INTERPERSONAL THEMES IN ENCOUNTER GROUP
PROCESS AS A FUNCTION OF STYLE OF LEADERSHIP
JEFFREY IRA BUTTER
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIRMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
UNIVERSlTY OF FLORIDA
3 1262 08552 4022
To C. KC.
I would like to acino;ledge tne a-ssitance cf th~e followingr
incirlJual: an agencies who~ hav; -cont-ibutedj to theo conpletiocn of hi~is
prjc.First;, my ife, wh~o has bezcrrs he unoffic l co-tdirctor of
myreeachprjets I she :we~~ re t arTcund to share~ ITU ;lu~rden, I
:ould~ noct have~;~ been~: able to do mny:n of thke things;; I hive d~ole. M
w~il~ling;?s tc. let me centerr" :ese? f "hemneso yc ite
th7ei Iltif=.';g on~Es Fau:1?l .krnett Elbl: j'c? .165111' Firle :al.:e, Ji.TGay
conv-~-rted t:,e en-cou~nter grups and helped make th-ir d~iffic7ult ad ell:ciLal
phaset of~il tM :.OjeCt po(Siiole. Ala n i~schert judy :-as~terok:, H;r-ris Jaffee,
-ad :ni~ce :ar~tin. const~:,itute ihe ratin, a-n~d scoring group~. i'hey; ""eserv
spec~ial tel-~ls ric; -one dEdi-catlt on "reyonrd the -all. of du~ty" they~ gave to
tha-t wear--i-c-re~ tas:. Dr. Joh7n Thio-nby and Richar Corb~in acted as
statisti~cel,~n an~crepute consultants, he7lpin~g no w~ithi ;-any of th!e pro-s
lesis th L alro~se inc thel Course% of the_ data Ef-alysis.3 Dis. Chrlz:les B.
Truax. anrd V:i.ce:-. T'iConnell a thank~edl or~ the practical suggstionis
theyi tzo:na e whn rlia'?G the dissertation p-cropsl PrIrcmi
thanki~ed for the! c2re se~7 rajve the p~reparatio of the f-ina~l. anuscript.
Finally, all the Etudent;'S who? par-lticipatced i~n ithe rricou~te Group, Resarc
project, are thanked for~ their' 3CcPer9tlion and: trusCt.
Th~ris pr'oj1c~t:i ws spprted~ii S by a gnt f~t'rin the Un~Ei eSity ojf
Florida Department ofT Psycchology. The cagouter analysis was run~ a the
University of F:Lcrida Comouting Center.
TABLE OF COMMENT
. .. .... .. .liii
. .. .. .. .. .. 2
Hy~poth~ese . .
Oc~comea measure .
1: D~~ESCRITI~ OF LrCOU-TER UBOPCCS AND
iFEi'E`''S COZFAC'" T CRC"ED iri7 'ID SOLICIT1
2: F3COE~ ri? 0800?-~J PESARC PROJECT:r~
3: .FII`VTIN OFi~~~. FCI \ ASLEGV
Trierr snalesig. ggfie ..
5: ITIEIWERSONAL THEF;-E ATiALYSIS TECHNaIQUE
6: A TEirIATIVE SCALE K)R RATINGl GEDUP~ LDrLjEDE
BERAVIOR: THE- LEADERi ~INTEPERSONA: L
7: INTER~P~ PERSCNA~L T iE4ES IN~ AN EARLY (I) MIDjT
rATE (II) SESSIONI: A, QUADRANDI ANALYSIS
ANI~D EiEE'UR STATEMEWOS~ .. .. ... .102
8: SUBJECTS' EILARDS~ PERSONL PFERENCET
SCEUESCO ES: PSTST, PO~STTS, Ry'D
DIFEal SRS. .. .. .. . .15
REEEEECEiS... ..............~~~ 125
E10ccnpi-crAPIA SET~CH .. ... .. ... ,. .. Ig4
LiSrT OF TABLES
1~: Per Centl of Judge A ;Wemente l on the Proce,,ss
$,atingE S-cales .. . ... . . . 36
2: IJudge7~s' Fo~cedLChoice Ranking of ieaders
on Activity;-Passiv~ity .. .. . . 37
3: Analysis of Curalated Firs;t i--inute of
Leeader Interv~enions from a~n Early and La-ce
Session U~sing the Leasdr I:nte_-,r.ersonal
BehaiVoL' Scale .. .. .. ..,~ .
L! I: Humber of Leader In-terventions inl Early(I)
and Late(II) Session Sarles U~sing~ the
-int-erpersonasl These, Anallyis Techniocu . 40
5: Int~erper-2sonal l The---; 1o--. -.
a: Intrp.?cl;;a.i l These An lysis Secanicue-:
Quadrant AnlysE1ris.. ........ 42
ci: E gyrtlj Ter-ongB1 PreferenceE Sche-.dule- Kean
and tanat eviaion b: Ss'Sex. .. 46
9: Va~luesr of F for Ed:-:aris Persronal Pleference
Scheule isorxinat Fnctin Anlyse 7
10: Edwards Per~sonal Pr~eference Schledu~le
Pre-test,-Pstte-, Differences ... 4
11~: n~l;,sis of V'ar~iance ,for Fret~est-?osttest
epenen~c-! c Scores of Figh and! Lows De~nendenrcy
s . . .. ~ ~~~ . . 51
Ab~:st~ract of Disser-tation Presented to t~he
Graduate Council of th;e Universi~ty of' Florida inr PartlaL Fulfillnhent
of thle Esquiirremens for the L~eire of Doctor of Pniilo~soph
LYTERPERS;;?~ONALJ THERE'S IN ENirCOUNTF3i~Z GROUP P.FD~CE~S:
~Jeffrey Ira. HuLtter
Chai~rar: Hrryn A. Grater, Jr., Phl.D.
Major- Dep~i~ar~tn Psychology
The purpose of this~ Study was to capare th~e e-ffEcts of active
and passive styles of encoun~ter group leadlership In terms~ of group process
and out~come, particularly with respect, to depsendencyq.
Th~e active leader ini~tiabes and: i- involved in interactions wilth
groupLiL moublers. He plays an important r-ole in strucstusrin an~d defi~rnin
the group. Additional characteristics ere n-:oted. Active leaders deal
with dependrencyi as an I~nividual urobjlem rather thanl a groupL penomencn.
The passive leader attenpt~s tc ma.intain the intera~ctionl w~ithin
the% marnbership, m:chilzze the t*reabel:-' pole in proviiiding their own strue--
tureis, and increase memb~errs' awareness of the group! process. Addition~ral
dia-ra-teristics we~l-~re oted. Passive leaders promote the exp~ression of
dlependePncy and aut~horityg conflicts, aSSuming thal~t bringiLng them to thne
for- e is he best way- to resolve- them.
The Inte-rp~e-jrsonl Tneme Analys~-is TechniquJle (ITAT) and thl-e Le-ader
Inte;rp-ersonall EehaJi~or Scale (LIBES), ;Ie~dif~ied versionss of the int'e~rpersonal~
process r~reserch? sci.ls developed by Leary and oth~ers, were us~ed to studyi
le-ader and membJer behavior. Actijve and cassi~ve leaders we"re~ assrd to
di.f-fer in thei-r inTcerersonral behnavior. Different: reciproca l behaviors
we~re theireorev expected fromT their re~spective groups. Specific diff''ernces-
Fassivelyi~ led groups, particularly short-t~erm goup~s, wer~e
expected to shlow more~ signs of celf-support and? less need f'or dependency
(etnvirontrenital support~) on te~irwintion thasnn ctively led gr~oups. Changes
in rzeyhers' niEed paLttens i.ere studied with- the Edward Personal1 Preference
~if'ty sjningl undergraduatess (28 rusles and 22 females) in six
encoiunter groups that !iret for six weektly sessions w~ere -tudied. The
leaders; were clinicll and counsolingT doctoral students (males), best
d~escribe^d s -ielativ~ely; inexperienced groupr leader:~, althougal their level
of expe~riense Is prchably typ~ica.1 of graduated i Stuident "tlherapIitSts" used
in psjdichotheray r:r-ear~ch. Leads;:ers weeesi,7ated a; act~ivet or passive
on th~e basis of' their styllstic prefre ences. Bo~th groups were~zi matched
for level of exper~i.Pnce.
Rela-iive fr-qulc~ies of type-s of in~tervention I-ere~ ailmosc~t
identical feel bo~th7 typ;es of leae~rshrip ex:cept fo~r suppor~t-seetin behavior,
which passivec leaders-- used five times mlore often!. jjo Significant dif-
fre~nces w-er~ toundi schn conmpari-~g ng memr I--terperonall behavior in th~e
second and fi!'t'h sessins.
Pre-tr-r eY:EPSdferneswr ngigbe Few: post-treatme~nt
changesc were~ found. ~cti~vely i~le -rouP)s deCre~ased in D-penrdeincy: scores
(DFpenden~cyJ equa;ls 3.:^erence plus SuTCcorance mlinus Au~Ltonomy~ Scale scoress.
PasivelyI led groups increased slightlyr (p < .05). Auto~rnomy an-rd Succorance
showd cangs farJin tne a!:tiv cond~it~ion~ (p < .10 an~d p < .01
rsspectiv~ely). Higa: dependency Ss decreased n! dlependncy in both
conditions. Low dependency Ss decreased in dependency onliy in the active
condition. mhere were no signLificanrt pre-group differences base~d. on sex.
Dependiency was the only significantly post-group diiscr~iminant, (p < .05).
Males in the active condition showced the greatest reduction in dependency.
Seieral methodological problems need resolution before the
hypotheses are rejected. Better quality control needs. to be maintained
to guarantee that the leaders are satisfactorily assuming their r~oles.
The relidoility and sensitivity of the judging should be increased.
Weekly attendance of merrters mu~st be stabilized. Stage of grou deve~lop-
rre-nt needs to be considered when coIap-arllrq grous. Repeated outcome
measure-s are needed. Suggestions ~ere m~ade to izplemenrt these
The research potential of the ITA1T and LISS !a~s aff~ire.
Modifications were suggested. The value of t~he EPPDS Dependency S-a~le as
an ou~tcome measur-e was also aff'irme~d.
Alithouga this study did not yield a sat-isfactory test of the
hypothesis that leader or therapist activity promrotes dependency and
non.-directivity: declreases d~epndncy, the results suggst that therapist
activJity p~er se dcas not increase dependency; it is t~he nature and
quality of the activity that matters.
Looking back on1 the h-istory~ of' group psyIchotherap~y, Hu~t (19611)
optirdistically~ rpor~ted that group ther~apy is nasi an accepted andi important;
rethnod of' tratme-nt andr th1at the gr~ou thPeray lit~eratursl has b'ecome more
me~aningful ovier the years, partly as the- result of thei development of a
more or less standardized descr~iptive voodoulary. Alth~cugh his first
conclusion would geerarlly goT, unch1llenged the second cone has not. e
variety of types of' groups has lincreased, crearting a ealt~-'h of information
and ter~riology that has not yet been assimla~~tedi by th~e professional com~-
munity. Ainderson (1969i), in a critical --view of the researchi literature,
said thatY the proliferai~on of research related to g~~rou cou~seling h~as
produced an "accumulation of bits of evidence" (p. 209)j with little of the
t~heoretical underp~ninnin needed to g~iv the practice or group counseling
a solid ground. On~e review of "theories"' of grocup development d~ra:wn fromn
the group therapy and gr-oup d~ynanics lijteraturt lst~ed 30) such: theor-ies,
wr~itt~en in term of each1 author's biases and _largelyr based on intuitivre
o~bservaition (Hill, 1961). There3sen (1969) de~scr-ibed the situation in
counseling threory, researuch-, and~ practice 3a turmoil an~d chaos. To mae
Iratters worse, he comp~lained, "Most studies; as they are conceptualized,
designed, execLedi aid analyzed, makle no differences to counseling theory
and practice" (p. 263). Hut's conclusion~1 is supported indirectly and in
an ?unflattering way bjy Bednar's (1970) recviewr which shcw~ed that of the
grou;p studies publIished since `1945, those applying minimal r-esearab
standars like controls, objecti.ve ouctcome~ criteria, andr prcise defini-
tions of vrariables were all published sine 19055, the average date being
With the enrgx-rence and groith of the "group dyjnandi.cs" movement
since World War II, a new dimension has been added to the prob-lem of
evauatrinlg the e-ffectiveness of gr'oups liith different patient, populations
and of evaluntin~g the varsoios styles of organizing them. Th~e population
for which gr~oupi experiences are now~ available has been e~jpanded to include
so-called "noirmals" outside of thle traditional thel-ray patient poCPUlaitio7.
Throug~j~:2 Naional Tiraining Laboratory program, "n~on-therapy" grows3 have
been madep a~vaildale to businesses, schools, civic groups and presurablyr
"healthy"3r' ind~ividuals who are interested in learniring more about the
dy~naics of gr'oups and increasing their in~terpers-onal sensitivityg (Coghill,
19657). The gr~owind acceptance and dema~d for groups that could saitisfy
tle: need for intense in~terpers~onal contact ndases f onrntyi
soci~ety7 increasingly mark-ed by rap4id social changF, disrupntion of thle
nuclear and extended faly, and feelingsj of a~lienat~ion has led! to the
evolutsionr of the encounter group move~ment (Burton, 1969, Egmar, 1,970).
Although thle encounlter group novement has little of the formal
orgai-zat~ion that the sensitivity training~-ld~oratory~ mov~enant has, its
currmb~T activityv generajlly seems t~o concern helping indiviiduals work
through the b3rr~iers to personal and interpersonal g~rowb and pr~omoting-
self-actualizaticn rather than with groupI p~henomena per se and their
o--ganizati onal applications~ (Murp-hy, 196'7). In actual practice these goals
tend to overlap. These divisions within thne group movernt have tended
to cross-f~ertlize eads oth~er. Irphy (1967) reportedly thalt th~e Esalen
movement has had a substantial ir43act on counseling and therapy, especially
in groups. Egus (197/0) reported several exarrples of the successful
mlixing of psychotherapy and training-1830ratory concepts together in
practice as justification for further experimentation. Th~e same profes-
sionals sometimes participate in mohre than one type of gro~up. Thusj,
group psyc-hotherapy1, trIaining laboratories, and~ encounter groups have
common roots, and often have common techniques and goals. Therefore the
questions regarding methodlology, group prorcesse;, styles of leadership,
and outcome~ that arise in regard to cne should be valid for the others
Th~e need for more r~igor and sophistication in psychotherapy
researd1, particularly in group research, is a well-docune~-nted, long-
stan~dig dobSrvation~ made by many reseanchers. The1 major cr'iticisms; seem
to be that even w~hen the typically uninterpretable, "pre-exp~eri-nntal"
case studies with N1=1 are eliminated from consideration, the diesign of"
most experime~nts and outcome studies is weak and inadequate; previous
findings and: relevanti research in other areas are frequently ignored; the
variables are often simplistic, vague, poorly operationalized or difficult
to replicat~e; criterion measures ir'e often vague and unre~lated to external
criteri~a in~ the subjects' life space; and results are reported in a f~onn
that lacks generaizaability or applicability, particularly to clinical
work (Ander-son?, 1969; Bednar, 1970; Carpbell & Dunn~ette, 1968; Dickenson
& Truax, 1.966; Edu;ards & Cronbach, 1952; Egan, 1970; Doldstein & Deam,
1966; G1eenwa~ld, 1967; Gundladar, 1967; McGra~h & Altmanu, 1966; Th~oresen,
19j69; Tuckman, 1965; Zax &Klein, 1960; anrd. otherS). Ander7son (1969)
contended that nsiv researchlers conr~iibute little or nothing to our pool
of knowledge because they rew still seard~hin for The Inuth. Hie offered
Gundlach's s~ummya of the literature as a counteractive to that motive:
"There is no simple universal patient; there is no universal treatmnent
name~d group therapy;. and there is no simple universal outcome- ranasure"
(Gundlach, 1967, quoted in Anderson, 1969, p. 223).
In the a:r-ea of indiv~dual and group therapy, as well. as the
burgeonin~g "non-ther~apy encounter and l~ior-atory group moverrents," the
relation of" structure to functions is not always clear-, yet this is one of
the two fund~amental issues underlying therapeutic activity: wrhat are the
goals (fu~nction~s) of therapy; how should therapy be structured to promote
its goals? Th~e elffct of style of leadership (a structural variable) on
group process is a proolemn that rarely ha~s been a~ddrssed in the group,
literature and is mrorec likely to be raised as a que~stionl than anIswered,
either in a theoretical or an ecrpirical wayr. Cursert (1968) remarked
that research on the behavior of T-group trainers is "not~iceablyr missing
fr~om the Ilterature on T--groups.... Z.Tis gapr exists despite widerlspread
acceptance of the- T-rou~p trainer as a key factor in group process"!
