Admission Criteria and Internal Assessment in a School of Social Work: An...
Admission Criteria and Internal Assessment
in a School of Social Work: An Analysis
P.K. VISVESVARAN
Examining data from a school of social work, the author has established that
candidates' past performance (Bachelor's Degree) determines the outcome of the
entrance test in a significant manner. Moreover, variations in grades received by
the trainees seem to stem from differences in the teachers' approach to evaluation.
rather than from true differences in performance. For greater transparency and
fairness, a regular and candid appraisal of the evaluation process by teachers
themselves is essential.
Mr. P.K. Visvesvaran is former Principal of the Madras School, of Social Work,
Chennai.
I N T R O D U C T I O N
It is customary to hold written tests and interviews along with group
discussions to assess the suitability of candidates seeking admission to
social work courses. The contents and components of these tests vary
among the different schools. So do the weightages given to these
components. The schools, however, are not always free to adopt
whatever entrance tests and procedures they prefer. The choice of
admission procedures is influenced by a variety of factors, for exam-
ple, the rules and regulations of the university to which the school is
affiliated and which sets standards for the conduct of the courses.
Additionally, the state government, which often provides funds for
the running of the courses, may also issue some directives in this
regard. This has taken place at least in one instance. Recently, the
Government of Tamil Nadu, through its Department of Collegiate
Education, has laid down guidelines for admission to MA in Social
Work which, inter alia, stipulate that the criteria would be the percent-
age of marks earned by the candidates in the Bachelor's Degree course,
combined with marks earned in the entrance test (40 per cent) and the
interview (10 per cent). This was in the year 1.999. The previous year,

256 P.K. Visvesvaran
the Department had forbidden the conduct of interview for admission
purpose (vide Tamil Nadu Higher Education Department G.O. Ms.
No. 168 dated 24.4.1998).
These directives give social work educators food for thought. Is
there a real need to conduct written tests and interviews at all? Do these
entrance tests merely serve as convenient devices to filter out a large
number of candidates and admit just a few, since the number of
authorised seats is small? To phrase it differently, will not the degree
marks alone serve as a criterion for admission, should there be a high
degree of correlation or association between these grades and perform-
ances in the other components of the entrance tests? Can not written
tests, group discussions and personal interviews be eliminated? Will
not we be, thereby, saving a lot of time and efforts, our own as well as
the candidates'?
THE NEED FOR APPRAISAL
Schools of social work would perhaps do well to study, independently
and jointly, the outcome of their entrance tests and inter-relationships
among their components. We can be sure that the outcome of such an
appraisal will render our systems of admissions more scientific than,
they have so far been.
It is equally necessary to make our procedures, for assessing stu-
dents' performance at various stages after admission, scientific. For
example, Flippo (1984:239) refers to 'common error's of traditional
rating' and says that some of these errors are the halo error (one aspect
of the person's performance influences the entire evaluation); the
central tendency (marking all personnel as average); 'tough' rating and
'easy' rating (rarely awarding high grades or awarding them to the
entire class).
Do 'tough' and 'easy' ratings exist in our departments? One should
indeed take a closer look at our own systems of appraisal so that,
ultimately, these could be standardised and improved.
REVIEW OF RELEVANT LITERATURE
It is interesting that even in the 1920s, the difficulties associated with
enrolment (to the social work courses) were thought of (see, for
example, Walker, 1928:147). Nanavatty (1968) has said that the
process of enrolment at the schools of social work makes an important
- contribution to social work and has emphasised the need to select those
with the right kind of personality.

