Journal of Abnormal Psychology

1979, Vol. 88, No. 2, 174-181

Susceptibility and Precipitating Factors in Depression:

Sex Differences and Similarities
Lenore Sawyer Radloff Donald S. Rae
Center for Epidemiologic Studies Statistical and Mathematical Applications Branch
National Institute of Mental Health Alcohol, Drug Abuse and Mental Health
Rockville, Maryland Administration
Rockville, Maryland

The well-documented sex difference in depression may be due to a sex differ-

ence in susceptibility, in precipitating factors, or in both. Data from a large
mental health survey were used to study precipitating factors. It was found
that women were exposed more often to more of the factors that relate to
depression; with a few exceptions, the factors related to depression similarly
for men and women. Matching on these factors did not eliminate the sex
difference in depression. This suggests that there may be a sex difference in
susceptibility. Current theories of depression are incorporated into a se-
quential model of how learned factors might contribute both to susceptibility
and to the sex difference.

There is strong evidence that depression is
more common among women than among men.
This has been found not only by studies of
psychiatric patients but also by household
surveys using self-report measures of depres-
sive symptoms (Weissman & Klerman, 1977).
However, there is also evidence that the sex
difference occurs only under certain condi-
tions. For example, in previous analyses of the
data used in this article, it was found that
married and divorced/separated women were
more depressed than their male counterparts.
Among the widowed and never married, men
were found to be more depressed but not al-
ways significantly so (Radloff, 1975). In their
review of the evidence, Weissman and Kler-
The field work for this research was conducted by
the Epidemiologic Field Station, Greater Kansas
City Mental Health Foundation, Kansas City, Mis-
souri and by Johns Hopkins Training Center for
Public Health Research, Johns Hopkins University,
Hagerstown, Maryland, under contract from the
Center for Epidemiologic Studies, National Institute
of Mental Health.
Requests for reprints and for additional details
on the method and results should be sent to Lenore
Sawyer Radloff, Center for Epidemiologic Studies,
National Institute of Mental Health, 5600 Fishers
Lane, Room 10C-09, Rockville, Maryland 20857.
man concluded that the patterns of sex differ-
ences cannot be entirely explained by sex
biases in symptom reporting and help seek-
ing or by genetic and endocrine factors alone.
They suggested that the effects of social roles
and social stresses should be studied.
It is helpful to distinguish between sus-
ceptibility and precipitating factors in analyz-
ing the sex difference in depression. A sex
difference could exist in susceptibility, in ex-
posure to precipitating factors, or in both.
Susceptibility is sometimes assumed to be
innate and biological. However, current the-
ories of depression suggest that learning may
also contribute to susceptibility. Since men
and women differ in learning history, there
may be sex differences in learned susceptibil-
ity.
Precipitating factors are generally assumed
to be outside environmental insults. These
factors could be biological, but they would
also include social situations and stresses. If
men and women differ in the degree of ex-
posure to these factors and if these factors af-
fect men and women in the same way, then
this might contribute to the sex difference in
depression.
The purpose of the present study is to -ex-