(p. 47). Goldnerg (1970) stated that:
The behavior of the group leader seems to be
the single mrost iaportant varidale in accounting
f~or the variance in events and outcomeps revsorted
by' the advocates of the various sensitivity
training groups.... In canvzassing the enlcounter
group literature for comparative studies of
leadership to anlswer such crucial auestions as
"cptimal procedures for giving f~eedback-, for
SThr~ous out this paper "therapy;'" "gowth,'! "triaining (T),"' and "?encu-
ter" are use~d interc.hangebly, as are "~therapyJ group;," trainingng group,"
and ":encounter group" Althoughl there a-re some differences, the
investigators sees t~hemn as essentially thle sam.
enlhancing feelings of psyichological safety, and
for stimulating individuals to tl~ry new behaviors"
one is i~vpressed by the paucity of^ such in~quiries
to date (pp. 2-3).
Egan (1970) called! leadership style one of the critical variables in
group outcome. When? trying to summazrize the effects of the leader he
offer-ed this qualifier: "It is difficult to specifyJ leadersjhip; qualities
in as -estricted an area as laboratory training, for little has been. done
on leadership in such groups and leadership styles vary geatlyr...
[italics added]" (p. 124). Gar.-ood (1967) called the trainer a m~ajor
force in wh~-at happers inl sensritivity training. Stock (1964) noted that
IFew studies have focused specifically on the role of the group leader and
asked, "Can t~rai~ner interventions be classified, and what is their Fimpact
on the gr-oup?" (p. 420)). The amount of stiucttu-e that should be pr-ovided
by the ~group leader is a subject for conltroversy in Foup counseling
(Anderson, 1969) and roup psychotherapy as wefll (Waliladi & St~rupp, 1964).
Based on th-eir atteirpts to study psych~otherapy variables, !:allach and
Strutpp (1964) called for clarification1 of w~hat therapi~st operations are-,
hm to identify them, and wrhat their relevance to therapy is. Culbert
(1968) concluded that "the absence of trainer re-search is no oversight,
it bespeaks the complexity of this research topic" (1p. 47).
The publiShedi research in the area of t~herap3ist and leader
variables as well as comparative group, processes anl-d outcores has been
liuited~. It has been dominated by the workli of Carl Roger and the
students of nonr-direct~ive, cli~ent-centered therapy le.g., see Carkhuff
& Truarx, 1967) whjo have been attenpting to define and validate the
":necessary~ and sufficient conditions of therap~euitic pers-ona~ity changee!
(Rogers, 1957, p. 95). There have been a nunt~er of studies published that
focus on aspects of the problem other than the relative presence or
absence of accurate empathy, warmth, an~d g~enuineness. Most of them,
however, deal with in~dividual, not group therapy ph~enome~na. Goldstein,
Heller, and Sechrest (1966) studied the relation of group size and group
interaction. They found that as roup size increases, thle ~interactions
between and among the group meabel-j and the leader dhange. They suggested
ending the all-tioo-commron speculation about the matter with syrstenatic
research into the relation of group size to leader intervention. Culbert
(1968) ~investigated the r-elation between leader self-disclosurec and
rember growthn in two T-groups. The high self-disclosing leader entered
into more "perceived therapeutic relationships" with gr~oupi lieberes and
the neabi~ers of the low self-disclosing leader's group entered into more
of' such relationships with each other. Although both groups attained
the sam level of self-~rawaeness by the end of the semester, those in the
high leader disclosure group did so earlier. Coons (195'7) coImpared two
types of group therapy w~ith a control group. The grow~- in wJhich want
interaction instead of personal difficulties wa~is empnhasized showed greater
improvemnt t-hanl the insight-oriented group a-n~d the control group. Land
(1963), alsao using hospitalized psychiatric patients, found an increase
in patient intieraction: and groupi responding when the group leader used
silence and redirection instead. of active verbl;ization. This conf'irmed
the earlier wrork of Salzberg (1961, 1962) and othelz. Ai study of college
male undra~-chievers (Gilbreatih, 1967) showed that students with high
dependncy needs react no~re favorably to a leader-strulctsurd than to a
grsop-structured counseling gr'oup-. A threke-year follow~-up, (Chiestnut &
GilbDreath, 1969) shc;ed that the difference (higher rade-po-int average)
persisted. In another abudy of nroup cournseingr w~ith, u~nd~erahievert
(Dickenson &r Truax, 1966), it wazs shown thlat the equiivocal results in?
many other studies Iray have becen the re-sult, o' confo~uning high and low
tiherapeutic conditions (i.e., coaib~ining the results of' therapists rated
highl and low on accurate empathy, warrant, etc.). Following an inlvestiga-
tion of trainer in~temrvenionis in seven T-groups, Psathes and Hardert
(1966) concluded that it; is possible to compare trainrs~- anld groups in
terms of the patterns and frequencies of imp~licit prescriptive and pro-
scriptive norm-messages enbedded in trainer conwan~icationls.
A number of' re~searrchers have dealt wilth the effect of thle leader
on a; group, by studyilng leaderless and alternately led and unled gTroups.
Exner (1965) found that irreglar leader attendance benefited out-oatient
therapy group Iweeers. Salzberg (1967) compared th~e verbalizations of
an actively; led hospital psychotherapy ~group w~hen the therapist was present
and abssent. Without the therapist the patients showed greater spontaneity,
less p-ralen~-releva~nt responding, less interest in personal prc'olemrs and
more interest in other grou.p mem~ers. Some patients spoke urp more and
some assumed leadership roles. Seligman and Sterne (1969) found. that
leaderless and alternate session groups showed more~ conventional, socially
oriented behavior as neasulred by the Hiill Inlteracticn! Phtrix. Harray~,
Aistrachain, Beckter, Eiller, and Schwartz (1?67) found irthat un~led groups
tended to be warrer an~d more supportive. In a study that clearly points
oult th~e lack of a simpirle direct relation between group process and outcome,
Tk~arua and Ca-khuff (1964) _foun~d, that whien using ralternate sessions with
juvenile delinquents somea of' the deepest levels of thelrapeltice process
occurred during unled sessions, yet they made the same amount of progress
on outcones measures as continuously led grouups. In a study with neurotic
ou~t-patients they found the alternate sessions techiniquve to be superior
to con~tinuous leader presence (Truax & Wnargo, 1969), while the reverse
proved tr-ue in a study with hospitalized patients (Truax, 1966). No
punlishied studies adequately account f~ol these differences.
These studies demonstrate a growing inlterest in research on grouJp
differences based on leader intervention or group structure. One of
their strengths is their recognition of the importance of' researet
paradigms that compare and contrast the effects of different structural
variables. Their weaknecsses include a lack of specification of how their
variables were operationalized. There is also a tendency to study vari-
ables one at a tirhe (unLivar~iate deslign) and ignore the potential effects
of others wh~en making generalizations about the results. For example,
data on group composition av rarely included (beyond a possible reference
to diap~ostic category or in/out patient status) yet this is probably an
important scurce of experinental variance. WJitn~ess the disparate results
anonlg Truax' studies of unled graps (Truax, 19~66; Truax & Car'xhuff, 1964;
and Truax; & War1go, 1969).
Although there does not seem to have been any3 deliberate attempt
to systematically study th~e relation of leader-initiat~ed structur-e anc
medaber personality, these studies, take~n as a group, seemn to be mo-ving in
that direction. Exner (1965), for exang-le, speculated that t~he effect
of u~nled group sessions could be emp~lainedc in terns of for-cing enters to
deal with their owYn dependency needs insteadi of transferring thlem to th-e
leader. Hie i.s attempting to replicated the study withli more attention to
this effect (Exner, 1970). Exner has not yet r-eportied results, so
Gilbr-eath's study~ (G~ilbreath, 1967; Chestn~ut & Gilb-e~athi, 1969) stands as
one of' the fewr, if not the only, grcup s~tudies in wihic~h the effect of
IX3ber dependency was investigate. This is surprisinga in light of the
attention clinical literal~ture has given to that phlenome~non.
A' survey~ of the literature showed tIhat most studies of dependency
involve dyadlic rather than~ polyadiic groups (i.e., "individual" vs. group
therapyr). A nurrter of th-ese seem relevant to this studyJ. Heller and
Goldstein (1961) cite theoretical literature that sugagsts th7at clients
enter therapy feeling dependent and 'oecome more riindependent, over time.
Th~ey found that initial client dependency facilitat'es th~erapy and nelps
maintain the relationship. An analysis of therapists' approac?-avoidance
responses tio client exprssion of dependncyr indicated that therapist's
rate of approach is related to conltinuation of therapy (:;inder, Ahmad~,
Bandura, & Raeu, 1962). Alexa~Ender and ALbeles (1968) found that clients
whose dependenlcy demands are n~ot or cannot be me~t (because of their nature
or intenlsity) terminate early. Hieilbrun (1970) found th~at di~ep~enden
females tend to receive non-diir ective~ r~sponses froml male thleracists.
These w~ho become suifficientlr f~lustratedd seeky dependenlcy gratification
else:here and terminate therapy early. Dependent maless tend to velrairr
in non-directive therapy.
Alexand-r and Abieles (19'69) expected fe~rale clients to express
more dependency than males initially and thr~oulgtout therapy with nale
therapists b~ut found no difference. Theyr also found that most types of
dependency relationships (e.g., client-therapist, client-famly) decrease
or level off by the time therapy is terudinated (Alexand~er &~ Abeles, 1968).
Snyder (1963) found an inlcrease in dependency through the ri~ddle of
therapy, followed byi a steady decrease to the original level by the end
of therapy, Schuldt (1964) found a small, constant decrease in expression
of' depend~ency, while Cartwright, Kir~tner, and F~iske (19163) reported no
change fromt beginning to end.
A nunlber of individual theracy studies also deal with pertinent
therapist variables. A~shby, Ford, Guerney, and Guerney (1957) studied
the effects of a non-d-irct~ive (relflectivJe) and a pleading (interpretive)
type of therapy on counseling center clients. Interp~retive therapy
clients produced bothl more open and more carded verbal behavior and wr~ee
rated by the thel-rapists in the study~ as showing greater mp~rovrement than
reflective th'erapyi clients. The effects of pre-th-erapay client personality
also proved to be moire imrportant in? interpretive than in reflective
therapyg. Usinlg the sam~e data Guerney (1956) founld no difference it
withinl-session dependency behavior. Bottsch-afer, also using college
students, foud that ref~lectiveI counselors tended to evokce less client;
depend~ency behavior while fostering greater client re~spondcing (Rottscbarer,
1960; Bettschrafer & Renzag~lia, 1962). After hypjnot~ially inducing a
"re~pressed" mnemory and potential negative co~untertransference, Gordon
(1957) studied the effects of leading and cfollowing styles of therapy in
ucncove~rin- g the repr~esion. The leading~ therapy resulted in grater mad
more accucrate uncovering. The reflective thlerapy evoked the negative
cou~nte-rtransferenc e more often. In another studyr of reflective and inter-
pret~ive therapy, in which a time limit; was imposed on the them~pists, it
was denonstrat~ed thnat good results-, could ben obained in about half the
time~ of 'therapy on which no limit was imposed. Both the Rogerian and.
Adilerian approaches shcwed a good deal of sim~larity Whnen operating under
the imosed structure of equal tir limits (Schllien, 1964).
Based on the a~ssu~tmpt that there are styles of therapy that out
across theoretical lin~es, several studies wesre condCucted that analyzed
psychotheraputic techniques. Using therapists' reports of their usual
practices, a group; of' inestigautors found three factors that diffier-entiatedl
therapists: (1) ana~lytic vs. exp~eriential (a general factor), (2) imrper-
sonal vs. personal affective approaches to patients, ;3) activle, directive
vs. non-direct~ive m~et~ioods (Mlclair & Lorr, 1964, Sun~dland~ & Barker, 1962,
Wdallach & Strupp, 1964). Segal (1970) studied? stylistic diffe~reces among
"A" and "B" therapists. He described A's as acre directive, In~terpret~ive,
and negative. They assum~e responsibhility for the structure and directionr
of thre therapeutic process. Hie coxare~-d th~em to th~e "leading~" typre in
Ashby et al. (1957). B's are mocre facilitative, more encouraging; and less
direct. Thy respond in a wlay that encourages greater expressrio~n by the
client. B's tend to folla.-r the client's lead rather than provide direc-
tion. Segal com7paed, them to the ''reflecting" type in Ashlby et al.
Th~e literature cited, in the lresent studyr calls a number of
traditionally beliefs about thePrapy into oulestion and also points to large
gaps in ouir understanding of group process. For exiample, thne role of
insighlt in the relief of emotional distress and7 beh-aviorc change has been
seriously questioned. Th~e finding that group in~teraction, per se can be
more effective than~ insight gainred (Coons, 1957) should be an adaarra~ssent
to those thlerapists who have not examined alt~ernativets to insight-oriented
therapy. One of the implications of Culbert's (1968) study is that high
self-disclosing group leaders can expect to obtain results equal to (and
perha~s better than) th-ose obtained by 10#? self-disclosing leader~s. If
this is a retliable rsult, then leaders; interested i~n short~e~inlg and/or
enriching their gr~oups could use this information to better imp~lemeint
their therapeutic goals. Gilbreath's (1967) approada to group~ counseling
with und~erachriever's is important since it dlemonstrates the potential pro-
ductivity o' th~erapeultic approaches that are molded to clients as theyr
are in~stead. of using the Procrulstean strategy of "adcjusiiting" clients to
confornl to the therapist's nethodi (and discari~ng the rris-fits as
"untreatab~le"). The e~ffcts of leaderless and alternat~e sessions lead
to questi~ons abcut the most effective ways of dealin-g w~ith some problems
and call for further ex~am~inaion by group leaders of their goals and
haJ pa-ticularu groups should be str-uctur-ed to m~axinally imrple~me-nt those
goals. The conftusion regarding the appropriateness of leaderless and
alternate sessions and the contradictoryr findings bocut client dependencyr
behavior over tce course of therapy emprhasize the need for more systematic
study andr repor~ting of population variable-s and more sophisticated, mur~lti-
vari-ate approaches to ope;rational definitions, and I_exprir~cntal~ design, as
has been pointed out by Anderson (1969), Ashby et al. (1957), C~artwirigot
et al. (1963), Heller and. Goldstein (1961), Thoresen (1969), and others.
Th~e role of and need for dependency --- both~ as a gcrop or "systems"
phenenoinon and as a personal phenenornonlr -- need to be clarified. Wha~t
are the conseluences ofr dependent behavior for the process and outccmre of
therapy? A last question, in keeping with the zeitgeist of contemporary
psychology, is wh~at do therapists actually do and w~hat are the consequences?
iopefully~, "myr authority kn~ew more than your authority" thinking Is waning
and thle eip~irical study of psychotherapeutic strategies is in its
DEFINITION OF THE PROBLEM
The focuLs of this study~ is th-e relation between group process and
ouitcome~ an~d contrating~n styles of leadership. Fromn a review of current
Qgroup andI psy:~chotherapy resear~Tch liter~ature, an attempt has been made to
identify~ thie salieant features of this r-elationship, as they are cu~rrenly
understood. It was concluded that the xui~sting lit.eraturte demonstrates
that different sty~les of conducting groups are expected to, and indeed
do, affect the process and outcorns of gr~cupu~s. IHotever these differences
h~avJe not bieen systematically exrplored and general agr-eement on their
effect-s and the-ir deirdb~aility? is lacking,. This studio is specif~ical~ly
concerned withi the~ similait~ies anid d~iffernces that result f^romr~r~r~r~^-^-^-^-^ conducting
groups in anl activei" and "passive" stylie especially with regard~ to thie
hanidilng o-f dependency~. The question of style cuts across theoretical
appiroaches to therapy and represents an impiortanti distinction. among
thearapists regardless of their professed aFffiliation,.
Because; of therapists' tendenlcy to con~ceptulalize wh~at has occurred
in- the-rapy~ !in terrs of their oun theoretical biases aLnd orof~esse arffil-
ationrs, it would seem advisable to lookli at whrat has happened in terms that
pernit Circ7parisonl hs Fiedler (1950a, 1950b) d~id In his nearly comarative
study~. Mic;air - and Lo (1964), Sundlan~d andj Barker (1962), adc W~al~lach~
and Strupp~ (1964) in~cluded "directLive, active meithlods" as one of three
facto~r-~anatic dMlrn-sion along rhi~ch therapeut~'ic activity fall~s. Segal
(1970) studiedi the styles of ai and B therapists andc con!cluded thnati A's
resentle the ":leading" h~era~ists st;udiedi by Ashby etr al. (1957) and the
"active" therapists described by Wvallach and Sctrapp (196il). The sarer
r-elation wa~s suggested for B's, "reflective," and "pcassive"' therapists.
