males low in this scale. Bailey, Harnett and Gibson (1972) also report that the distancing of male subjects was significantly correlated with heterosexual interest1 however, this occurred only under the condition of a male approaching the subjects. For females, anxiety was significantly correlated with distancing when approaching or being approached by a male.
Frankel and Barrett (1971) report that individuals hi gh on a s c ale of authoritarianism and low in self-esteem maintained the largest interpers onal distance fro m others . Tipton , Nailey and Obenshain (1975) found that "traditional" women placed greater di s tance between t hemselves and males than did " feminis t s . " The investigators point out that various studies have shown "feminists" to exce ed "traditional" women in autonomy , aggression , dominance and self-confidence .
In They report no significant relationship betwe en anxiety , body image boundary and any of the measures of spatial behavior .
Similar negative findings have been reported by Meisels and Canter (1970) with introversion and schizophre n ic tendency , Boorem and Flowe r s (1972) found the relationship between a behavioral measure of pers onal space and self-reported anxiety to be quite low . Church (1975) reports no significant correlati on between actual distancing behavior and several personality me asures , including social desirability (as measured by the Ma r low-Crovme) and three factors of s elf-esteem (evaluation concern , self-re gard and interacti on anxiety) . 20

Sex Differences
Effect of Sex of Subject, There is a substantial body of research using both simulated and behavioral measures indicating that males and females differ in their overall interpersonal distancing behavior. Lett, Clark and Altman (1969). in their review article on interpersonal distance, state that it appears "moderately well established" that females in general maintain less interpersonal distance from others than males, Yet Maccoby and Jac· klin (1974) in their extensive review on sex differences state that there is no conclusive evidence which would indicate that there is a general tendency for girls to be "proximity seekers", At least in this one area of possible sex differences, Maccoby and Jacklin reviewed a very limited sample of the literature on interpersonal distancing. Sommer (1967), for example, notes that sex differences in distancing behavior have often been found, He suggests that in our culture females tolerate others at a closer distance than males. The fact is supported, according to Sommer, by the general observation of women holding hands and kissing each other, which is relatively rare among males in our culture.
One . aspect of Guardo•s (1969) study explored where elementary school subjects would place a cutout silhouette figure which was to represent themselves in relation to a pre-printed figure on a page, In a schoolyard setting, the experimenter described different degrees of acquaintance, liking or thr~at and then asked, "Where would you be standing?", 21 The average age of the 60 subjects was 11 years, 7 .months.
Guardo reports that girls place significantly less distance between the figures in situations described as with a "best friend" and "someone you like very much." However, girls had significantly greater distance between the figures in a situation with "someone you're afraid of." These results were explained in terms of sex appropriate behavior. Pedersen (197Jc),using a . simulated measure of personal space, asked 1J2 children in grades one through six to place a movable profile which was to represent themselves from another profile representing another person "as close as possible so that the subjects still felt comfortable." The profiles represented a man, woman, boy and a girl. He reports that across all grade levels and stimulus persons, girls placed the profile representing them~elves significantly closer to the other profile than boys did. Pedersen stated that his findings indicate that females have smaller personal space zones than males · and that this sex difference emerges during the years they are in elementary school.
Interpersonal distance in White•s (1975) study was measured by the "hori zontal nose to nose distance between two seated speakers." A subject and a confederate were asked to seat themselves and to discuss the counseling program at the college. The confederate would place his chair in a predetermined position thus allowing the subject to place his chair at a distance of his choice from the confederate. The actual measurement was taken by determining chair to chair 22 distance and adding one of three correction factors, depending upon whe ther the subj ec t was leani ng forward, s itting upright or l eaning back. For this sample of 80 college fres hmen the s ex of the s ubject was an import ant factor in interpers onal dis tancing. Female subjects sat closer to the confederate than the males, regardless of the sex of the confederate.
Effect of Sex of Confederate. Several studies using behavioral measures have indicated that regardless of sex of subject , people maintain less distance from females than males. Horowitz , Duff and Stratton (1964) found that both male and female adults maintained greater interpersonal distance at "body buffer zones" as they called it , when approaching men than women . Lomranz, Shapira, Choresh and Gilat (1975) investigated the personal space of J-, 5-, and 7-year old children as they approached both boys and girls in their own age range. Each child entered a room where another child was already seated on a bench and drawing, and was told to sit next to the seated child. It was reported that for all age s both boys• and girls' personal space zones were smaller when they approached girls than boys • .
