Journal of Personality and Social Psychology

1980, Vol. 39, No.3, 472-480

The Revised UCLA Loneliness Scale: Concurrent and

Discriminant Validity Evidence
Dan Russell
University of Iowa
Letitia A. Peplau and Carolyn E. Cutrona
University of California, Los Angeles

The development of an adequate assessment instrument is a necessary pre-

requisite for social psychological research on loneliness. Two studies provide
methodological refinement in the measurement of loneliness. Study 1 presents a
revised version of the self-report UCLA (University of California, Los Angeles)
Loneliness Scale, designed to counter the possible effects of responsebias in the
original scale, and reports concurrent validity evidence for the revised measure.
Study 2 demonstrates that although loneliness is correlated with measures of
negative affect, social risk taking, and affiliative tendencies, it is nonetheless a
distinct psychological experience.

Social relationships are at the core of hu- the fact that lonelinessis a common and dis-
man life. Not surprisingly, problematic as- tressing problem for many people. In one
pects of relationships have been a major national survey (Bradburn, 1969), 260;0 of
focus of psychological research. Psychologists Americans reported having felt "very lonely
have undertaken careful analyses of aggres- or remote from other people" during the past
sion, competition, crowding, and other nega- few weeks. Loneliness has been linked to a
tive factors in social relations. Someproblems variety of other serious individual and social
of social relations have, however, beenempha- problems, including alcoholism (Nerviano &
sized to the neglect of others. Researchers Gross, 1976), adolescent delinquent behavior
have investigated instances where there are (Brennan & Auslander, Note 1), suicide
"too many" people, and individuals feel sub- (Jacobs, 1971; Wenz, 1977), and physical
jectively "crowded" (e.g., Freedman, 1975; illness and overutilization of health care ser-
Stokols, 1972). However, little attention has vices (Lynch, 1976).
been given to the other end of the continuum Empirical research on loneliness has been
where social relationships are "too few," and hampered by a variety of problems (see re-
people feel subjectively "lonely." views by Peplau & Perlman, 1979; Peplau,
The importance of research on loneliness Russell, & Heim, 1978). A major hindrance
lies not only in its potential for shedding light is that loneliness,unlike aggression,competi-
on basic aspects of social relations but also in tion, and crowding, cannot be readily manipu-
lated by researchers.Thus, the crucial task
for investigators is not the developmentof an
The authors would like to thank Robin Lewis, experimental paradigm to produce loneliness
Jeff Tanaka, Debbie McGrew, and Phillip Malamuth in differing degrees under controlled condi-
for their help in collecting the data for Study 1;
Warren H. Jones for providing subjects from the tions but rather the development of instru-
University of Tulsa for Study 2 j and Richard R. ments to detect variations in loneliness that
Lau for his comments on the manuscript and help occur in everyday life.
with data analysis. Our research on loneliness led initially to
Requests for reprints should be sent to Dan Rus-
sell, Division of Educational Psychology, Measure-
the developmentof a 20-item, self-report mea-
ment, and Statistics, College of Education, University sure, the UCLA (University of California,
of Iowa, Iowa City, Iowa 52242. Los Angeles) Loneliness Scale (Russell, Pep-

Copyright 1980 by the American Psychological Association, Inc. 0022-3514/80/3903-0472$00.75

