DOES FIRO-B RELATE BETTER TO INTERPERSONAL OR INTRAPERSONAL BEHAVIOR?

JOHN R. HURLEY

Michigan State University

After meeting for 33 hours over 7 weeks, 64 undergraduates from 1 1 small interpersonal skills groups rated themselves on Schutz’s (1958) Fundamen- tal Interpersonal Relations Orientations-Behavior (FIRO-B). Three weeks and 17 more group interaction hours later, they also described each same- group participant, including self, on Lorr and McNair’s (1963) Interpersonal Behavior Inventory (IBI). Correlations between self-ratings on 6 FIRO-B and 15 IBI scales yielded 25 significant statistically (p < .05) values, but merely 5 among FIRO-B’s 90 parallel correlations with individual’s mean IBI ratings from pooled small group peers. Of all 30 significant correlations, 19 linked FIRO-B’s overlapping affection and inclusion measures positively, but narrowly, with 4 IBI scales that address affiliativeness/sociability. The findings challenge Schutz’s (1958) paradoxical claim that FIRO-B validly assesses interpersonal behavior by an intrapersonal method.

Schutz (1958) presented the Fundamental Interpersonal Relations Orientations- Behavior (FIRO-B) as a tool for assessing interpersonal conduct. This seems clear in several introductory statements: “The primary purposes for developing FIRO-B are (1) to construct a measure of how an individual acts in interpersonal situations [italics added]”; “for this form of FIR0 it was decided to concentrate on how a person behaves rather than how he feels”; and “[it] simply represents the thing being measured, how an individual characteristically relates to other people [italics added]” (p. 58). Because FIRO-B is limited to self-reports, these interpersonal aims seem paradoxical. A recent study (Hurley, 1989a) of the behavior of participants in small groups found that FIRO-B’s six scales had surprisingly weak correlations with the principal factors from two other interpersonal inventories. Both latter instruments, Lorr and McNair’s (1963) 140-item Interpersonal Behavior Inventory (IBI) and Hurley’s (1989b) scales of acceptance vs. rejection of self and acceptance vs. rejection of others for rating conduct in small groups, yield self- and peer-based measures of affiliation and dominance, the two most central dimensions of interpersonal conduct (Conte & Plutchik, 1981; Foa, 1961; Hurley, 1980; Merenda, 1987; Wiggins, 1982). It also was found that five of FIRO-B’s six scales formed an isolated factor composed of measures more strongly interconnected than any linked with a factor from either of these other inventories (Hurley, 1989b). The present study more broadly surveys FIRO-B’s validity by examining its correlations with ratings of individuals by self and by their small group peers on the 15 substantive scales of IBI’s third revision (IBI-4).

METHOD Subjects and Procedure

For two consecutive terms FIRO-B was administered in the seventh week of 10-week terms to all 76 students enrolled in an elective psychology course at a large midwestern university that featured about 50 hours of small group participation. More fully described elsewhere (Hurley, 1986, 1989b), this course was concerned with enhancing students’ awareness of their own interpersonal style, so reviewing their FIRO-B responses seemed pertinent. Enrollees were permitted to respond to FIRO-B anonymously, but only 1 made this choice. Each person subsequently received detailed feedback about her/his FIRO-B

454

FIRO-B’s Intrapersonal and Interpersonal Correlates 45 5

scores in the context of classroom discussion of these measures’ meanings, limitations, and possible implications.