Admission Criteria and Internal Assessment... 257
P. Ramachandran, and A. Padmanabha, in their pioneering study of
professional social workers in India (1969), have mentioned that the
professionals do not always come up to the expectations of the em-
ployers. To set right this situation, some changes in the enrolment
procedures are indicated, besides modifying the training appropriately.
H. Nagpaul (1965) pointed out the need for an inter-disciplinary
approach to training in the Indian context. If this idea were acceptable,
then it should be borne in mind even at the entry or enrolment point.
R. Prasad (1987) bemoaned the paucity or the virtual absence of
research activity in professional schools. Obviously, no study regard-
ing enrolment and evaluation had come to his notice.
Armaity S. Desai, in her monumental study (1994: 235,601), found
that most social work students, in her sample, had a background in the
arts faculty. This is one of the few available studies that have touched
upon the background of social work recruits. She also found that the
present performance of the social work students was significantly
associated with their past (Bachelor's Degree) performance.
R.R. Singh (1985: 102) emphatically highlighted, more than ever
before, the need for a careful evaluation of students' performance in
field work and goes to the extent of suggesting that a five-member
panel should conduct the evaluation in each case to avert 'the possi-
bility of bias affecting evaluation'.
The situation in the United States of America (USA) has not been
very different. Morris, Dana, Glasser, Marks, Rein, Schreiber and
Saunders (1974: 263) have bewailed the fact that the educational
research studies (in social work) have been rarely undertaken in that
country. Evaluation research, they say, is virtually non-existent.
Ramachandran, as far back as in 1961, suggested that a separate
council of social work research ought to be created and evaluation
research should be encouraged.
Misra (1994: 324, 325) referred to Philip Klein's five-fold clas-
sification and Friedlander's fifteen-fold classification of social
work research, but the training aspects and enrolment do not find a
place therein. On the other hand, Palaniswamy (1994) refers to ten
different emerging aspects of social work such as curriculum teach-
ing and research and suggests that the Association of Schools of
Social Work in India conduct a survey about all aspects of social
work education.
Two distinct, but related, trends are derived from the review of
literature presented above. These are:

258 P.K. Visvesvaran
1. One set of authors have given specific suggestions regarding
admission, recruitment and the training process, including the
assessment of candidates' performance during training. How-
ever, they have not cited any research studies to support their
contention (except in one case).
2. A second set of authors have laid a strong, but general, empha-
sis on the need for more research on the profession including
professional training. But they have not indicated the specific
sub-areas that should be investigated.
To sum up, the opinion of most experts quoted above is that research
on recruitment to and training in social work is a neglected area and
virtually no research based data exist that could be cited as a precursor
of the present attempt. Nor has my query, sent to six different schools
of social work in and outside the city of Chennai, yielded any infor-
mation on similar studies carried out elsewhere. Prof. Armaity S.
Desai's study, already referred to is, of course, an exception.
CONCEPTUAL FRAMEWORK
The present study may be deemed to be a partial response to the call
implicit in both sets of works cited above. The conceptual framework
for the present study is as follows.
Two specific areas have been chosen for systematic investigation,
namely the admission criteria and the assessment of students' perform-
ance in the course of the training process. An important link connecting
these two aspects is the fact that raw data are readily available for
statistical analysis in every school of social work in these two areas.
There is no need for any special efforts to be made by the teacher-re-
searcher for obtaining original data for a fruitful research study. The
required data are practically 'up for grabs'. There is no need to prepare
a separate instrument for data collection except to determine the
aspects of admission and internal assessment on which the researcher
wishes to focus attention.
Yet another reason for covering admission criteria and internal
assessment of students' performance after admissions in one and the
same study, is the focus on consistency. In other words, attention is
being drawn to the same issue, in two different contexts. More specifi-
cally:
• Are the candidates aspiring to join a post-graduate training
programme consistent to their performance at the degree level,
and later at the entrance test for the course they aspire for?

Admission Criteria and Internal Assessment... 259
• Do teachers consistently use uniform criteria to assess the per-
formance of students who were selected on the basis of entrance
tests and who are currently undergoing a training programme?
In the present study, the term 'consistency' is being more or less
used in the same sense in which Jahoda, Deutsch and Cook (1959:168,
172) have used it to describe reliability, stability and equivalence.
Stability is 'consistency' of measures on repeated application, they
say. Further, they have opined that 'equivalence' is established when
different investigators produce 'consistent' results.
Thus, the present study focuses attention on consistency in candi-
dates' performance as well as consistency of the evaluation criteria
employed by teachers. It is worth mentioning that the present study
does not purport to establish a correlation between performance in
entrance tests and performance during the actual training at later
stages. In other words, the present writer has not undertaken the task
of validating the entrance tests. Rather, an attempt has been made to
see if the candidates are consistent in their performance and further, to
see if the internal assessment criteria and standards employed by the
teachers are uniform. Consistency and uniformity belong to the realm
of reliability rather than validity.
STATEMENT OF OBJECTIVES AND THE ENUMERATION
OF RELEVANT VARIABLES

A major objective of the present study is to find out whether, and to
what extent, candidates' performance in the Bachelor's Degree is
related to their performance in the various components of the entrance
test conducted for admission to MA in Social Work. The marks scored
by the candidates in their Bachelor's Degree is treated as the inde-
pendent variable. The dependent variables are marks awarded in the
entrance tests (written test and personal interview).
Another major objective is to find out whether there are significant
variations in marks awarded to students of social work for their
performance in subjects assessed internally, namely field work and
research project work. The variables relevant in this case are subjects
in which students of social work specialise, namely Personnel Man-
agement on the one hand and all other social work practice-related
subjects on the other. These may be treated as independent variables,
since each of these two groups claims roughly 50 per cent of the
total student strength. The marks awarded to them in only the two