472
 THE REVISED UCLA LONELINESS SCALE 473
lau, & Ferguson, 1978). In studies using structs. A final concern is the potential con-
collegesamples,the scale showedhigh internal founding of loneliness scores with social de-
consistency (coefficient alpha of .96). Con- sirability. Since there is a social stigma
current validity was indicated by relation- attached to loneliness (Gordon, 1976), indi-
ships between scores on the loneliness scale viduals who want to appear in a positive light
and otherand
tionships, indicators of states.
affective loneliness,.social rela- might underreport their experience of loneli-
ness.
Researchby several other investigators has The two studies reported here addressthese
also supported the adequacy of the UCLA potential problems with the original UCLA
LonelinessScale as a measureand has begun LonelinessScaleand provide clarification con-
to provide a more detailed description of the cerning the nature of loneliness. In the first
experience of loneliness. Loneliness is related study, a revised version of the UCLA Lone-
to a number of personal characteristics, in- liness Scale is developed, incorporating new
cluding low self-esteem, shyness, feelings of positively worded items. The concurrent valid-
alienation, external locus of control, and belief ity of the revised scale is established by relat-
that the world is not a just place (Jones, ing lonelinessscores to the experience of af-
Freemon, & Goswick, in press). Lonely stu- fects that have been linked both empirically
dents report experiencing problems of inhib- (Russell et al., 1978) and theoretically
ited sociability (Horowitz & French, 1979) (Weiss, 1973) to loneliness. A second study
and, in dyadic interactions, rate both them- providesa further test of the concurrent valid-
selves and their partners more negatively ity of the revised scale by examining relation-
than do nonlonely students (Jones et al., in ships between loneliness and social behavior.
press). Among both studentsand older adults, In addition, Study 2 addressesthe discrimi-
lonelinessis linked to negative affects, includ- nant validity of the revised scale by demon-
ing boredom, restlessness,and unhappiness, strating that loneliness scores are distinct
and to dissatisfaction with social relationships from measures of social desirability, social
(Perlman, Gerson, & Spinner, 1978; Russell risk taking, negative emotional states, and
et al., 1978). affiliative motivation.
Although the UCLA Loneliness Scale is a
reasonably adequate measure, several poten- Study 1
tial problems with the scale are apparent. The first study had several goals. First, it
First, all items on the scale are worded in the was designedto revise the UCLA Loneliness
same direction, with high scores reflecting .':'caleby incorporating new, positively worded
feelings of social dissatisfaction. Any sys- items. The internal consistencyof this revised
tematic response bias toward high or low measure was assessed,and the correlation
scores, irrespective of item content, would betweenscoreson the original and the revised
influence the total scale score. A second po- £cale was calculated. Finally, the concurrent
tential problem concerns the discriminant validity of the revised loneliness scale was
validity of the scale. Substantial correlations investigated.
(ranging from .4 to .5) have been found
between loneliness scores.and the Beck De- Method
pression Inventory (Bragg, 1979) and the
The 162 students (64 males and 98 females) who
Coopersmith measure of self-esteem (Jones voluntarily participated in Study 1 were all tested
et al., in press). Conceptually, it is reasonable in the spring of their first year at UCLA. They
that loneliness might co-occur with depres- completed an extensive questionnaire; pertinent to
this research were the measures of loneliness and
sion and low self-esteem, and such findings
emotional states.
support the validity of the UCLA Loneliness Loneliness measures. The original UCLA Lone-
Scale. At the same time, however, these find- liness Scale (Russell et al., 1978) was given, fol-
lowed by 19 new items written by the authors.
ings indicate a need to demonstrate the dis-
These new items measured satisfaction with social
criminant validity of the scale by showing relationships and represented, as nearly as possible,
that loneliness is distin.:t from related con- opposite wordings of the original scale items.
474 D. RUSSELL, L. PEPLAU, AND C. CUTRONA