Land? (1763), continuingi thle worked of Salzberg (1961, 19i62), contrasted, an
active, verbaizing~ therapist to one whjlo emplcys3 silence and -reirection
as basic therapeutiic intervJentions. Gilbreat~h (1967), hiad compared a hiigh
authoritariian leader who structured hris group with a le:: authi!oritar~ian
leader ho allowed thle embers to structure- the grouo. In h-is en~thology
of active psychotherapies, Greernweld (196j7) concluded that therapist
activity cannot be defined quantitatively as the number of thlerapict comr-
munzicationls or the noni-directive thera~pist whno tals almost as~ nuch as
his client would be~ classified as active, H~e fe-lt thlat tha-t could d contlz-
dlcct h'sj uirs;ndetndin of activity since thle co!cannications air lim~iterd
t~o reflectica of reelings. To GJI-r-eeald active." is noct synoi-rccus with
"directive," althoughl there is a good. deal of overlap. He ulnder-tand~s
the difference to be in terns of degree and attitude. The ac"tive therapist
"tends to utilize consciously devised interventicns for a specific pre-
conceived go~al.... This goal Imay be increased ins~igt, dcreased anxuiety,
the modifications of overt behavior...[etc.]" (p. x).
L~eaeshipi styles are classi~fied as being either active or passive
relative to ,he interaction patterns typical of group nuenbers. Leadership
style could be judged active or passive by such considerations as the
zmounrt of' time t~he leader is the focus of attention, the ~inuner and
duration of tires thle leader is inv-olved in interactions wirth group mm-
bers, tne degree to whichd the leader establishes grrrou n~ors, the extent
to which th~e leader~ acts to i'ulf'ill. memers' n~eeds, the ext'ent, to whirich
the leader de~libErately int~erve~nes in the group~E process to at~tain a
specific goal or result, etc.
Tw~o termsh~ that cleairly neead defining are active an~d pissive.
Th~ey; havje been de~fi-ed in this studcy relative to eachr other and to thle
be~havior of' a hypocth~ecical average gro~up member.
Active means causin or in~ibiati~ ng omn~lt or
actili7vi. It~ irwplies involvemrent. Ain active
l-eader is one w~ho by his behaviorr initi~ates and
directs the course of group inte~ractions andC is
involved in the interactions. Th~e act~iv- leader
plays an important role in structuring and.
defining the group by setting the rules of 1inter-
a-c~tion, setting contijngencies on rember behavior,
initiatingg and entering into interaction with
group centers, ;rnd 'y pro~viding~ one or more
behavioral moiels for gorop ment~ers. Th~e acti\Ie
le-ader is dirctive in the sense th-at he makes
suggestions and "set~s the stage"' so~ that the
group process will be productive. BSeflect~ive
s~tatements by thle leader awre ot consi~ded
indica~tor~s of high leader- activity when? they
serve to mizntain the interacticn w~ithn the
group reaer~shp rather thanr involving thie leader
~in stru~cturing: and enter~ing the c~going grcup~i
Passive means not acting, not~ engaging in frequent,
open action. A passive leader is one whio attempts
cry n~ally miia@to didact the course of. groupL inter-
action~s anda whose involvement, In the in;teractions
is usually rrant to focus attention on thle -r~oup
p7rc-ess. The passive leader tr-ies to minimrize his
role in structu~ring, and defining thle g~~iram and mnx-
imize the nembers' role in providing their ownl struc-
tre~-. H-e prcviides few t'ehavioral models for the
group members. Hlis role is largely in contrz-
distincti~on to the role the typical group nernbers
are expected to assune. ~elflective statements, ques-
tions (e.g., "Howi do you~ feel bocut that?"), and
quiet reinforcement ar atypical strategies aimed
at naintaining the interaction wlth~in th-e group,
Nothing here should be take2n to imply that
eithe;rr style of leadiershipr is at all tires
appropriate or of highe-r value than the other.
There is corrn risk in assigning leaders blanket labels likre active
or nassi~ve since consistency in style is niot always attain~able or desirable.
Active lea~ders will occasionally behave- passively: and vice versa.Te
classicicationl of' a leader or therapist as active or passive- is rmde,
th-eef'ore, in? ttern of the relative proportions of these behlaviors that
he exhibits over the courlse of his grcup.
AlthougaS active and passivie lack stanidardl, conventional meanrings
in psychoth~erapy andc gr~ou~ work, they and their approximate" synonyms are
frequently used in discussions of th-er-apy. 'The strategies and goals
associated wi~th both of them have their supp~orters and detractors. To
Rogers (19,12), the directive expertr" harpers the redevelopment of in~de-
penden;t and self--relant client bJehavior and i~ncreas~-es client dependncy.
Th~e client-centered! arpp~oach, whlich consists of reflecting or clarifyiing
client ~feelings simple ack~nowledgrants, and silence, is alleged to give
thle client greYate-r freedom to develop himself.
In psa--sively led groups, the leader initially lets th rupko
verbel~ly or non-verb~ally that it is their responsibility t~o decide the
surbstanc~re of the contract within whiich they are w~illingS to workc. Bjy this
method th-e goals of the gr~oup m~d the role of th~e leader becnead que3stions
cor the group, and the well anecdoted: process of deaing with the leader
and8 establishingl a contract ensues (see e.g., Bennis & Shrepalrd, 1956,
Non,-direct~ive straitegi~es ar~e not seen by all grLoupj leaders as the
most productive~ in Illsi~tuations. Egan (1970) believes that we: are only
stazrting to find our what techniues are effective inl wJhat cir~cumstannces.
Groups should bce organiized in tenrs of the participants' goals. He sees
the relation between the leader and' grou mem~rbers as being a contractual
one in which the contract can. be vagle an implsied or expl~licit- mad public
If one i~s concerned writh! diemonstrating wh~at, hapenls when a groups of people
collect without clear goals or organization, then he believes that the
unstructured, leaderliess group is aippropr~iate, for the natural dynamics
of groups wlill become most evide~nt and available for study in that wayr.
If onea is not interested in studying goup process as an end in itself,
then some variant of the contr-act group is called for. Anir initial con-
tract (a tang~ible document in some cases), clearly specifying the _rules
and responsi~blit~ies of ~group, participation, helps to bypass the conflicts
about the contract that ma~rk passivel.y led groups. When the provisions
of the contract I-late to the inlterpersonal tgrowth process, Egan cl~a~ims
this approach is the most appropriate and reward~ing Fg 190
...sin~ce this [meabser reaction to goallessness]
is a systems effect, the leader cannot claim
credit for having produced any special r'esullts,
anld the value to the participant; is dubious.
hi~~le the person whose electronic supply is dis-
rupted rany be able to get along with candles
and a fireplace, this demonsitrati~on of self-
sufficiency is not what he is playing the power
comp-4any to produce (p. 95).
In actively led gr~oups, th-e a~tterpt is typically made to decrease
the! intensity of this eff~lect bjy admrowlledg~ing the problem or otherwise
rirhibiting its ovrert expression. It, is allowed to arise as a personal
issue fo~r individual nemb~ers rather t~han being mrade a group issue. The
lead'e- meets his p-rimary contractual cb~igatio 'o the group by provi-ding
thle stru~cture it will needl to facilitate its work andi cy settling the
affective tone of the group~ (i.e., by being a model). Active leadership
is based on thle assurptionrr that the group is more concerned writhl the
interpersonal growth process than with systems phenomeina, andl mor con~-
cerned with the outcome of the group than thle experience roer se.
Th~e dev;elopmen~tal gravth of a passively led group is based on the
mentrer~s' exp-er-ience of and ability to deal with ad~iguity and goallessnecss.
Initially, feelinlgs of helplessness, frustration and dependence become
prominent tin the gr~ou. For somer individuals, this. is a rare opp~ortunity
to become aware of those feelings anrd deal with them. Th~e resolution of
conflicts regading dependent and countendcependent behavior is a signi-
LCicant (some leaders would say the major) goal of groups (Bennis &
Shepar~d, 1956). Initial goallessness is th~-erefore essential to the pur-
pose of these grous, which is, in par-t, to; create its own goals (Benn~e,
1961r; Bennis, 1964; Gibb, 1964). To promoti'e this arbiGuity thre formal
leader renounces control of the Groupi, creating a leardership vacuum that
e~vokces any unresolved aiu'cithoit and dependcyrs conflicts existing in the
group. Berns~tein (1965) regrd resolutionr of the transference pecblem
(i.e., member dependency on: the nominal authority figure in the group,
the leader) as critical to growth and w~a~ns that unless the runbier-leader
polarization is worked through, transference resistan~ces rray continually
crop up or act; as a crak~ie on the gr~ou's progress. He advocates the
stra-tegy of aggrava~ting the prcblex so that it can ccre to the fore to be
Th~e q:uestionl of hav to facilitate re~solution of dependency in
active g~rous has not, been give~n the amount of attention that; it deserves.
Ihe consensus of active leaders; seems to be that passi~ve leadership
magn~rifies the problem and that depend~ency can boe handled in the same way
any other problem is -- on an individual basis and in terms of that
individual's conflicts (Egan, 1970). Thl-is is an unproven assumption tha~t
my3; be based on such diverse factors as leader preference for an active
style, op~timrism based on success w!ith other problems, or an unwill-i~ngess
to deal with merber dependency. There hes apparently been no research
done on the relation of these f'undam~ental styles of group leadership to
th~e resolution of dependency. T~he leader !jho does niot deal in! some w~ay
with the possible complications anid limitations of his style may find
himself engaging in, anti-therapeutic, counterp~rodluctive activity that
defeats the contract he is trying to fulfill. Sh~epherdi (1970) warns that
"a major hazard...iLs the therapist's assuming excessive responsibility
for the direction of the group by too much activity, thus fostering
patient passivity and defeating his csms goal of patient self-sup~port"
(p. 237'). To Egan (1970), one of the~ dangers of' active leader-ship,
especially when the leader has overinvested in being~ a "parent," is that
the leadership skills of the members will not, develop, increase, anld get
social reinforcemesnt. Blake (19611) warns that the leader who cling~s to
his role w~ill pro~mote excessive modeling of his behavior instead. of the
developrrent of authentic response styles. Hiis level and style of inter-
vention. is Ilikely to become~ the stanar for the group. Th~us he could
inh~ioit growing out of dependent modes of beh-avingr and actually foster
strong dependency on himself (for approval, permission, solutions, etc.).
The diffu~sion of the leadership function th-~roughouti the group~ can, there-
fore, be a special problem for active leaders.
The? characteristics of active~ an-d pa-ssve leadiership have not
been sujffricient:ly studied to pe~rmit reliable stasemmats to 'oe made con~-
cer~nin~g their similarities, diffre~rences, benefits, and sho~rtoomings. Nor
h-as it beeni- made suffcien~tly clear has the benefits could be Ic.uiniized
andi the pr'oblems m~niized. Since a nunber of' issues raised regading
leade-reaer rlation~s, leader behnaviorr and group outcome~ cocern the
typ~e and frequen:-cy of transactions- tha~t occur- in groups, an anaysis of
those transact-ion; (il~., the group's inter~perrsena behav~iier) sh~ould pro-
vide the basis for the conparat~ive study; that has b~een suggest~ed here.
It is hylpothesizedl -,hatL ?tive- and pascive~ group~ ;aders differ in their
interpersonal beh~navior and that these iif'ferencess affect m~eme initer-
per~sonal behavior.l~ ita the in~terpers-ona~l be:havior of gorous in the two
condilticns is expected to differ.
It rdce dfeecsi the itrpersonal talivior of' ac~tive
and passlve leaders; havec been infel-~ rred fro the foural def~initicas of
active and oassiv~e 3iated earner. The Pas31ive leader interic~ts w~ith
groups meters in; a plardoinanly nonr-di'rctiive~ ::ay.~ Hie reflects feelings
and concerns, cr?:veying h:cis un~~_~tnderstandi of' a situa'ion, attempting to
expand the neres I a::a~ B.reneBs of w~t th~ey see-m tio ie e~xper~iencing or
ccn:r~veing! to others-~, an7d encoura~ging them~~, to exlore.~3~1 ther ir? on eelings~j
an~d reactions. Thi~s involve~S behaviorjCi that. could bS called teac~hingr and
supporting. One of the wayis passive leadels f-,-strate the folrnation of
depen~dentr riaslation with them is joy responding2 to; support se~eking by3 wilth--
hoilding the expected re~sponse. th~is cou~ld be in r~ai~tion to cr int~ead
of reflfcting feelings, etc. Act~liv leaders trfv to strulctur~e the group
and are th~erefore more likely to respond in, ways that convey structure.
They set downrr guidelines for the :interactions an~d remind mrember~s whlen the
guidelines are not being followed. Th~ey suggest things to doc an~d not to
do. One of the w3y's active leaders achievei their goals an~d influence
the gSroup process is through the use of teachring, strulcturing, and punishling
behavior (in the learning theory sense). Positive cons-~-equencce are put on
desired resp~onses and negative consequences (diisapproval, aff~ctlive dis-
tance, sarcasm, etc.) are put on undiesired responses, thereby~ shapi~ng the
group process. Therefore, it is hypothesized that active leaders us~e
teaching, supporting, and punishing behlavior to structure and influence
group process. Passive leaders use teaching, supporting, and withholdingg
behavior. Active leaders use more teaching~ than supporting anid passive
leaders use morl supporting than teachin behavior. Active ileades use
more activelyr punishing~j beh-avi~or and passive leaders use mno re ithholding
Th~e active leader, by vrirtue of his role in stiructuriig groups and
helping group meab~ers, is more liely to engender~ dependency upon himself
that is not likely to be resolved in the course of the group. Th"e passive
leader,w-ho, by his behavior, is more frustrating of the attenmpts of' grouLp
menberj to create a dependent relationship, is less likely to have dif-
ficulty in this area. Thlis does not lead to the conclusion that passive
leaders are more effective overall, since time~ spent dealing with
dependency is time tak~en exiay from other probrleTm. Te active leader is
expected to deal more- successfully with other difficulties i.n which his
activity i~s fa~ci~ltative.
The passive leader is more likely to evoke hostile, rejecting,
andc support-seekcing thenles at first since he is f'rustrating, not meeting,
the suppcort needs of his group. Wae~n member; s find tihat they can support
therselvess and each other, they do not need to seekr the support of the
leader and instead. are more~ supportive of each other. Thle active leader
is less likely to evoke h~ostility and resentment early in the course of
his group since hils ac'tivityr will be meeting the group's support n~eeds.
Howlevier, the situation changes later in the group: errzbers eventually
derland mer~e supspor~t than the leader is willing to give; merrbers beccme
anx~i.ous about their dependency; members begin to feel mr-e comfor3tabl~e
and less dopendenlt; meab~ers workc throuSh their needl for dependency.
This leads to what Bern~stein (1965) described as tr;ansfernce rasistances,
wh~ereini resentrent and hostility is cover-,ly or over~tly expressed. Th~is
is sometires seen as whait could 'e called "attemots to dethrone the
fatiher/1eader." 'Therefore, actiively led groups are~ morre likely to see
an increase In, hostile and rejeicting behavior later in their development
tha~ln assively ledj groups. Because of the potency of m-deling effects
and se-lective leader ;creinforcemen~1t, rrnters of active roups also have
the potential tc exhibit mote positive, suppor~tivec behavicr if and when
thl-ey reduce their need for dependency -- either as a group or singly.
An oultcome cr~iteri~on that is relevant to th~is study is change: in
what Perls. (1970) calls selfc-suport -- that is, doing for oneself whriatever
oneis apaleof, ir-stead of manpulating others into doing it; tak~in
responsi;bility fo cr onself. Hel descr~ibes matura~tion (emiotional and bio-
logilcal) as "th~e developretnt, from e~nvir-onnentalt support to self-cuopIort"
pr. 17)'. Elsewhlere, Har';ley and! Xosentaium (196,3) found great un~anity in
the use of self-ac~cept~ance, self-confidence, and self'-relianc~e a~s cri~teri~a
for patient; illp~oviement anmng the group p;sych~oth!erapistss the-~y sur'veed.
Need f~r~ envriroilznmenal suppor-t wouldJ appear to oe a good defintiti~on of
dependency and~c self-support would appear to b- a~n~ appropriate and accepted
criter~ion of a favorablee goup outcome. Since it h~as been suggested that
passive leadership frustrates and reduces the need for dependent beh-avrior,
outcoma-~ rrssures of need for dependency or environmental support should
shoni a greatie_ reduction in those needs in passively led gr~oupjs than in
actively led grnoup~s. This would be particualarly true of shor-t-tnr~m goups.
