SUBJECTS SAW A PRELIMINARY TAPE, THEN A SECOND TAPE SHOWING A SPEECH DISRUPTED BY A FIGHT. THEY COMPLETED QUESTIONNAIRES ABOUT EVENTS AT ONE OR MORE DIFFERENT TIMES: IMMEDIATELY AFTER VIEWING, AND/OR AFTER 1, 7, OR 28 DAYS. THE RESEARCH DESIGN PROVIDED AN OPPORTUNITY TO COMPARE THE MEMORIES OF SUBJECTS WHO REPEATEDLY REHEARSED THEIR STATEMENTS OF EVENTS WITH THOSE WHO DID NOT. STORAGE THEORY AND RECONSTRUCTIVE THEORY OF MEMORY ENTAILED DIFFERENT PREDICTIONS CONCERNING HOW REHEARSAL AND THE PASSAGE OF TIME WOULD AFFECT DETAILED MEMORY. THE CONCLUSIONS THAT SUBJECTS FORMED ABOUT THE DISRUPTION REMAINED STABLE OVER 28 DAYS. DETAILED MEMORY RAPIDLY DETERIORATED FOR SUBJECTS WHO WERE NOT REQUIRED TO REHEARSE THEIR STATEMENTS OF EVENTS. SUBJECTS' CONCLUSIONS ABOUT MEANING AFFECTED THEIR SELECTIVE ATTENTION TO DETAIL. SUBJECTS SELECTIVELY ATTENDED TO MEANINGFUL INFORMATION WHEN THEY DID NOT EXPECT THE DISRUPTION THAT OCCURRED DURING THE TAPED EVENT. SUBJECTS WHO HAD BEEN LED TO EXPECT THE DISRUPTION DID NOT SELECTIVELY ATTEND TO MEANINGFUL INFORMATION. THE FINDINGS SUGGEST THAT OBSERVERS WILL SELECTIVELY ATTEND TO MEANINGFUL INFORMATION IF THEY HAVE DIFFICULTY IN DRAWING GENERAL CONCLUSIONS ABOUT THE MEANING OF AN EVENT AND THAT REPEATED REHEARSAL MAY PROMOTE AND EVEN ENHANCE ACCURATE MEMORY FOR MEANINGFUL INFORMATION. FINDINGS ON DETAILED MEMORY SUGGEST THAT A LINEUP OF SUSPECTS CAN CAUSE A WITNESS TO IDENTIFY A SUSPECT OF SIMILAR APPEARANCE, MEMORIZE THE DETAILS OF THAT PERSON'S APPEARANCE, AND THEN ERRONEOUSLY IDENTIFY THAT SUSPECT IN COURT AS THE GUILTY PERSON. A LIST OF REFERENCES IS PROVIDED. (AUTHOR ABSTRACT MODIFIED--DEP)