PAPERS OX PSYCHOLOGY AND EDUCATION 391 



was made that the investigator should have considered 

 these limitations in preparing his report of the study. 

 In a reply the author of the original article criticised his 

 critic. Among other things he said. "Nor can there be 

 any question about the reliability of gathering the data. 

 The tests were given by two experienced examiners and 

 scored by trained scorers under supervision." 



This statement expresses what appears to be a pre- 

 valent attitude toward the measures yielded by standard- 

 ized educational tests. If the tests were administered by 

 experienced examiners and if there is reason to believe 

 that no errors were made in marking the test papers, 

 then the scores may be considered accurate measures of 

 the traits or abilities which the tests were designed to 

 measure. If specifically interrogated, most test users 

 would probably admit that our present standardized 

 tests are imperfect, but a large number disregard pos- 

 sible limitations when they are using these instruments 

 of educational research. Variable errors are always 

 present in test scores and constant errors are frequently 

 introduced even when the tests have been carefully ad- 

 ministered. Critical studies of standardized tests have 

 demonstrated that the possibility of errors in test scores 

 is sufficiently great to make the investigator assume the 

 responsibility for proving that his data are accurate 

 when there is any reason for suspicion. Failure to do 

 this means that the investigator is building upon a sus- 

 picious foundation which may result in the collapse of 

 his conclusions. The conclusions reached by educational 

 research cannot be more dependable than the weakest 

 step in the study. 



In view of the frequent failure of investigators to be 

 critical of their data, it is then not inappropriate to raise 

 again the question, "Are we receiving adequate divi- 

 dends for the time and money which is being invested in 

 educational research in the United State- 

 There are literally hundreds of persons putting time 

 and money in educational research, but with few except- 

 ions they are working independently and with little 

 reference to what other workers have already done. If 

 one examines the voluminous literature in the field, he 