If the masters of an actively led group are able to successriblly deal w:ith
their dependency (either collectively or singly), then the re~verse relation
would be true: both~ tyes of roups would show a decre~ase in dependedncy
and the actively led group would show the greater change. 'This later
effect would priodobly be most evident in long-term~ groups.
H.A.r; PArrayj (1938) and mother devised a categorizatiojn ,for me
Inl studyingr normal personalities. Edwards (1959), using Iurray's defini-
tions of univiersal human needs, develope~sd a personatit su.-rvey de~siged
to mea2sure tose needs in a standardized, objective mann-ser (Erwars 'Persona-l1
Preference Scheduile). Since the participa~nts in groups are assumed to dif-
fer in the degree to which they manifest the various i~nterp~ersonal needs,
and a good de3al of roup3 activity centers around~ increasing awareess of
in~terpersonal needs andl finding aLpp~roriat~e sati~sfac-tion andi e .plession; of
thlem, it is felt that examiininrg changs in neced pa~ttern ilc i:l be a useful
method of studyingi-i the3 effects of pazrticipating in e~ncountr orl therapy
groups. Pr-eviculs re~search has vindicated th~a'; thiere are' ch~ange in self-
report of interpersonlal needs fiollowing successful. ther-apy. Inlcreases ar~e
reported on the AutonomyF and INurturance Scales, wlnile d~creaFises are
reported on~ the Succorance and Dep~endcency Scales (H~eller & Goldstein,
1961). Since several~ of the Edward~s Pe-rsonal Prefer~ence! ScheduJle (EPPS)
variables could be said to neasure a need for environmental Su~Pport and
a need for self-support, the EPPS woulld appear to be a suitable sel~f-
administered measure of therapy or group outcome. Th~e Dependencyy Scale
in parti~cular seems suitced to this project. Its use w~as suggested by
the work~ of several researchers (Heller & Goldstein, 1961; Bernardn &
Jessor, 1957; Gisvold, 1958; Zuckerman & Gross, 1958). The Depen~denrcy
score is; the algebr-aic stan of the Succorance and Defebrence scores less
the Aut~onon score.
Th~re research work of the Kaiser Research Foundation during the
early 1950's has been described as "!the most imrpre~ssive~ body; of emp~3irical
researds about m~e network~ of interpersonal effects stirculated by the
behaviors of Ss in interaction..." (Miueller, 1969a, p. II). Th~is study
follcr.ed thle lead of Learyi, Freedman,u La~orgre, and others (Freedman,
Leary, Ossorio, & Cofcfey, 1951; La"orge, Leary, ICaboisek~, Coff~ey, &
Freedrman, 1954; La~orge & Suczek, 1955; Leary, 1957) in a~pproadbingw
process reSeanrc thr~uigh the study of interpersonal behavi~or-. This
research group developed and applied! int~el-ersonal research schemata and
Instruments to a variety of sub~jects. Other researchers have applied
their work to famly interactions and psych~oth~erapy (M~ueller, 1969a,
196'9b; Maeller & DillinIg, 1968, 1969; Terrill & Terrill, 1965; and others).
The basic assuap~tion of this anpmrach is that one individual's
behavior will elicit predictable responses from the other parxticipantls)
in the system (origin~ally, a dyad). The affective component of
interpersonall behavior is also referred to as the int~eorpersonal there
Interpersonal ther:>es are defined as those attempts of an i~ndividual to
establish emotional. states in his interactions bhat tendl to elicit
predictdable responises from others (Mueller & Dilling, 1969), in other
w~ords, the "me~dium" r-eally is the "message.:'
An additional assumption of interpersonal th~ene analysis is that
when individua~als are in interaction (e.g., in therapyr), they wi~ll learn
which behaviors are rew~narded and which are~ punshed. In the thierapy or
group situation) the effect of such learning on the client or nonbrler is
to modify his behavior in the directions, of m~ore successful response pat-
terns (i.e., successfully with the therapist, group leader, and r~ou~p
mr~Jeabers (The thersapist is acknowrledged to change also; however his
con~tinu~ing experince with the therap-y situation wrill reduce the effect
of a-nyi partic-ular exp~erince as a function of the reinforcementi schedules
he is subjected to in! hi~s work.)
Acco~rding to th'is metthodl, behavior; are described
as interpersonally-oriented. responses which- can
be plotted a~rou a circ~umplex and defined in
terns; of two majol- ax~es: a domrinant-sub~n~ssive
ax~is and ;_n affiliat~ive-di~s afiliative axis
(La~orge & Su:c-ek, 1955). The basic proposit~ion
of the systemx Is tha~t all ;responises can? be
plottecd in terrE5 of these two H~dor axs and
u'nat tnese axes ase sufficienti to exalnain most
interpersonal behavior (MuReller, 19?69a, p. 8).
The intersections of the domi~nant-submir~ ssive and af~filiative-
diisaffiliative ax~es yield~s four rrajor quadrantsa that can be further
divided in~to` ctan3ts and six~teenths (i.e~.J single categories). Thne four
quadrants are: (I) Do~~ninant-Disalrfiliative, descri~ptively refe~rred to as
coupetitive~-hostile; (II) Submissive-Disaff'iliative~, descrip~tivrely. Irferred
to as passive-resistanlt; (ITII Sulamissiv~e-Aftfiliative, de~scriptively
referred to as support seeking; (IV) Dominan!t-Affiilaiativ, descriptively
referred to as sup~portive-interp:retive.. The four octants that are
expected to describe the th~eres of most leader interpersonal behavior are:
Puni~sh-Hate, relabeled Piunishin-g; Boast-Reject, rela'eled Dist~ancing;
Dominazte-Teach, relabeled Tesching; Give-Support, r-elabeled Supportin~g.
PunishingF and Distncin~g coliprise Quiadrant I; Teaching and Supporting
comprise Quarianti IV. Th~e 16 interpersonlal behavior categories are
described by a circusplex. shosing their relation to each other and the
major- axes (see Acppendixe I for an adaption of the orig~inal circ-mplex
and definitions of the t~henhes).
One o" th~e advantages of the Interpersonal Ther, Analysis
Technique (ITA~T) is the flexibility it offers. For thle purposes of"
this study, behasvior waes an~alyzedi by quadrant categories and, in some
cases, by octanlt. Th~emles could also be judged by; hemisph3ere~ or by single
categories. Th~ems cou-ld be juded. according to level of accessibility
(pdalio. conscious, and private levels), iegree of int~ensity (Freedm-an,
et al., 1951), or wh:-ethler they :were self-self, self'-therapist, or self-
ot~her interactions (Mu~eller, 1969a) .
Ine hypotheses that have been stated info~rmally earlier were
oerartionalized thrlou20 rhe use of the Initerpersona~l The~n~reAnl~ysis
Technique~ (ITA-T), the Leader Interpersonal Behavior Scale (LIBS) -- a
rmiodifed version of the ITA1ZT designed to study learder-appuropriate themes,
and the Edsa~rds personal Prefere~nce Schedule (EE-PPS). Th~e first group of
hypotheses conIcern1s the difference between the types of interperyonal
tihemnes expressed by~ active leaders (AD) and passive leaders (PLs) and
their respective~ grouips. 'The second group of hypotheses con~cerns
differences in the oiutcome~ of groups in each conditions!, especally with-
regard to dependency and sel~f-supIport.
Hyrpothesis 1: Leaders are expected to differ
in the number of times they intervene in their
respective groups. For the nurbier of leader
themes recorded, the specific hypothecsi~zed
AL > PL, for !~nunbe of leader int'e~rventions.
Hypothesis 2: Leaders are expected to differ
in the types and frequ~encyr of interpersonal
themes expressed, ajs me~asured by the LIES.
Th~e specific hypoth~esized! outcomex is
Teaching Themres > Surpport~ing Th~ers, for AL
Teaching Themes < Supportin; Thel-es, for PL
Punishing~ Themes > W~it~hholding Themes," for AL;
Punishing Themes < W~ithholcling Theme~s, f~or PL
AL, > PL, for Teaching Thene~s
A~L < PL, for Supporting Thremes
AL- > PL, for Funishing Themes
AL < PL, for W~ithholding heres.
Ryp~otheisi 3: Gjroup~s are expected to differ in
thle freqluncy of interpersonal theuesc expressed,
as measured by the ITAT, wihen stage of group
develomed-nt is considered. M~ore disaffiliative
and~c support-seeking themes and fewer sup~portive,
affiliat~ive theme~s will be found early in passively
led groups. Th~e re~verse w~ill be tru~e for actively
led giroups. The specific hI-ypothesized outcome is
Ai < P, for proportions of themes in Quad-rant~s
I, II, III during a-n early session
A > P, for proportion of themes in Quadrant
rV during an early session
A > F, for proportions of themes3 in Quatdrants
I, II, III during a late session
A < P, for pr~oportion of themes in Quadrant
TVr dur-iz4 a late sessionl.
:iypothlessi 4: Groups are expected to diiffer in
thle amou~~nt of change they s'ohow on measures of
dep~endency and environmental versu~s self-sucpport.
Groups ;in the active condition wuj.il show less
SWithho~ldingE themes ar classified as "C" or "F" ITATC theme~s dependiin~g
on the degree and over~tness of the wJithholding behaviorr. These corre-
spendi to "Distancing" and "Submissive D3isaf'filiation" on the LIES.
Th~re relationshipF should hold true in toth categories.
changp~E than groups in the passi~ve condition
on the EPPS Djependencyg Scale (Defere~nce +
Succorance Autonomy T score-s) and on several
scales that are hypiothesized to rieasure the
need for envii-tronmenal su~pprt. Scales that
me-asule other interuerscrud. needs will show no
liffernces between groups or will showl greater
change in the active condition?. The specific
hy~pothesized! outcomeh is
A < P, f'or change ~in EPPS Depenency Scale
A < P, for change inl EPPS Achievement, Autono~my,
Nurturance, and Diominance Scale scores
A < ?, for- change in E3PP-S Deferencee Succorance,
and Abasementl~ Scale scores
A > P, for change in EPPS Or-der, Exhib~lition,
Affiliation, In~traception~, Change,
Endlrance, Hete~rosexuality, and
Agresion Scale scores.
Hypothesis 5: In addition to overall differences
in outcome between subjects (SCs) in the active
and passive conditiion, indiiv~idu~als rated high and
10:: as dependencyv cn initial testing are expected
to show d-ifferent rates of daangez in Depend'ency
Scal~e (DPY) score on second tiesting. Hig dependenlcy
is dief;ned as a T score of 51 or above on thie
Dep~endenlcy Scale; Low dependency is a T secre of
h9 orbelow. The specific hypothiesize outcans is
A < P, for ch~age~ in DPY in High" DPY group
A > P, for change in DIPY in Low; DPY Group.
The hyspoth~eses w~ere tested using the .05 level of alpl~ha as thle
criterion for statist ical signficanrce. Rslsotie tta ee
(p <.05) have beenI markedt~ wLith an asterisk; other si~gniflcanlce levels
rrentioned belowI ar used for descriptive puroses an~d to suggest trends
worthlry of fuLrther study.
Ten encounter groups, consisting of one leader and 12 rmber~s,
were orgaizii~ ed. The participants were single rsale and fernale under-
graduates who had expressed interest in being in an encounter group and
met~l these cr-iteria (see Appendix 1): wrillingness to participate in th7e
research aspects of' the project,, not concurrently in therapy, and not in
more than one encounter group previously. Participants were told that
any beneefit~s they received from thie groups wJould be personal, no exper-i-
mental credit was involved. An equal number of males and ferrales wlere
assigned to each group,. The g-orops met f~or six w~eekly sessions that
typically lasted fr;Omi two to twlo-mod-one-half hours. Sone newy membersl~
wrere added at the second session but none were added after that. The
six groups that had mai:ainned the most regular and stable me-rbers~hip
(average n = 8) were, included in the data analysis. Tae total simple
(N = 50) consisted of 28 mal1es and 22 females.
Ihe group~is were to be run in accordance with the guidelines laid
downri in the leader's noanual (Appjendix; 2). Nlon-verbal exercises were diis-
couraged unrless t~hey sp~ontaneouslyi emerged from the group process in
order that the groups cl~oselyr reset 01e therapy goups, be as unliform as~
possible within each exrperime~ntal condition, and. that the raxcirmar amount
of interaction be recorded and availablee for later study.
Th~e group lEead'rershe second and third year male doctoral students
i~n clinlical or counseling psyhologyv who had had prior exp~l~erice with
groups (mootly as ambner-s) plus exper-~ience wlith ind~ividuail therapy or
counseling. They are best described as r-elatively inexperienced group,
leaders althoughl th-eir level of experience is prioabd3y typical of gradu-
ate student "therapists"! used in psychotherapy research. Prior to and
concurrently wiith the encounter groups, the- leaders participatied in a
didactic encounter group designed to famriliarize them with the techiqiues
and styles they w~ere to emp~loy in the groups, dleal with the prc'olems; that
arose, and fani~lize them with the styles of some experienced en~counlter
groq leaders. The information they needed was also stmnnarized in a
nlanual. prepared by' the present invecstigator with w~hich~ they w:ere expected
to become famniliar (App~endices 2 and 3). After thie initial sessions ,
during: wich the active-passive distinction w:as exp~i:.laed, th~e lead~ers
were designated as active or pcassive. ThIjs was dones on the basis of
their stylistic preferences anld mount of psrevrious extcerien~ce so that
both groups were app~rox-imately mated for level of experiences. Although~
the leaders were adare that active anid passive leadership was being
studied, they we-ei not told the hypotheses or method of analysis used.
The Edia~rds Personal P1fere~-nce Schedule (EPPS) wais administered
to participants before their first session and after the sixth. Only
the data of those participants who had completed both EFPS anld had attended
at least three sessions (never missing two' consecutive ones) were included
in the data analysis.
Th~e sessions were all audiotaped. The process analysis wras based
on pozrtions"f of the tapes of an? early and! late session (the second and
M i~ller aind Maley (1969) found that consecutive time samles and~ shorter
fif~th). Th~irtv minute sampnles were taken from each tape (approximat~ely
25% of each session). The samples were comp~osed of ten-miinutre segme~nt~s
takenr at pre-_fixed: intervals (ten minutes after the beginningg, ten
minutes before the end, and five minutes before and after the midpr3~oint).
Thle t~apest were analyzed using the Interpersonal Tfheme Analysis Technique
ITA'P) described by Muueller (INleller, 1969a, 1969b. Mueller and' Dillrin,
1969). Thne tap~e wlas racted In 60-second units. Leader in~ter~ventionrs were
analyzed within their original context and separately. A tapae consisting
of the ciumul~at~d first minute of leader inlterventions frromn eachl segmesnt
(six nidnutes per leader) plus a brief, contexutual lead-in and lead-out
was composed and analyzed using the Leader Interpersonal Behavior' Scale
(LIES), a variation of the ITAT designed for this study (see Appendix 6).
Thpe tape was rated lin 30-second units.
The tapes of the six grou~ps were~ rated by three judges as to
which categol-y the ,noups' interpersonal themes fell into. Althouyp
ratincsj were mrade indiependentlyi, thie lack of a written transcript ma~de it
necessry fo the udges to is agree on the nluder of scordale units
they hear-d, which actors) they were rating, and in what sequence. Th~e num-
ber of ccorable units was not always immediately and universally obvious
since the! judges had to ccasider the number of speak-ers in the tlher serple,
and the@ intensity an~d duration of their verbializations. The ruile of thatrl
followed was that~ context superseded affect, affe'tct surperceded content and
intensity supercepded duration. If the judges disage-ed ovn the nujer of
different uniLts, the usadanmn nusrber proposed was used and. the judg~e~s) wlho
but r-anomized time samples both canl generate the same~ representatives
informationn about the total g~roup session. Consecutive samping1 at pre-
fixed inervals was f~elt to be most appropriate in this study.
proposed smaller nuter' s repeated their scoring to indicate thet. all, the
unit~s welre s~imilcar i, their estimnaticn. Disagreement~s were typically be-
twreen, the uIse of" one or two:r ni:-its w~hen it was qauestioned: whiether or not a
new themie had. emrgd. TIhis runthod made It quite clear wrhat judges wherTe
rating and agreeing cn. 'Ihe filnal ratings consisted of speement; by two
of th~e three judges. Units for wjhi~ch no consensus could be rtached were
rated "'unclassifi~able" (Appendix 5).
Th~e criterion level for rather re~liabilityi, based on the results
of previous studies (Freedmanl et al., 1~951; Mueller, 1969a; Terrill &
'jerrill, 1965), wa~s set at, a mninimLm of 60%, for two-judg~e agr~eemnt.
