SIDDIQUI-SHAH DEPRESSION SCALE (SSDS)

Development and Validation

Salma Siddiqui & *Ashiq Ali Shah

Running Head:

Siddiqui-Shah Depression Scale

National Institute of Psychology

Quaid-i-Azam University
Islamabad, Pakistan

----------------------------
*
The second author is now working at the Department of Psychology, International Islamic University
Malaysia. All correspondence should be addressed to the second author.
The purpose of the study was to develop and validate a self-report scale to measure depression in both clinical
and non-clinical Pakistani populations. The 72 items generated from the university students were judged for
their relevance to depression by the psychiatrists and clinical psychologists. An approximate 50% consensus
among judges was taken as selection criterion. The 36 items so obtained were split into two equivalent halves
and tested on the clinical as well as non-clinical populations. The spilt half reliability of the scale with
Spearman-Brown correction were r= 0.79 and r= 0.84 for the clinical sample and r= 0.80 and r= 0.89 for the
non-clinical samples respectively. The Co-efficient Alpha for the clinical and non-clinical samples were 0.91
and 0.89. The scale correlated significantly with the Zung's depression scale, r= 0.55, (p<.001) and
psychiatrists' ratings of depression r= 0.40, (p<.05). The scale also showed a significant correlation with
subjective mood ratings for the clinical group r= 0.64, (p<.001) as compared to the non-clinical group r= 0.14,
(p= n.s). The scale also demonstrated high sensitivity and specificity. The percentiles and cut off scores for the
clinical as well as non-clinical groups have been determined.

The psychological concept of depression has been variously described as having the blues; feeling sad,
guilty, hopeless, helpless and melancholy and as reacting to the grief of losing some loved object. It is also
described as a feeling state or symptom, a syndrome or reaction, a characteristic or life style, and/or and illness
(Schuyler, 1974). Some researchers have adopted an operational approach to the definition and subsequent
classification of depression (Depue & Monroe, 1978, Spitzer, Sheehy & Endicott 1977). The most prevalent
definition is that of a dysphoric (chronic) feeling of illness and discontented mood and/or a pervasive loss of
interest which is characterized by certain symptoms (Spitzer, Sheehy & Endicott, 1977). These and other
similar definitions of depression are the product of work on the conceptualization of depression and the
development of instruments for its assessment primarily carried out in the West. Assuming the universality of
psychological disorders, such as depression, one can use these instruments in other cultures, which have a
different outlook than the Western one. However, one can not ignore the errors in assessment as a result of the
disregard for local values and norms. One way of taking care of such errors is to adapt and standardize the
instrument before using it in culture other than where it was developed. Such attempts too, carry the limitation
as cultures differ not only with respect to their norms and values but also in terms of their lexical categories for
emotions (Russell, 1991). Emotional experiences and their expressions are determined to a great extent by the
words available in a particular language. The basic categories of emotion may be pan-cultural but the
expression varies with the degree of permissiveness present in a culture alongwith the available distinct lexical
categories. For instance, Muslim societies impose religious restrictions on the expression of sexual desires. The
lack of sexual experiences, therefore, in the case of unmarried persons, especially women, renders the items
attempting to measure a decrease in libido or sexual feelings as an indicator of depression, irrelevant.
Cross-cultural research has yet to emphasize the peculiarity and complexity of the cultural meanings
associated with psychological disorders by lay persons. There appears an overwhelming concern to adapt
2