A second set of measures assessed explicit self- for the final ~cale,along with 10 of the origi-
labels of loneliness. Examples of such questions are
"During your lifetime, how often have you felt
nal negatively worded items. The criterion for
lonely?" and "During the past two weeks, how selectingtheseitems was their correlation with
lonely have you felt ?" Six such questions were the self-labeling lonelinessindex. The 10 posi-
asked, all involving the student identifying himself tively worded and 10 negatively worded items
or herself as lonely. Responses to each of these six
that had th~ highest correlations (all greater
questions were summed to form a single self-label-
ing of loneliness index (coefficient alpha = .78). than .40 in magnitude) with the self-labeling
Emotional state. The questionnaire contained index were selected for the final instrument.
three mood measures assessing anxiety and depres- The revised UCLA LonelinessScaleitems and
sion. The Beck Depression Inventory (BD! j Beck,
scoring format are given in Table 1. Note
1967) consists of 21 symptoms or attitudes charac-
teristic of clinical depression. Each item has several
that the original and new items are randomly
alternatives describing manifestations of each symp- intermixed in the revisedscale.1
tom that vary in intensity. Scoring on each item The internal consistency of the revised
ranges from 0 to 3, depending on the severity of measure was high (coefficient alpha of .94)
the symptom manifestation that is selected by re-
and compared favorably with the alpha co-
spondents as being self-descriptive. The BD! was
found to be quite reliable with clinical populations, efficient of .96 obtained for the original scale.
with a Spearman-Brown split-half coefficient of .93 The correlation between the revised and
being reported. Validity for the measure has been original scales for the present sample was .91.
indicated by relating BD! scores to clinical judg- Tests for sex differences were also con-
ments of the severity of depression: Correlations of
.65 and .67 were found in two studies. The BD! has ducted, comparing the scores of male and
also been shown to assess validly the severity of female students on the revised measure. A
depression in college populations (Bum berry, Oliver, significant sex difference was found, t(157)
& McClure, 1978). = 3.20, P < .001, with men scoring signifi-
The Costello-Comrey Depression and Anxiety cantly lonelier than women (Ms = 36.23 and
scales were also administered (Costello & Comrey,
1967). To develop these scales, a variety of factor 31.12, respectively).2 However, including sex
analytic studies were conducted with both normal
and clinical populations. The goal of these analyses
was the construction of factorially "pure" anxiety
1 From the current data, we have also developed
and depression scales, which minimized the inter-
correlation of the two measures; the final anxiety a 4-item survey version of the UCLA Loneliness
and depression scales were found to correlate .40 Scale, consisting of two positively worded and two
for males and .50 for females. Validity for the two negatively worded items. Using optimal subset re-
scales has been indicated by relationships with gression procedures, the set of four items that best
clinical diagnoses and other measures of anxiety and predicted scores on the self-labeling loneliness index
depression. Split-half reliabilities of .90 for the de- were selected. The items chosen were Numbers 1,
pression scale (14 items) and .70 for the anxiety 13, 15, and 18 from Table 1. This four-item lone-
scale (9 items) have been reported. liness scale had a coefficient alpha of .75 in the
In addition to these mood measures, students also current study. We recommend that investigators who
rated on 9-point scales the intensity of their current want a shortened version of the loneliness scale use
experience of 25 emotions. Thes~ included such af- these four items. Normative data and a bibliography
fects as bored, empty, hopeless, and satisfied, along of research using both versions of the scale are also
with a set of affects believed to be unrelated to being compiled by the authors. Investigators using
loneliness, such as resigned, embarrassed, and con- the measures are urged to send us summary data
fident. from their samples as well as a brief description of
their research and findings.
2Since previous research has not indicated any
Results sex differences in loneliness, this result suggests a
possible sampling bias. The participants in Study 1
Data analyses had two goals: to developa were originally tested.-in the fall of their first year at
revised loneliness scale and to assessthe con- UCLA, and returning versus nonreturning students
were compared on their loneliness scores from this
current validity of the revised scale. earlier testing. No significant differences were found
Revised UCLA Loneliness Scale. Prior to for women. In contrast, men who returned for re-
data analysis, it was decided that the new testing were lonelier than nonreturning men, t(143)
= 2.45, P < .02, which could have created the ob-
scale should consist of 20 items, half reflecting
served sex differences. Because of this possible sam-
satisfaction with social relationships and half pling bias, data from the participants in Study 1
reflecting dissatisfaction. Accordingly, 10 of were not included in the normative statistics pre-
the new positively worded items were selected sented in Table 2.
 THE REVISED UCLA LONELINESS SCALE 4i5
Table 1
The Re!-1sed UCLA Lol1eline.~s Scale

Directions: Indicate ho\v often you feel the \vay described in each of the follo\ving statements. Circle one
number for each.

