Prospects for evaluation of learning

Title

Creator

Ebel, Robert L.

Language

English

Source

Panorama Volume XXI (No. 4) April 1969

Year

1969

Rights

In Copyright - Educational Use Permitted

Fulltext

■ Can we depend upon objective tests and other methods of valuating educational achievement and knowledge? PROSPECTS FOR EVALUATION OF LEARNING What of the prospects for educational evaluation? Will the present practices be re versed? Will the present problems be resolved? The long history of education sug gests that enduring changes are more likely to evolve slowly than to explode sud denly. But changes do come. One of the current and anticipated changes has to do with the increased em phasis on education and its evaluation. Since World War II, the rush of students to college, in greater numbers than most good colleges could accommodate effective ly, has led to enormous ex pansion of admissions testing programs. The flow of dol lars to aid students who are able but not affluent has led to the development of scholarship testing programs. The needs these testing pro grams have served will con tinue, and no better alter native seems likely to dev elop. But we ought to hope and expect that tests will improve and their results will be used with increasing wisdom. New concern for quality in education and for equality of educational opportunity, with resultant increases in government expenditures and involvement, have led to re cognition of the need for re liable assessment of the re sults of our educational ef forts. The national assess ment is one attempt to meet this need, and some states have enacted law relating to mandatory testing and re porting of test results in the public schools. The growth of the widescale programs for testing educational achievement and for college admission and scholarship testing has led to another major change: the development of high Panorama speed, high capacity, highly automatic machines lor scor ing objective tests. It is reasonable to suppose that the years ahead will see ra pid growth in the utilization of these diverse and versa tile devices. Despite the current popu larity of objective tests and mass testing programs, how ever, there are those who believe that all is not well with the evaluation of learn ing today. Among the con cerns expressed are these: 1. That the tests currently used to evaluate learning are inadequate to the task, mea suring only imperfectly the less important educational outcomes. One’s opinion on this mat ter is, of course, likely to depend largely on whether he agrees with what most schools spend most of their time trying to do, that is, to help students gain com mand of useful verbal infor mation. The subject matter of most studies — history, literature, science, geography, even mathematics (if its svmbol are regarded as essen tially verbal symbols) — is verbal information. If verbal information is extracted from formal education, there is very little if anything left. But many educators are unwilling to admit that their aims are so prosaic, prefer ring to claim objective that are more spiritual than ma terial, and hence largely im measurable. Verbal know ledge is certainly not all that matters where man is con cerned, and the school can not afford to ignore muscu lar skills, or attitudes, or values, or character, or overt behavior. But neither can the school afford to give any of these things priority over command of knowledge in specifying its mission. And if it should choose to give other things priority, it will almost certainly find that cultivation of command of useful knowledge is the best, if not the only, means it can use to attain the ends it seeks. If this is true about man and the process of educating him, tests can do much of the job of evaluating learn ing. Many tests in current use are inadequate, it is true. But their faults lie less in the direction they point April 1969 7 than in the distance they travel. OBJECTIVE TESTS 2. That objective tests are spuriously attractive because of the ease with which they can be scored en masse, but are seriously deficient as tools for the evaluation of learning because of their in herent ambiguities, their ten dency to emphasize superfi cial factual information, and their reward of successful guessing. The supposed deficiencies of objective tests, however, are not inherent in the form. Objective test scores are typi cally more reliable than essay test scores, both because each student’s performance is judged against the same standard, and because of ex tensiveness in sampling var ious aspects of achievement. It is true that objective test questions appear to be tri vial more often than do es say test questions, but this is a matter of numbers. If a test can include only a few questions, as an essay test ordinarily does, the ten dency is to make each one general and comprehensive. Objective test questions also tend to be more “factual,” but it is important to re member that a fact in this sense is a verifiable truth, which need not be trivial. If a subject is not loaded with important factual truths, the value of studying it would seem open to serious question. It is true that answers to objective test items apparent., ly could be learned by rote?’ without real understanding, but this seldom happens. For one thing, it is always possible to pose questions that the examine has never encountered