260 P.K. Visvesvaran
components of training that are internally assessed, namely field work
and research project work, will be treated as dependent variables.
METHOD OF STUDY
The units of observation will be the marks earned by applicants
attending written tests and interviews held by the selected institution
for admission to MA in Social Work. These will include the Bachelor's
Degree marks.
The relevant universe or population consists of all the candidates
for whom records are available. In the institution under consideration,
such records were available for a five-year period, namely, 1995—
1999. This being a preliminary study, it was decided to randomly select
the data for one of these five years and see the trends. Thus, it is a
probability sample of information pertaining to one particular year,
namely 1996.
Actually, it had been planned to compare the 1996 data with those
of 1997, 1998 or 1999, but the admission criteria kept changing during
these successive years and hence no comparison was attempted. For
example, in 1997, admissions completed the announcement of many
candidates' degree results. In 1998, interviews had been banned and
hence were not conducted, as mentioned earlier. In 1999, the weigh-
tage given to Bachelor's Degree marks (in regard to admission to MA
in social work) had been raised from 40 per cent (in 1998) to 100 per
cent (out of a total 160).
In regard to objective, the universe or population consists of all the
records of teachers' internal assessment of student's performance in
field work and research in the selected institution. The records pertain-
ing to two years were randomly selected, namely 1997 and 1999 and
examined for purposes of the study.
This study is confined to one institution (a school of social work)
with which the researcher in familiar. The name of the institution is
not revealed for reasons of confidentiality.
THE ANALYSIS
The chi-square test, the 't' test and the median test were applied to see
the association between the relevant variables, to test the difference
between means and to determine the significance of these differences
and associations.
The results of the analysis are presented in the form of cross-tables
as follows:

Admission Criteria and Internal Assessment... 261
TABLE 1: Admission to MA in Social Work, 1996: Relationship between Basic
Degree and Entrance Test (Written) Performance

Chi-square: 32.3
Significant at 0.001 level
Degree of freedom: 2
Note: The Entrance Test marks are classified as follows.:
Low : 14 or less out of a maximum of 40 marks.
Middle: 15 to 24.
High : 25 and above.
It is clear from Table 1 that candidates with 56 per cent marks and
above in the qualifying examination (namely, the Bachelor's Degree)
have performed better in the entrance test. This association is statisti-
cally very significant. This may be due to several reasons such as
higher Intelligence Quotient, greater self-confidence, superior general
knowledge and a higher level of motivation on the part of the high-
achievers.
TABLE 2: Relationship between Basic Degree and Admission Interview
Performance
Chi-square: 16.78
Significant at 0.001 level.
Degree of Freedom: 2
Note: The Interview Marks are classified as follows:
Low : Upto 9 out of maximum of 20.
Middle: 10 to 14.
High : 15 and above.
It is seen that high-achievers in the basic (Bachelor's) degree also
perform better at the admission interview for the MA in Social Work
course (Table 2). Adequate speaking skills and self-confidence are
required in a greater measure for a successful interview performance,
than perhaps for the written test. High achievement at the degree level

262 P.K. Visvesvaran
may indeed prepare the candidate for a successful interview. This fact
may account for the results stated above.
TABLE 3: Comparison of the Means of Marks Awarded by Teachers
The marks awarded by two teachers in the same Department have
been compared in the Table 3. The said teachers were selected because
they had awarded, respectively, the highest and the lowest marks in
field work in the year under reference (1997).
It is seen that the difference is statistically significant and is not
accidental. The difference is of such an order that all the students under
one teacher might very well have passed with distinction while those
under the second could have secured only mediocre grades. Such
results should be deemed unfortunate and unfair, if the grades awarded
by the respective teachers do not truthfully and accurately reflect the
actual abilities of the students and represent only the individual idi-
osyncrasies of the concerned teachers.
Even if the above sample is deemed non-representative, since it
pertains to a single year (though randomly selected), the median test
could still be applied (see Levin, 1983: 186) and the significance of
the variation estimated as follows:
TABLE 4: Extreme Variation in Marks: The Median Test
Chi-square (with Yate's correction): 6.1 Significant at 0.02 level.
Degree of Freedom = 1
Note: 'Stringent' — Most grades awarded are Below Median.
'Liberal' — Most grades are Above Median.
Yate's correction: Yate's correction has been applied only in 2 x 2 tables, wherever the need
arose. This is in consonance with views expressed among others, by Garret (1981) and
Porkess (1988: 233).