Similarly, Eberts and Lepper (1975) found when measuring the distances that preschool children approached an experiment er seated at a table, that they more closely approached the female than the male experimenters. There was no effect for sex of subject nor a sex of subject by sex of experimenter interaction .
These findings were explained by the investigators as resulting from both boys and girls at this early age receiving more 23 nurturance and intimate contact from female than male adult figures. Barrios, Corbitt, Estes and Topping (1976) examined the effect that social stigma has on interpersonal distancing as well as the effect of sex of the interacting persons.
Using chosen seated distance as the measure of interpersonal distance, they reported that both males and females chose to sit closer to the female confederates than the male confederates.
Interaction of Sex of Subject bY Sex of Confederate. The findings from studies using simulated measure of interpersonal distance provide evidence that opposite sex pairs. maintain smaller distances between themselves than same sex pairs.
The series of studies by Kuethe and associates have consistently found that, when subjects are instructed to place felt cutouts of human figures anywhere they wish on a board, malefemale felt-figure pairings were placed closer together than same sex pairings (Kuethe, 1962as 1964(Kuethe, 1962as , 1964aKuethe & Stricker, 196Ja Kuethe & Weingartner, 1964). Meisels and Guardo (1969) instructed subjects in grades three to ten· to trace a manipulable cutout figure representing themselves to a series of preprinted figures. They report a consistent developmental pattern of children in the later grades, beginning about grade six, of having overall smaller personal space zones. There is also a shift to closer distance being maintained with the opposite sex. Tolor and Salfia (1971) asked 160 male college students to place either two male or one female silhouette figure anywhere they wished on a felt 24 board. The figures were described as having either positive or negative attributes (e.g. high intelligence, low intelligence). Male-female figure pairs were placed significantly closer together than same sex figures when . given a positive set.
In contrast. studies using behavioral measures have generally not supported the above findings of smaller interpersonal distances being maintained by mixed sex pairs. the studies appears due to the differing types of stress induced and their probable differential impact depending on sex of subject.
The inclusion of "stress" as a variable in these studies certainly makes it mere difficult to determine sex differences in interpersonal distancing. In Dosey and Meisel•s (1969) and Bailey, Hartnett and Gibson's (1972) studies, the effect of "stress" on interpersonal distancing differed while Meisels and Canter (1970) found no effect of stress on distancing. It is apparent that the often expressed view that mixed sex pairs have smaller personal space zones than do same sex pairs, have no. t been clearly demonstrated in studies using behavioral measures. This may be due to confounding variables included in such studies as well as methodological weakness which will be more fully discussed later.

Body Orientation
There is convincing evidence from many investigations using behavioral measures that body orientation (i.e. degree communicator's body is turned in the direction of the person he is addressing) is a major factor influencing interpersonal distancing. In one of the earliest studies incorporating this variable. Sommer (1959) asked male and female subjects to sit down at a table and discuss a topic with an already seated confederate. Females preferred sitting alongside a female confederate significantly more than the male subjects or than either sex with male co_ nfederates. Males tended to sit opposite both male and female confederates. Subsequently. Sommer (1967) in a general review article on spatial arrangements of people stated that "side-by-side seating, which is generally considered to be the most intimate of all seating arrangements for people already acquainted, is comparatively rare among males if they are given the opportunity to sit across from one another" (p. 149). Horowitz, Duff and Stratton '(1964) instructed subjects to simply walk towards either a male or f emale experimenter.
The subject approached the "object person" at eight different angles of approaches. When the subject stopped moving forward, at his or her accord, the distance between the subject's feet and the object person was noted. The investigators report that while females more closely approached others in a frontal orientation (i.e. face to face) than sideways, males more closely approached others sideways than frontally. This 27 difference was found regardless of the sex of the person being approached. Mehrabian (1969) reviewed the research findings on the effects of posture and gesture in the communication of affect.
Generally, the findings indicate that for females the least direct orientation occurs in situations where the person being addressed is intensely disliked, the most direct orientation for neutral persons and moderately direct orientation where the person being addressed is intensely li'ked.
The only consistent finding for males is that when a person being addressed is intensely liked, a less direct body orientation tends to occur.