In the public domain

174
 FACTORS IN DEPRESSION
plore the sex difference in depression through
a series of questions. First, are women, more
than men, exposed to the circumstances that
are likely to precipitate depression? Second,
are these circumstances related to depression
in the same way for both men and women
across household status (based on marital
status, children, and relationship to head of
household)? If so, then does adjusting for
these factors reduce or eliminate the sex dif-
ference in depression? Finally, the interaction
of susceptibility and precipitating factors is
discussed in the context of social learning
theories of depression.
Method
Subjects and Design
A mental health interview survey was conducted
in Kansas City, Missouri, in 1971-1972 and in
Washington County, Maryland in 1971-1973.1 Sam-
ples of households were drawn to be representative
of each community, excluding residents of certain
institutions, such as hospitals and prisons. One re-
spondent was randomly selected for interview from
the household residents who were age 18 and over.
Interviews were obtained from 74.8% of the se-
lected sample in Kansas City and from 80.1% in
Washington County. Because preliminary analyses
suggested that the whites and the nonwhites may
differ in relations among certain variables, the present
article covers analyses of the 2,515 whites only (876
from Kansas City and 1,639 from Washington
County).
Apparatus
The questionnaire used in this survey included
over 300 separate questions covering psychological
and psychophysiological problems (especially symp-
toms of depression, including symptoms listed on
the Center for Epidemiologic Studies Depression
Scale [CES-D]) and other variables hypothesized to
be related to depression (including physical illness,
need for and use of treatment, demographic vari-
ables, recent life events, and social functioning). The
measure of income was total household income
rather than personal income of the respondent.
The CES-D was developed to measure symptoms
of depression in epidemiologic studies of general
populations. It consists of 20 items selected from
previously existing depression scales (e.g., Zung Self-
Rating Depression Scale, Beck Depression Inventory,
Minnesota Multiphasic Personality Inventory), se-
lected so that each major factor (or cluster of symp-
toms) in the clinical syndrome of depression (as
identified by clinical judgment and factor-analytic
175
studies) was represented by a few items. Items were
also chosen on the basis of validity and discrim-
inatory power when this could be determined from
the literature. Responses were on a 4-point scale of
the frequency with which the symptom was experi-
enced in the past week. The score is the sum of the
response weights (0 = never to 3 — most of the time
for the negative items and the reverse for the posi-
tive items). (See Appendix for CES-D items.)
Properties of the scale (its internal consistency
and validity) have been reported in detail elsewhere
(see Radloff, 1977). The scale discriminated well be-
tween psychiatric inpatients and the community
samples. Among depressed outpatients, the CES-D
correlated highly with the Hopkins Symptom Check-
list (SCL-90) and the Hamilton and Raskin clinical
rating scales for depression; there was also a signifi-
cant reduction of scores over time as depressed pa-
tients under treatment recovered (Weissman, Sho-
lomskas, Pottenger, Prusoff, & Locke, 1977). Within
the survey, the CES-D showed the relations with
other variables that would be expected of a depres-
sion scale according to the literature on depression
(e.g., see Comstock & Helsing, 1976; Radloff, 1975).
The scales representing precipitating factors were
selected because they were reasonably objective re-
ports of current life situations and activities, be-
cause they could be linked to depression on the basis
of previous research, and because they related at
least moderately (p < .10) to depression in the
present data. Combinations of items into small scales
were made only if the items were conceptually re-
lated and if they related to depression in the same
direction (even if not significantly) for both sexes.
The 17 scales were classified into five categories:
Background characteristics of the person. Age,
education, and early loss (whether or not the person
lived with own parents throughout childhood).
Economic situation in the household. Household
income, any welfare (whether the source of income
includes welfare or unemployment), and knowledge
of welfare/unemployed (whether the person knows
anyone personally who is either unemployed or on
welfare).
Job or duties of the person. Occupation (Hol-
lingshead level 1-7 of occupation if employed; house-
wives and retired coded 8, unemployed and student
coded 9), number supervised (by respondent on job;
0 if not employed), and household activities (leave
the house, spend money, and work around the
house and yard).
Other activities, not job related. Get together
(with friends or neighbors), excessive bedrest (stay
1
Note that these results cannot be used to assert
the null hypothesis of no sex differences even in the
limited number of precipitating factors studied. How-
ever, since the selection and scaling of the variables
was done to maximize chances of finding differences,
it is safest to be conservative in interpretation.
Cross-validation is the only real test of the gen-
erality of the results.
176 LENORE SAWYER RADLOFF AND DONALD S. RAE
Table 1
Average Depression Scores (CES-D) by Sex and Household Status
1 : spouse, no
Sex children
Male
n 307
M 6.01
Female
n 337
M 8.36
in bed, oversleep, nap), leisure activities—general
(reading, hobbies, public entertainment, games),
leisure activities—specific (more specific such as
reading the newspaper, going to movies), and re-
ligious activities (attending church, reading the
Bible).
Physical status. Current illness, recent illness, and
handicap (interviewer rating, need help in self-care).
Because the precipitating factors might vary de-
pending on the person's living situation, the total
male and female samples were subdivided into house-
hold status subgroups. Those who defined themselves
as head of household, cohead of household, or