2
Western models of psychological disorders while disregarding the more local nuances of emotional and other
experiences. Such an outlook poses both methodological as well as problems of validity due to the difficulties
in linguistic and conceptual translation in representing illness episodes as meaningful social events. Therefore,
to make the analysis and conceptualization of a disorder more universal, credence must be given to the
conceptual organization of cultural knowledge of that disorder. That is, to discover how a lay person talks about
his illness in social as well as personal context. This emphasis becomes critical with reference to the assessment
of depression. Despite its universality, depression may manifest differently across various cultures as
expression of emotion is determined both by the language and conceptual organization of the disorder. For
instance, psychological and mental symptoms are reported to be less prominent (and/or less differentiated) in
certain non-western societies than somatic features (Marsella & White, 1984). In non-western cultures, there is
a tendency in the depressives to somatize their illness (Nijdam, 1986; Sethi, 1986) either due to illiteracy or
lack of awareness as well as lesser acceptability of the psychological disorders in these cultures (Shah, 1993).
Not only this, a difference in value orientation may determine specific predictors of depression as well (Aldwin
& Greenberger, 1987). For instance, difference in intrinsic cultural values resulted in higher scores of Japanese
University students on self-report of depression as compared to their American counterpart (Baron &
Matsuyama, 1987). With reference to traditional Muslim societies one significant difference manifests in terms
of suicide ideation, which is generally regarded as a taboo (Shah, 1993) and therefore, may not be an
appropriate measure of depression.
Moreover, the suicidal ideas of depressed may have no correspondence with the actual proportion of
suicide committed (Venkoba Rao, 1978) which may be a mere desire to be rather dead. Such differences in
value structure across cultures and the documentation of culture specific contents of emotion have highlighted
the need to develop indigenous norms and culture relevant operational definition of psychopathology.
The present study was carried out to realise this need in the context of Pakistani culture, which is rather
unique in its composition. The complex intermingling of religious dictates and social values influences the
individual's attitude and thinking pattern toward distressing situation. Often religious considerations come in
conflict with the more pressing social values. In a culture where long standing values are giving way to more
materialistic pursuits, the class structure has been going through changes. The majority of women still have
their traditional roles, fending and feeding the family, though they can have a peep into the changing society
through the television. This often leads to conflicts and helplessness. The conflicting value structure leads to
emotional distress but its expression is handicapped because of low literacy level, authoritarian family set up
and scarce psychiatric help. In Pakistani villages, men and women still go to faith healers when they get
depressed fearing it to be a possession by evil spirits. The rapid urbanization and the disregard of basic
amenities in villages give rise to a prevalent sense of deprivation and frustration in their efforts to change their
lot that in turn paves way for a feeling of lack of control. The authoritarian structure of society inculcates an
attitude of self-blame. In such a cultural context, if an instrument measuring depression, developed in the West,
3

3
is administered, it is possible that individuals may spuriously score higher, whereas, many cultural peculiarities
will be left unassessed. Therefore, in the absence of any local empirical data the development of a scale which
is based upon the experiential definition of depression stands relevant.

METHOD
The development of an indigenous depression scale was carried over three phases, comprising of
generation of items for the scale, assessment of the relevance of items and determination of its reliability and
validity. Three different samples were taken for each phase according to the specific purpose entailed.

Phase I
The purpose of the phase I was to generate the items for the depression scale.

Sample:
Sample consisted of 80 (40 male and 40 female) students from the University of Punjab, Lahore and
Quaid-i-Azam University, Islamabad. Their ages ranged from 20-25 and they were all studying in postgraduate
classes. The sample could be regarded as representative of the Pakistani university students as those enrolled
here belonged to geographically different parts of the country.

Procedure:
The subjects were given a semi-structured questionnaire, which consisted of three steps.

i) subjects were given instructions to recall and enlist those situations when they felt depressed.
They could write as many situations they felt relevant.

ii) when the subjects completed the listing of situations the researcher gave them specific
examples in order to explain to them the relation of cognition, feelings and behaviour to a
depressing situation.

iii) they were then asked to write their cognitions, feelings and behaviour regarding the situations
earlier listed.

The individual protocol (containing situations, cognitions, feelings and behaviour associated with it)
were compiled and statements were drawn from them. This resulted in 72 constructs pertaining to depression.
These were then selected for the second phase of the study. The statements extracted from the protocols helped
generate the item pool for the selection of items for the scale. Most of the subjects expressed in Urdu, therefore,
the remaining protocols were translated from English to Urdu. Excluding the repetitive statements 72 most
representative constructs were selected which pertained to either, cognitions, feelings or behaviour related to
depression. Out of 72, 48 constructs stated cognitions, 16 described feelings and 8 constructs were related to

4
behaviour, more specifically to the somatic complaints. These 72 constructs were then used in the second phase
of the study to establish the intensity of each item with regard to depression.

Phase II
As the items for the scale were generated by the student sample, it was considered necessary to determine the
intensity of the items with regard to their relevance for clinically depressed group. Therefore, consensus among
the judges was obtained to select the most representative items.

Sample
Forty-two judges participated to help assess the relevance of the items. These judges had the practical
clinical experience and were working in different psychiatric setting of the hospitals and other institutions as
psychiatrists as well as clinical psychologists in Rawalpindi, Lahore and Peshawar.

