ED 236 51*4
CG 017 066
Mueller, John H. ; And Others
Distinguishing Me from Thee.
May 83 ,
12p. ; Paper presented at the Annual Meeting
Midwestern Psychological Association (55th,
IL, May 5-7, 1983) .
Reports - Research/Technical (143) — *
Speeches/Conference Papers 1 150 )
MF01/PC01 Plus Postage. , >
Cognitive Processes; Mndividual ^Characteristics;
/^Personality Traits ; Recall (Psychology )•;./ *Self
Concept ; *Self Evaluation (Individuals)
*Self Definition " .
"Trait adjectives which people use to describe
themselves will have features both unique to the individual and
shared with or common to many people. To examine the uniqueness
descriptors of onrejs self, and how unique desti^i^ptor s might be ^
organized in memory, subjects (N=40)^made self -descr ipt iven^ss and
other-descr ipt iveness rat ings for 'the same set' of 120 trait
adjectives representing three levels of likability. Uniquely
descriptive items took longer for self -descr iptiveness dedisions than
for items that were descriptive of both self and other. Although fc
unique features may be generated as descriptive of one's self, it
appears they are accessed more slowly. This result is more consistent
with a viewfthat sees trait distinctiveness* as computed rather than
prestored. In terms of endorsement, un iquely descriptive i'tems showed
minimal Hikability eff ects , whereas likable items were predominantly
seen as descriptive of both self and other and\ unlikable items were ^
.rejected afe mutually nondescr iptive . Recall differences among the
subtypes of items were n^t^pronounced ove'rall, \hough there was some
variation by likability level. (Author/WAS)
* Reproductions supplied by EDRS are the best that cafri be made *
* 4 from the original document. *
DISTINGUISHING ME FROM THEE
John H. Mueller
Michael J. Ross
University of Missouri, ^Golumbia
1 • - V
* — i
j U.S. DEPARTMENT OF EDUCATION
NATIONAL INSTITUTE OF EDUCATION
EDUCATIONAL RESOURCES INFORMATION
This document has been rnproduced as
received from the person or organisation
' originating it.
^ Minor chantjes havr; b«en madu to improve
reproduction qudlity. ^
• Points of vmw or opinions stated in thi^docu
mt>ntdo not necessarily represent official NIE
position or policy
"PERMISSION TO REPRODUCE THIS
MATERIAL HAS BEEN GRANTED BY
TO THE EDUCATION A\- RESOURCES
INFORMATION CENTER (ERIC})."
Paper presented at the Annual Meeting of l&e Midwestern Psychological Association,
Chicago, IL, May 5-7, 1983.
DISTINGUISHING ME FROM THEE
John H. Mueller, Michapl 1 Or,-, J\u
, mcnaei J. R 0iS , andWtin Heesackei
UniversftySf Missouri, CoTumbi,
■ • / < • ■ >
^ • *
Subjects made sel f-des-cripli veness and otherrdescrlptlveness ratings
" for the same set of 120 trait' adjectives representing three levels of
1 i kabi Hty. Uniquely descriptive items took longer -for sel f-
descriptiveness decisions than for items that we're descriptive of both
self and other. AJthough unique features ma^ be generated as
descriptive of one's self, it appears they are accesL more slowly.
This result is more consistent wi.ttt a view that sees trait
distinctiveness as computed rather than pr*estored. In terms of
endors'ement,jjniquely descriptive items shewed minimal likability
effects, whereas; likable items were predominantly seen as descriptive
of both self and other and iinl ikable items were rejected as mutually
nondescriptive. Recall differences among "the subtypes of items were* :
not pronounced. overall., though .theVe was some variation by liability
level . - . • s
Address ^11 corr^epondenc^ to: \
John x H. Mueller
t Psychology Department
210 McAlester Hall
University of Misuari
Columbia, MO 65211 I
[3J4/882-3084, or 882-6860]
Paper presented at the' annual meetings of thS
Mtdwest^rn Psychological' Association, Chicago, May 6, 1983.