Intended to assess distinctive expressed (toward others) and wanted (from others) facets of Affection, Control, and Inclusion, high (.93 to .94) internal consistency was claimed for FIRO-B’s six scales (Schutz, 1958), although their 4-week test-retest stability later was reported to average about .69 (Gilligan, 1973). Prior reports of the construct validity of FIRO-B’s scales are conflicting (Froehle, 1970; Gluck, 1979; Kramer, 1967; Ryan, Maguire, & Ryan, 1970). Because four of its six scales, Affection and Inclusion expressed and wanted, overlap notably, as shown by their median correlation of .55 in Schutz’s (1958, p. 80) initial report and .56 in DiMarco, Kuehl, and Wims’ (1975, p. 210) larger sample ( N = 467), FIRO-B has a special problem. The ambivalence of Bloxom’s (1972) Mental measurements yearbook review of FIRO-B was continued in Lifton’s (1 985) cautious commentary in that publication’s later issue, “FIRO-B’s scores should be interpreted judiciously . . . due to the less than certain psychological mean- ing of the ‘FIR0 types’ derived from the various FIRO-B subscales” (p. 579). FIRO-B remains widely used, however, and averages about 25 Social science citation index cita- tions annually by my count.

At the end of these small groups, about 3 weeks after collecting FIRO-B data, IBI-4 also was offered to volunteers as an alternative to required written assignments. Of all enrollees, mostly college seniors and juniors, 1 1 chose to not complete IBI-4. Intended to comprehensively assess the interpersonal behaviors addressed by Leary’s (1957) interpersonal circle, the construct validity of IBI-4’s 15 specific scales appears well sup- ported by factor-analytic studies (Bochner & Kaminski, 1974; Lorr & Suziedelis, 1969). Wiggins’ (1982) general review of interpersonal inventories noted that, “on both substan- tive and psychometric grounds . . . [IBI-41 appears to be a useful clinical device for assessment of patient characteristics” (p. 15). Differing from FIRO-B’s exclusively self- report format, IBI-4 often has been used to collect others’ perceptions of target persons. In the present work, each person was described by self and by at least three other participants from the same small group, although an average of six such peers contributed to these mean ratings. Pooled ratings from knowledgeable peers are usually more inter- nally consistent and stable than self-ratings and, after a review of the pertinent literature, were characterized by Lewin and Zwany (1976) as “superior to all other measures available” (p. 430).

RESULTS In addition to descriptive statistics, Table 1 gives all product-moment correlations

of FIRO-B self-ratings with IBI-4 self-ratings and, separately, with group peers’ mean ratings for 64 persons. On prosocial measures, these small group participants rated themselves rather favorably; self-ratings exceeded peers’ rating significantly (p < .01 by the two-tailed test for paired measures) on IBI-4’s three most blatantly prosocial scales (Nurturance, Agreeableness, and Affiliation). In contrast, only on the asocial Inhibi- tion scale did peers rate individuals above self-ratings significantly (p < .05, t-test for paired measures). Correlations between the ratings of individuals by peers and by self on IBI-4’s scales averaged .38 and ranged from .61 for Sociability to - .12 for Deference. Only for Deference, Dependence, and Mistrust did IBI-4 ratings by self and peers not correlate significantly (for p < .05 by the one-tailed test, r < .21).

The IBI-4 scales are arranged in Table 1 in factorial sets including unipolar Control, Nurturance, and Dependence and bipolar Sociability as identified from ratings of 290 “ ‘normal’ men and women representing a diversity of occupations and social classes” (p. 125) by acquaintences of at least 1 year (Lorr & Suziedelis, 1969). Within each factor, the scales were ordered by factor loadings; all surpassed 3 3 except for .38 for Deference. The one other exception concerned Inhibition, a scale that loaded equally on factors

Tabl

e 1

Des

crip

tive

Stat

istic

s an

d Pr

oduc

t-mom

ent C

orre

latio

ns of

FIR

O-B

's Si

x Sc

ales

with

IBI-

4 R

atin

gs b

y Se

rf (S

) an

d Pe

ers

(P)

IBI-

4

~

M:S

D

5.2:

2.5

6.4:

2.3

4.6:

2.1

5.6:

2.9

3.2:

2.3

3.2:

1.9

A

FF:e

xp

AFF

:wan

1N

C:e

xp

1NC

:wan

C

0N:e

xp

C0N

:wan

Se

lf Peers

SP

S

P

SP

S

P

SP

S

P

Con

trol

C

ompe

titiv

enes

s D

omin

ance

Ex

hibi

tion

Hos

tility

M

istru

st

Nur

tura

nce

Nu r

t u r a

n c e

Agr

eeab

lene

ss

Aff

iliat

ion

Def

eren

ce

Dep

ende

nce

Dep

ende

nce

Subm

issi

vene

ss

Aba

sem

ent

Soci

abili

ty

Det

achm

ent

Soci

abili

ty

Inhi

bitio

n

21.4

:4.0

20

.4:3

.9

16.9

:3.6

18

.1:4

.0

19.9

:3.4

30.9

:4.9

24

. k2.

9 21

.0:3

.5

~ -

24.0

:4.0

21.1

:4.0

20

.2:4

.6

19.3

:4.8

22.5

:5.4

25

.7:6

.0

14.8

:4.0

-

20.4

:4.5

19

.6:4

.2

16.4

:4.4

17

.7:4

.4

20.4

:3.6

28.6

:4.4

23

.7:2

.8

19.3

:3.3

23

.4:3

.3

__

-

-

20.4

:2.6

2 1

.0:5

.O

18.6

:3.7

22.6

:4.6

24

.7:5

.7

16.0

:3.9

-

- .0

5 .0

7 .1

0 .09

.22

.19

-.18

.0

5

-.12

-.11

.12

.08

.27*

.0

1 .4

2*

.16

.09

.02

.03

-.12

-.14

.04

-.I2

.oo

- .3

5* - .3

9*

.50*

.2

5' - .3

4* - .2

5*

.10

.16

.03

.02

.ll

.15

.13

.22

.22

- .0

3 .2

4 .1

2 .1

3 .1

9 .1

6 .OO

.18

.20

.03

.12

-.19

-.lo

.0

1 .1

3 .O

O -.

02

-.21

-.24

.09

.09

.18

.ll

.06

.23

.10

-.08

.23

.08

.22

.25

.ll

-.11

.3

7*

.I5

.27*

.2

0 .2

7*

.04

.I9

.01

-.09

-.01

.06

.02

.ll -.08

.oo

.oo

.14

-.02

-.07

-.04

-.13

.02

.08

.02

-.01

-.03

-.17

-.03

-.04 -.04

- .2

9* - .2

7*

- .3

7* - .4

0*

- .1

5 - .1

6

.32*

.2

2 .4

9*

.21

.38*

.1

5 -.

25*-

.23

-.43

*-.2

0 -.

20

-.11

.45*

.2

1 .4

2*

.12

.32*

.0

8

.28*

.0

8 .1

4 -.04

.02

.04

.06

.06

.12

.05

-.oo

.01

-.OO

-.21

-.13

-.04

-.03

-.09

-.03

-.18

.29*

.2

1 - .0

7 - .0

7

.10

-.23

.08

-.07

.19

-.05

.13

-.12

.21

-.09

.12

.02

-.05

-.05

.08

-.04

.25*

.04

.48*

.l

l .5

0*

.03

.27*

.1

6

.26*

- .1

8 .1

5 -.1

7 .0

3 .0

3

'Und

erlin

ed s

elf-

and

pee

r-ba

sed

mea

ns d

iffer

ed s

igni

fican

tly (p

< .

05) b

y th

e tw

o-ta

iled

test

for

pai

red

mea

sure

s. *p

< .0

5 by

the

two-

taile

d te

st.

FIRO-B’s Intrapersonal and Interpersonal Correlates 457

Dependence (S7) and on Sociability ( - 37). Because Inhibition’s strongest correlation (.65) was with Detachment, another bipolar partner of Sociability, Inhibition was assigned to factor Sociability.