Procedure
Judges were asked to rate the items in the light of their clinical experience, keeping in mind the actual
occurrence of these cognitions, feelings and behaviours in the clinical sample. They were given the list
containing 72 items, along with a 3-point scale denoting "1" as normal sadness, "2" as mild depression and "3'
as severe depression. If an item appeared more characteristic of normal sadness, the judges rated it as '1', if it
was found to be reported more frequently by mildly depressed persons, the item was given a rating of '2', and if
the item was considered as characteristic of severely depressed patients, the rating assigned was '3'. This helped
determine the classification of items as characteristics of normal sadness, mild depression or severe depression.
The second pilot study helped select the items for the final scale. Frequencies and percentages were obtained
for each item, which indicated judges' consensus for assigning items into one of the three categories (see Table
1). The items, which were assigned to one of the three categories with above 50% consensus among the judges,
were selected for the final scale. However, in the category of normal sadness and `severe depression' the
criterion had to be lowered to 47% for few items to make the number of items equal in each category. In all, 36
items were retained, 12 in each category. (i.e., normal sadness, mild depression and severe depression).

Phase III
In this phase of the study the validity and reliability of the scale was determined for the psychiatric and
non-psychiatric groups.

Sample:
5

5
 The non-clinical sample for this phase of study comprised of 206 male and female university students
from five universities of major cities of the country. They were: Quaid-i-Azam university, Islamabad; Punjab
university, Lahore; Karachi university, Karachi, university of Peshawar, Peshawar; and Balochistan university,
Quetta. An attempt was to ensure more approximate representation of non-clinical group. The subjects were
students at post-graduate level and their ages were in the range of 22 to 28.
The clinical sample for this phase, which comprised of 60 patients (23 male and 37 female) was
selected from two major hospitals of Rawalpindi and Islamabad. They were mostly tested at the psychiatric
outpatient department of the hospital, few of them, however, were contacted in the psychiatric wards as well.

Instruments:
The non-clinical group was assessed on the following measures:-

(a) A self-rating 7-point mood scale (the subjects judged their current mood by selecting any
number from -3 to +3 on this scale. The +3 was labelled as `very pleasant' and -3 as `very
unpleasant'. The scale denoted `zero' for indicating a mood which is neither pleasant nor
unpleasant.

(b) The 36-item obtained from phase II that related to various degrees of depression.

(c) The Zung's depression scale (Zung, 1965), which has been designed for use with general
population, is brief and can be self-administered. Moreover, it covers affective, psychological
and somatic symptoms and has been used in cross-cultural research (Marsella, 1980).

(d) A questionnaire about personal information for instance, age, occupation etc.

The clinical group was also assessed on these instruments, with the exception of Zung's depression
scale as it was in English and its administration on them was not feasible. Instead psychiatrists making the
referral of the patient for the study were asked to evaluate the clinically depressed on an assessment form based
on ICD-10. There were 30 such patients from the total sample of 60 who were diagnosed by the psychiatrists
before being assessed on 36-items.

Procedure:
The non-clinical group was studied in groups and instructions followed the introduction of each
measure, whereas, the clinical group was studied individually in a one-to-one situation. Patients diagnosed by
psychiatrists as depressed and capable to communicate were selected for the study.

RESULTS
The data of phase III of study provided the differential analysis of 36 items scale between the clinical
and non-clinical group. The clinical group had a higher mean = 51.93 and SD = 18.33 than the non-clinical

6
group, mean = 27.93, SD = 12.7.

Item Analysis
The item analysis was performed to determine the relationship of individual items with the scale. The
result showed that all the item total correlations were significant at 0.05 level and above.
-----------
insert table 1 about here
-----------

Split Half Reliability

As the items of the scales were randomly ordered (in terms of the relevance of the items) into two
equal halves, the split - half reliability was computed, which indicated a high significant value of r = 0.79
p< .001 (Spearman Brown correction r = 0.84, p< .001) for the clinical group and an equally high significant
value of r = 0.80 (p< .001) for the non-clinical group (Spearman Brown correction r = 0.89, p< .001).

----------
insert table 2 about here
----------

Internal consistency reliability

Coefficient alpha was computed to determine the internal consistency of the scale. The value obtained
for the clinical group was 0.91, (p< .001) and for the non-clinical group 0.89(p< .001).