spouse of head of household were designated as liv-
ing in their "own household." These were further
classified as living with spouse but no children
(Group 1), with spouse and children (Group 2), or
with no spouse (Group 3). Group 4 contained those
not living in their own household (about 68% were
child of head of household, 22% were other rela-
tives, and 10% were unrelated to head of household).
Results
Basic Sex Difference in Depression
Previous analyses of these data (Radloff,
1975) have shown that marital status and re-
lationship to head of household interacted
with sex in their relations to depression.
Women who were married, divorced/sepa-
rated, and never married and who were not
heads of their own households were found to
be more depressed than their male counter-
parts. Men were as depressed as women (or
slightly more so) among those widowed and
those never married who were heads of
household. Table 1 shows the analysis of
variance of depression scores by sex and
household status. The sexes differed signifi-
cantly, F(1, 2481) = 21.23, p < .001 (women
were higher), as did the household status
Household status
2 : spouse, 4 : not own
children 3 : no spouse household
461 145 94
8.04 9.92 10.40
S7S 472 98
10.11 10.75 12.90
groups, ^(3, 2481)-20.20, p < .001. Al-
though the interaction was not significant,
separate tests within each household status
suggested that the sex difference in Group 3
(no spouse) was not significant. This group
combined never married, widowed, and di-
vorced/separated respondents so that the Sex
X Marital Status interaction was obscured.
Marital status was entered as a separate vari-
able in later analyses of this group.
Are Women More Exposed to Precipitating
Factors?
On 12 of the 17 scales representing pre-
cipitating factors, there was a difference (p
< .10) between the means for males and fe-
males. On 8 of these scales, the females scored
more in the direction associated with depres-
sion. These 8 scales (education, income, oc-
cupation, number supervised, leisure activ-
ities-general, leisure activities-specific, cur-
rent illness, and recent illness) were considered
candidates for explaining away the sex dif-
ference in depression. The four scales on
which males scored in the direction associated
with depression (age, bedrest, religious ac-
tivities, and handicap) must also be con-
sidered, since adjusting for them might in-
crease the sex difference in depression.
Do the Same Precipitating Factors Relate to
Depression for Males and Females?
Fourteen of the 17 scales were related to
depression in the same direction (at the 10%
level or better) for males and females (not
subdivided by household status). Three vari-
 FACTORS IN DEPRESSION
ables reached the 10% level only for the
females. These were early loss, knowledge of
welfare, and leisure activities-general.
When the males and females were further
subdivided into household status groups,
there were many sex differences in the correla-
tions of the scales with depression. In Group
1 (spouse, no children), the men and women
were alike (related significantly in the same
direction) on only 4 scales. In Group 2
(spouse and children), the sexes were alike
on 10 scales. In Group 3 (no spouse), the
sexes were alike on 4 scales, whereas on 1
scale (divorced/separated), they were both
significant but in opposite directions. In
Group 4 (not own household), they were
alike on only 1 scale and significant in op-
posite directions on 1 scale (religious ac-
tivities). Further analyses were conducted to
interpret the importance of these differences.
If the scales within each of the categories
are considered alternative measures of the
broader concept (e.g., income, any welfare,
and knowledge of welfare as measures of
household economics), then the differences ap-
pear much less important, especially among
the married respondents (Groups 1 and 2),
in which numbers were largest. In Groups 1
and 2, at least one scale from each category
was related to depression (in the same direc-
tion) for each sex, with only one exception
(no background scale for Group 1 males). In
Group 3, household economics and job/duties
related to depression for females but not
males; in Group 4, they related to depression
for males but not for females. In Group 4,
there was no background scale important for
males.
Since the intercorrelations among the scales
may have obscured some relations, several
multiple regression methods were used to in-
terpret the data further. When all scales were
used in a multiple regression to predict de-
pression, the regressions for males and fe-
males did not differ significantly in Groups 2
and 3; there were borderline differences (.05
< p < .10) in Groups 1 and 4. Another way
to compare the sexes was to let the computer
pick the best subsets of scales specific for
each subgroup and then compare them across
the sexes. (A procedure called All Possible
177
Subsets Regression (BMD-P9R) was used.)
Both the overall multiple regression and the
best subsets method confirmed the impression
from the simple correlations: The specific
scales within the broader categories could
usually substitute for each other so that sex
differences were minimal.
Finally, the sexes were compared on a
universal subset of scales (i.e., the same sub-
set of variables for each group). With future
cross-validation in mind, these scales were
chosen for their simplicity, for their clarity,
and for the likelihood that they would be
found in other studies as well as for their
strength and their generality in the present
study. The scales selected were age, current
illness, education, income, get together, re-
ligious activities, and bedrest. Widowed and
divorced/separated categories were also used
in Groups 3 and 4.
Using this universal subset of variables, the
multiple regressions were compared across the
sexes within each household status. The sexes
did not differ significantly except in Group 4
(p < .03). The difference in Group 4 seemed
to be mainly due to the reversal of sign for
religious activities for Group 4 males. Com-
paring the betas for specific scales suggested
that education was more important for mar-
ried females than for married males (Groups
1 and 2); income was more important for