Statement Xevcr Rarely Sometimes Often

1. I feel in tune with the people around me" :! 3 4
2. I lack companionship 2 3 4
3. There is no one I can turn to 1 2 3 4
4. I do not feel alone- 1 2 3 4
5. I feel part of a group of friends" 1 2 3 .f
6. I have a lot in common \vith the people around me" 1 2 3 4
7. I am no longer close to anyone 2 3 4
8. My interests and ideas are not shared by those
around me 1 2 3 4
9. I am an outgoing persona 1 2 3 4
10. There are people I feel close toa 1 2 3
11. I feel left out 1 2 3 4
12. My social relationships are superficial 2 3 4
13. Xo one really knows me \vell 2 3 4
14. I feel isolated from others 2 3 4
15. I can find companionship \vhen I want ita 2 3 4
16. There are people who really understand mea 2 3 4
17. I am unhappy being so withdrn\vn 2 3 4
18. People are around me but not \vith me 2 3 4
19. There are people I can talk to" 2 3 4
20. There are people I can turn to" 2 3 4

Note. The total score is the sum of all 20 items.

a Item should be reversed (i.e., 1 = 4, 2 = 3, 3 = 2, 4 = 1) before scoring.

as a variable in the analyses indicated that undertook a secondstudy to investigate issues

sex did not mediate any of the findings. . of validity in greater detail. The goal of Study
Loneliness and emotional state. "To assess 2 was to demonstrate that loneliness,as mea-
the concurrent validity of the revised scale, sured by the revised scale, is discriminable
the relationship of loneliness scores to mea- from measuresof other, conceptually related
sures of emotional states was examined.Lone- constructs such as depression, lack of affili-
liness scores were significantly correlated ative motivation, and low social risk taking.
with scoreson the Beck DepressionInventory Study 2 also replicated the internal consis-
(r = .62) and with the Costello-Comrey tency analyses reported in Study 1 with a
Anxiety (r = .32) and Depression (r = .55) new sample and provided additional tests of
scales.Lonelinessscoreswere also significantly concurrent validity.
correlated (all rs above .40) with feeling
abandoned, depressed,empty, hopeless, iso-
lated, and self-enclosedand with not feeling Method
sociable or satisfied. Loneliness scores were
not significantly correlated with such concep- To increase the generality of the findings, a
broader sample of college students was recruited.
tually unrelated affects as feeling creative, One third of the students were enrolled in intro-
embarrassed,sensitive, surprised, or thought- ductory psychology classes at the University of
ful. Tulsa, one third were enrolled in introductory psy-
chology classes at UCLA, and the remaining stu-
dents were drawn from upper division psychology
Study 2 classes at UCLA. All students received partial course
credit for their participation. The total sample of
237 students (107 males and 130 females) each com-
Having obtained evidence concerning the pleted a questionnaire containing the following
psychometric properties and concurrent valid- "materials, with the order of the materials randomly
ity of the revised UCLA Loneliness Scale,we varied.
476 D. RUSSELL, L. PEPLAU, AND C. CUTRONA