Of Table 1’s 90 correlations of self-ratings on FIRO-B and IBI-4 and 90 others with IBI-4 peers’ ratings, 30 were statistically significant (p < .05, two-tailed test). Most of these linked FIRO-B to bipolar Sociability’s scales of Sociability, Inhibition, and Detach- ment, as 47% (17 of 36) of this trio’s correlations with FIRO-B were significant. Of Table 1’s 144 remaining correlations, merely 13 (9v0) were significant. FIRO-B’s notable linkages with Sociability’s scales included all five of the significant peer-rating bonds, each accompanying a significant self-based correlation. Always for self-ratings and often for peers’ ratings, factor Sociability’s three scales generally correlated significantly with FIRO-B’s Affection expressed, Affection wanted, and Inclusion expressed scales.

This rich redundancy contrasts with only three significant bonds among the 36 cor- relations of FIRO-B with self- and peer-ratings on the Dependence, Submissiveness, and Abasement scales of IBI-4’s Dependence factor, all positively linked with Control wanted (from others). Because wanting control from others is an operational definition of dependence, these connections support the construct validity of all four scales.

Among FIRO-B’s 60 correlations with the five scales of IBI-4‘s Control factor, merely four attained significance; each positively linked FIRO-B’s Control expressed (toward others) with self-ratings on IBI-4’s scales of Competitiveness, Dominance, Exhibition, and Hostility. Because each scale of this quartet involves some form of controlling con- duct toward others, these positive links supported the construct validity of each measure.

FIRO-B’s six remaining significant correlations all involved scales Nurturance, Agreeableness, Affiliation, and Deference of IBI-4’s Nurturance factor. Only Affiliation and Agreeableness had common links to a FIRO-B scale, Affection expressed. Affiliation, however, also connected positively and significantly with FIRO-B’s three remaining Affec- tion and Inclusion scales. Thus, both Affiliation and Sociability always correlated significantly with FIRO-B’s quartet of overlapping Affection and Inclusion measures. Bipolar factor Sociability’s two other scales, Detachment and Inhibition, each had three significant FIRO-B links that also paralleled Affiliation’s except for having contrary signs.

To summarize, 4 IBI-4 scales, Sociability’s trio plus Affiliation, accounted for 70% (21 of 30) of all significant inter-inventory correlations. Of the 11 remaining IBI-4 scales, Mistrust and Nurturance never linked significantly to FIRO-B, and merely 1 of each’s 12 correlations was significant for IBI-4’s 9 other scales.

DISCUSSION Support for the validity of FIRO-B’s scales from IBI-4 self-ratings was clearest for

Control expressed and Control wanted, as each had four distinctive and significant cor- relations with appropriate scales. Although FIRO-B’s Affection expressed also correlated appropriately and significantly with IBI-4’s scales, the meaning of these connections is obscured by IBI-4’s identically patterned quartets of significant correlations with two other FIRO-B measures, Affection wanted and Inclusion expressed. FIRO-B’s remain- ing index, Inclusion wanted, also connected significantly, and positively with self-ratings on two IBI-4 scales. Extensive overlap between FIRO-B’s Affection and Inclusion scales was confirmed by their median correlation of .53 in the present study and also was reflected by their rather nondistinctive correlations with IBI-4. This left only FIRO-B’s Control expressed and Control wanted scales with unambiguous support from IBI-4 self-ratings.

That merely 4 of IBI-4’s 15 scales contributed to more than two-thirds of all sig- nificant correlations with FIRO-B suggests that the latter addresses only a narrow seg- ment of the range of relevant behavior. Given the strong prior evidence of marked overlap

458 Journal of Clinical Psychology, July 1990, Vol. 46, No. 4

among four of FIRO-B’s scales, this constriction is hardly surprising. Further evidence of FIRO-B’s limitations seems apparent in its wholly nonsignificant bonds with IBI-4’s Mistrust and Nurturance scales because they represent such critical interpersonal processes.