In Pellergrini and Empey•s (1970) study, a confederate was already seated in a room when the subject was brought in and told he was to describe himself to this other person.
The subject was then told "pull up a chair and go ahead whenever you're ready." While there was a general tendency for all subjects to turn a~ay from a face-to-face orientation, the closer they sat to the confederate, female subjects to a significantly greater extent, sat-further away from a face-toface orientation than males. Patterson, Mullens and Romano (1971)  were defined as those "which would serve to lessen the presence of the intruder," such as placing elbow or hand between himself and the intruder. There was no effect of sex of subject on any of the dependent measures; however, females demonstrated more blocking than males at the closest adjacent positions while males showed more blocking than females in the face-to-face position.
In a similar study, Fisher and Byrne (1975). examined sex differences in response to spatial intrusions in both adjacent and across seated positions. A major methodologica;i_ improvement from Patterson et. al•s (1971) study was the use of both male and female col)federates in initiating spatial intrusions. Sex differen~es reported were related to the spatial relationship with the "invader,•• Males felt more negatively toward the "invaders" who sat across from them than did females. But when the "invaders" sat adjacent to the· subjects, females felt more negatively towards them than males. 'Males were also found to erect more barriers . (e.g. books, personal effects) in the across position while females · more often erected adjacent barriers. Contrary to prediction, the sex of the "invader•• did not have a differential effect on any of the dependent measures. The lack of a sex of "invader" effect was · felt to be a result of the asocial library setting of the study. Since the library generally has a work orientation, the sexual identity of someone approaching a person 29 may not be relevant. The investigators suggest that in a more social situation, sex of an "invader•• would be significant.
Bryne, Baskett, and Hodges (1971) provide further evidence that a side-by-side as compared to a frontal seating orientation have different meaning for males than females. Differences in attraction, defined ~s likeability and desirability as a work partner as measured by a paper and pencil test, were related to distancing in a side-by-side seating situation for females1 for males. it was related to choice of a face-toface seat versus a non-facing seat.
One of the few studies using both a simulated and behavioral measure of personal space and incorporating spatial orientation as a variable was that of Pedersen (197Ja). Subjects were asked to place a gum-backed -profile representing another person, in relation to a profile, representing the subject as close as would be comfortable for the subject. These were to be placed on a line radiating fr~m the profile at nine different angles. Hartnett, Bailey & Gibson, 19701 Pedersen, 197Ja). However, there is evidence that subject awareness that interpersonal distancing behavior is being measured ~lters his spatial behavior (Eberts, 1972). There is also evidence that eye contact (Argyle andDean, 1965r Goldberg, Kiesler andGollins, 1969), body orientation (e.g. Horowitz, Duff & Stratton, 19641 Bryne, Baskett & Hodges, 1971;Fisher and Byrne, 1975), sex o.f subject (e.g. Sommer, 1967;Pedersen, 197301 Guardo, 1969) and

HyPotheses and Predictions
The central hypotheses of this study area (1) Responses on the simulated measure of interpersonal distancing are related to personality variables.
(2) Locus of control mediates the extent that personality variables are related to the behavioral measure of interpersonal distancing.
(J) Locus of control determines the .extent that the simulated and behavioral measure of interpersonal distancing are comparable.
(4) Locus of control has a mediational role in determining individuals' interpersonal distancing behavior.
Specifically, the following predictions are proposeda (1) Needs for affiliation, autonomy and dominance will be related to responses on the simulated measure of interpersonal distancing.
(2) Subjects with an internal locus of control will reflect needs for affiliation, autonomy and dominance to a greater extent on t he behavioral measure of interpersonal distanc ing than subjects with an external locus of control.
()) The simulated and behavioral measure of interpersonal distancing will be related to a greater extent for sub j ects with an internal locus of control than for subjects wi th an external locus of· control.
(4) Subjects with an internal locus of control will demonstrate closer interpersonal distancing behavior on both the simulated and behavioral measure than subjects with .an external locus of control.
In view of the somewhat exploratory nature of this study, the fol l owing suppositions are presented in a non-specific formats (5) Distancing on both the simulated and behavioral measures will differ between side and frontal orientations.