Loneliness measures. The original UCLA Loneli- favorable in their attitudes toward a boring task,
ness Scale was administered, along with the 10 to be socially conforming, and to be very susceptible
positively worded items included in the revised to persuasion. A test-retest correlation of .88 for
scale. The self-labeling loneliness questions from this measure has been reported over a I-month
Study 1 were also given, and a self-labeling loneli- period. Internal consistency for the measure is also
ness index was again created by summing the re.; high, with a K-R 20 coefficient of .88 being found.
sponses to these six items (coefficient alpha=.72). The measures of introversion-extroversion and
Social activities and relationships. Students were lying developed by Eysenck and Eysenck (1975)
asked how frequently they had engaged in a variety were also administered. The Introversion-Extrover-
of solitary activities (e.g., eaten dinner alone) and sion scale is designed to assess whether the re-
social activities (e.g., done something with a friend) spondent is a sociable and friendly person versus a
during the previous 2 weeks. Students also indicated quiet and introspective person. Validation of this
how many close friends they had and the nature of measure has involved demonstrations of relationships
their current dating or marital status. between scores on the scale and conditionability,
Mood and personality measures. Seven measures level of aspiration, vigilance, and time judgment,
of mood and personality were administered. To based on Eysenck's personality theory. The Lie
assess depression, the Beck Depression Inventory scale is designed to assess whether individuals are
(Beck, 1967) was given. State anxiety was mea- distorting their responses. This measure has been
sured using the State-Trait Anxiety Inventory validated by comparing scores on the scale under
(STAI; Speilberger, Gorsuch, & Lushene, 1970). conditions where the respondents should be moti-
This scale consists of a set of affective statements vated to distort their responses versus neutral con-
(e.g., "I feel calm") that are used by respondents ditions. Test-retest reliabilities for both measures
to describe their current feelings. Validity evidence are high, with a correlation of .89 found for the
for this measure consists of correlations with other Introversion-Extroversion scale and .84 for the Lie
anxiety measures and scores of individuals exposed scale over a I-month period. A coefficient alpha of
to anxiety-provoking situations. Internal consistency .85 is reported for the Introversion-Extroversion
for the STAI is high, with coefficient alphas rang- scale and .80 for the Lie scale.
ing from .83 to .92 in different studies. A final measure that was administered was the
Self-esteem was measured using the Texas Social Assertiveness scale developed by Rathus (1973). On
Behavior Inventory (TSBI-Form A; Helmreich & this scale, respondents indicate how self-descriptive
Stapp, 1974). This scale is designed to assesssocial a set of assertive and nonassertive behaviors are.
self-esteem; items concern the respondent's feelings Validity for the measure has been indicated by sig-
of social self-confidence. The TSBI has been vali- nificant correlations between scores on the scale and
dated by demonstrating relationships with inter- peer ratings of assertiveness. Significant relation-
personal attraction in laboratory settings and with ships were also found between scale scores and
the endorsement of positive self-descriptions on a verbal reports of assertive behavior in different social
measure of masculinity and femininity. For the ver- situations. Test-retest reliability (over a 2-month
sion of the scale used here, a coefficient alpha of interval) of .78 is reported by Rathus, along with
.85 was found for both males and females. a split-half correlation of .77. For the present study.
Mehrabian's (1970) measures of Affiliative Ten- half of the 30 items on the Rathus measure were
dencies and Sensitivity to Rejection were included used, consisting of every odd item from the scale.
to assess approach and avoidance orientations
toward social relationships. Validity for the mea-
sures has been indicated by low correlations with a Results
measure of social desirability and (as theoretically
predicted) by nonsignificant correlations between Sex differences. Comparisons were made
scores on the two scales. In laboratory studies the between the mean loneliness scores for males
Affiliative Tendency scale has been related to af-
and females. In contrast to the findings re-
filiative behavior in social situations and to a mea-
sure of dependency. Sensitivity to Rejection has ported in Study 1, no significant differences
been related to dependency and to conformity in were found, t(228) = .72, ns. In Table 2,
social situations. The reliabilities for both measures normative loneliness data from this sample
are sufficiently high; K-R 20 (Kuder-Richardson are presented separately for males and fe-
formula) coefficients of .80 and .83 are reported by
Mehrabian for the Affiliative Tendency and Sensi-
males. Including sex as a variable in the
tivity to Rejection scales, respectively. analyses reported below indicated no medi-
To measure social desirability, the Marlowe- ation of the relationships by sex of subject.
Crowne Social Desirability Inventory (Crowne & Internal consistency. To cross-validatethe
Marlowe, 1964) was administered. This scale is internal consistencyfindings from Study 1, the
designed to identify individuals who tend to describe
themselves in an overly positive or desirable fashion.
same procedures were used to analyze data
Individuals who respond in a socially desirable from the secondstudy. An alpha coefficient of
fashion on this measure have been found to be more .94 was again found for the revised loneliness
 THE REVISED UCLA LONELINESS SCALE 477