Beyond the self-report domain, FIRO-B’s connections with individuals’ IBI-4 ratings by pooled group peers were rare. Merely 5 of 90 such correlations reached statistical significance, and all involved only the three scales of factor Sociability. Thus, IBI-4‘s Detachment scale correlated negatively and significantly with three of FIRO-B’s overlap- ping scales, and Affection expressed also correlated positively with Sociability, but negatively with Inhibition. Peers’ rating from all scales of IBI-4’s Control, Nurturance, and Dependence factors never linked significantly to FIRO-B’s three remaining scales.

The 255 imbalance between FIRO-B’s significant linkages with IBI-4 ratings by self and peers seems incongruent with Schutz’s (1958) repeated emphasis upon FIRO-B’s function of appraising overt behavior. Due to the greater reliability and probable greater validity of the multisource ratings by peers than of self-ratings (Horowitz, Inouye, & Siegelman, 1979), FIRO-B’s firmer connection with self-ratings than peer-ratings on IBI-4 seems more influenced by considerations of self-presentation (Johnson, 1981) than by self-disclosure or psychometrics. A restricted range of the ratings by pooled group peers cannot account for their fewer connections with FIRO-B because their standard devia- tions exceeded those of self-ratings on 6 of the 15 scales.

That Sociability was identified previously as one of the two principal factors of Borgatta, Cottrell, and Mann’s (1 958) exhaustive earlier analysis of interpersonal percep- tions in five small groups suggests that FIRO-B’s predominant connections with Sociability and Affiliation are probably manifestations of the broader underlying bipolar dimension of acceptance vs. rejection of others or affiliativeness. Carson (1969) regarded Sociability, one of only three major factors that Carter (1954) found to influence behavior in diverse groups of adult males, as an index of the “affiliation-disaffiliation dimen- sion” (p. 99). Affiliativeness also was identified recently as the dimension most salient in diverse inventories of the socioemotional climate of personal growth and/or psychotherapeutic groups (Hurley & Brooks, 1987).

Because all of the five significant correlations of IBI-4 peer-ratings on Sociability’s three scales with FIRO-B had parallels in the self-ratings, the absence of peer-rating counterparts of the 20 additional significant linkages among self-ratings is puzzling. The explanation that behaviors of the affiliation - sociability genre were so paramount in the small group experience that only similar conduct could be appraised reliably by both self and peers conflicts with prior evidence that self-ratings for self-acceptance consistently correlated positively and significantly with IBI-4 peer-rated Control in the same groups (Hurley, 1989a). Another possible explanation for this split is that FIRO-B’s predictive validity is uniquely circumscribed. This fits the heavy overlap of four of FIRO-B’s scales, and it would not be surprising if this limitation of FIRO-B also handicaps other self- report personality inventories. Extensive item-overlap among the various scales of both the MMPI and in Millon’s (1983) Clinical Multiaxial Inventory are well known facts. Nor are the present findings unique, as Jansen, Robb, and Bonk (1973) found that different factor structures undergirded self- and peer-ratings of 173 counseling graduate students on 12 bipolar semantic differential items.

These results caution against regarding FIRO-B scores as useful for predicting behavior beyond the “affiliativeness” aspect of self-presentation. A recent study of FIRO-Bs child-oriented counterpart, FIRO-BC found that its six scores accounted for only 5 to 9% of the variance in such peer-nomination measures as popularity, “best leaders,” and “best friends” (Burton & Goggin, 1987). They commented, “because the validity of peer-nomination data has been previously supported, these findings suggest that the FIRO-BC scale scores are influenced by factors other than what they purport

FIRO-B’s Intrapersonal and Interpersonal Correlates 45 9

to measure” (p. 167). This consistent evidence of weak linkages between FIRO-B and overt behavior, in the company of evidence that FIRO-B scales are generally dissociated from the central dimensions of interpersonal conduct (Hurley, 1989a), suggests severe limits to its predictive power.