DISTINGUISHING ME FROM THEE
John H. Mueller, Michael 'J. Ross, and Martin Heesacker
University of Mi ssouri ,a Col umbia
— ->- JL--
The 'study I'm going to* report today examined the- un iqueness of
descriptors of one's self, and how unique descriptors might be
organized in memory relative to less distinctive features. As- nany y
authors have observed (e.g., Snyder & Fromki n , 1980) , weQpend a great
deal of cur lilves trying to. estahj i sh individual identities, though —
judging from the standard of living achieved by various, fashion
designers — relatively. few of us, are completely successful!
For various reasons, some trait adjectives may not distinguish
* \ : ^ o. V
oge person from another very well. At one extreme, for example, some
characteristics may &e so. desirable or befiign that we attribute them .'
to others as freely^as^to ourselves (e.g., ciecent, friendly,
honorable, witty), or' perhaps the corresponding adjectives are so
generally descriptive that' they apfply to a large number of people
(e.g., American, masculine, right-handed, Protectant). At .the other
extreme, some characteristics may be sd dastardly that we hesitate to
attribute them to anyone (e.g., evil, incestuous, lascivious,'
rapacious, /Hthles^ wanton), or perhaps i n yictuari-al ^terms we just
rarely 'encounter anyone whtim we would label with*a particular
adjective (e.g., acrobatic, dainty, messianic, weird). Of course,
many factors will determine just which terms are distinctive,
including reference grouo^and perhaps even veVbal facility, not^just
( * . ^ I
base rate. \ \ \ 1 s
Mueller, Ross, Heesacker . 7 2
v . . . ' • V
Assuming that we would thus "find both unique and shared features
in our self concept,- the question arises as to their relative
importanceand accessibility. In one relevant study * McGui re knd
Pidaw.er-Singer (1976). had subjects generate self descriptions. Tbey^
fpund that 'the features mentioned tended to be those .that set the
subject aparv from other people, hot those aspects that were, most
common. Of course, .this doesn't- mean shared 'features are not a part
.of our self concept, but /unique features seemed especially relevant in
the" course of unpaced spontaneous descriptions . Q However, order in an
^unpaced description' coul d reflect processes other than. speed of. access
to a'-sinqle' specific aspect of >he seTf concept. For one thing,
unpaced descriptions maximize the "editing" of rSsponses^jjU^wi ng the
unique featured to'domi nate 'output whether -they^e actual ly accessed
first or not. - *
Still, it is intuitively appeaTHg^hat unique traits might be
accessed* rapidly, and Figure 1 shows' two simple models consistent with
this notion." Mpde?l A in Figure 1 represents a scheme where traits are
organized w'ithin the self concept by%i s ti ncti veness , so that a 1
top-down search process would provide rapid affirmative decisions for
distinctive, nonshared traits. Slower affirmations would occur for
shared- 'features, because these characteristics are accessed only aHer
the most definitive trafts h'a^ been reviewed. Model. B shows' a •
different arrangement, where- each feature^has stored With it a "tag"
denoting 'some^egree of distinctiveness, In this case, If the feature
tag exceeds some high criterion, a rapid affirmative response results,
whereas a slowe'r affirmative response results for less distinctive
Mueller, Ross, tfeesacker . ' "3
features, wi'l/h negative responses being slowest of al 1
■ < . \
'Insert Figure 1 .about here ■. . •
As sensible as these, two model s jirfojht seem, the data we will
present^don 1 t fit either one, nor any conceptual ization that makes the
same basic prediction, namely thatv unique featur.es'are accessed fa?ter n
than shared. Instead, some process Kke that shown *in. Mode,! C seems
more likely, and it is in general consistent with McGuire and V".