Table 2 nificantly from thosewho were dating steadily

Normative Statistics for Male a1td Female or who were married.
College Stttdents in Stltdy 2 Discriminant validity. The discriminant
validity of the revised scale was assessedby
Statistic Males Females
examining the relationship between loneliness
n 102 128 £cores and scores on the other measures of
M 37.06 36.06 mood and personality. Initial analyses exam-
SD 10.91 10.11 ined the correlation of loneliness scores with
Mdn 35.38 34.17 the mood and personality measures and the
Mode 30.00 29.00
Range 20-68 20-66 £~lf-labeling lonelinessinde.,,<.
An inspection of
these correlations (seeTable 3) indicates that
loneliness scale scores correlate more highly
measure; and once again the correlation be- with the self-labeling loneliness index than
tween scores on the original and revised lone- with any of the other measures.Nonetheless,
liness scales was .91. The similarity of the loneliness was strongly correlated with many
findings from these two independent studies of the mood and personality variables, raising
is striking. the possibility that these variables might, if
Concurrent validity. The relationship be- combined, account for much of the variance
tween scores on the revised loneliness scale in loneliness scores. For example, loneliness
and measuresof social activities and relation- might be a function of low social risk taking
ships were examined as a test of concurrent combined with high levels of anxiety. There-
validity. For solitary relationships, significant fore, a second discriminant validity test was
correlations (all ps < .001) were found be- conducted to examine whether scores on the
tween loneliness scores and the amount of UCLA Loneliness Scale were related more
time students spent alone each day (r = .41), closely to the other measures of loneliness
the number of times they had eaten dinner (i.e., the self-labeling loneliness index) than
alone during the previous 2 weeks (r = .34), to an optimal linear combination of the mood
and the number of times they had spent a and personality variables.
weekend night alone during the previous 2 Multiple regression analysis was used to
weeks (r = .44). Lonely students also re- combine the mood and personality measures
ported doing fewer social activities with as predictors of loneliness. To eliminate cor-
friends (r = -.28) and having fewer close relations among the predictors, a factor analy-
friends (r = -.44). The relationship between
loneliness and dating or marital status was
examined, using analysis of variance to com- Table 3
pare three groups: those students who were Correlation of Scores on the Rerised UCLA
not dating at all, those students dating Loneliness Scale TVith Self-Labeled
casually, and those students either dating Loneliness and the Measltres of
Mood and Personality
steadily or married. Using scores on the re-
vised UCLA Loneliness Scale as the depen- Revised UCLA
dent measure, significant differences were Measure LonelinessScale
found among these three groups, F(2, 187) =
22.97, P < .001. Studentswho were not dating Self-labeling loneliness index .705
.505
at all had a mean loneliness score of 43.1; Depression
Self-esteem -.493
students who were dating casually and those In troversion-extroversion -,457
who were romantically involved had meansof Affil.iative tendency -.452
34.0 and 32.7, respectively. Post-hoc compari- Anxiety .359
-.342
sons (using the Scheffe procedure) indicated Assertiveness
.276
that students who were not dating were sig- Sensitivity to rejection -.203
Social desirability
nificantly more lonely than the other two Lying -.001

groups,F(l, 189) = 35.23,p < .001; students

who were dating casually did not differ sig- Note. Based on 239 students in Study 2.
478 D. RUSSELL, L. PEPLAU, AND C. CUTRONA

