A standardized silent reading test should be given at intervals
 to all students. The results must be interpreted with great care! At
 the introductory stage the tests tend to overestimate the student's
 reading ability. They should not be used in determining instructional
 level. But students want a record of their progress. Governmental
 agencies often request a record of the student's progress and achieve-
 ment. In many states elementary grade certificates are dependent in
 part on the student's achievement on a standardized reading test. There-
 fore, an objective measure of progress is required. Since the standard-
 ized reading test is objective and its error of overestimate tends to
 remain rather constant, it can and should be used for this purpose. It
 is recommended that a test with several forms be used. The test should
 be appropriate to the stage of reading where the student is operating.
 A testing interval of about fifty hours may be feasible. A different
 form should be used for each testing.

 Standardized silent reading tests that may be used at the three
 stages of adult basic education include the Gray-Votaw-Rogers General
 Achievement Tests, (Steck Co.), and the Stanford Achievement Tests
 (Psychological Corporation). Others of value, when carefully inter-
 preted, are the Metropolitan Achievement Tests (Psychological Corpora-
 tion), and California Reading Tests (California Test Bureau), and the
 Gates Reading Survey, Grades 3 5-10 (Bureau of Publication, Teachers
 College, Columbia University).

 All standardized tests have a band of error. This band is des-
cribed in the test manuals. It is important that teachers keep it in
mind when reviewing test results. It is possible, just by chance, for
a person to earn a score much too high or much too low any time he
takes a test. Some students who have made real progress will show a
decline in test scores after fifty hours of instruction! Many teachers
panic when they see such a result. However, the loss in score can
frequently be explained by examining each item carefully. Sometimes a
student who reads poorly guesses at every answer and gets one fourth
of the answers right. Upon taking another form of the test at a later
date, he may only attempt those questions to which he is sure of the
answer. Thus, his total score and grade-norm regress. The best way
to prevent such cases is to choose the standardized tests carefully. The
test chosen should be one where the student's expected score is near the
middle of the range of scores possible for that test.

 An item-by-item analysis should be made of each test. Perhaps
the least important information that a test gives is the total reading
grade level. The great value lies in the diagnostic information that
it can give you. A careful examination of the test may reveal specific
deficiencies in word attack skills, in meaning vocabulary, in spotting
key words, in reading critically, and even in test-taking. The last
item is of particular importance for adults. Many adults have never
seen or taken a standardized test. The first exposure to this type test-
ing is apt to be unsettling. Remember that separate answer sheets


-13-