Padawer-Singer 1 s interpretation that we determine the most salient
features of our selves by a process of comparison to the context of .
the decision/. For example, being an American in* Paris is much more
sal ient than., being an American in Peoria — not to mention more fun!
Therefore, in contrast' to Models A and E* comparative models such as
» Model' C would predict"'that access would be faster for shared'traits,
with the determination of unique traits ^equ^'ring more.tir^e/ Self §
descriptions (viz. McGuire t & Padawer-Si riger , 1976) might still contain
predominkntly.dVstincti ve. feature.s, given an unpaced task, but such
'traits .are available only after .a process that cortsumss some time. In
ModetfC, trait Uniqueness is computed , so to speak, rather than
prestored as in Models A and B.
The predictions seem taifly clear then, and the data we collected
were intended to .'provide some initial information about the
organization of features as a function of distinctiveness. In the \
Interest of time, we will have to focus on just the one question: are
unique-trait decisions made faster than shared-trait decisions? The-
V 4 ' •
Muel ler,/ Ross, Hees.acker « ■■ ■ . - 4 -
methodology is simple, so^I will describe it only briefly. I will be
happy' "to provide further details to anyone who wants them.
1 t Method &
Forty subjects Vated 120 trait*, adjectives* selected fram-^hree
likability 'levels in the Anderson (1968) norms: 40 likable items, 40
unlikable, and 40 "neutral" items, all from the high meaningf iriness
subset in the' norms. A subject first i;ated all 120 items for
sel f-descriptiveness and then for descripti veness of their /"best
friend/ 11 or vice versa. Responses were indicated on an 8-point scale,
on the basis of a rapid f i rsOynpressi on . The two rating phases were
followed by an unannouncfed^recal 1 test for the 120 items.
In addition, the Self Consciousness Questionnaire (Buss, 1980)
was also administered, nt had been expected that high self;-aw y are [
subjects might be especially aware of their unique traits, and access
them. f,asteic than less self-aware subjects, particularly under the
prestored assumption. However, this seemed not to be the case, so
these data won't be considered further. - j
Results and Discussion
\ .■ I'tems were tabulated 'on the basis >of their descriptiveness
of each target pers(jn. (For purposes of analysis, items rated 1-4*
were considered nondescripti ve', with 5-8 being descriptive.) Thfs
results in -four subtypes of items, shown as column headings in Tabl'e
1. Some items were judged descriptive of BOTH targets, sor/e
descriptive of SELF ONLY, some OTHER ONLY, while the remainder were,
descriptive of NEITHER target. Three aspects, of the data are shown in
Table 1: how many items were classified o£ endorsed into each of the '
Mueller, Ross, Heesacker
four categories, how rapidly deci sions were made for each subtype of
item,, and the incidental recall of items of each type.- Let me quickly
summarize the h^h points of the result's. u ~
\ ' . Insert Table 1 about here /
'■ The way subjects classified the items is shown in t}ie top section
of Table !. Perhaps the most pertinent, result in terms of item
endorsement was a significant Self by "Other by Likability triple
interaction (F (1 ,78) = 34.82). Most adjectives 'were considered
descriptive of neither or both targets (see' thp fourth* row of Table 1,
labeled "All"). This would be expected,, assuming we share many traits
with our 'best friend. "While these mutually descrl^i ven^ns show
pronounced likaMity gradients (first and fourth column in Table 1),
neither of the unique descriptors show such likabilTty effects (second
and third column in.Jable 1). & x <J ( - /
- Going to the next section of the table, in the' data of greatest
interest, analysis of the la-Sffncjes for self descripti veness decisions
•revealed a significant Self by Other interaction (F (1,39) = 7.11).