Thlis represents moderatete' ag~eneant. Aigreeenet cf 80% or greater
represents "'nigh" agreeme1-nt.
Prior to the~~ actual rating, several students w~ere farJiiarize
w~ith the rating tecni-que and the scoring ca~tegcrri~e using a slightly
olaborat~ed. version of the scoring ma~nual deeloped by Crowder (1970)
under ~ue!ller's supefrvision (Anppndix Li). Fractice tapes wer~e rated and
difficult unlits werei discussed. The three students w~ith the highest level
of int~eragireeme-nt _n~d agr7eement with the investigator were selected as
judg~es. One- was a doctoral student in Educational Psychclogyi, one a doc-
-toral stuldent .in P"eorsoel Sernvices, and one a senior in Psychology. ?Two
were~ frL!1e and one was mle. The judges listened to the tapes in a
-randolcized border and they were not in~formed of thre hyp~otheses cr experi-
men~tal de3sign du~rin the judgbyi period. For theirT last task, proLdulcing
an empirical ranking of the leaders on activity-passivlIity, theyr wer~e
All4; parirs of leader segSrnans were ranked ucsing the forced-hoice rn3thod~.
familiarized with the "Definitions" handout (Appendlix 3). This was their
only exposure~ to the overall design of tihe project.
Thne ratings rGEwhich t~ne data analysis resi~ted! are believed to be
reliable since the per ce~nt of agreement among judge ranged(1 Ifrom moderate
to high. A 25% sazple of the tapes rated u~sin~g the IT-AT consistingn of
three tapES, on4 ratedl in each thirdi of the rating Sroup:'s life) showed
that t~he average agdreement betweea~cn any tw!o Utute3 :as 91.5% andC the
ave~rage ae-emen~rt anocng all threeF judges~~ was~ 48.12;. Using; the LIES, the
ave~rags judge agre:..2Ent ove-r a 900~ sa-ple (ve~ry other session?) of leader
tape was 69.79 for~ to j~udges ad 39.72 for thiree judges (Table 1).
Thie judges ranked th~fe segments of thie leader tape on activity-
pasivty Tere wa~S conplete agr.eement on i;hich leadiers were activJe and
which were passivre. Th~e ranke order was agred to be identical in th~e
early' and late session exrcet forn the inversion by one judge of two
rankingis within the passive sondition (Table 2). Comments by the judges
suGgested that their rankings wiereF baed in la~rge measure on differences
in voice qlualities (vollum and for'cefulncss) and f'requency of" verbal
intenrventions rather t~ha on differnces in? som of th~e formal defi.nin
qualities of active and paissive leadership.
Sanp~le $ 2-Judg~re Agr~eement % 3-Judge Agereeent
Interpersonal ThemeR An~alysis Ttchni'que
Al-I 87.7 44.9
Pl-I 93.0 42.0
P3-IIr 93.7 57.4
I;an = 91.5 Iean = 48.1
Leader Interpersonal Behavicr Scale
_____~ __I _~__~___~
Per Cent of J'udgy Ageement on
the Process Rat~ing Scales
Notie. -- Disagreecents unde-rlined.
Th~e 15 pairs of l-aer? see~ets
w;eiu anL(yzedi for leach session
__ ______ ~_ I~_
of ~eaders on Activity-Passivrity
Judg~oes ForcedB-Choice Bankings
1 Judge 2
Judge 1 Judy~ 2 Juidy~ 3
A3 A3 A3
Al Al Al
Table 3 indicates that the active leaders were ju~dged to have
made- 50%~ mor~e in~terve~ntions during the cumul~ated six, ri~nujts of leader
tape than the passive leaders. That is, within the same spae of time,
the active leaders used a greater numer of interpersonal behaviors in
their transactions with g~rou~p members. Using the average nuvb~er of intier-
ventions calculajted over complete sessions activJe leaders intervened
almost 80%~ more than passive leaders (Table 4).
Although the active leaders used a greater nurb~er of interpersonal
behaviors than? passive leaders, they had alnost identical relative ,fre-
quencies in f~ive of the six LIBS categories (Table 3). Passive leaders
were rated as using five tirs as many Subnriissive-Aff'iliativec themes as
active leaders. Visual inspection of T'able 3 indicates that the writhin-
grouip variance would exceed the betvween-group, variani~ce in this analysis.
All1 leaders used Teaching more than Supporting. This wras not pre-
dicted for passive leaders. Both groups used Distancing (the o~nly obtained
sample of Withh~olding) more than, Punishing, which was not pr-edicted! for
A mdreixe model analysis of variance u~sing condition (active-passive),
session (earlyr-late), and g~roup~s (reFpli~cationis nested in condition) as
variables showed little difference in the secup process. (TaDles 5 and 6;
Appedix7).Since the data weret gthered as prop~oirtion of responses in
_ ~_~___ __
- --~------I ~---
Al 37 3 1 16 15 2
A2 25 0 7 15 3 0
A3 28 3 4 16 5 0
N I= 90 12 47 23 2
Pl 9 0 0 3 1 5
P2 23 2 3 14 4 0
P3 29 1 4 15 7 2
NJ = 61 3 7 ?2 12 7
~ _~~__ ___
Analysis of C~umlated Fist Minulte of Leader Interve~nt~ions frmcn an
Ea;rl and Late Session U~ing the Leader Interpersonal Behavior Scale
Group n Punlishi~g
Distancing Teach;in-g Su~upportlng
Propor~tioin of Nj
Niunber of Leader Int~erventions in Early(I) and Late(II) Session
Samples Using the Inlterpersonal There Analysi Technrique
Group) I II __Group I =II ~
Al 11 21 Pl 8 5
A2 25 19 P2 13 11
A3 12 19 P 13 10
n = 48 n = 59 n = 34 n = 26
Mrean =16 Mean =19.67 DNean =11.33 Mean= 8.67
Active Group: N = 107 Passive Group: N = 60
Mean = 17.83 Mlean = 10
I. K K;
Active Passive Earrly Inte Ear~ly Late
Quadrant I 1.05i 1.28 1.11 1.22 9 1.12
Quad~rznt II .61 .76 .?2 .65 .72 .51
Quad--ant III .99 1.03 l1.10 .91 l.0i6 .91
Quadrant IV 1.37 1.04 1.17 1.24 1.32i 1.42
i ~rier Stateme~nts
Quadrent I 12 1.8 1.18 1.40 1.04 1.36
Qu~Lad an II .~79 .80 .80 .78 .87 .72
Quaran II 120 1.12 1.27 1.09 1.30 1.18
Quadrant IV .82 .711 .76 .80 .86 .78
Interpersonal Theme Analyrsis Technlique: Qucarant Analysis
All (Entser & Leader) Stateme~nts
co as \a r
r to a
o~n a r
LO E-- 01014
01 mm -tH
to @- n Lo\ C
Oa 0 E-
OO LL LA 0
01 L o\ r- u t--
0 E--s e-
each quadirant, they were subjected to the square root cf the are sine
transformation reconmnended by W;iner (1962). Propor~ti~ons were calculated
using all scorenble responses within th~e time sales and using the mremb~er
th~emes only. This all~s~ed for analysis of the total pattern of leader-
mrarber interaction as well as for mceaber patt~ernl alone.
Analyzing~ member and leader themes t~ogther, no differences were~
found comp~aring a-,tive and passive conditions in thie early and late
session in Qudrants I, II, or IV. There was~ a decre-ase in Submissive-
Affiliative (support-seekting) themes (p < .10) that closely approached
significance. Analysis of member th~enes alone showed no signficant
results. Th~e analysis of Quadant III supports the hypothesis that groups
decrease in support seeking over time, although no difference could be
attributed to style of leadership. In fact, the leaders rmust have
decreased sup~port-seekcing behavior also.
A compaison of F values for scores on thle 17 EPPS scales (Taole 9)
shows that pre-treatment diff~eretnces between active ande passive groups were
negligible, writh the exception of Exhlibition, on which group members in the
passive condriition tended to score hliger (q < .05), and Aggession, cn
which group members in the active condition tended to score higher (E <
.10). TIhe stepwise discriminanlt function analysic (Table 9) indicated that
the active and passivie group could. be diffrerentiated beyond p = .05 primarily
tased! on differences in Exhibition Scale scores. To control for this
difference as well as other minor sample variations, difference scores wyere
used in analyzing post-treatment cha~ngs.
Note. -- Scale abbreviations stand for' (1) Achiievemient1',
(2) Deference, (3) Order, (li) Exh~ibition., (5) Auton-
ony, (6) Affiliation, (7) Intracep~tion, (8) Succcra~nce,
(9) Dom~inanice, (10) Abaserlent, (11) N~urt~urance, (12)
Change, (13) Endura~nce, (14) Heterosexcualityi, (15)
Aggression, (16) Consistency, (17) D-pendence
[(2) + (8) (5) = (17)3.
Active Condition Passie~ LCondition All Ss
(n=24) (n=26) (N=50O)
Scale L-nS.D. FRan S.D. Grand M~ean
(1)A.CH 46. 29 12.911 46.38 7.76 46.34
(2)DEFI 40 .46 9.59 41.112 9.65 50.96
(3)OFD 42.46 8.47 41.50 8.59 41.96
(4)iEM S67 .04 52.62 10.18 L17.8C
(5)UT 58.7 10.68 56.92 8.80 57.76
(6)AF~F 52.17 10.17 54.81 8.72 53.5'!
(7))EE~ 55.04 9.88 51.96 8.99 53.44
(8)sUC 55.96 9.67 54.35 10.26 55.12
(9)DOM 42.42 9.78 42.73 10.06 42.58
(10)AB3A 48.67 9.23 50.58 11.21 49.66
(11)NUR 5'1.58 11.30 511.81 8.85 54.70
(12)CHiG 54.92 7.89 53.31 8.88 54.08
(13)EVD 42.71 8.59 44.12 10 .90 43.44
(14)BE-Tc3 53.62 9.33 55.15 7.10 511.12
(15)AGG ~ 52.79 8.911 17.81 10.48 50.20
(16)CON 53.25 6.65 53.00 9.98 53.12
(17)DP-Y 37.75 19.47 38.85 16.86 38.32
(1)ACGE 45.71 10.80 49. 04 9.34 47.44
(2)DEF 38 .80 10.38 42.58 10.77 40.76
(3)ORD 38.92 5.39 42.38 10.46 40.72
(4)KET~ 49.33 9.511 47 .62 8.09 48.44
(5)UT 62.37 11.02 58.54 8.28 60.38
(6)kaF 53 .04 11.221 54.23 8.31 53.66
(7!nJT 5'1.87 7.77 52.86 10 .23 53.62
(8)sUC 52.75 10.37 57.38 8.74 55.16
(9)DOM 45.08 9.38 42.46 11.19 113.72
(10).cBA 47.58 0. 48.04 11.07 47.82
(11)lNUr3 56.46 10.49 57.65 10.12 57.08
(12)FrGS 53.29! 9.19 30.42 11.47 51.80
(13)ENDlz 43.67 8.37 44.12 9.55 43.90
(14HET53.42 10.88 53.65 6.81 53.511
(15)AGc, 52.79 9.24 47.58 11.27 50.08
(16)CON\ 53.92 12.97 49.23 14.75 51.48
(17)DPY 29.92 17.69 112.58 1!.41 36.50
Edwards Personal Pre~ference Schedule-
e~an~S an~d Standard Djeviations:
H NO OCaA ~ ~ m
mm- am m eom a m
d~ AddU d O dd BN j NC7 d O dA' OC
orI man ~~c~monmemame=
O iMOC 3MT 0~C 3U UO~r
Oac I L~Jonmnnr;m mmamm; mH-- e
agen Lnt 3rlT0 L\CI mHCa embomo mm OH
in LO n o NO ~ e m 010 n~L~ c \Ot--r
COr i 0 0 az em mn inr-\ o i tn O \
O -d C O~ 110 No CO ium I-H O 03 Cr
Hl HHHl Hllr mHHH
I C~ mom~o-am.~memmmomm~
I On H HHH a\ HH- Liri ~5ML
*HON~~C ~ ~~
'i @HC U~o moom ro oomata
0 4 m mm b LO ~ mea e
>HQ moo om Cm~ C U O omm \( Lnamm e
O 0 3000OOOOOrmewH00 a
0 H~ *D HENMn~ mobe in u7 me
CA IIQ ~ LnC\ mam em oS m i3 emmHN C
a mo~~;~ommne mm mo e
i. . . . . . .
400 NHO c LrmN L03CCr Nr- ?D C
V` a V 0 0, a me air lr l
.90(i3,"0) = 2.23; 1~95(3,40) = 2.84; P.99(3,00) = 0.31.
p < .10
p" 2 .05
piHI < .01
Values of' F for E1::arris Per~so-nal Preference Schedule
Discrim-inant Fuinction An~alyses
A-tive x Passive Conditionl
F.9,o(1,40) = 2.8;; (~9 (1,0) = 4.08; F(99(1,0) = 731.
Condi'tion x( Sex (Active-?ass~ive x Zale-F'emale)
Analyrses of va~rince for~ all of th~e EPPS scales (Table 10) using
pre-grouF~p/p~-~ ost-rou difference scores indicate few changes in need. struc-
ture immediately followring the encounter gr~ous. The actively led groups
showed a decrease in Dependncy Zcale scores and theF pass'ivelyr led groups
showed a sight .inclease. This d~ifference, significant beyond the .05
level, was contrary- to expectation. A2tutonomy and Sucocrance, components
of' the Depende~nc Scale, showNed changes flavoring thle active -ondiition
(g < .10 anid a ` .071 rCIespctivelj), also contr~aryi to exp~ectation~. The
only com~ponenrt of the D-ep~endecy Scale that showed no change wjas Die~ference.
Exhnibition, on wh~-ich n~!ieater of passivelyr led gI~crous originally hlad the
higher mean, 5show~ed1 a decrease for that grou an a slight; crease for
mrbRiers of actively ledi groups. Th're not change, by which both~ groups
movedl closer to a mea T score of 50, was significant slightly above the
.01 level. Thie grea-te' amo3unTl of chalng ini the passively led! groups was
contrary to e-,ec taticin.
Astp::ise discrimi~nnt; finntionr enalysis of EPPS posttest scores
(Table 9) showedl tliet the youpsJ could no lo~nger be differentiated on
Exh-~ibition but co~uld be on the basis of" Succo-rance (p_ < .10), Aggession
(11 < .10, as beford:, and De~pedency (p < .01), the best post-encounter
Thle EPPS variables were also analyzed by sex-w:tin- thin-rou to
determine if any sySten71Ct~ic mle--femal e dif~ersence erxisted (Table 9).
Nc significant~ pre-group differences wNere found. Depede~ncy was the only
varidole that, siifican~tly differentiated thie groups (p < .05) Both
roles and fema~les in? the passive condition showed slight i~ncreses In
Edwards~ Pesonal Prefere~nce~ Schedule Pretest-_Posttest Diff~erences
Analy~ses of Variance
within F atf P
763 .4 '
-.9 0(1,4) = 4.54; F.95(1,
4) = 7.71; F( (1,4) = 21.20.
Dependency score. Females in tlhe active condition already had thle lowest
mean Dependency sco-re (T = 27.6), which decreased only sl~ightly. Th~e
greatesl enange was in thle mrales in the active condition. Their man~u
drpped from 45 to 34.
A bio-wayi analysis of degree of dependency by group, (Table 11)
showed that thle differential effects of the active and passive conditions
failed to reach the .015 level of significance. The relative change among
h~igh and los dependency group nmber~se is significant beyond the .10 level.
Righ dependency group nsrbfers acppaently de-creas~ed in Dependency score
no natter what group they were in, offsetting the differences among low~
dependency group, ~nhners, who showed an overall decrease in the active
condition and an increase in the passive condition. Th~e second part of
this hypothesis received somne confirmation in that Ilos dependency giroup
~ireabers showed more constructive change wnmen in actively led groups.
Source Mean Sq~uare F df P
Condition (I) 825.32 3.73 1 N.S.
LevJel of DFT(Ki) 121!9.33 7.011
Groups in Cond~iition I(J) 221.27 4
IK 424.60 2.38 1 N.S
I(J)K 178.13 4
Within 292.50 35
Analysis of Viarian~ce for Fr~etest-Posttest
of High~ ad Low De3pendency
F 00(1,4) = 4.54; 1.95(1,4)
a"p < .05
= 7.71; F 9(1,4) = 21.20
D3ISCUSS~ION AIJD CONCUJISIOS
Theg results of the data. ana-lysi- ge~neally5 d~id not. sport the
hypothes~e of ~this exp-erime-nt. Before thecretical assunpgtions are calledc
into question,, a nueber of zethodoslogical problems must be r-esolv~ed.
