The Indian Journal of Social Work, Vol. XXI, No. 3, (December, 1960). T...
The Indian Journal of Social Work, Vol. XXI, No. 3, (December, 1960).
T H E A T T I T U D E S O F T H E R E F U G E E S : A P I L O T S T U D Y
Introduction.—With the partition of India on August 15, 1947, more than half of the
area of Bengal was truncated off the original state to form a new state, East Pakistan, under
a separate government. Since then an unending trail of grief-stricken, panicky human beings
have been continuing to cross the border of East Pakistan to reach West Bengal. They are the
refugees. In the beginning their numbers were not much; the rehabilitation did not seem to be
a serious problem. In some quarters it was felt that this was a temporary migration and that
the refugees would go back to their home land after a few days. This did not prove true.
All these years they are pouring in, sometimes like an avalanche. Compelled by circum-
stances they left their homes and faced an uncertain future and created a major problem for
the new government of West Bengal. With meagre resources the state government fought hard
to cope with the situation. Later, the Central Government rendered all necessary help but
the problem was not wholly solved. Some schemes were implemented more or less successfully.
Dandakaranya scheme was chalked out to rehabilitate a vast number of refugees.
Inspite of the governments' help in these
Method.—Of all methods of measurement
years there came a constant opposition from of attitudes, the most prominent, the most
the majority of the refugees against almost widely used and the most carefully designed
all plans and programmes of the government. and tested is the attitude scale. In essence,
A state of tension prevailed in the state the method of scaling requires that the indi-
province and created almost a baffling vidual reacts verbally with expressions of
problem for all. An analysis of the problem approval or disapproval, agreement or dis-
revealed that the economic insecurity played agreement to a set of carefully standardized
an important role in their plight, but that items on propositions.
was not the only factor responsible for the
Various types of scales were devised and
tension. O n e is led to the conviction that mere
used by different psychologists to measure the
economic solutions of the problems of the attitudes towards a number of things. Among
refugees may not always lead to a permanent these most commonly used scales are
settlement of the refugees. Along with eco-
Thurstone's and Likert's. T h e relative merits
nomic problems one must consider the mental of Thurstone's and Likert's methods of scale
aspect of the refugees. With severe frustra-
construction have been discussed in detail by
tions and hardships of existence their attitudes
m a n y workers in this field.1'2 While the
have been changed to a great extent. To advantage of using one scale over the other
understand their attitudes is to understand is still a disputed fact, the researches on both
their minds which would definitely be an the types of scales continue to flourish to a
important contribution to any planning for great extent. Here, Likert's methods of scale
their betterment.
construction has been followed.3
Considering this point in question, the pre-
To get a first h a n d knowledge about the
sent work was planned to have an idea and conditions of the refugees from East Pakistan
measure the attitudes of the refugees. T h e and to make acquaintance with their attitudes
results of the study, it may be expected, would
a general opinion survey was conducted.
help in the task of smooth rehabilitation.
Accordingly, refugees were interviewed
1 D . Krec h and R. S. Crutchfield, Theory and Problems of Social Psychology, 1948, p p . 218-19.
2 L . W. Ferguson, Personality Measurement, 1952, p. 123.
3 R . Likert, "A Techniqu e for the Measurement of Attitudes, " Arch. Psychol., 1932, No . 140.

2 8 8
separately and were asked to express their ed more or less true indicator of the
views on different aspects of their difficulties success of the final programme, the question
in rehabilitation. This interview continued of sampling was of m u c h importance.
uninterrupted for one to two hours in each
Here according to the suitability, the
case and the whole proceedings were noted. method of stratified sampling was used.4 T h e
This had been used as one of the resources in population was divided into two homogeneous
constructing the items of different scales. groups or strata based on socio-economic
Other than this, the daily newspaper reports status, (the refugees resided in camps
of different activities and statements of the sponsored by the Government, who were to
refugees helped immensely to that purpose. depend on the monthly doles and the refugees
On critical examination of these, it was who were comparatively better off could build
revealed t h a t in finding out a way to smooth their houses with partial financial help of the
the rehabilitation work it was essential to Government and got some occupation, etc.)
know the attitudes of the refugees towards In short, inmates of camps formed one group
five different aspects, namely, Islamic religion,
and the residents of colonies composed the
India government, Pakistan government, in-
other. T h e r e was difference between these
habitants of West Bengal and themselves.
groups in standard of education also.
T h e questions were framed according to the
Next, a simple r a n d o m sampling from each
rules provided for it and especially the direc-
of these segments or strata was done. From
tive laid by Likert was kept in view. T h e the list of serial names obtained, every tenth
experienced colleagues of the author were name was selected as a subject in one camp,
where they numbered about 350. In the case
invited for suggestion in framing the questions of another colony a similar procedure was
and relevant modifications were made. As the
followed. For this pilot study out of 65 data
task of putting right and accurate questions 60 collected for the final treatment.
in the scale was important matter in this case,
the questions were repeatedly read by many
psychologists and were asked to forward their
comments. Finally, seven items for each
scale of Islamic religion and India Gov-
ernment, five items for Pakistan Govern-
ment, eight items for the inhabitants
of West Bengal and eleven items for
the scale meant for measuring attitudes
On the basis of median values of scores of
towards themselves, were selected to measure different scales, the percentage of favourable-
the attitudes towards all these five aspects.
ness and unfavourableness has been calculated.
An information schedule was prepared with
necessary queries in it. It was revised with
m u c h care.
As the information of this kind of researches
was not available five open-end questions, one
for each scale, were included in order to use
the results to find out the validity of the scales.
Sampling.—As this pilot study was consider-
4D. W, Pade n and E, F, Lindquist, Statistics for Economics and Business, 1956, pp . 136-38.