Admission Criteria and Internal Assessment... 263
It is once again proven that marked variation in grades, evident in
the above data, is not a product of chance, but of other factors, possibly
variation in standards of evaluation adopted by the respective teachers.
TABLE 5: Specialisation-Wise Variation in Marks Awarded for Field Work
(1999)
Chi-square: 5.16. Significant at 0.05 level.
Degree of Freedom: 2.
It is seen that a relatively smaller proportion of students of Personnel
Management have received higher grades (Table 5). The difference is
statistically significant. In other words, some factor, other than mere
chance, is probably responsible for the variation.
TABLE 6: Specialisation-Wise Variation in Marks Awarded for Field Work
(1999)
Chi-square (with Yate's correction) : 5.38. Significant at 0.05 level.
Degree of Freedom: 1
Note: The categorisation of marks in this Table differs from that in the previous table
because of variation in maximum marks in the two cases.
Once again it is found that the marks differ according to specialisa-
tion to a significant degree (Table 6). The implications will be dis-
cussed at a later stage.
The variation in marks awarded by the eight teachers is statistically
significant (Table 7). In other words, some teachers do seem to have
a tendency to give relatively low grades, while many other teachers
have the opposite tendency. The trend evident here is not a product of
chance, but of other systematic factors. >

264 P.K. Visvesvaran
TABLE 7: Field Work Marks Awarded by AH Teachers Compared with One
Another (1997)

Marks Awarded by Teachers
No. of Teachers
Out of 150
70 to 79
1
80 to 89
1
90 plus
6
Total
8
Chi-square : 6.17 Significant at : 0.05 level.
Degree of Freedom : 2.
TABLE 8: Variation Among All Teachers in Marks Awarded for Project Work
(Research), 1997
Marks Awarded by
No. of Teachers
Teachers (Out of 25)
1 3 - 14
2
1 5 - 16
4
17 Plus
2
Total
8
Chi-square : 0.98 Not significant at 0.05 level.
Degree of Freedom : 2.
In the above case (Table 8), any variation that is seen in marks
awarded for project work (research) is not substantial and is not
statistically significant.
TABLE 9: Variation Among All Teachers in Marks Awarded for Project Work
(Research), 1999
Chi-square : 6.5 Significant at 0.05 level.
Degree of Freedom : 2.
In contrast to the past situation (two years earlier), where no
variation was seen, in the current year, the Personnel Management
students have received higher grades (Table 9). This is also in contrast
to the trends seen in field work.

Admission Criteria and Internal Assessment... 265
DISCUSSION, SUGGESTIONS AND LIMITATIONS
Past Performance
Schools of social work in India have routinely been conducting inter-
views, written tests and group discussions as requirements for admis-
sion to social work training programmes. It is seen from the data made
available in one institution that marks scored in basic degree seems to
play a crucial role in securing admission to social work courses (Tables
1 and 2).
This should not be interpreted to mean that the schools should rely
entirely and only on the Bachelor's Degree grades. However, it may
be worthwhile stipulating a passing or admissible minimum in the
various other components, namely the written test, the personal inter-
view and group discussions.
Nevertheless, it ought to be mentioned that should, for some
reason, the government once again ban the holding of interviews,
the schools placing a sole reliance on performance at the Bachelor's
Degree level will not be committing a grave error or injustice to the
candidates. More research in this area will help the schools to evolve
standardised admission procedures and entrance tests for admission to
social work courses.
Extreme and Unaccounted Variations
Extreme variations in marks by two teachers in this sample was not a
product of chance. The variation was real and significant and probably
arises from variations in teachers' attitudes and approach to evaluation,
one of them being too liberal and the other too stringent.
Flippo (1984) offered a possible remedy. Harshness and leniency
are both constant and systematic errors and can be corrected by
manipulating the ratings by 'translating the raw score into percentiles',
for example. However, the best solution, according to him, is educa-
tion. Variations in scores should be discussed by the faculty members
and the possible reasons for the same should be gone into. An honest
self-appraisal on the part of the teachers is necessary to see how far the
grades awarded by them reflect true differences among the students
and not differences among the teachers themselves, namely their
individual idiosyncrasies.
There are noticeable variations in grades awarded to students spe-
cialising in different subjects. All these variations, once again, cannot
be attributed to true differences among the trainees themselves. Unless