One problem! th~at needs more atitention is improving the quality of the
Judging. Th- judges in~ this proje-ct operated under a consider1ablle handi-
cap by having to rate the auiotipes withou.t thei benefit, of a type~script.
The qua~lityv of thle tapes wras not always good du~e to ex~cessive backiroul~nd
noise and on,.j ly moerate fidelity voice rep~~roduction. Thi- l:d to :I cer-
tai aoun o iiosncatc "roec Ive judgi~nz of thle tap;es that"i wasl
probablyis unav;;oiddle~l under the~ circums~tanlces r' -*.:criso
bet-ier quality and pr~ov-iding: trscJsripts ::uld redue thlat sigi;:.filcanly.!.;
Th~e use of 5jdge.s writi grreate- eperieEnce in .Interperacnl diagnoc~i~sl- and
more traiing with- then specific scales inviolvJed would prjcably.S resultl in-
greatcer unanilmity Emang~~1~ judgs and greater validty of their 1t-ratcins.
Although thel_ prdbl:en of thle validityi of theP ratings is no~t eaily sol1vd,
using sophist~icat-d jud~ges wilth cons-iderable experience in? psyJ~rchtnerap and
int~el-pe-rson2l d~iaga--csi~s seems to be a reasonable~ e y to assure- the; validit;y
as well as the relisbilityi of the rati~gs~.
Another problems, suggested by thle re3sults of thE: LiES analysis: is
airntaining quality control over t~he operati~onalized8 1inde'Condeni variables5.
Alth-ough th~e judges could dli'ffrentiate between thle tw~o tlypes of leade-~rs,
the different.'iiati~on w~as apparently based on beha-vior that w~asj for the
most part, irrelevant or peripheral to th~e defining features of activity
anld passivit~y. Th~e passive leaders frequently failed to intervene when
such an intervention would have facilitated the group's progress. They
tended to interpret their role to rean being quiet, sof~t spoken, inactive
leaders as opposed to being active, asse~rtive leaders. Thus, instead of
comparing member-leader interaction-oriented leaders-- with group pr'ocess-
orinte leader,, relatively active leaders w!ere apparently conpared with
relat'ivgly inayfe.Ly le~aders. The LIBS showed that passive leaders parrti-
cipatedi in the groups less frequently as leaders than the active leadersj
~i thast thley intiervened less and used mnore suppor~t-seeking behiavior. ThEe
fol~-rmer my have b~een? hppropriat~e at times, andi indeed was anticipated as
a gener-al difference, but the latter is que1StIonale. cSubmlssive-
Af~iliative behavior is generally nprrher-appropriate beh~avior, related to
a7ssuming a meaber role in the group (as it was iLn most; instances recorded
in this study~).
One~ of the ,fundamental assaupqtions of thle interpersonal theory of
beha~vior is that behavior results in con~t~er-b'ehravior (raeuller, 1969a).
If the leaders do not behazve differently, then the- nebers- of their groups
would not; be expected to respond differently. ~rTh diffrences that did.
ene~'~~ rge ay be related to the finrdin5 that participation in groups mayU
increase a pa-rtic~ipan~t's need. for dependency, particular-y~ if the leader
is inarct~ive (as opposed to active or passivre).
Petter control over the leade-rs' behavior could be oj'oained by one
or a cain~ration of thiese methods: better tmrainng of the group leaders,
in~cludng nlre efrec~tive role-induction techniq~ues; using experienced
leaders w~ho alreadyb f.t; the~ cri~teria? for the role models being studied,
rather thanl at'rrqteptin to train leaders to assume roles, closer optimallyl,
session-by-session)j monitoring of each leader's behavior to provide him
with inm~edia'te feedacck on his performance, using a test that could pre-
dict the type of leade-r an individual is likely to be as a sel~ectingZ and
scr~eenin device. In view of the behavioral portraits Sega~l (1970)) has
been ab1e to dre uscing the A--B Scale, that, might be anl appropriate pre-
dictor. W~ile, Bron, and Pollack (1970) have developed a test to detect
the orientations of potential groups therapy leaders that might also be
Another methodological problem is the length of the group.
Although~ meny Groups do not last as long; as six sessions, a nunber of
the predicted differences mray need additional time to b~comle apparent.
Without a clear rieasure of wher-e a group is in terms of the process of
group7 development, it would be questionable to assume that groups being
corrpareid are at the same stage, or that any given group is at the stage at
which it, is presumed to be, as M~cGrath and Altma (1966jt) pointed out.
Also, weekly variations in atten~dance, which no doubt diluted the effects
of the groups and slowied their progress," would have less influence on
the process and Dutatev if the groups Ket longer. A. great effort sh-ouild
be mrade to assure regular attendance by everyone. Unf'ortun~ately this is
not always feasib'le in an acadenzic setting.
A prnoblem slated to the length of thle group is the problem of
Th~ie Kent ~S-ltaet~e student-iNatina Guard confrontation aind conseq~uent stsu-
dent st~rik~e occu~rred near the end of the gr~oups and disrupted at least
one -ession~ of every group~- directly, and. indirectly~by leading to a
reacheduling~ of mitjtermn exan~intions.
whe7rn outcome~ should be measured. Id!eal~ly, a post-group evaluation should
be mrade shortly after the enxd of the grou and at several int~eral.s thee--
after. There are no guidelines as to how toiese intervails should be spaced
or has many would be useful. The example of' Karson and Wi~edershine (1961)
is instrulctive. They administered the 16 PF to rreabers of anarlytically
oriented psychotherapy groups. In a group started six. mnnthls prior to
testing, no Nhange was measured; in a group' thlat had been in p~rogres for'
18 mojnth~s, sigifican; changes wiere! found after another six months of
Th~e analyrsis of' interpersonal themes has show~n preaise as a
reseasrch strategy,ad the Interp~ersonal herl3 Analysis Techi-iquea wl
as its vJariant, the Lead~er InterPrsonal Behavior Scale, have been usefulI
tools in analyzing and coraparing group processes. Further wcrke is needed
to improve them as research instruments. Judges in this studyv enlcou-tered
sone of the same probslemsr Terrill and Terrill (1965) reported I~n their
study, YLSugting that there rway be particular deficiencies in the scoring
system that n~ee correction. Improving the scoring man~Fual and cla~rifying
t~he d~efinitions of' certain t~hemes (particularrly In the Domninan~t hemiisphere)
would go a lon~g way toward corrcting; those deficiencies.
Th~e results obtained have some= ilrplications for further research
in this area. The effects of in~active leaders resemble those reported by
Seligman and Sterne (1969), Salzberg (1967), and6 others~ for leade-rless
groups. If the rese:rblance is ore~x than superficial, then a similar
analysis of th-e outocan of sudh groups mayj showJ detrimental changes in
participants. The EPPS Dependency Scale should receive more attention as
a potential outcone measure since Lt is conlvenient, re-levnt, and is
apparently one of the few psychlometric. mea~sures that has shown sensitivity
to the effects of' therapy, particularly group ther-apy (e.g., see Anlker &
Walsh, 1961; Coffey, Freed-nan, Leary~, & Osso~rio, 1950; Pattison?, 1965) .
Although thi~s has not been a faLir test of the contention of
Rogers (1942) and others that thlerap~ist activity promotes dependency
while non-directiri~ty decreases dependency, the results su~gget that
therapist activity per se does not increase dependency; it is~ the quait
of the activity that could potentially do so. Thus it is not contradiictoryi
to advocate therapist activity to reduce meab~ter dependency. However, this
activity should be aimed at increasing amb~rers' awareness of their behavior
and reducinlg~ the need for dependency (tihrougia interaction with groa mrem-
bers or -- perhas -- focusing attention on the group process) rather than
h~ie analysis of the outcome by hidrh and low~ inlitial de~pendency
indicated that there~ myy be som-e value in considering the need-structure
of' potential merrbers wh~en organizing$ groups. This suggstion is only
tenltative. The- effect needs mor conclusive repDlic-ation.
A, final conclusion i~s in regard to the need for replications.
While ccapara~tive studies using two groups represent a vast improvement
over N = 1 studies, the adoption of the N1 = 2 design? as the standard in
group research will lead to endless controversies in the literature based
onl co~ntra~dictory _fin~ings. TIhe differences observed in this study among
leaders and ~their groups suggest that o~nly~ by summou~nting the difficulties
of large scale investi~-igatio will group r~esearchiers discover reliable
methods by wh~-ich groups can be made m~or e!ffective and increase our
understa~dndin of group processes.
Thc purpose of this studiy was to Inve;rstigateE the Iselation betwieen
tw~o co~ntrsting styles of' Lncoun~ter or therapyi goup lea~dership, labeled
active and passive, alnd! group process an~d outcore. Particular attention
has been gbr-ran to d'ependiency- as a process endi outcomeo variabjle. From a
reviewi of the Iliter-ature it was con~cluCded hat differesnt styles of condu~vcti~in
groups areF exps-ec! e to -- and dro -- affect gorou pr~ccs andA outrtcoe.
These d~iffercn, 7es have2 not becen syst~e:nati;calyi: exPlored an~d gneral agre~e-
rnent on their eff-ects Mnd their desirability is lackinrg.
Active ~leadrship nar~ bee-n deinpired he,-e as a styl~e of neveto
in which thle leader initiates~ an is involved in intelactions rwith the
merbers of ther Grou~. Th~e active leader plays an~ imporwtant role in struc-
turing andB dfiningr~ thea grou by setti ng rues for= interaction~, setting;
contingencies o; n after' 3ehavior, providiing behaviicral !:odels for the m~emr-
bers~, and Lgenerallyl "setting thie stag" ior prod~uctivei group p-ccesjs.
Active leadersc tt~::p~t to deal vithr dPependenc-y :.henpr it a;rises rs an_ ind-Iivi_-
dual problern rther thaLn as a grouJp p~henomanonr a the assau;7tion that
lintentionally :::sgnifying t~he probl~emn is of ddlicu-~s varlue (Faan~L, 1970) and
takes R~ay -rom the work~ of thTe g~oup: interpetrsonal growth (~gan, 1970).
P-ssive leadership hasl been, defined 2s a style of interar~ctin in
which ~ ~ ~ th cdratmt o minimize hris role in str~uctu~r~in an~d definin
the gr`oup and namin~ize U.e me'-~IDoers' role in providiing th'reir owyn s1;:tructur
The passive leader is group prlocesse-o~iented; his strat~egies area ainacd at
increasing members' awar~ueness of th~e group process anid ma~intain-ing the
interaction within the group membership. Passive leaders promote the
expression of dependency and authority conflicts on the assumption that
by bringing: them to the fore, thercan best be dealt w~it~h and resolved.
Thne process rese-arch technique, knownm here as the Interpersonal
Then Analysis Teichniqjue (ITAT), and a mlodified version devised for this
study, the Leader Interp~ersonal Behavior Scale (~LIES), were used to study
group leaders' interp~ersonal behavior anld its effects on group ramaters.
Interpersonal Theies, the affective components of interpersonal behavior,
were defined as those attempts of an individual to ectablish emotional
states in his i~nteractions thlat tend to elicit predcictable responses from
others (M3ueller & Dilling, 1969). By their defining c~haracteristics,
acti~e and passive leaders wjere assumed to differ in their interpecrsonal
behavrior. These differences should result in the elicitation of different
reciprocal beh~avJiors f'rom their respective group er~rbers. Specific
differences were suggested.
M~errers of passively7 led! groups, particularly shor~t-ter~m g~roulps,
are expcted. to showl more signs of self-support and less need for dependency
and environmental support following their groups than rnecters of actively
led groups. Selfspport~ has been! defined as doing for oneself instead
of manipulating and emotionally blackmalin others into doing it (Perls,
Changes in the! need paiterns of" group members were studied through~
the use of the Edwards Personal Preference Schedule (EPPS). Dependency
wJas op~eration~ally defined as the algebraic, stm of the EPPS Defere~nce and
Succorance Scale scores less the Aultonomy~ Scale score, following the
suggstion of Heller- and Goldstein (1961). Several EPPS scales were also
used as indicators of need for environmentally and self-support.
Ten encoilnter groups, each consisting~ initially of 12 members and
a leader, were oarganzed. TIhe participants were single male and ~feale
undergrauates w~ho h~ad expressed interest in being in an encounter group
and a~t these additional criteria: willingess to participate inl the
research aspects of the project, not concurre-ntlyr in therapy, and not
previcuslyr in more than one encounter group. N~o experimental credit was
involved. An equal1 number of males and females were initially assigned
to each group. Th~-e groups rret for six wieekl;y sessions that typically
lasted f~rom two to tw~o--and-one-half hours. Tihe si~x groups that had mrain-
tained the mos-t regular and stable mem;bershnip (average n = 8) were
included in the data analysis. The total sample (IN = 50) consisted of
28 males an~d 22 _females.
Th~e group leaders w-ere second and third year ma~le doctoral
students in. clinical or counseling psychologyr who would be best described
as relatively inexperienced group leaders, although their level of exper-
ience is probably typical of the graduate student "'therapists" used in
psychotheraapy research. Leadiers w~ere desigated as active or passive on
the basis of their stylistic preferences and amount of previons experience
so that both groups were approximately matched for level of experience.
The process analysis wa~s based on 30-mrinute sarpfles f"rom? the tapes
of an early and late session (the second and fifth). A tape, consisting
of six. minutes of leader interventions, was comp~osed and analyzed. using
the LIBS. The sampling procedures emp~loy~ed were felt, to adequately
represent the total groul process. Miller and Dlaley (1969) were cited
as establishing the representative nature of consecutive timei samples.
The tapes w~ere rated by three judges as to which category the
groups' interpersonal themes fell into. Scoring: was based on agreernant
by two of the three judges. Thle judges also ranked tne leader tape
samples on activity-passivity .
The average agreement of any two judges on the ITAT was 91.5%
and 69.72 on the LIBS, which was felt to represent high and moderate
agreemnt respectiv-el~y. There was complete agreement on which judges wrere
active and which were passive. Commnts by the judges suggested that
their rankcings w~er based in large measure on differences in voice qurall-
ties (vojlumre and forcenl1ness) and frequency of verbal interventions
rather than on differences in some of the formal defining qualities of
active and passive leadership.
Active leaders erlel- found to have intervened more frequently
t~han passive leaders -- almost 80% mo~re over al~l tapes sarrpled. The
relative frequencies of types of intervention were atmo~st identical for
both types of leaders with the exception of sup~port-seeking behavior,
which the passive leaders were judged to have used! five~ times more fre-
quently~ than active leaders. All leaders used Teaching themes more than
Supporting themes, which was not predicted for passive leaders. Both
grouCps used more D~istancing than Punishing themes, which w~as not predicted
for active leaders.
No sig~ifican~t differences (p .05) were found ~h~en comparing
active and passive conditions in the early and late ssin sn h
Pre-treakent differences on the EPPS scales were negligible,
with the exceptions of Exhibition, on which individuals in the passive
conditions tended to scol~e higher (JE < .05), and Aggressi.on, on which
-individuals in the active condition tended to score higher (p < .10).
To control for this as well as other sample variations, difference scores
were used in analyzing post-treatment changes.
Few chains in need structure were found immdiately ,followi~n
the en~counrter groups. 'IWe actively led groups showed a decrease in
Dependency Scale scores and the passively led groups show~ed a slight
increase. Thiis difference, sialificant beyond the .05 level, was contrary
to expectation. Autonomy~ and Succorance, conponents of the Dependency
Scale, showed dnanges favoring the active condition (p_ < .10 and p < .01
respectively), also contra-ry to expectation.
A stopwise d~iscriminant function analysis of" EPS posti-test scores
showed that the groups could no longer be differentiated on Exhibitijon,
but couLlld be on the basis of Sulccorance (p < .10), Aggression (p < .10,
as beforeE), and~ Dependency (q < .01), the best post-encounter group
dis criminant .
Nou significantt pre-group~ differences based on sex of participants
were fou~nd. Dependency was the only post-group1 variable that sligniicantly
differentilated the participants when grouped by sex (p_ < .05). Thne
greatest changes in Depoendency score was among males in the active condition,
whose mean T score dropped from 45 to 34.
When, difference scores were analyzed according to level of
dependency (High-Low)~, high~ dependency participants shaded a decrease
in Dependency scores no matter which condition they were in (active or
passivee, and low dependency participants showed an overall decrease
in the active condition and increase in the passive condition (p_ < .10).
Lowv dependency participants apparently showed more constructive change
when in actively led groups, givingS some support to the hypothesis that
leadership style has a differential effect on participants according to
their level of dependency.
Althouga the data analysis generally did not support the
hypotheses of' this experiment, several methodological problems must be
resolved before the theoretical assuanptions are called into question.
Thie reliability and validity of the judging need to be increased.