-----------
insert table 3 about here
-----------

Construct Validity
The correlation between the scores of non-clinical group on our and Zung's depression scale indicated
a value of r =0.55 (p< .001). The scores of the clinical group significantly correlated with psychiatrists ratings
for depression (r =0.4, p< .05). The scores of the clinical group on the scale correlated significantly with their
self-reported mood (r -0.64, p< .001), whereas, the correlation of the scores of the non-clinical group with their
subjective mood ratings was not-significant (r = 0.14, p > .05).
-----------
insert table 4 about here
7

7
 -----------

Factorial Validity
Principal component analysis was performed to determine the factor structure of the scale. The
majority of the items except item no. 4, 5, 12 and 35 had factor loadings of .35 and above on the first factor.
The first factor had an eigenvalue of 8.4 and explains 23.3% of the total variance. All the items are positively
loaded on the first factor, Which is regarded as depression related factor. The obtained one factor solution
explaining major proportion of variance suggests unidimensionality of our scale. All other factors had
eigenvalues less than 2 and explained 5% and less of the total variance.
-----------
insert table 5 about here
-----------

Sensitivity and Specificity

The sensitivity and specificity were determined for three different categories of affect i.e., normal
sadness, mild depression and severe depression. The sensitivity of the scale i.e., correct identification of
depressives is quite high for normal sadness and mild depression i.e. 95% and 77% respectively, whereas, it is
50% for severe depression. The specificity of scale, i.e. correct identification of non-depressed is 55% for
normal sadness, 84% for mild depression and 94% for severe depression. A 50% sensitivity of the scale in our
opinion is due to the inclusion of a large proportion of non-depressed in the sample and the inclusion of items
of normal sadness, which add to the score of depression. Moreover, the clinical group included individuals with
different degrees of depression as indicated by a high standard deviation.
Cut-Off Scores
As the scale measures the frequency of the indicators of depression, subject's score is determined by
the category of endorsement of his responses. The frequency distribution of the score value was computed for
both the clinical and non-clinical groups separately to determine the cut-off points for the scale. The two
frequency distributions were used to locate an optimal cutting score that would minimize the sum of false
positives and false negatives. This procedure has been used by other scale developers as well (Westhuis &
Thyer, 1989). A score of 26 specifies 56% of the non-clinical sample as not depressed, whereas, the same score
classifies less than 5% as non-cases in the clinical sample. Taking a score of 26 as the lowest score and the
score of 36 as the upper limit indicative of "mild depression", provides us the first range of clinical cut-off
scores. The frequency distribution of the scores explains that 83% of the non-clinical group score below 36,
whereas in the clinical group 21.7% cases are being classified as mildly depressed. Further, 93.7% cases of the
non-clinical group scored below 49, whereas in the clinical group this percentage is 46.71. This provides us
another set of clinical cut-off score ranging from 37-49, interpretable as "moderate depression", whereas, a
score of 50 and above signifies presence of "severe depression". The classification of scores can further be

8
facilitated with the help of given percentile ranks for the clinical and non-clinical group.

Discriminant Validity
A discriminant index for each cut-off score (i.e., 26, 37 and 50) was obtained by dichotomizing the
frequencies of false positive and false negatives around each cut off scores. This resulted in three 2 x 2
contingency tables (Tables 6, 7 and 8). Table 6 shows that below the cut-off score of `26', the frequency of non-
cases in the non-clinical sample is 113, whereas, 93 individuals of non-clinical group are being classified as
false positive at this cut off point. In comparison, only 3 cases of clinical group are being missed out as non-
cases and 57 are classified as depressives at the same cut off score. The phi-coefficient demonstrates a
significantly high discriminant validity for the cut off score of 26, phi = 0.42; chi square = 46.9, df = 1, p< .001.
Table 7 shows that below the cut-off score of `37', the frequency of non-cases in the non-clinical
sample is 173, whereas, 33 individuals of non-clinical group are being classified as false positive at this cut off
point. In comparison, only 14 cases of clinical group are being missed out as non-cases and 46 are classified as
depressives at the same cut off score. The phi-coefficient demonstrates a significantly high discriminant validity
for the cut off score of 36, phi = 0.55; chi square = 81.8, df = 1, p< .001.
Table 8 shows that below the cut-off score of `50', the frequency of non-cases in the non-clinical
sample is 193, whereas, 13 individuals of non-clinical group are being classified as false positive at this cut off
point. In comparison, 30 cases of clinical group are being missed out as non-cases and 30 are classified as
depressives at the same cut off score. The phi-coefficient demonstrates a significantly high discriminant validity
for the cut off score of 50, phi = 0.48; chi square = 60.45, df = 1, p< .001.