REFERENCES BLOXOM, B. (1972). [Review of FIRO scales]. In 0. K. Buros (Ed.), Theseventh mentalmeasurements year-

BOCHNER, A., & KAMINSKI, E. (1974). Modes of interpersonal behavior: A replication. PsychologicalReports,

BORGATTA, E. F., COTTRELL, L. S., & MANN, J . H. (1958). The spectrum of individual interaction characteristics: An inter-dimensional analysis. Psychological Reports, 4, 279-3 19.

BURTON, S. A., & GOGGIN, W. C. (1987). Factor structure and concurrent validity of the Fundamental In- terpersonal Relations Orientation Behavior scale for Children. Journal of Clinical Child Psychology,

book (Vol. I, pp. 169-170). Highland Park, NJ: Gryphon Press.

35, 1079-1083.

16, 164-167. CARSON, R. C. (1969). Interaction concepts of personality. Chicago: Aldine. CARTER, L. F. (1954). Evaluating the performance of individuals as members of small groups. Personnel

CONTE, H. R., & PLUTCHIK, R. (1981). A circumplex model for interpersonal personality traits. Journal

DIMARCO, N., KUEHL, C., & WIMS, E. (1975). Leadership style and interpersonal need orientation as

FOA, U. G. (1961). Convergences in the analysis of the structure of interpersonal behavior. Psychological

FROEHLE, T. C. (1970). Construct validity of the FIRO-B questionnaire: A failure to replicate. Journal of

GILLIGAN, J. F. (1973). FIRO-B: Norms and reliability revisited. Journal of Clinical Psychology, 29, 374-376. CLUCK, G. A. (1979). The Kramer-Froehle controversy: A contribution to the construct validity of the FIRO-B

HOROWITZ, L. M., INOUYE, D., & SIEGELMAN, E. Y. (1979). Averaging judges’ ratings to increase their cor-

HURLEY, J . R. (1980). Two interpersonal dimensions relevant to group and family therapy. In L. R. Wolberg

HURLEY, J. R. (1986). Interpersonal behavior, range of ratings, and personal security. Psychological Reports,

HURLEY, J.R . (1989a). FIRO-B’s dissociation from two other interpersonal inventories. Manuscript sub- mitted for publication.

HURLEY, J . R. (1989b). Self-Acceptance and Other-Acceptance scales for small groups. Generic, Social, and General Psychology Monographs, 115, 485-505.

HURLEY, J. R., & BROOKS, L. A. (1987). Group Climate’s principal dimension: Affiliation. International Journal of Group Psychotherapy, 39, 141-148.

JANSEN, D. G., ROBE, G. P., 8 BONK, E. C. (1973). Peer ratings and self-ratings of twelve bipolar items of practicum counselors ranked high and low in competence by their peers. Journal of Counseling

JOHNSON, J. A. (1981). The “self-disclosure” and “self-presentation” views of item response dynamics and

KRAMER, E. (1967). A contribution toward validation of the FIRO-B questionnaire. Journal of Projective

LEARY, T. (1957). Inferpersonal diagnosis of personality. New York: Ronald Press. LEWIN, A. Y., & ZWANY, A. (1976). Peer nominations: A model, literature critique, and a paradigm for

LIFTON, P. D. (1985). Review of the FIRO Awareness scales. In J. V . Mitchell (Ed.), The ninfh mental

LORR, M., & MCNAIR, D. M. (1963). An interpersonal behavior circle. Journal of Abnormal and Social

Psychology, 7, 477-484.

of Personality and Social Psychology, 40, 701-71 1.

moderators of changes in leadership dimension scores. Personnel Psychology, 28, 207-2 13.

Review, 68. 341-353.

Projective Techniques and Personality Assessment, 34, 146-148.

questionnaire. Journal of Personality Assessment, 43, 541-543.

relation with an external criterion. Journal of Consulting and Clinical Psychology, 47, 453-458.