Table 4 as a lower bound estimate of the revised lone-

Regression A nalysis Predicting Scores on the line~s scale's reliability, an additional 520;0
Revised UCLA Loneliness Scale From Scores of the variance in loneliness scores is poten-
on the Affiliatit'e Motit'ation, Social Risk
Taking, Negatit'e Affect, and Social
tially explainable. This permitted a second
Desirability Factors discriminant validity test for the loneliness
scale-a test of whether the unexplained
Standardized variance could be accounted for by the self-
Predictor beta F valuea R2 labeling lonelinessindex.
To investigate this possibility, a hierarchi-
Affiliative cal regression analysis was performed. The
motivation -.523 70.30* 12 four mood and personality factors were en-
Social risk
taking -.504 59.32* 17 tered first into the regression equation, fol-
Negative affect .375 41.50* 14 lowed by the self-labeling loneliness index.
Social After eliminating the variance explained by
desirability -.073 1.43 .00
the mood and personality factors, the self-
.This is a test of the significance of each beta weight, labeling lonelinessindex was still a significant
with df = 1, 170. predictor of loneliness, F (1, 169) = 81.01,
* P <.001. P < .001. An additional 180;0of the variance
in lonelinessscale scoreswas explained by the
sis was first conducted on the mood and per- loneliness index. This result provides clear
sonality measures. Using principal factoring evidence of the discriminant validity of the
with a varimax rotation, a four-factor struc- revised UCLA LonelinessScale.
ture emerged.:iThe first factor was labeled A final discriminant validity test examined
Social Risk Taking; measuresof introversion- whether the previously reported concurrent
extroversion, self-esteem,sensitivity to rejec- validity evidence for the UCLA Loneliness
tion, and assertivenessloaded highly on it. A Scale was based solely on the influence of the
second factor was labeled Negative Affect; the mood and personality variables. For example,
depression and anxiety scales defined this is the relationship betweenloneliness and the
factor. The third factor was labeled Social number of close friends uniquely attributable
Desirability., since the measuresof social de- to loneliness, or does it reflect the fact that
sirability and lying loaded highly on this loneliness is correlated with low affiIiative
factor. A final factor was labeled Affiliative motivation and low social risk taking? The
Motivation; the affiliative tendency and in- ,general question is whether the relationships
troversion-extroversion scales had high load- found betweenlonelinessand social behaviors
ings. are due to the influence of a third set of
In the subsequent regression analyses, variables (i.e., the mood and personality mea-
scores on these four factors were used in lieu sures) on both the person's loneliness and
of the nine separate scales to predict scores social relations. If scores on the revised lone-
on the revised loneliness scale. All four factor liness scale are related to the concurrent
scores were entered simultaneously into the validity criteria independent of the mood and
regressionequation. The regressionresults are per~onality variables, then further discrimi-
summarized in Table 4. Affiliative motivation, nant validity evidence for the scale is pro-
social risk taking, and negative affect were vided.
all significant predictors of loneliness,whereas The concurrent validity analyses were
the social desirability factor was unrelated to therefore redone, using partial correlations to
loneliness. In combination, these four factors control statistically for the effect of the mood
predicted 43 C;cof the variance in loneliness and personality variables. Significant relation-
scores. ships (p < .001) were again found between
Although these factors were predictive of
loneliness, in combination they were able to
account for less than half the reliable variance 3The factor analysis results are available from the
in .lonelinessscores.Using the alpha coefficient authors.
 THE REVISED UCLA LONELINESS SCALE

loneliness scores and the amount of time would provide empirical evidence on this
alone each day (r = .27), the number of issue.
times dinner was eaten alone (r = .31), the Although the results from the current
number of weekend eveningsalone (r = .31), ~tudies support the utility of the UCLA Lone-
and the number of close friends (r = -.27). liness Scale as a measure of loneliness, it
However, the relationship between loneliness should be emphasized that the validity of a
and the number of social activities now only measure is never "proven." The validity of
approached significance (r = .12, P < .07). the lonelinessscale in other populations needs
Analysis of covariance was used to test the to be established,and the ability of the mea-
relationship between loneliness and current sure to detect loneliness in such "at risk"
dating or marital status, controlling for the groups as newcomers and the recently di-
effect of the mood and personality variables. vorced should also be investigate~. We hope
This significant relationship also persisted, that the UCLA Loneliness Scale will provide
F(2, 144) = 7.35, P < .001. a starting point for a greater understanding
of the widespread and distressing experience
General Discussion of loneliness.