DISCUSSION
The indigenously developed scale - henceforth called Siddiqui-Shah Depression Scale (SSDS) has
been subjected to different measures of validity. The scale has demonstrated significantly high split-half
reliability and internal consistency. It has also shown a significant relationship with Zung's depression scale,
which indicates its relevance to measure depression. The significant correlation of the scale with psychiatric
ratings indicate that the scale is sensitive to the clinical manifestations of depression. The cut off scores with
separate frequency distribution tables for clinical and non-clinical group allow one to evaluate individual's
score in a differential manner. Depression here is taken as a continuous construct, normally distributed in the
population, therefore, the non-clinical group can be considered an appropriate reference group for the normal
population. This assumption of continuity of depression is substantiated by the phenomenological evidence of
many who feel depressed on different occasions in their lives. However, care must be exercised in using the cut
off points as classifying indicators. They are not to be interpreted as a precise estimate of true cut-off score.
Such a definitive diagnostic status requires further validation studies to establish the stability of cut-off scores,
which depends heavily on the size of the sample. Therefore, a larger sample would provide more certainty to
9

9
the cut off scores as diagnostic indicators. Moreover, the scale has been designed to assess the severity of
respondent's self reported depression and not to provide definitive diagnostic judgement. Therefore, the
corroborative data from other key sources would still be needed for a diagnostic or treatment decision. This is
one kind of interpretation, which may be correct. However, the findings in our case suggest (n.s.correlation)
that in the case of normal persons the subjective mood (which is surely assumed to be pleasant) has no
relationship with any measure of depression. In fact one should have expected a negative correlation but as it is
not the case, then our above interpretation may be true.
The development of SSDS attempts to quantify and assess depression in Pakistan. As the scale depends
on the cultural interpretation of depression, it will be useful both for theoretical investigation on depression as
well as an assessment tool in clinical set-up. A perusal of the items of the scale gives the idea, that most of the
items pertain to the hopelessness aspect of depression (Items 1, 3, 5, 9, 10, 16, and 25). This is consistent with
the prevalent conceptualization of the depressive disorder that regards hopelessness as one of the chief
component of depression. Interpersonal conflicts with friends, parents and other family members also constitute
a significant component of depression (Items 6, 17, 24, 30, & 34) indicating that in Pakistan significant others
influence the life of an individual in a psychologically significant manner. The component of guilt manifests in
terms of being punished for some deed (Item 10) and prayers not being answered (Item 5). This is different
from the expression of guilt as measured by depression scale developed in the West. The guilt here is more in
connection with perceived transgression of religious laws than social mores. The cognition that the prayers are
not answered also refers to the same where the person thinks that God has stopped listening to him. Thus, in
contrast to other scales of depression SSDS explicitly relates the feelings of guilt to perceived or actual
transgression of divine laws emphasizing the religious orientation of our people. The punishment has a divine
connotation. Feelings of personal worthlessness and incompetence, similar to Western features of depression,
are also reported (Item 9, 15, 16, 22, 23 and 27). The fact that no item pertaining to sex was obtained
substantiates the assumption that due to cultural restrictions in this regard such complaints would not come up
as an index of depression. Complaints regarding bodily functions constitute a significant portion of the scale
(Item 4, 11, 21 and 32). It is generally accepted that depression is expressed in similar terms more in the non-
Western cultures than in the Western one (Marsella & White, 1984). Death wish has also been reported (Item 13
and 36). This is in contrast with the suicidal wish generally measured by the scales developed in the West. This
again may be a reflection of religious orientation of Pakistani people as their religion, Islam, forbids suicide. It
is generally believed that in the hereafter those who commit suicide shall be deprived of Allah's blessings.
Moreover, the one who commits suicide is believed to be going through a perpetual torture till the day of
judgement. Therefore, the depressives here verbalize a passive wish for death desiring relief from their
miseries. Such a content that is truly indigenous would not have emerged had any adaptation been opted for the
development of scale.
SSDS will be useful both in clinical assessment and research. Being a measure of depression based on
10

10
culture relevant expression of depression may help assess the very features missed out, thus facilitating the
clinicians in more accurate diagnosis. Nevertheless, SSDS is in the primary stages of its development. The
authors understand that many more validation studies are in order to attain a status of definitive diagnostic
instrument. The focus was more to keep the items relevant to the culture where it is being developed. Therefore,
for equivalence of its classification with Western standard diagnostic categories calls for new series of research
which may entice some of the future researchers.