2 8 9
Interpretation of the Result.—After tabu-
just l i t e r a t e ( I ) , under m a t r i c u l a t e ( I I ) , and
lating the scores the median value of each matriculate and above ( I I I ) . In the first
scale had been calculated. In Table 2, the scale, the difference between the means of
percentage of favourableness and unfavour-
categories I and I I , and I and I I I was signi-
ableness of the attitudes of the refugees had ficant, while the difference between the means
been found out. In the first scale, the attitude of categories II and I I I was not significant.
towards Islamic religion was unfavourable, the
In the second scale, in no case the difference
percentage being 61.4. It was evident from between means was significant, hence the
the table t h a t the percentage of unfavourable-
refugees of various educational levels were
ness in the 2nd, 3rd and 4th scales was greater
unanimous in their attitudes towards India
than the percentage of unfavourableness. T h e
Government. In the 3rd scale the only signi-
attitudes of the majority of the refugees ficant difference occurred between the cate-
towards Islamic religion, India Government, gories I and I I . T h e results in the 4th scale
Pakistan Government and the inhabitants of indicated no significant difference present
West Bengal were not favourable. In the case
between the means of categories towards the
of the 5th scale, attitude towards themselves inhabitants of West Bengal were alike. In
was favourable. In all the scales the the case of the 5th scale, which may be called
difference between the percentage of favour-
more or less self-rating type, the difference
ableness and unfavourableness varied widely. between the means of categories I and I I , I
and I I I was highly significant. It was evident
Whether different educational levels of the from the table that the educated refugees were
refugees made any difference in the attitude more self-confident than the uneducated
scores that had been calculated and the results
were placed in Table 3. T h e total number
of refugees was classified under the categories,
Internal-consistency reliability.—The inter-

2 9 0
nal-consistency reliability5 of each scale had between each item score and the total score.
been found out by calculating correlations The result is given below.
The suitability of each item to be included end direct questions were kept, one for each
in the final scale was dependent on the scale, in order to get their frank attitudes
amount of correlations present between that towards Islamic religion, India Government,
item and the total score. Accordingly, the item
Pakistan Government, inhabitants of West
which did not hold good correlations with Bengal and towards themselves. The answers
the total score must be omitted from the scale. of these questions were classified into two
Higher the correlation higher will be the groups, favourable and unfavourable. Then,
chance of retaining the item in the final scale. the biserial correlations were calculated bet-
Here in a very few cases the correlations were ween their response in each question and the
much higher. But considering the paucity of total score of each scale. The validity of the
data, it would be better to omit only those scales could be determined after considering
items which correlate negligibly with the total the amount of correlations are given below.
SCALES correlations (biserial) between
the responses to the open-end questions
Validity of the Scales.—It is very difficult
and the total scores in the scale.
to determine the validity of any attitude scale
perfectly owing to lack of suitable criteria and
especially, the difficulty assumes greater when
no parallel test of any contemporary research
is available in the field. In this case, there
are no large scale researches in India concern-
ing the refugees available, the materials of
which can be kept as a criterion in deter-
In no case the amount of correlation can
mining the validity of the scales.
indicate the good validity of the scale. Here
the factors responsible for the drawback in
So there was no alternative but to follow getting better amount of correlation might be
a method which was inherently weaker in lack of a truly representative sample and
comparison with other methods. The present poor criterion. To achieve much better
method required to put in the same question-
result the nature of criterion should be
naire, certain open-end direct questions to be improved, which is, in the present case, not
answered by the subjects. Here, five open-
5W. J. Goode and P. K. Halt . Methods in Social Research, 1952, pp . 271-76.

2 9 1
Apart from this, during construction of the important point to be considered. However,
scales, the items were gone through by some in this case, there is no alternative but to
experts. The validity of each item was expect better result when the final question-
increased to a considerable extent.6
naire would be applied to a large sample.
The percentage of unfavourable attitudes
is not greater, as it was expected. In rehabili-
The main purpose of this pilot study is to tating the refugees, one must ponder about
find out the effectivity of the items and the the question of favourability of attitudes.
scales in measuring the attitudes of the Otherwise, whatever opportunities the
refugees. Accordingly, the aim is to pick up refugees may get, they will not be at ease with
only the most valid and reliable items to be themselves permanently in any situation.
put in the final questionnaire and to discard Of course, there is a good point in the result
the unreliable ones.
that the refugees do not lose their confidence
in themselves. The result also shows that
The results reveal more or less satisfactory (Table 3), in this case, education has got some
indication of the fulfilment of our purpose. influence on the formation of attitudes.
As far as the reliability of the items is con-
Though the difference in attitudes between
cerned, there are only a few items which do the educated and uneducated persons is not
not correlate much with the total score; so always significant in this small sample, it is
these items would be discarded. The improve-
expected that the difference will increase when
ment of the validity of the scales is another the sample will be more truly representative.
6 H . Gullisken, "Intrinsi c Validity", American Psychologist, 1950. 5. (Fro m Lindzey, G.,
Handbook of Social Psychology, Vol. 1, 1954, p. 340.)