(6) Sex of subject and sex of confederate will differentially affect interpersonal distancing on both the simulated and behavioral measures. According to locus of control scores and sex, four groups were formed& male internals, male externals, female externals , and female internals. Subjects within each of those four groups were randomly assigned to either male or female confederate conditions, yielding the final groups (N=10 for each group).
The groups were male -internal-male , male -internal-female, male -external-mal e , male -external-female , female -internal-male , female -internal-female , female-external-male, and femaleexternal-female . The mean age in months and I -E scores for eac h of the experimental groups are presented in Table 1 .
There were no significant differences in I -E scores or age among the four groups with an internal locus of control nor among the four groups wi th an external loc us of control . However , there was a significant difference between the male -externalmale and the female -external-fe male group .
The total mean I -E score of 11.77 for the initial subject pool of 241 students appears representative of the present college population. W hile the mean I -E score s reported by In order to control f or possible confounding effects of confederate variability , the final c ho i ce of two male and two female confederates was done on the basis of an observermatching procedure . The potential. confederates were rated by two observers on five 5-point scales . Taking into account   (Rotter, 1975). Comprehensive reviews of these studies are supportive of the construct validity of this instrument with the internality-externality concept operating in many different situations (Joe, 19711 Lefcourt, 19661 Phares, 1976;Rotter, 1966Rotter, , 1975 The instructions for the I-E were the standard ones provided by Rotter (1966). Subjects were subs. equently contacted 4J by phone and scheduled in groups of six to complete the PRF.

5.55
Instructions for the PRF were the standard ones provided by Jackson (1967). While t he subjects were completing the PRF, Each subject thus received a score for manipulation of felt figures in a frontal and a side orientation. When the experimental procedure was completed, each subject was asked (1)  Behavioral Measure--Side Orientation, An attempt was made to develop an unobtrusive measure of how subjects distance themselves from strangers. Measures were taken in a room 17! feet (5.3 m) by 15 feet (4,6 m) and empty with the exception of a bench 15 feet (4,6 m) long at one end of the room.
Behind the bench was a two-way mirror stretching across the entire wall, In an adjacent room it was possible by way of the two-way mirror to observe where subjects seated themselves.
A tape marked off in centimeters was placed across the entire length of the two-way mirror.
Each subject was met by the experimenter outside what was called the "waiting room" and given the following instructions• "I would like you to fill out this biographical data sheet in the waiting room. I will return in a few minutes to bring you to another room where we will complete the last phase of this study. One of my assistants is in this room working on some papers related to the experiment," For subjects who were given same sex felt figures on the simulated spatial measure, the confederate seated in the waiting room was also of the same sex as the subject, For those subjects given mixed sex felt figUres, the confederate was of t~e opposite sex. Thus, the behavioral measure paralleled as closely as possible the simulated spatial measure.
The confederate was seated one-foot from the end of the bench (me.asured from the mid-line of his or her torso to the edge). When the subject entered the room the confederate was instructed to smile., glance at the subject, and then to resume "marking" papers. The confederate avoided eye contact with the subject while he or she was seating themselves. Once seated, any questions the subject asked the confederate were answered directly, but conversation was discouraged. Since For subjects who were given same sex felt figures on the simulated measure., the confederate was also of the same sex as the subject. For those subjects given mixed sex felt figures, the confederate was of the opposite sex. The confederate was instructed to sit upright with his feet flat on the ground between the legs of the chair. Both the confederate •a chair and the one the subjects used were armless wooden chairs normally found in school settings. The "experimental room" was 11 J/4 feet (J.6 m) by 9 1/4 feet (2.8 m).
The floor of ·this room was composed of tiles 12 inches (J0.48 cm) square. It was therefore quite easy to place markings on the floor in it-inch (J.81 cm) intervals. While these markings were clear enough for the experimenter to use in estimating subject distance; they were sufficiently camouflaged so that the subjects were not aware of them. Besides the chair the confederate was sitting in, the only other chair in the room was placed in a far corner.
Each subject was brought by the experimenter from the "waiting room" to the doorway of the "experimental room" and tolda "For the next phase of this study, please pull up a chair in front of and facing (confederate•s name) and sit down." If the subject asked how far should he place the chair, the experimenter responded ''wherever you wish." The subject thus had to retrieve the only available chair in the room and place it some distance from the confederate. The confederate was instructed to glance up to the subject when he or she entered the room, smile, and the-n resume "marking" papers. The confederate avoided eye contact with the subject while he or she was placing their chair. In order to prevent subject awareness that distance measurement was the variable of interest, once the subject was seated the following instructions were given. "I am going to hand you a list of word pairs 50 which you are to attempt to memorize in one minute. When I say time is up, hand the list over to (confederate•s name).