Several metthods were suggested to do this. Stricter quality control
needs to be maintain~ed over the operationalized independent variables.
Methods for assuring that leaders fulfill their roles well were suggested.
It wjas also suggested that greater efforts be mae to stabilize weekly
attendance and take level of group development into consideration.
Longer groups were recommended when feasible. It ws recommended that
repeated oultcome~ measures be unde w~hen feasible.
The research potential of the Interpersonal Theme Analysis
Techniques and its variant, the Leader Interpersonal Behavior Scale, was
reaffirmed. Suggestions er~ew nade to impnrove~ them as research instruments.
The v-alue of the EPPS Dependency Scale as a convenient, relevant, and
apparently sensitive outcomei measure was also n~oted. The need for further
research compcaring more than two groups at a time~ was noted. Large-scale
invrestigation~s are needed to discover reliable naethods by which groups
can be mlade more effective,
Alt~hough th!is study did not yieldi a satisfactory test, of t~he
hypothesis that the~rapist activity p~romotes depe-ndency and non-dtirectiv~ity
decreases depend~ency, the results suggest that therapist, activity per se
does not increase dependency;; It i~s thel n~atulre anld quality of th7e activity
DESCRIPTION OF: E:C~OUNiERl-i GROUiPS AND IIB~ER('S
CONTHAFl~CTi CIRCULATEiD TO SOLICIT VOLUNTEEr~RS
Encounter Giroup R~esearh Project
ENTCOUNiTER GROUPS FORMIIING FOR SPRING QULARD~ER
Me purposes of encounter groups are: to allow~ individuals to
meet other individuals as wjell as themselves in ways that are not,
usually possible in social situations, to increase personal and inter-
personal s~ensitivity, and to assist the participants in the discovery
of their own personal, academic, and social goals.
Between eight andl 12 encounter groups wrill be ru*n during the
Spring~ 1970 Quarter by qualified graduate students in clinical and
counseling psychol~ogy. All th~e gr~ou~p leaders will have had previous
experience in group worki. Each group will consist of an approximately
equal number of students of both sexes.
The g:roups will. meet weektly during th~e quarter- for a twv~o-hour
session, at times and places to be arranged.
Thenre- will not be a fee for participating in a group.
Students w~ho are interested in~ applying should me~et the following
1. Potential participants mulst not have previ~ously b~een in more
than one encounter group (of any3 type) or therapy groupu~.
23. Since one of the purposes of the project, is to evaluated
encounter group experiences, participants are asked to agree to the use
of several data-igathelring procedures (e.g., a pre- and post-group
questionnaire). You will be idenltified! only by your first name and
only authorized project work~elrs w~ll. have access to this confidential
3. Pat~icipan~ts should un~derstand that an encounter group is
nct; a substitute for th~erapy. If you feel youi need counseling or
therapy, please speak; to somone at the Counseling Center (311 Lit.)
or the Scudent M\E~ntcal Health Servrice (350 Inf.).
LI. If you arne in therapy, you muPst have the consent of your
therapist to participate i~n ma~ encounter group.
5. Participants should understand the inrportanlce of attending as
mnany group sessions as possible for their own benefit as well as for the
EGRP's- and should not volunteer if thley expect their attendance to be
ENTCOUR~IEIR GROUPi RESEARCH PROJECI':
1. Data Gamhering andJ Confi~dentiality
Ideally data gathering should not interfere with the proup, process
or leader-enter. relations, Whmether or not this is the case depends on
how intrusive the neasuring procedures are and the leader's attitudes
towar-d it. In areas of uncertainty for the group mrembers, the attitudes
of the leader, conveyed non-verbally as w~ell as verbally, provide the
members with a model to emullatie, an example of certainty and confidence.
Thnerefor~e, it is important that the leader not feel conflicted about the
research or ethical aspects of the project.
Aill group participants (Ss) have agreed to fill orut the pre- an~d
post-group measres and allow the sessions to be taped as a condition for
being in one of th~e groups. Wle ask participants to engage in this fair
exchange of services with us which should not cause a leader to feel
eaerrbnassed in anyi way. Research such as th~is (wh~ich could not be con-
ducted ethically without such an agreement) increases our un~derstan~ding
of group processes and helps us make groups more eff~ective.
One of the reasons successful groups occur is thiat the group
mrbxers perceive the group setting as being a safe, private place. Thnis
is promoted by committing everyone to the ethic of "dhiscre~et silence"l
about the group proceedings. Miost everyone will wanut to discuss what
has happened to them in the grouup, particularly if it wras a significant
encounter. I always begin a group by asking the participants not to
gossip about other n~nters of the group and not to nEntion anyone by
name outside the group without that person's permission. I acknowledged
that to expect secrecy is unrealistic but that they should use discretion
in1 discussing the group. I ask that anyone wrho feels unable to do this
should leave and that anyone who has any doubts about th~e ability of
others to do this should discuss this with the group. Dealing with this
frankly and seriously means that it probably wrill not become3 a problem
and. also sets the tone for the group. Th~is is doubly important when
research is involved.
The data-gathering consists of the EPPS and tap~e-recording. The
first EPPSj should be handed outsat the first session and is due back by
the second session. Remind people by phone to re~turn, it on timne. Ask
Ss to take the EPPS honestly, for themselves, without trying to meet
anyone's expectations of "correct responses."' Tn~ere are noine. All
sessions should be r-ecorded. The second EPPS wi~ll be handed out at
the last session and collected later. Ask Ss to usje pencil and return
booklet and anwer sheet. To assume confidentiality, every leader should
assign a nurrber~ to every participant. Keep a record of this number. I
wail eventually add some prefix to it~ so that every partici~pant, will have
a unique nurnber. Although a record w~ill be kept of nama-s and numbers
for~ comnication purposes, each participant (and groups leader) will be
identified only by t7urber for re-search purposes. This nurnber should
appear on the EPPS protocols -- no names. During the group sessions,
individuals should be introduced and referred to only by first names.
At any time, if group menters start to refer to persons and incidents
outside of the group, they should be asked to deal with the issue in the
here-and-now of the group situation. One of several alternate strategies
wvill be used to mak~e sure that the judges are n~ot f'arilir width anyone
whose voice they hear on the tapes. Group members can make this job
easier by folla.sing the above-metntionedi suggestions.
As far as feedoack on the t~esting: is concerned, I am in favor of
sharing infonmaion with _Ss but feel that this must be left to the
discretion of the group lea~der. If you feel you will have the time and
desire to explain the scores to your group rKembers, I will give this
information to you when it is aviailable.
A last researdT problem that needs mentioning here is the use of
tape recorders. W;e are p~ersonally responsible for th~em so use caution in
all your relations writh the eachin~e. Fake sure you are r-ecolirdin at
1 7/8 in~ches/second~ and on the correct track. YouL wiiill be given recording
tape desine~d. to hold four two;-houlr sessions, one cn ealch track;. Use
thne enclosed index card to note the: date, session nu~mber1, tracks number,
and group leader for each recording. Record this infcremation at the
start of every~ track. !hen star~ting a traCk, leave enough1 unused tape
so th~at thle tap~e won't slip off if rewround. :lird it to 5, then reset
counter to 0. DC! NOT iiEWINDr TAP~E. :ake sure the mlicrophone is properly
plugged in, the monitor switch is on "speaker off," th~e playback volume
is at "0," the tone and volume level are pr~operly adjusted anid the record
button is depressed. ( hien volure is "automatic," the volume r~ete does
not operate.) If you fail to do any of these, th~e session wrill be lost
to the project, so please allow; yourself tire to adijust the mrachine before
the group starts. Find a place for the microphone that allows it. to
pick up~ everyrone's voice clearly writh a niniz~-rtmi of backgr~toundi noise and
Start thle nachinle when the group is scheduled to begin and let it
;-un until the session ends, even if it goes a little over two i-ours. Do
not, encourage sessions that last longer than that, but use your judgment
as to whether something should be completed during a given session or
can be carried over to the next. If new material is presented during
the last 15 minutes, you wsill probably run over, U~se this time to allow
everyone to get their "last; licks" in. If a session does not seem to
have enouds momentum to sustain itself for two hours, you might ask the
group if they wan~t to end early. NJo session should las~t less than an
I must, return the recorders each morning or I'll be turned into a
pumpkin, so please make sure arrangaements have been mrade to pick up the
2. Leadership Style: TySpe 1
Th~e first thing that all leaders shlouldi do is tak~e care of the
administrative matters explaini' confidentiality. data-gathering) described
above. AIfter that has been taken care of, the first session begins.
(The TR goes on nowJ.) lyipe 1 (Active) leaders should have reviewedl the
"Def'initions" handout before coming to the group. The f'ollowing Infor-
mation is meant to amnplify the maiter~ial discussed during the training~
groups and in the earlier handout. You might begin by exp~lajinig to ouvar
group that your role i~s to facilitate interactions among geoup1 merrters
and that the responzsibility for what happens rests with the members
themselves. Expx~lain sone of the basic rules that you w~an~t them to
1. Stayi in the here-and-now.
2. No gossiping.
3. Alwrays ta0C~ to someone, not about them as if they were invisible.
Don't talk "in general,"' talk in specifics to someone or eve~-ryonle.
Mrake personal statements only (1st person pronouns).
4. Don't attempt to speak for others or tell them whiat they-j really
5. If you don't want to answer anything, say so. Y'ou my be asked
what your objection is, but you don't have to answer the question
or state your objection if you don't want to.
6. Question~s should be honest questions, not statemnts ini disguise.
7. Tryr to deal with issues on a feeling instead of intellectual level.
8. "You should" statements as movral prescriptions should be rephrased
as "I wrant you to" statercents.
9. Confidentiality: don't discuss group business outside of group --
especially with group nencers -- while the group is still m~eeting.
Ujse discretion later. "Discre~et Silence."
Obvriousliy they waill not abide by all of these rules immediately, but by
stating them at the start, you will be perceived as st~ructuring the
group. Riemlind nant~ers when they are vioclating specific rulies. AfTter
the rules have b~een explained, "M~ake the rounds." Th~at is, start by
asking each individual to tell the group wheree he is" anid wh~at he's
feeling here nd ow (Whnenever it's not clear what's going on in the
group (e.g., prolonge~d silence), this is a ulseful~ technique to employ.)
You might want to go around once before asking~ anyone to elabsorate or
you might want to encourage someone to elaborate as he brings out some-
thing of interest. Group participation can be encouragd by asking,
"How does the group feel about that?" or someh~ su1ch question.
I'd like to introduce a distinction not made heretofor, namely
the difference between the leader as leader and the leader as memb~er.
In "Definitions," I described the active lader, errphasizing his rol in
str-ucturing an~d guiding the gr-oup toward successful encounter. Although
the Type 1 leader is expected to be warm, empathetic, self-congruent and
genuine (etc.), he is not a membiSer of the group in the sameR way that the
Ss are. these e yu~c-wth-enhanin qualities should be apparent in what is
said and done, buti the leader will not seek to draw attention to hlis
person -- get in the "hot seat" -- as membrs will. Although he may
share his feeling ~swhn the group "ma~es the rouindis," he may n~ot if that,
seems inappropriate. In other wo-rds, ~Tpe 1 leaders will participate
actively as leaders but passively as memb~ers. Group m~embers should be
able to remeir~ber your personal qu~alities without k~nowing a lot of the
content of your personal life. It is possible, for examle, to identify
with a meatber's difficulty ("I kn~ow how~ frus~trating that can be") without
getting deeply involvd in your difficulty. Should it happen that you
have an insight or somnehowy grow during a group experience, this certainly
can be shared with the group mrerbers, but you should not try to consciously
deal with a hianig--up of yours in t~he group. Th~at is the purpose of th~e
Leader (emp~athic with lowi
(emoathic wJith A AL/AMI PL/AM
P AL/PM~ IPL/PM
Th~e fiur~e above shows four possible cochinationss of leader and member
roles. ne are using th~e twro lowesr cells in this phase of the ERGCP.
Iember role is beingF held constant and Leader; role is being studied.
Being active does not mea~n always being~ in th;e spot light or
being a part of everyi (or even most) interact~ions. Everyyoe must do
their own growing. Thle leader tries to promote growth situations and
help reanve stumbling blocks to growth (e.g., correct projections, mis-
perceptionrs, focus on feelings). Ashen an individual wdould, in your
estimation, benefit, by your intervention in h:is effort to work through~
an impasse, then center the process but remrember that you can only igelp
someone do hris owrn work,; you cannot do it for hlim. Sometires, the best
thing a leader can do is stand aside, ready to assist, but in the
3. Leadership Style: Type 11
Type 11 (Passive) leaders should familiarize themselves with the
"Defin-itions" flyer and handle adainistrative matters in a way that is
not likely to conflict with their leadeirship, role. Th~iis may be diffi-
cult, but please tryr to do this. It may help to chang posture o-r tone
or something to indicate that what you are doing as administrator is
different from your role as leader. You miglt, ref~er to yourself as a
"facilitator" cr "enabler" rather than lead~er. After the administrative
duties are taken care of, start the tape recorder and say scaething like
Encounter groups allow individuals to relate to
themselves and others in ways that are not always
possible or easy in normal social situations. Wh~at
happens here depends on what you want to happen.
So why don't we begin.
The leader then lapses into silence and lets the group start
developing. The leader should resist attempts to get him to structure
the group. Possibly you could reflect the intent of such behavior
("You~'d like me to tell you what to say." "W:ould you like to talk about
your anxiety?") or ask "Hiow do you feel now?" Verc3al interventions
should be .infrequent and brief. If youl find yourself talking as much
as the average member, you are probably saying too much. Howr much the
group will push you to activate them will depend in large part on how
verbal they are. If the silence lasts more than 10 minutes and the
group seems to be having trouble getting off the ground, you flight wonder
out loud about what the difficulty in talking is or what seems to be
happening. ('Ee samer can be done for groups that chit-chat instead of
really di~alogure.) TIhe latter is a question the Type 11 leader will use
as ane of his more frequent types of interventions, since it asks the
unbSership to look at themselves and each other and verbalize what is
happening, wihat they are doing, or avoiding (a frequent problem), or
trying to do. After these process questions are posed, the leader can
usually draw back and. let thne group handle them.
Then same ground rules apply for Type 11 groups. HowPeve, thle
rules are introduced slowly. Th~e rule of confidentiality and discreet
silence is prchably the only rule that needs to be clear from the start.
Bring up additional rules as they seem appropriate. Thlat is, when someone
seems to be violating a rule, ask him if he would do whatever you'd like
to enlcourag; e.g., "could I ask you to speak in the first person" or
"in the here and nosr." Try to get the groups to discover the rules
themselves before you try as direct an approach as that. Foor exranple,
when someone has the habit of speaking for everyone or in generalities
("It's a known fact," "_People say," "Everybody thinks that"), you might
interrupt and! ask the group how they feel about being spoken for, or
included, in such a statement. If you don't feel that it would be pro-
ductive to wait ftor the group to become aware of your purpose, ask an
individual to consider your suggestion (as the earlier examplee. Type
11 leaders generally will not address suggstions to the roup; they
will ask an individual to do something.
4. Clinical Judgement and SomeR Limits
Th~e major responsibility of all leaders is to see that a growJth
enhancing environmeRnt is provided and that no harm is done to anyone.
Ihrere is always a risk that someone does not have the self-support (in
this case, ego defenses) to handle a given encounter or the encounter
process itself. It is one of the functions of the leader to use his
,judgement and intuition to estimate~howz saf'e everyone is. Whren it seems
apparent that scueone needs additional environ~mental support in a given
situation, it is your responsibility to provide it. Sometimes this neans
assuming a protective attitude towr~d somecn who is being att~acked.
Sometimes a neRmber needs to be taken, off' the hot seat. Askz him if he
wants to get off and in those situations where you feel that it would
be unwrise to risk continuing, tell toe group that you think it would be
best to let tne mastter drop. You mray want; to ask someone to leave for
a few minutes or longer. You rrny want the whole group to take a short
Occasionally, someone gets upset during or after a group. TIry
to anticipate wrho this might be so that you are not caupt unaware. A4t
some point, perhaps after a very intense session, let it be knownl that
you are available to talk to anyone who feels they would like to discuss
any u~ncoinfCortahee feelings that they might have. ('1his is to be used to
deal with personal anxiety and should not be used to continue the group
or develop cliques.) SaySingj this diplomnatically is very imp~ortant, since
you will be dealing with a very delicate m~atter. You can refer them to
me if you feel this would be advisable. Mo0st people can w~ait until the
next day to be seen. Determine if th~is is the case. If a crisis seems
to be developing. you can take or send the person (p~referably with an
escort,) to the Infi~rzar for medication, crisis therapy, or admission.