266 P.K. Visvesvaran
this trend is checked, it might very well result in one or more groups
continually being denied reasonably high grades which they might
really deserve.
The teachers earlier compared (Tables 3 and 4) were both handling
subjects pertaining to the same speciality which goes on to show that
there exists not only inter-group variations in grades (that is, between
specialisation groups), but intra-group (that is, within the same group)
differences as well.
However, it seems highly unlikely that these variations reflect the
true differences among students' abilities and performance, unless one
is willing to argue that all the high-achievers have deliberately been
assigned to one particular teacher and the rest to the other. Such a stand
would seem untenable, for obvious reasons. Unaccounted variations
of this kind are seen not only in regard to field work but in marks
awarded for research as well.
Suggestions
To reduce the extent of unexplained variations, teachers should discuss
the issue among themselves, both generally and department-wise (or
specialisation-wise). Additionally, it will be worthwhile displaying all
the mark lists for the benefit of the student community. Such a display
will facilitate general and specialisation-wise comparisons and the
prompt settling of grievances, if any. If particulars regarding mean
grades and standard deviations are also provided, it will lend credibility
and transparency to the entire system of evaluation.
Limitations
This preliminary and exploratory investigation has obviously suffered
from a number of limitations. For example, data pertaining to just a
few selected years have been examined. Also, these data have ema-
nated from one particular institution. Continuous observation of
trends, and inter-school symposia and discussions on the same are sure
to pay greater dividends.
This has been an exploratory study designed to see certain trends
and in regard to admission criteria and evaluation of social work
students' performance. It has shown that it is possible to use the
available statistical procedures to appraise admission criteria and
award of grades. Social work teachers would do well to adopt these
procedures regularly and discuss the outcome among themselves

Admission Criteria and Internal Assessment... 267
dispassionately and objectively and institute due remedial measures
wherever necessary.
ACKNOWLEDGEMENT
I wish to gratefully acknowledge the kind, willing and able assistance I have received
from J.S. Gunavathy, Research Scholar, Madras School of Social Work in connection
with the preparation of this article.
REFERENCES
Desai, A.S.
A Study of Social Work Education in India (mimeo), New
1994
Delhi: Ministry of Welfare, Government of India.
Flippo, E.B.
Personnel Management, New York: McGraw-Hill Book
1984
Co.
Garret, H.E.
Statistics in Psychology and Education, Bombay: Vakils,
1981
Feffer and Simons Ltd.
Jahoda, M, Deutsch, M.
Research Methods in Social Relations, New York: Henry
and Cook, S.W..
Holt and Company.
1959
Levin, J.
Elementary Statistics in Social Research, New York: Har-
1983
per and Row.
Misra, P.D.
Social Work Philosophy and Methods, New Delhi: Inter-
1994
India Publications.
Morris, R., Dana, B.,
Encyclopedia of Social Work, Washington, DC: National
Glasser, P., Marks, R.,
Association of Social Workers.
Rein, M., Schreiber, P.
and Saunders, B.
1974
Nagpaul, H.
Social Work Education in India. In S.K. Khinduka (Ed.).
1965
Social Work in India, Allahabad: Kitab Mahal Pvt. Ltd.
Nanavatty, M.C.
The Social Work Profession. In Encyclopedia of Social
1968
Work in India, New Delhi: Planning Commission.
Palaniswamy, S.
Social Work Education in Tamil Nadu. In K.K. Jacob
1994
(Ed.), Social Work Education in India, Delhi: Himanshu
Publications.
Porkess, R.
Dictionary of Statistics, London: Collins.
1988
Prasad, R.
Courses in Research. In A.B. Bose (Ed.), Encyclopedia of
1987
Social Work in India, New Delhi: Ministry of Welfare.
Ramachandran, P.
Social Work Research and Statistics. In A.R. Wadia (Ed.),
1961
History and Philosophy of Social Work in India, Bombay:
Allied Publishers, 446-456.

268 P.K. Visvesvaran
Ramachandran, P.
Professional Social Workers in India: A Study of Their
and Padmanabha, A.
Employment Position and Functions, Bombay: United
1969
Asia.
Singh, R.R.
Field Work in Social Work Education, New Delhi: Con-
1985
cept Publishing Co.
Walker, S.H.
Social Work and Training of Social Workers, Chapel Hill:
1928
University of North Carolina Press.