When I say begin, try to recite, in any order, the word pairs.
(Confederate•sna.me) will prompt you if you forget a word.
You are to go through the list as many times as necessary until you successfully recite it without prompting" (see Appendix for word list). While the subject was memorizing the word list the interpersonal distance from the experimenter was determined by estimating the horizontal dist ance bet ween the front legs of the subject and confederate's chair through the use of the floor markings.
The reliability of these estimates was pretested by having an individual role play the subject and place his chair at different distances from the confederate. After several practice sessions, it was possible to accurately · estimate distances in approximately five to ten seconds. The experimenter and a confederate taking measurements of 20 different placements of a volunteer role playing the subject obtained a correlation of .99 between estimates of distance and actual · distance taken by a tape measure within 1/2-inch ( 1. 27 cm).

Post-Experimental Questionnaire
When subjects completed the memory task, they were asked to complete a questionnaire (see Appendix D). It included rating both confederates that they had met on the attributes of size, strength, aggressivity, bearing and attractiveness.
The experimenter stressed the necessity for the sub j ect to give honest ratings and promised that the confederates would not s ee their ratings. They were also asked the question , "What did you think this experiment was all about?" Upon completing the questionnaire they were partially debriefed. Nearly all the subjects stated that they had no idea. None stated that the purpose of the study dealt with interpersonal distancing behavior.

Intercorrelations Among All Variables
A series of correlational analyses were computed between the measures of interpersonal distancing. personality variables and subjects• ratings of confederates. The following abbreviated eorrelational matrix tables were constructeda Table J for all subjects (n=80), Table 4 for males (n=40), Table 5 for females (n=40), Table 6 for externals (n=40), and Table 7 for internals (n=40) (see Appendix for complete correlati onal matrix tables).     The behavioral measures of interpersonal distance were also significantly correlated with va.rious personality factors .
The strongest relations~ip between the behavioral measures and personality variables are seen in Table 7 for the internals, The behavioral measure at the side orientation, for subjects with an internal locus of control , was negatively correlated with Affiliation (£=-. 43 ) and positively correlated with Autonomy (£=.40) , In contrast , as seen in Table 6 for subjects with an external .
locus of control , the behavioral measure at the side orientation was significantly related only to Impulsivity (£=.37). Table 3 indicates that for all subjects the behavioral measure of side orientation was significantly correlated with Impulsivity (£=.37) and externality (_r=.26). Table 4 indicates that for males this measure was also correlated with Impulsivity (_r=.40). For females, as shown in Table 5, the behavioral measure for side orientation was negatively related to Affiliation (.£=-.J8) and positively related to aggression (£=.J4) as well as Impulsivity (r=.J6).  personal distance scores is presented in Table 8.
As can be seen in  Table 8 Summary Table of Analysis   The  Table 9.    Table 10.
The mean distance scores for side and frontal orientations were 174.21 cm and 88 . 89 cm respectively. The interaction of sex of subject by sex of confederate by orientations was found to be significant (F (1 , ?2) = J . 99, E_~.05). However, Table 10 Summary     There was no significant correlation between Affiliation and Dominance and responses on the simulated measure.
The second hypothesis was concerned with whether locus of control determines the extent to which personality variables are related to actual interpersonal distance. Specifically, it was predicted that for subjects with an internal locus of control needs for Affiliation, Autonomy and Dominance will be more rela ted to t he behavioral measure of inte rperso nal d i stancing than they will be for subjects with a n ex ternal locus of c ont rol. The results gene rally supported t his hypothes is.
Interpersonal distancing on the behavioral measure at the side orientation for internals was positively related to Aut onomy , ne gativ ely related to Affiliation , and no significant relatio ns hi p to ominance. There were no significant relationships betwee n externals' i nterpersonal distancing on the behavioral meas ure and Affiliation , Autonomy and Dominance .
