94 SOCIAL SCIENTIST where the selection to a programme of study or for a career is based on a process whose validity has not been conclusively established, the practice of accepting candidates solely based on the ranks in the selection process must be abandoned. Instead, as described here, the final selection must be on the basis of a lottery among all those considered acceptable. Such a practice while opening the possibilities to more number of deserving candidates, will also have a sobering effect on those who are finally selected since they will be aware of the role of chance in their selection. However, efforts must continue to refine the screening procedures such that the cut-off score can be gradually raised to ensure that the number of candidates found acceptable (and thus becomes eligible to participate in the lottery), is not much larger than the number required to fill up the seats. One of the ways in which the number of the acceptable candidates can be reduced is to use multiple criteria in each of which the candidate should cross a threshold level to become eligible for the lottery. While such a method is not strictly in accordance with our principle of rejecting only those who are definitely inappropriate for the course, the probability of error against those who are appropriate may be greatly reduced. The development of a variety of tests, validation of the same, and evaluation of their performance on a continuous basis, is a process we should accept as a necessary component of research in the field of higher education.10 It is a very slow process and before enough confidence can be gained in the discriminatory power of the tests, it is better to accept the role of chance too, rather than accept the role of inherent biases and misguided objectivity. NOTES AND REFERENCES 1. Donald Crowley, 'Selection Tests and Equal Opportunity' ^Administration and Society, Vol 17, No.3, Nov 1985, pp. 361-384. 2. See for example: Baldev R. Sharma, 'What marks a manager: merit, motivation or money' Economic and Political Weekly, May 27, 1972, pp. M51-M66. This article reports the findings of a study based on admission (to IIM, Ahmedabad) data for five years. The findings suggest, among other things, that over the year 1967-71, the number of lowest income group (less than Rs. 3000/= p.a) applicants have reduced ^ while that of the highest income (over Rs. 15000 p.a.) has increased. Written tests have aggravated the problems for poor students. 3. See, for instance. Green B., 'A Primer of Testing', American Psychologist , Vol.36, 1981, pp. 1001-1011. 4. See, for instance, an almost continuing debate on the issue in the 'Letters to the Editor' columns of English newspapers such as The Hindu and the Indian Express, recently and in the past. 5. For example See E.J. Mcgrath, The Liberal Arts College and the Emergent Caste Systems; Teachers College Press, New York, 1966, p. 22. 'Selective Admission on limited measures of ability to pass admission tests may dose the door of opportunity to those who have the capacity to innovate; to recognize interrelations, or who have dedication to human welfare that will result in singular contributions to the well-being of mankind'. 6. An example of the dependence on the infallibility of those scores is the study by Shantamani V.S., 'Selection of Students', Indian Journal of Industrial Relations.,