before, and thus require answers he could not have learned by rote. For another, rote learning is a difficult, ineffective, and un satisfying method of learning most things that students study. That test questions, either in objective or essay form, are sometimes ambiguous is also beyond dispute. But with reasonable skill and care in test construction, this can be reduced to the point where it no longer interferes seriously with the evaluation of learning. Panorama Like ambiguity, guessing is not a genuine menace in the use of objective tests. Well-motivated students do very little blind guessing on tests that are appropriate for them. The correctness of their informed guesses is re lated substantially to the amount of relevant informa tion they command. Thus their “guesses” provide valid indications of achievement. A student who does a great deal of blind guessing is likely to get a very low score on a good test. Fin ally, both ambiguity and guessing would result in in consistent results from re peated measurements, and so if a test constructor succeeds in building a test that yields reliable scores, it is safe to conclude that defects related to ambiguity and guessing are not serious on that test. Thus despite the criticisms of objective tests, it seems likely that their popularity will continue to grow. 3. That wide-scale testing programs and the use of standardized tests place teachers in curricular strait jackets, preventing them from meeting local needs or mak ing use of unique local op portunities, suppressing their creative ideas and their indi vidualities as teachers, and rewarding routine, mechani cal teaching. It is true that if students and teachers know in ad vance the general nature ol questions to be asked and content to be covered in a test used to evaluate learn ing, they will direct their study and teaching toward these kinds of capability. But if the tests are good tests, with appropriate cur ricular coverage and empha sis, and if they are not the sole basis for evaluation, they are likely to do much more good than harm. After all, the test-makers, in most cases, are themselves master teach ers, and the tests they build aim to follow rather than to lead curricular innovation. The teachers most likely to make the review of old tests a major part of their instruc tional program, as if they had been placed in a cur ricular straitjacket, are those who are least secure in their positions because they are least competent. April 1969 External tests have been influencing what is taught in particular classrooms for nearly 40 years; yet is it not true, in view of the increas ing mobility of our people, that a greater degree of uni formity among classrooms than we have today could well be tolerated? 4. That testing places stu dents under undue pressure and exposes them to unne cessary experiences of failure, diminishing their self-confi dence and destroying the joy of learning. It is not the measure of achievement but the aspira tion to achievement that places students under pres sure. Test scores simply re port levels of achievement; if the reports are disappoint ing, the blame may rest on ineffective learning or teach ing, or on unrealistic expec tations. The suggestion that the way to deal with excess pressure is to stop paying so much attention to achieve ment makes very little edu cational sense. Instead we need to pay more attention to the setting of realistic goals, and to the recognition of individual differences ill interests, abilities, and ave nues for self-fulfillment. 5. That testing, particular ly intelligence and aptitude testing, leads to the labeling of pupils as bright or dull, in both cases adversely af fecting their expectations, their efforts, and their self concepts; denying and thus tending to destroy the almost infinite potential for devel opment inherent in every human being. Although the items in most intelligence and apti tude tests are clearly mea sures of developed ability, too many educators have been willing to believe that they provided direct and de pendable measures of innate capacity for learning. On too many occasions, a child’s low IQ score has been used to explain his failure to learn instead of being used to help him to learn. But these tests have some times been interpreted pro perly and used constructive ly. It is hard to beat a good intelligence test as a convenient measure of a young child’s general educa tional development. Since 10 Panora ua all learning builds on prior learning, effective teaching requires information on each child’s level of educational development. More schools may join those which have abandoned intelligence testing because of abuses and because of local pressures, but it is not likely that intelligence testing will disappear. We can hope and expect, however, that intelligence and aptitude tests will be interpreted more realistically and used more constructively. In all, to teach without testing is unthinkable. Teachers are likely to do more testing in the future, and to do it better as they become more skilled in the techniques of their craft. Above all, they are likely to use the results of testing more wisely and more cons tructively. — By Robert L. Ebel, from The Education Digest, March, 19.69. AWARENESS OF LIMITS As we advance in life, we learn the limits of our abilities. — James Anthony Fraude April 1969 11

pages

6+

Media