If anyone becomes anxious e~noug1 to need crisis th-erapy or admssiion,
please call me. Ylou can sugest a person leave thle group. You can,
suggest- (and arrange) counseling for anyone.
While there is no real growthn without some growing pains (sufferin,
anxiety, deprescsion, conftaion~, etc.), w~e know. better tha~n most grop
mem~~ber-s what theyi may be getting into an~d unless w~e wanSt to take sone of
the responsibility for turning a "nontail n;eurotic" inlto a "patient,"' we
must set sce limrrits on wh~at can occur in our groups. Some additional
Limits are: no using alcdhoolic bevierages or drugs during; groups, although
talking about them is O.K;. Nuo sexual contacti, al~thougi touchiing is O.K.
(Just watch those hands.) These shoulldnl't need to be mentioned ~unless
they specifically come up.
Odds and E~nds, After-Tloughts
I have a limited number of EPPS booklets, so I dust have them back
to distribute to other groups (and retulrn to their righntful owners) as
soon Es possible. Ask participants to fill it in and bring or send it
to rE at the Counsseling Center, 311 L~TTL within a few days or return
it to you by the second session at the latest.
Please keep the reel of tape and empty tak~e-up reel in a safe
place -- away from children, heat, an~d magnetism. I'll supply you with
a tape recorder and adapter plulg for every session. Make sure you keep
an accurate record of which tape track you record each session cn.
Newv Inf'ormration Policy -- all participants willl be given a
department~al feedlack forn on which they can voice general satisfaction
or dissatisfacticaLo with our activities. Thley can also request nore
feedback at that time. If anion~e asks, tell theml arrangements wll be
mrade for feedback to those interested when the project has ended.
I'll wuorki out the logistics later.
[NuOTE~. -- Thne "General Defin~ition" section of the
following process scales were: also included;
A Tentative Scale for- the M~easureme~nt of
Accurate Empiathy, A Tentative Scale for the
M~easure~ment of Nocnpnossessive !armth, A
Tentative Scale for the M~easurement of
Therapist Genuineness or Self-Congrunce
(Carkhu~ff h Tr~uax, 1967).]
DE I'ITIONIS OF ACiT~IVE ALD P:SSnEE
GIVEN'i TOFI GROP LiDES NO~ TAP ANALYSTS
Tw;o ternrj that need defining are active and passive, neither of
which has a standard, accepted meaning in therapy. Both terves ar being
defined here relative to each other and to the behavior of a hypothetical
average group meat~ier. Active means causing or initiating movement or
activity, It implies involvement in activity. An active leader is one
who by hi~s behavior initiates and directs the course of group interactions
an~d is involved in the interactions. The active leader plays an important
role in struct~uring and defining the group by setting~ rules for inter-
action, initiatingr and entering into interaction with the group remoeprs,
and by providing one or more behavioral models for the group members.
Ihe active leader is dirctive in the sense that he makes suggestions
and "sets the stage" so that the group process will be productive.
Reflective statements by the leader will not be considered indicators
of high leader activ~ity since they generally serve to maintain the inter-
action within the group membership rather than involve the leader in
structuringi anrd entering the ongoing process.
PassivTe means not acting, not engaging in open action. A passive
leader is one who atte3Tpts to direct only minimally the course of group
i~nter-actions and whose involvement in the interactions is usually not
cbaracterized by overt action. Th~e passive leader tries to minimize his
r-ole in str~ucturing and defining the group, and in interacting with group
mr~Jembrs. He provides few behavioral models for the group membSers. In
fact, his r1ole is largely in contradist~inction to the role th-e typical
group marr~bers ar-e expected to assure. Reflective statements, questions
(e.g., "How~ do you feel about that?"), and quiet reinf'orcemnent will be
typical of passive leadership since the passive leader airns to maintain
the interaction within the gSroup medaership.
Th~e group leader's approach to encounnter is grounrded in his
personal development -- i.e., i~n terms of the qualities that have been
found to be cormmon to effective therapists (Rogers, 1961) -- as wrell aE
in his. development in terms of his particular theoretical position.
Leaders will be expected to demonstrate their positive therapeutic
qu~lit~ies within the context of their style and technique of leadership.
There is somne risk in assigning leaders bolank(et labels like
active or passive since consistency in style is not always attainable
or desirable. A~ct~ive leaders w~ill on occasion behav~e passively and vice
ve~rsa. However, I do not at this time wan~t to analyze the group process
in ternrj of wh~at follows specific units of active or passive behavior;
I wiant to ex~r~ine the overall effects of these styles. Though I lecog-
nlize a diffewrece between units of leadership behavior and style of
leadership, I will designate a leader as de facto active or passive
according; to the overall impression his behavior makes -- in terms of
the relative balance of active nid. passive behavior over the course of
iiothing here should be taken to imrply: that either style of
leadership is at all times appropriate or of higher value than the
ENJCOUNTE~R GROUP EESEARFCH PROJECT':
Scoring Manual for the Interpersonal
Th~erre Ar~alysis Tlechn~ique
Note. -- This scoring~ manual was adaptedi from the one assembled by James
Crowder (1970) Th-e information contained herein was originally
published by Fr~eedma~n et al. (1951) .
Th~e interpersonal circumplex consists of 16 reflexes (categories)
of interpersonal behavior, into which all interpersonal behaviors may be
rated. It is divided into quadrants by orthogonal axes. The horizontal
axis covers the dimension of dominance-submission, while the vertical
axis represents the afiiiative-disaffiliative (or love-hate) diceRnsion.
In rating behaviors into categories, the behaviors are first
ju~dgd in termsr of the ax~es, and thus the behavior-s are placed into
quadrants of the circ~unplex. Tnen, a behavior is judged into specific
category within the quadrant by matching it with the descriptive terre
of those categories. Statemrents some~tinrs include behaviors of more than
one categoryr, in which case amltiple scoring should be teed.
Problern arise because  the categories are not mut~ually
excluisive,  the meaning of behaviors is determined partly by the
context in which they occur,  affect and content (i.e., words) are
som~etiress incongrent, an~d [II] r-aters mnay use different levels of intler-
pretation. These problems are demo-nstrated beloar by the use of a few
Con~side~r the client statement: "I like you." If this statesmen
were genuine, it would be rated "M." If it were said sarcastically, it
would be rated "D." If it cars after an interpretation which the client
did not w~anlt to deal with, it would be rated "F."
For another exarple, consider the following client statement,:
"You looke tired today." If this statements connoted genuine sympathy,
it would ce rated. "N," If it came~ out of the client's uilt, for seeking
help from the therapist, it is possible to argue that it should be r-ated
"Hi," but. this rating would require deeper inter-prtation than the
The client statement, "I don't trust you," impolies distrust
"G" and rejection "C." It is necessaryJ to choose one or the other in
this rating system.
In rating the client and therapist behaviocrs, the following
priorities are listed so that the above problem w~ill be minimi~zed:
 _Contet taes preicedence over affect;  affect- takes precedence
over content; and  interpretation does not go beyond the irmrediate
Three ty~pes of reported client-to-ather behavior are scored.
These arre  client's reports of actual interaction with others,
 clienrc's fant~asized interaction writh others (includes wishes, de-
sires, should-haves, and fears), and  clien~t's feelings about others
as reflected in his statements about them. The following examples
illustrate these categories:
 C: ";My parents told no that I shouldn't get serious about
anyi girls while I'm here. I told them to stayr out of
 C: ":I wish I had some close friends,"
C: "I'm afraid that people will reject me~."
C: "I should have told her off."
 C: ":I distrust ngr parents."
C: "They are selfish people."
Below, exmol~es of behalvior- for each category are listed, a~nd,
where deeme~d helpful, exp~lanatory~ statements are included. It is
impossible to provide exasp~les for some of the mraningp of some re~flextes,
because the meaninbs are soreatimes very dependent cn the tone of voice,
e.g., sarcastic behavior (reflex "D"),
Categoriles of Interpersonal Beha~vior
Aim: To reproduce accurately the emct-ional state the actor is attempting
to establish; to identify the theme of the actor's interpersonal
behavior. Enpathize with t~he individual whose behavior is being
r-ted from the position of the object or obJects of the activity.
Priorities: 1. Context takes precedence over affect
2. Affect takes precedence over content
3 Interpretation does not go beyond the immedRiate context
Quadrant ELDE~ (Domrinant-Disaffiliative; Competitive-Hcstile):
B. Boasting, Self-Stirmulat-ing, H~arcissistic, Intellectual-
C. Rejecting, Withholding, Conpeting, AccusinA, Arguing
D. Punishi~ng, Being Sarcastic, Threatening (p.6)
E. Hating, Atta~cking, Disaffiliating (p.6)
Q~uadlrant FGHI (Submissive-Disu~,5~u5,u~,5affiliative; Passive-Resistan~t)
F. Co::plaining, Rebelling, Njagging, Sulking, Passively
Hesisting or Ag~gressing(p)
G. Distrulsting, Suspecting, Beinlg Skeptical (or Show~in
Inlcomplete Acceptance) (p.8)
H. Condemiening, Self, Wiit~hdrawing (p.9)
I. Slbmnitting, Deferringi, Dbeying (p.10)
Qua3drant JKIN; (Subm:iLssive-Aff~liati vle; Suport-Seekcing)
J. Admiring, Praising, Asking Opinlion (p.10)
K. Trusting, Depending, Asking for Hielp, Being Depndent
L. Cooperating, Agreeing, Confiding, Collaborating (p.11)
M. Loving, Affiliating, IdentiIyinrg With (p.12)
Quadranlt, NOPA (Domrinanlt-Af filiative ; Suipportive-Interpretive )
N~. Supporting, Sympathizing, Reflecting Feelings, Reassuring
Ceneralizing Conscious Feelng, Approve, Nurt~ure,
Th~erapeutic Probing (p.12)
0. Giving Hielp, In~terp~reting Beyond Conscious Feelings,
Offering A~ssistance (p.14)
P. Teaching, Advising, Giving Opinion~, Informning, Acting
as an Authority on Thng p.s
A. Domin~ating, Directing, Commnanding, Inforrmation-rathering,
Diaggosic Probing, Behaving~ Independent~ly, Being Bossy
(Civiing Unsolicited Orders or Advice) (p.16)
B. Boast |
E. Hate I I
F. Complain 11
IV NJ. Support
III M3. Love
H. Condemn? Self K. Trust
I. Submit J. Admire
Exarples of Behavior for Eada Cat~egorsr
Rieflex "B" (Boasting, Self-Stimnulating, Narcissistic, Inltellectualizing
Therapist and client "BI."
1. 'Oerapist or client is boastful. Examples:
C: "I made the highest score on the final examination."
C: "Lockts like I really helped you."
2. Wandering, free-associating, conversation in which the
speaker provides his own stimulation. This category
usually applies more to the client than the therapist.
E>2nples would include client sta~terrents in hichir a
"list" of activities since the prevrouLs session is
covered without emotion, and without a previous therapist
eliciting question. Th~is is generallyi a long, rambling
statement, which may have been started by a therapist
question, but lh~ich continued w~ithn the client providing
"L" vs. "B" his own stimulation. In this case, the client's stat~e-
ce~nt would be rated in two parts, thN answer to the
therapist's question would be rated an "L," and the reist
of the client's statement a "B."
3. Therapist or client intellectualizes.
C: "I really feel affectionate toward you."
T: "That;'s because yrou once had that f~eelin toward
your- fa~ther. "
T: "WhSat is it that's troubling you?"
C: "I haven't worked out ny1 Oedipus complexx.'
Client-to-oth~er "B. "
1. Client reports boasting to others.
C: "~I told him howJ wonderful I ami."
2. Client reports having been narcissistic wvith others.
C: "I took ad~vantagp of her."
Reflex "C" (Rejectin, Wi;thholding, CoIpeting, Accusjing)~
Ter~apist and client "C."
1. Client or therapist rejects previous statement (r-egardless
of' whether previous statement was true'). Examp~les:
C: "Ndo, that isn't right. Wh.Cat bothersj re is that no
one seems to really care for me." In this example,
the "No, that isn't right" woulld be rated "C." Tihe
second part would be rated "P" if no strong emotions
were attached. to it. Of course, if the client
expressed feelings of hurlt or sadness, the second
parst rry be rated "K~." A "no" statem-ent following a
therapist question with no point of viewj attached
(i.e., where therapist does not make a positive state-
ment that is subsequently rejected) should be rated
"L" instead of "C."
2. Client and therapist are arguing, ccrpeting, usuallyr with
an un~der-current of hostility. E~x~mples:
"C" vs. "P" T: "You can find people like that in Nejw York."
C: "I've looked and there are no people like that here."
T: "You haven't looked in the righlt places. You've me~t
only a frew people here."
C: "I knoxn I canrt find people like that here. I need
to go somewhere else."
The first therapist statement in this interchange may not
be rated a "C," depending on the previious client statenaent
th~at elicited it. For instance, if the previous client
statement had been "I need to find some~ people that I
could trust" the first therapist statement doove mn~-iht
be rated "P."
3. Client or therapist refused a pre~vious suggstion,
T: "'I will not see you twice a w~eek."
C: "No ma~tt~er what you say, I wson't stay here."
Client-to-other "C. "
1. Client reports rejection of others.
C: "I don't like him."
2. Client reports competing with others.
C: "I tried to beat him at his owun game."
Reflex "D": (Sarcastic, bzreatening, Punishing Behavrior)
Therapist and Client "D."
T: "If you don't get out of that relationship, I'll
stop seeing~ you."
C: "People are going to keep buggr~i~nge until I kill
Client-to-other "D. "
C: "I told him that if he continued to hara~ss m~e that
I wouldn't see him anymore."
Heflex "E~" (Hate, Attack, Disaffiliate)
Thera~pist an~d Client "E."
T': "G~et ou~t of my~ off~ice."
C: "G~o to hell."
T: "You're anidiot."
C: "She's nothing but a wrhore."
C: "I broke up wi~th him."
C,: "I h~ate myJ motherr"
Reflex "F"" (Complain, Rebel, N~ag, Sulk, Passively Resist)
Thieracist and Client "F."
1. Client passively resists therapist's interpretation put
in the form of statements or question. Exanp~les:
a.T: "Sounds like you get anxious around ccapetent females."
C: "I don't knrow."
b.T: "Is it that youi" boyfriend remindz you of your father
in sos way~js?"
"F,"I "A"r C: "I don't kcno. [Pause.] One thing tihat really dris-
turbss me is that I can't concentrate when I study."
c.T: "D~o I hear some resentment in there?;"
no "F?," "L" C: :'I don't knocw. [Pause.] You mayr be righlt. Yeah, I
w~asn't aware of it but I really do resent himl for that."
Note: In example a, the client's "I don't kn~ow" is
rated "F," because it indicates passive resis-
tance to the therapist's statement. In these
cases, the client is demonstrating an unwilling-
ness to even consider the validity of the
stat~enant, but at the same timne is not flatly
re~jecting it, either. In example b, the "I
don't kn~owr" is followIed by the change of sub-
ject. In this case, it is rather obvious
that the change of subject is a defensive
maneuver, seemingly unrelated to the therapist's
question. The "I don't kanow" should be scored
"F," and the change of subject should be scored
"A." In example c, the "I don't knowu" was
intended to indicate thlougtfulness, an
attenpt, to deal with the therapist's question,
which is validated by the rest of the client's
statement. In this example, the "I don't
know" is not scored, but the remainder of the
statement should be enclosed in parentheses
and scored "L."
2. Sometimes the therapists or client angrily withdraws
(sulks), with some such coin~ent as "I don't kn~ow~."
These should be scored as "F."
Client-to-other "F. "
C: "I resented his saying that, but I didn't say
C: "Wa~en Dad yelled at me, I w~ent~ to up room and
didn't come out for hours."
Heflex "G" (Distrust,, Suspect, Be Skeptical)
Th~era.oist and client "G."
1. Therapist, or client expresses sktept~icism at the previous
statement of the other party. Exarrples:
"What do you mean?"
The first tw~o exa~oles should be scored "G" when the
previous statement and its meaning was perfectly clear.
Th'e "maybe" expresses inccu~plete acceptance, or, better,
neither rejection nor acceptance, but does express
2. Therapist or client is suspicious of feelings, motives,
etc., expressed by the other party. Exaxples:
C: "I don't think~ youl really like me~."
T: "Are you sure you're dealing wlith the thing that's
really bugging you?"
"~C" vs. "G"TI Note: If the statement is an unconditional rejection
or accusation (e.g., "You don't like rre!"), it
should be rated "C,"' not "G."
Cli~ent-to-0th-er "G. "
C: "I didn't believe her."
C: "SomEtime~s, it seemir like no one can be trusted."
Reflex "H": (Condemn Self. Withidraw)