Projective drawing tests have had limitations for scientific evidence including reliability and validity even
though they have been used for a long time to evaluate emotional-behavioral problems in the field of
psychotherapy. It is important work to clarify the process of confirmation bias of the test evaluator,
because the confirmation bias is usually unavoidable. In this study, the confirmation bias by previous
imformation was tested using a randomized controlled trial. A total of 171 fifth and sixth elementary school
students were participated in the EBDT drawing test, and six art therapists participated in the evaluation of
their emotional-behavioral problems. All students completed a positive automatic thinking questionnaire and
were allocated to three groups equally. Previous imformation, such as negative, positive, and no information
were given to each group to evaluate of the drawing. The the results, the negative previous imformation
group had a significantly higher score for emotional-behavioral problems the than other groups. According
to the results, the time and ways to provide personal information about clients should be regulated. The
implication of educating art therapists in order to become aware of confirmation bias was also discussed.