Group 1 males than for Group 1 females;
get together was more important for Group 3
females than for Group 3 males; divorced/
separated was positively associated with de-
pression in Group 3 females but negatively so
in Group 3 males; and religious activities
was positive for Group 4 males and negative
for Group 4 females.
Considering all of the analyses, there were
a few intriguing hints of sex differences in
the correlates of depression that might merit
further study. For example, economic and
occupation variables seemed more important
for married men whereas the education vari-
able was more important for married women.
This difference should be tested separately
for housewives and for wives employed out-
side the home. However, there seemed to be
more similarities than differences between the
sexes: Most of the conditions associated with
178 LENORE SAWYER RADLOFF AND DONALD S. RAE
depression for women were similarly associ-
ated for men (see Footnote 1). Age was
strongly related to depression for all groups:
The younger people were more depressed.
Physical illness was also important for all
groups. Early loss of parents (as measured
here) was not important, except for one
group. At least one of the economic scales and
one of the occupational scales was important
in six of the eight groups. One or more of the
"other activities" scales was important in all
groups. There seemed to be enough similarity
to return to the original question: Will ad-
justing for these variables eliminate the sex
difference in depression?
Application of the Findings to the Basic Sex
Difference
For each subgroup, the variables of the
universal subset were introduced as covariates
in analysis of covariance by sex. Analyses
were done using one covariate at a time, then
combinations of a few covariates, and finally
all of the covariates together. (Marital status
and religious activities were omitted from
the covariance analyses for Group 3 and
Group 4, respectively.) For the married re-
spondents (Groups 1 and 2) and the not-own-
household (Group 4) respondents, there was
no covariate or set of covariates that elimi-
nated the sex difference in depression. The
main effect for sex was significant at the .01
level or better in every analysis. For the non-
married respondents (Group 3), the sex dif-
ference was not significant initially, and with
income alone as a covariate, the mean differ-
ence in depression for men and women was
essentially zero. However, it should be noted
that marital status was an important variable
in this group: Among the divorced/separated
respondents the women were more depressed
than the men, but among the widowed and
never married respondents, the men were more
depressed.
Discussion
In the data reported here, it was found that
women were exposed more than men to more
of the situational (precipitating) factors that
relate to depression, although men were more
exposed to some of these factors. With the ex-
ception of marital status, the situational
factors studied had similar relations to de-
pression for men and women even when house-
hold status was taken into account. Con-
trolling for these factors (both singly and in
combination) did not substantially change
the pattern of sex differences. Women had
higher depression scores than men among mar-
ried people (with and without children) and
among people not living in their own house-
holds. Among the nonmarried people in their
own households, the sex difference depended
on marital status.
These findings suggest that the sex differ-
ence in depression is unlikely to be due en-
tirely to a difference in the kind of precipitat-
ing factors studied here. The data do suggest
that there is probably some sex difference in
susceptibility that may be partly biological
(the mechanism as yet is unclear) and partly
learned.
Possible mechanisms for learned susceptibil-
ity are suggested by current theories of de-
pression, including the reinforcement model
(e.g., Ferster, Lewinsohn), which focuses on
low rates of rewards or rewarded behavior in
depressed people; the learned helplessness
model (Seligman), which focuses on lack of
control over rewards and punishments; and
the cognitive model (Beck), which focuses on
pessimistic beliefs, including (but not limited
to) the belief that you have no control over
rewards and punishments. (For summaries
of these, see Friedman & Katz, 1974.) Mc-
Lean (1976) has included these three factors
(and others) in a sequential model which
suggests that "depression is the consequence
of ineffective coping techniques used to
remedy situational life problems" (p. 69).
McLean also suggested that over time, re-
peated goal frustration leads to feelings of
lack of control, which lead to anticipation of
chronic failure, which leads to depressive
symptomatology. Depression itself is seen as
a sequence of symptoms in order of severity:
cognitive and mood disturbances followed
first by behavioral disturbance and then by
somatic complaints.
A slightly different sequential model (in-
 FACTORS IN DEPRESSION
fluenced by all four of the above theories) is
proposed here, taking account of susceptibil-
ity and precipitating factors, with applica-
tions to the sex difference in depression. The
cognitive dimension of depression (beliefs
about control over reinforcers, similar to the
concepts of learned helplessness and negative
cognitions) is here seen as a learned suscepti-
bility factor. These beliefs could be learned
directly, by experience in uncontrollable sit-
uations (as in the learned helplessness model;
see Friedman & Katz, 1974) or by verbal/
symbolic means, such as hearing other people
advocate such beliefs. More indirectly, a past
history of overall low rates of reinforcement
(as in the reinforcement model; see Fried-
man & Katz, 1974) would lead to an im-
poverished response repertoire so that prob-
lem-solving skills would not be available. Past
history of conflict (responses simultaneously
rewarded and punished) may also interfere
with learning or at least with overt respond-
ing. Both of these would reduce successful
problem solving. Repeated lack of success
might lead to a generalized expectation of
failure. Repeated goal frustrations (see Mc-
Lean, 1976) would include a history of non-