As Table -1 shows (in line 8, labeled "All"), decision speed was faster
ipr scared descriptors (Ms = 3297 and 3395 msec, for Both, and Neither,
respec\i vely) than* for unique descriptors (Ms = 3486 -and 3622 msec,
'for Self-Oaly and Other-Only, respetti vely) . In the individual
comparison^ of greatest interest, the Self-Only items (3486 msec) and
/ 0 c*
Both items (3297 msec) were significantly different (Tukey test)/-
* Thus, it appears that rapid self-reference decisions are made
Mueller, Ro^s, Heesacker
only for items that are generally descriptive, and that decisions
about unique features require more time. As I said "earl i er , .when
editing output in an unpaced test (McGuire & Padawer-Singer, 1976),
unique items might still "occur' first and more often as the most
definitive descriptors, .but the self-reference descision per-se seems
* b > 7 •
slower for unique-items. On this basis, it appears that Madels.A and
B and others of that class can be rejected, and that ( something like
Model C serves as a better description of the organization of features
in the self concept. V # .
- The probability of recall data indicated a marginally significant
triple interaction of Self by Other by Likability (F (2,78) = 5.57, £
< .07). Neutral items were, not recalled very well at all, likable ^
item*: tended to be recalled better when not self-descriptive, and /
unlikable items'were recalled ^est f or commonly descriptive items'.
The interpretation of this interaction is unclear, and likely risky in
view of the marginal level of significance.
In closing, o it/appears that self-referent decisions involving
distinctive traits take more time than is the case for shared traits.
We have also observed this in some similar-experiments that I don l t
have time to cover here. ThisS^suggests. that distinctiveness
information - is not prestored in our sel f, concept , unless one argues
that we "search through le§s distinctive features first, as in a
bottom-up search in-Model A. While not proving Model C-specifically,
the data are at least consistent with Model C, and in accord with a
•comparative process as outlined by people sucly as' McGoire and
MuelleY, Ross, Heesacker
Anderson, N. H. Likableness ratings of 555 personality-trait words."
Journal %f Personality and Social ^ychology, 1968, 9, 272-279.
Buss, A. H. Sel f consciousness arrd social anxiety. San Francisco,
CA: W. H. Freeman, 1980. ' -
McGuire, W. J . , &* Padawer-Si nger, A. Trait salience in the,
spontaneous self-concept./' Journal of Personal i ty and Social
Psychology, 1976, 33,' 746-754.
Snyder, C.R., & Fromkin.'H. L. Uniqueness: The human pursuit of
difference, ^few York: Plenum Pres^s, 1980.
Mueller, Ross, Heesacker
^ ,. TABLE 1
Number Rated by Mutuality of Self- and Other-Descri ptiveness ,
Latency of S£l f-RTf erence Decision, and Probability of Recall
Self-Other Descri ptiveness Types
& ■ ■ . m
Number of each type
46 . 3 !
8 Latency (msec)
' Likable '
3577 _ '
LJnl i kabl e
All * ,
• * Probability of recall
• Likable' \
• Neutral |
. 20 N
Unlikable '* \
All • j
Note: Both refers to a word rated as descriptive of self and
other, Se?1f-0nly refers to \ word rated \s descriptive of
but not descriptive of other, Other-Only refers to,a word
was not self-descriptive but was\other-descriptiy^
and Neither, refers to a word descriptive of neither self
nor v other .
MODEL A: Strength Hierarchy
MODEL B: Distinctiveness Tags *
"Am I (X)?"
fast '"yes" (unique)
slow "yes" (shared)
• Am i^ixp
tag - c
tag ■ K?
fast 'yes' 1 (unique)
slow "yes'" (shared)
"no * (slowest)
MODEL C: Feature Set Comparison
"Am f(X)" injplies contrast:
'Am 1 more (X) than others?"
Retrieve SELF feature set
and OTHER feature set
Is X in both-sets?
Is XJn SELF set?
fast "yes (shared)
slow "yes" (unique) t
~ 'Figure 1. Three theoretical conceptions about the relative speed
a sel.f-descriptiveness decision about a trad t QO* that is either
shared with other people or which mgre uniquely describes us.