& M. L. Aronson (Eds.), Group and family therapy, 1980 (pp. 65-78). New York: Brunner/Mazel.

59, 219-228.

Psychology, 20, 419-442.

personality scale validity. Journal of Personality and Social Psychology, 40, 761-769.

Techniques and Personality Assessment, 31, 80-81.

research. Personnel Psychology, 29, 423-447.

measurement yearbook (Vol. I, pp. 578-579). Lincoln: University of Nebraska Press.

Psychology, 67, 68-75.

460 Journal of Clinical Psychology, July 1990, Vol. 46, No. 4

LORR, M., & SUZIEDELIS, A. (1969). Modes of interpersonal behavior. British Journalof Social and Clinical

MERENDA, P. F. (1987). Toward a four-factor theory of temperament and/or personality. Journal of Per-

MELON, T. (1983). Millon ClinicalMultiaxial Inventory manual (3rd ed.). Minneapolis, MN: National Com-

RYAN, B. A., MAGUIRE, T. O., & RYAN, T. M. (1970). An examination of the construct validity of FIRO-B.

SCHUTZ, W. C . (1958). FIRO: A three-dimensional theory of interpersonol behavior. New York: Rinehart. WIGCINS, J. S. (1982). Circumplex models of interpersonal behavior in clinical psychology. In P. C. Ken-

dall & J. N. Butcher (Eds.), Handbook of research methods in clinicalpsychology (pp. 183-221). New York: John Wiley.

Psychology, 8, 124-132.

sonality Assessment, 51, 367-374.

puter Systems.

Journal of Projective Techniques and Personality Assessment, 34, 4 19-425.

SEX, SEX ROLE IDENTIFICATION, AND COLLEGE STUDENTS’ PROJECTIVE DRAWINGS

DEREK N. ARONOFF A N D NAOMI B . McCORMICK

State University of New York College at Plattsburgh

A study of 109 undergraduates found that sex and sex role identification, as assessed by the Bem Sex Role Inventory (BSRI), were related significantly to sex sequence, sexual differences, and quality of drawing on the Draw-A- Person (DAP). Specifically, males and masculine persons tended to draw males first, female and feminine persons tended to draw females first, and androgynous persons were equally likely to draw a male or a female first. The greater the sexual differentiation of the female human figure drawing, the lower the subject’s degree of masculinity. The single best predictor for identifying a female subject was the quality of the female figure drawing. Results were consistent with Machover’s hypothesis that drawings of the human figure reveal facets of the subject’s self-image.

According to Machover (1949)’ projective drawings provide the clinician with in- formation about an individual’s self-image and sexual identification. Sex sequence on the Draw-A-Person (DAP) technique has been used widely to indicate the extent to which people identify with their own or the opposite sex. Traditionally, men and women have been expected to draw figures of their own sex first (Paludi & Bauer, 1979)’ with failure to conform to this expectation interpreted as implying ambivalent sexual identification. Similarly, Swensen (1955) has asserted that a lack of sexual differentiation between an individual’s drawings of a man and a woman indicates impaired sexual identification.

In contrast with Machover’s earlier formulation, contemporary research with the DAP suggests that both women and men may draw male figures in their initial projec- tive drawing of a person, simply because “person” and “man” have become synonymous in our culture (Paludi, 1978). Moreover, studies that investigated the relationship be- tween sexual differentiation and “sexual identification’’ have yielded equivocal results

Derek N. Aronoff is presently at the McGill University Student Counselling Service, 3637 Peel Street, Room 301, Montreal QC, Canada H3A 1x1.

This article is based in part on the first author’s 1987 master’s thesis, supervised by the second author. The authors wish to express their appreciation to M. Weisz, M. Marrache, T. Zandi, and G. Brannigan for their assistance and methodological advice.

Please address all correspondence to Professor Naomi B. McCormick, Department of Psychology, State University of New York, College at Plattsburgh, Plattsburgh, NY 12901.