11

11
REFERENCES

Aldwin, C., & Greenberger, E. (1987). Cultural differences in Predictors of Depression. American Journal of
Community Psychology, 15, 789-813.

Baron, M.R., & Matsuyama, Y. (1987). Symptoms of Depression and Psychological Distress in United States
and Japanese College Students. The Journal of Social Psychology, 128, 803-816.

Depue, R.A., & Monroe, S.M. (1978). Learned helplessness in the perspective of the depressive disorders:
Conceptual and definitional issues. Journal of Abnormal Psychology, 87, 3-20.

Marsella, A.J. (1980). Depressive experience and disorder across cultures. In H.C. Triandis & J.G. Dragum
(Eds.), Handbook of Cross-cultural Psychology. Vol.6. (pp.237-280). Boston: Allyn & Bacon.

Marsella, A.J., & White, G.M., (1984). (Eds.). Cultural conceptions of Mental Health and Therapy. Boston:
Academic Publishers.

Nijdam, S.J. (1986). Depression: A diagnosis sometimes missed and sometimes mistaken. Psychopathology, 19,
225-230.

Russell, J.A. (1991). Culture and the Categorization of Emotions. Psychological Bulletin, 110, 3, 426-450.

Schuyler, D. (1974). The Depressive Spectrum. New York: Aronson.

Sethi, B.B. (1986). Epidemiology of depression in India. Psychopathology, 19, 26-36.

Shah, A.A. (1993). Depression: Theoretical and methodological considerations in cross-cultural perspective
(pp.1-14), Vol.IV. Proceedings of the Eighth International conference of Pakistan Psychological
Association, Islamabad: National Institute of Psychology.

Spitzer, R.L., Sheehy, M., & Endicott, J. (1977) "DSM-III: Guiding Principles". In V.M. Rakoff, H.C. Stancer,
H.B. Kedward (Eds.). Psychiatric Diagnosis. New York: Brunner/Mazel.

Venkoba Rao, A. (1978). Some aspects of psychiatry in India: Over-view: Transcultural Psychiatric Research
Review, 14, 7.

Westhuis, D., & Thyer, B.A. (1989). Development and validation of the clinical anxiety scale: A rapid
assessment instrument for empirical practice. Educational and Psychological Measurement, 49, 153-
163.

Zung, W.W.R. (1965). A self rating depression scale. Archives of General Psychiatry. 12, 63-70.

12

12
 English Translation of Siddiqui-Shah Depression Scale (SSDS)

This questionnaire describes thoughts and feelings of an individual. There are four columns against each
statement, reflecting the various degrees of these thoughts and feelings. You are requested to read each statement
carefully and indicate how much they apply to you with the help of given columns. For instance, if a statement
`never' applies to you indicate it by putting a ( / ) correct sign in the first column. If the statement stands true for you
`all the times', put a mark in the last column. Like wise use other columns as well.

Never Sometimes
Often All the times

1. I have become very hopeless.

2. I feel myself confused.

3. I am very unfortunate.

4. I have almost lost my appetite.

5. My prayers do not get answered.

6. I have differences with my parents.

7. People always criticize me.

8. I feel like crying aloud.

9. I can not do anything properly.

10. I am being punished for my deeds.

11. My heart starts pounding suddenly.

12. Success and failure depends upon

Kismat (Luck).

13. My life is reaching its end.

14. I am haunted by the feeling of having

lost something.

15. I lack something.

16. I feel myself as worthless.

17. Other always dominate me.

18. I get anxious easily.

13

13
19. I hate my life.

20. I feel lonely.

21. My body seems tired.

22. I am a useless person.

23. I have many flaws.

24. Friends do not understand my feelings.

25. I loose heart very quickly.

26. I am incapable.

27. I am inferior to others.

28. I have lost zeal of life.

29. I am a detestable person.

30. My friends seem selfish to me.

31. The memories from the past make me sad.

32. My sleep is disturbed.

33. I am very hopeless about my future.

34. I donot come up to my parents' image of

an ideal child.