reinforcement, in any of these contexts.
McLean (1976) and Beck (see Friedman
& Katz, 1974) suggested that depressive af-
fect follows directly from the negative cog-
nitions, whereas the reinforcement model sug-
gests that it is the result of too few posi-
tive and too many negative reinforcers (see
also Brenner, 1975). It is suggested here that
the cognitive dimension is merely a suscepti-
bility factor that is latent until activated by
the relevant precipitating factor, which is a
goal-oriented problem situation (a reward to
be gained or a punishment to be avoided or
escaped). At that point, the cognition would
lead to the expectation of failure of all at-
tempts to solve the problem. Presumably,
this expectation would reduce the probability
of adequate goal-oriented behavior (i.e., the
behavioral deficit of depression). This in
turn would reduce the probability of rein-
forcement, leaving the individual with a defi-
cit in rewards and/or excess of punishments,
which results in depressed affect (absence of
enjoyment as well as presence of sadness).
179
The somatic dimension of depression (ap-
petite and sleep disturbance and possibly
fatigue and lack of energy) is assumed to be
the last (and most severe) stage (see Mc-
Lean, 1976). How it develops is not clear, al-
though there is some theoretical and empirical
basis for suggesting that it may follow from
the affective and/or behavioral disturbances
(Brenner, Note 1).
This model of depression has its own
built-in vicious cycle. The expectation of fail-
ure leads to actual failure, which confirms
and strengthens the expectation. The de-
pressed affect interferes with social interac-
tions. This further reduces reinforcers. Lack
of sleep and appetite would result in fatigue,
and lack of energy may also be a direct so-
matic symptom. These will aggravate the be-
havioral deficit. It may be that severity of
depression is related to the length of time
that this cycle continues before being inter-
rupted. Theoretically, interruption at any
point would help, but only a change in the
cognitive dimension would prevent relapse.
If learned susceptibility (as in this or
similar models) does contribute (along with
other factors) to depression, then it probably
contributes to the sex difference. There is
evidence that men and women differ in the
learning histories relevant to the development
of the cognitive dimension. (For a review of
this evidence, see Radloff & Monroe, 1978.)
Men are given more contingent reinforce-
ment for their behavior (e.g., Maccoby &
Jacklin, 1974; Serbin, O'Leary, Kent, &
Tonick, 1973) and more direct verbal train-
ing on the importance of effort and control
(e.g., Dweck, Davidson, Nelson, & Enna,
1978). Women are more often simultaneously
rewarded and punished (e.g., Unger, 1976).
If measures of locus of control are relevant
to the cognitive dimension, then the fact that
women are (on the average) more external
than men suggests that this difference in
past history has had the expected effect.
The model presented here suggests that if
no precipitating factor is present, there should
be a sex difference in the cognitive dimension
of depression but not in the behavioral, affec-
tive, or somatic dimensions. In problem situa-
tions, however, women should be higher than
180 LENORE SAWYER RADLOFF AND DONALD S. RAE
men on all four dimensions and on the full
syndrome of clinical depression. This is a
testable hypothesis.
More generally, this model suggests that
the sex difference in depression may not be
due to gender per se; the factors that con-
tribute to depression in women would prob-
ably have the same effect on men who are
exposed to them. If men and women (or any
other two groups) could be matched on both
susceptibility and precipitating factors, they
should be similar in rates of depression. Speci-
fication of the relevant factors is the challenge
for further research.
Reference Note
1. Brenner, B. The nature of the relationship be-
tween depressed affect and its somatic correlates.
Paper presented at the Society for Epidemiologic
Research, Iowa City, June 1978.
References
Brenner, B. Enjoyment as a preventive of depressive
affect. Journal of Community Psychology, 1975, 3,
346-357.
Comstock, G. W., & Helsing, K. J. Symptoms of de-
pression in two communities. Psychological Medi-
cine, 1976,15,551-563.
Dweck, C. S., Davidson, W., Nelson, S., & Enna, B.
Sex differences in learned helplessness: II. The con-
tingencies of evaluative feedback in the classroom
and III. An experimental analysis. Developmental
Psychology, 1978,14, 268-276.
Friedman, R. J., & Katz, M. M. (Eds.). The psy-
chology of depression: Contemporary theory and
research. Washington, D.C.: V. H. Winston, 1974.
Maccoby, E. E., & Jacklin, C. N. The psychology of
sex differences. Stanford, Calif.: Stanford Univer-
sity Press, 1974.
McLean, P. Therapeutic decision-making in behav-
ioral treatment of depression. In P. O. Davidson
(Ed.), The behavioral management of anxiety, de-
pression and pain. New York: Brunner/Mazel,
1976.
Radloff, L. S. Sex differences in depression: The ef-
fects of occupation and marital status. Sex Roles,
1975, 1, 249-265.
Radloff, L. S. The CES-D scale: A self-report de-
pression scale for research in the general popula-
tion. Journal of Applied Psychological Measure-
ment, 1977,1, 385^01.
Radloff, L. S., & Monroe, M. M. Sex differences in
helplessness: With implications for depression. In
L. S. Hansen & R. S. Rapoza (Eds.), Career de-
velopment and counseling of women. Springfield,
111.: Charles C Thomas, 1978.
Serbin, L. A., O'Leary, D. K., Kent, R. N., &
Tonick, I. J. A comparison of teacher response to
problems and preacademic abilities of boys and
girls. Child Development, 1973, 44, 796-804.
Unger, R. K. Male is greater than female: The so-
cialization of status inequality. The Counseling
Psychologist, 1976, 6, 2-9.
Weissman, M. M., & Klerman, G. L. Sex differences
and the epidemiology of depression. Archives of
General Psychiatry, 1977, 34, 98-111.
Weissman, M. M., Sholomskas, M. P., Pottenger, M.,
Prusoff, B. A., & Locke, B. Z. Assessing depressive
symptoms in five psychiatric populations: A
validation study. American Journal of Epidemiol-
ogy, 1977, 106, 203-214.
 FACTORS IN DEPRESSION 181