The results from Studies 1 and 2 indicate Reference Note

that the UCLA Loneliness Scale has been
successfully revised. The revised loneliness 1. Brennan, T., & Auslander, N. Adolescent loneli-
ness: An exPloratory study of social and psy-
scale has high internal consistency,with a co- chological pre-dispositions and theory. Unpub-
efficient alpha of .94 found in both studies. lished manuscript, Behavioral Research Institute,
Concurrent validity for the new measurewas Boulder, Colorado, January 1979.
indicated by demonstrating that lonely people
report experiencing emotions theoretically References
linked to loneliness and do not report experi-
Beck, A. T. Depression. New York: Hoeber, 1967.
encing emotions unrelated to loneliness. Bradburn, N. The structure of psychological weU-
Lonely individuals also reported more limited being. Chicago: Aldine, 1969.
social activities and relationships. Discrimi- Bragg, M. A comparative study of loneliness and
nant validity for the revised loneliness scale depression. Unpublished doctoral dissertation,
was indicated by evidence that scoreson the University of California, Los Angeles, 1979.
Bumberry, W., Oliver, J. M., & McClure,.]. Validity
measure were not confounded by social de- of the Beck Depression Inventory in a university
sirability. Scoreson the scale were also found population using psychiatric estimate as the cri-
to correlate more highly with other measures terion. Journal of Clinical and Consulting Psy-
of loneliness than with the measuresof mood chology, 1978,46, 150-155.
Costello, C. G., & Comrey, A. L. Scales for measur-
and personality variables that were examined.
ing depression and anxiety. Journal of Psychology,
The revised loneliness scale passed a very 1967,66, 303-313.
stringent discriminant validity test, with the Crowne, D. P., & Marlowe, D. The approval motive:
demonstration that relationships betweenlone- Studies in evaluative dependence. New York:
liness and the concurrent validity criteria Wiley, 1964.
Eysenck, H. J., & Eysenck, S. B. G. Eysenck Per-
examined in Study 2 were independentof the sonality Questionnaire. San Diego, Calif.: Educa-
influence of the other mood and personality tional and Industrial Testing Service, 1975.
variables on loneliness. Freedman, J. L. Crowding and behavior. New York:
An important issue raised by altering the Viking Press, 1975.
Gordon, S. Lonely in America. New York: Simon
UCLA Loneliness Scale is whether empirical
& Schuster, 1976.
relationships found by researchersusing the Helmreich, R., & Stapp, J. Short forms of the Texas
original scale are still valid for the revised Social Behavior Inventory (TSBI) , an objective
version. In Studies 1 & 2, the high correlation measure of self-esteem. Bulletin of the Psycho-
of .91 between scores on the original and nomic Society, 1974, 4, 473-475.
Horowitz, L. M., & French, R. de Sales. Interper-
revised lonelinessscalessuggeststhat previous sonal problems of people who describe themselves
findings would still hold true for the new scale. as lonely. Journal of Consulting and Clinical
Future research replicating earlier findings Psychology, 1979, 47, 762-764.
480 D. RUSSELL, L. PEPLAU, AND C. CUTRONA

Jacobs, J. Adolescent suicide. New York: Wiley, Perlman, D., Gerson, A. C., & Spinner, B. Loneliness
1971. among senior citizens: An empirical report.
JoneS, W. H., Freemon, J. R., & Goswick, R. A. Essence, 1978,2(4),239-248.
The persistence of loneliness: Self and other re- Rathus, S. A. A 30-item schedule for assessing as-
jection? Journal of Personality, in press. sertive behavior. Behavior Therapy, 1973, 4, 398-
Lynch, J. J. The broken heart: The medical conse- 406.
quences of loneliness in America. New York: Russell, D., Peplau, L. A., & Ferguson, M. L. De-
Basic Books, 1976. veloping a measure of loneliness. Journal of Per-
Mehrabian, A. The development and validation of sonality Assessment, 1978, 42, 290-294.
measures of affiliative tendency and sensitivity to Spielberger, C. D., Gorsuch, R. L., & Lushene, R. E.
rejection. Educational and Psychological M easure- Manual for the State-Trait Anxiety Inventory.
ment, 1970, 30, 417-428. Palo Alto, Calif.: Consulting Psychologists Press,
Nerviano, V. J., & Gross, W. F. Loneliness and 1970.
locus of control for alcoholic males: Validity Stokols, D. On the distinction between density and
against Murray need and Cattell trait dimensions. crowding: Some implications for future research.
Journal of Clinical Psychology, 1976, 32, 479-484. Psychological Review, 1972, 79, 275-277.
Peplau, L. A., & Perlman, D. Blueprint for a social Weiss, R. S. Loneliness: The experience of emotional
psychological theory of loneliness. In M. Cook & and social isolation. Cambridge, Mass.: Massa-
G. Wilson (Eds.) , Love and attraction. Oxford, chusetts Institute of Technology Press, 1973.
England: Pergamon Press, 1979. Wenz, F. V. Seasonal suicide attempts and forms of
Peplau, L. A., Russell, D" & Heim, M. Loneliness: loneliness. Psychological Reports, 1977, 40, 807-810.
A bibliography of research and theory. JSAS
Catalog of Selected Documents in Psychology,
1978,8,38. (Ms. No. 1682) ReceivedJuly 30, 1979
.