35. Most of the people are not trustworthy.

36. I intensely wish for death.

14

14
Table 1 Item-total score correlation for indigenous depression scale

--------------------------------------------------
Items Non-clinical Clinical
(N = 206) (N = 60)
--------------------------------------------------
1 .5427** .6580**
2 .5746** .5570**
3 .4463** .6308**
4 .2236** .3122*
5 .3006** .5235**
6 .3907** .3232*
7 .4651** .3784*
8 .4546** .4318**
9 .3996** .3930**
10 .4333** .3354*
11 .4480** .4295**
12 .3695** .4369**
13 .5307** .4959**
14 .5417** .5821**
15 .5847** .5475**
16 .6170** .6530**
17 .4860** .4182**
18 .4864** .4409**
19 .6076** .5919**
20 .5686** .4341**
21 .4626** .3461*
22 .4093** .7167**
23 .3711** .5763**
24 .5420** .4742**
25 .5702** .5854**
26 .4018** .6468**
27 .6024** .6491**
28 .5856** .6217**
29 .3484** .6744**
30 .4023** .6135**
31 .4714** .4802**
32 .3678** .4485**
33 .5768** .5478**
34 .5468** .4414**
35 .2776** .4802**
36 .5310** .4702**
--------------------------------------------------

15

15
Table 2 Correlation coefficients for split-half reliability and Spearman-Brown correction of the scale
for clinical and non-clinical groups

-------------------------------------------------
Groups Split-half Spearman Brown p
Coefficients Correction
-------------------------------------------------

Clinical 0.79 0.84 0.001

Non-clinical 0.80 0.89 0.001

-------------------------------------------------

Table 3 Internal consistency of the scale for clinical and non-clinical groups

-------------------------------------------------
Groups Alpha-co-efficient p
-------------------------------------------------

Clinical 0.91 0.001

Non-clinical 0.89 0.001

-------------------------------------------------

Table 4 Construct and concurrent validity of indigenous depression scale

-------------------------------------------------
r p
-------------------------------------------------
Correlation between the scores on items
of indigenous depression scale and
Zung depression scale
(non-clinical group N = 206) 0.55 .001

Correlation between current mood and

items of indigenous depression scale
(non-clinical group N = 206) 0.14 n.s.

Correlation between current mood and

items of indigenous depression scale
(clinical group N = 60) 0.64 .001

Correlation between items of indigenous

depression scale and psychiatric rating
for depth of depression
(clinical group N = 30) 0.40 .05
-------------------------------------------------

16

16
Table 5 Eigenvalues and percentages of variance explained by the extracted factors
-------------------------------------------------
Factor Eigenvalue Pct of Var Cum Pct
-------------------------------------------------
1 8.39801 23.3 23.3
2 1.94145 5.4 28.7
3 1.76956 4.9 33.6
4 1.60113 4.4 38.1
5 1.31419 3.7 41.7
6 1.22751 3.4 45.1
7 1.21081 3.4 48.5
8 1.16808 3.2 51.8
9 1.11418 3.1 54.8
10 1.05994 2.9 57.8
11 1.01177 2.8 60.6
-------------------------------------------------

Table 6 Discriminant validity of the scale for the cutting score of, below and above, 26 for depressed and
non-depressed groups.
-------------------------------------------------
Depression Score Non-Depressed Depressed
-------------------------------------------------
26 and below 113 3

27 and above 93 57
-------------------------------------------------
Phi = 0.42, X2 = 46.9, df = 1, p < .0001

Table 7 Discriminant validity of the scale for the cutting score of, below and above, 37 for depressed and
non-depressed groups.
-------------------------------------------------
Depression Score Non-Depressed Depressed
-------------------------------------------------
37 and below 173 14

38 and above 33 46
-------------------------------------------------
Phi = 0.55, X2 = 81.8, df = 1, p < .0001

Table 8 Discriminant validity of the scale for the cutting score of, below and above, 50 for depressed and
non-depressed groups.
-------------------------------------------------
Depression Score Non-Depressed Depressed
-------------------------------------------------
50 and below 193 30

51 and above 13 30

17

17
-------------------------------------------------
Phi = 0.48, X2 = 60.45, df = 1, p < .0001

18

18