Appendix
Center for Epidemiological Studies Depression Scale (CES-D)

Instructions for questions on the CES-D: Below is a list of the ways you might have felt or
behaved. Please tell me how often you have felt this way during the past week. Hand Card A.

Rarely or none of the time (less than 1 day)

Some or a little of the time (1-2 days)
Occasionally or a moderate amount of time (3-4 days)
Most or all of the time (5-7 days)

During the Past Week Rarely A little Moderate Most

1. I was bothered by things that usually don't bother
me 0 1 2 3
2. I did not feel like eating; my appetite was poor 0 1 2 3
3. I felt that I could not shake off the blues even with
help from my family or friends 0 1 2 3
4. I felt that I was just as good as other people 3 2 1 0
5 . I h a d trouble keeping m y mind o n what I w a s doing 0 1 2 3
6. I felt depressed 0 1 2 3
7 . I felt that everything I d i d w a s a n effort 0 1 2 3
8 . I felt hopeful about t h e future 3 2 1 0
9. I thought my life had been a failure 0 1 2 3
10. I felt fearful 0 1 2 3
11. My sleep was restless 0 1 2 3
12. I was happy 3 2 1 0
13. I talked less than usual 0 1 2 3
1 4 . I felt lonely 0 1 2 3
15. People were unfriendly 0 1 2 3
16. I enjoyed life 3 2 1 0
17. I had crying spells 0 1 2 3
18. I felt sad 0 1 2 3
19. I felt that people disliked me 0 1 2 3
20. I could not get "going" ' 0 1 2 3

Received June 30, 1978