The third hypothesis was concerned with the extent to which a subject's locus of control determines the relationship between a simulated measure and actual interpersonal distancing behavior , It was predicted that the simulated and behavio r al measure of interpersonal distancing would be related to a greater extent for subjects with an internal locus of control than subjects with an external locus of control . The results did not support this hypothesis , There were no significant correlations between the simulated and behavioral measure s for either internals or externals .
The fourth hypothesis stated that locus of control has a mediational role in interpersonal distancing. It was predicted that subjects with an internal locus of control would demonstrate closer interpersonal distancing behavior on both the simulated and behavi oral measures than subjects with an external locus of control . The results partially supported this hypothesis. On the simulated measure , internals placed the felt fi gures significantly closer together than did the externals .
Although there was a trend on the behavioral measure for 76 i nternals to distance themselves clos er t o others than externals , t his f i ndi ng did not reach stati stical s ignificance.
Due to the somewhat exploratory nature of the present s t udy , hypothes es five , six and seven were presented i n a nonspec ific format. According t o hypothes i s five , it was predicted that distancing on both the simulated and behavioral measure would differ between side and front al orientation.
The results supported this hypothesis. Distancing at the frontal orientation was significantly closer for both the simulated and behavioral measure than for the side orientation.
Hypothesis six predicted that sex of subject and conf e derate  given to the subjects).
Another possible limitation was the use of the Rotter I-E Scale. Major reviews of the research literature have indicated that there is support for the view that locus of control needs to be studied at a multi-dimensional rather than unidimensional level (Joe, 1971;Phares, 1975). Factor analysis of the Rotter I-E scale has generally revealed a dimension relating to personal habits, traits and goals and another dimension concerned with social and political events. It would certainly seem that a measure which could differentiate between these two dimensions would increase its predictive power for certain situat.ions. Optimally, the more specific a measur~ of expectancy for particular type of situation, the greater its potential predictive power.
Constructing a separate measure for every specific purpose would be a most difficult task. A measure of generalized expectancy which could distinguish between the two dimensions of personal versus political control would be easier to develop and a decided refinement of the tool. It would seem that factor analysis of the I-E is only a first step in the development of an improved instrument. There is at this time little support for the predictive utility of these separate factors 80 and good evidence supporting Rotter's contention that most of the variance is accounted for by one general factor (Rotter, 1975 The preponderance of evidence from previous· studies indicates that interpersonal distancing is influenced by sex. The present finding that mixed sex felt-figure pairs are placed significantly closer together than same sex pairs is consistent with studies using simulated measures (e.g. Kuethe, 1962a;Meisels and Guardo, 1969;Tolor and Salfia, 1971).
That the results of the behavioral measure did not support the above finding is also consistent with previous investigations (e.g. Dosey and Meisels, 1969;Bailey, Hartnett & Gibson,1972;Eberts and Lepper, 1975). It would appear that a generally held belief among people in this culture is that a man and a 81 woman belong closer together than either two men or two women.
This belief is clearly expressed in the simulated measure.
There are, however, many factors which would determine whether this cultural view is expressed behaviorally. For instance, the degree of acquaintance between individuals and the setting of their interaction appears to be crucial factors. While the behavioral measures were taken among strangers and in a task~oriented setting, one can only guess as to the assumptions subjects held concerning these factors on the simulated measure.
The present finding that the simulated and behavioral measure of interpersonal distancing were not related has also often been reported in previous investigations (Dosey and Meisels, 1969;Love and Aiello, 1976). One of the major methodological problems with previous studies which have found significant correlations between simulated and behavioral measures has been in their not controlling for subject awareness of the task (e.g. Little, 19651 Haase and Markey, 1973 it appears to probably be tangential. Hall (1966) has pointed out that in our distancing behavior we may utilize depth perception, olfaction, heat radiation and audition. These are very different sensory cues then that are available in a simulated measure of interpersonal distancing. This lack of congruence between the simulated and behavioral measure of interpersonal distancing is similar to the often cited problem of t r aditional personality assessment and behavioral prediction (Mischel, 1968). It is generally recognized that projective instruments have so little predictive validity that their sole use in predicting future behavior is completely unjustified (Anastasi, 1961). A similar conclusion regarding interpersonal distancing may also be warranted. The phenomenon of i nterpersonal distancing may best be investigated by recognizing that there is a cognitive and a behavioral dimension. Each of these dimensions require separate researc h efforts before they can be adequately defined and integrated into a comprehensive ·theoretical framework.