TWO QUESTIONS DOMINATE THE DISCUSSION. (1) AT WHAT PIVOTAL POINTS IN THE RESEARCH PROCESS CAN THE EXPERIMENTAL STUDY GO ASTRAY, GIVING RISE TO MISLEADING RESULTS AND CONCLUSIONS? (2) WHAT STEPS CAN RESEARCHERS TAKE TO AVOID THESE PITFALLS? THE DISCUSSION FOCUSES ON THE WORK OF BOTH INVESTIGATORS AND EXPERIMENTERS. THE INVESTIGATORS' FIRST PITFALL CONCERNS STANDARD PARADIGMS, THEIR TENACITY, AND INVESTIGATOR RELUCTANCE TO DISCOVER NEW PATTERNS. MOREOVER, A FAMILIAR PARADIGM MAY BLIND INVESTIGATORS TO THE EVIDENCE PRESENT AND CAUSE THEM TO 'SEE' NONEXISTENT EVENTS. INVESTIGATORS MUST ALSO BE AWARE OF THE DIFFERENCE BETWEEN GENERAL PARADIGMS ACKNOWLEDGED IN THE FIELD AND PERSONAL PET THEORIES AND HYPOTHESES. FURTHER PITFALLS MAY BE ENCOUNTERED IN THE EXPERIMENTAL DESIGN, WHICH AFFECTS RESULTS ACCORDING TO ITS STRUCTURAL COMPLEXITY OR SIMPLICITY, AND TO ITS SUBJECT GROUP DIVISIONS. LOOSE PROCEDURES AND THE FAILURE TO PRESCRIBE STEP-BY-STEP IMPLEMENTATION MAY ALSO AFFECT RESULTS. THE FINAL INVESTIGATOR PITFALL INVOLVES DATA ANALYSIS AND THE NEED FOR UNBIASED APPLICATION OF STATISTICAL METHODS FOR OBJECTIVE RESULTS. PITFALLS OF EXPERIMENTERS HAVE TO DO WITH THEIR PERSONAL ATTRIBUTES (E.G., RACE, SEX), FAILURE TO CONTROL FOR OTHER ATTRIBUTES, FAILURE TO ASSESS INTERACTIVE EFECTS, AND FAILURE TO SAMPLE PSYCHOLOGICAL TASKS. IN ADDITION, EXPERIMENTERS MAY FAIL TO FOLLOW PROCEDURES, MAY MISRECORD THE DATA, AND MAY FUDGE DATA. FUDGING RESULTS THAT WERE NOT ACTUALLY OBTAINED IN THE RESEARCH IS BLAMED ON SCIENTISTS' DESIRE FOR RECOGNITION. BY MAKING EXPLICIT THE PITFALLS OUTLINED HERE, INTEGRATING THEM INTO GRADUATE TRAINING PROGRAMS, AND KEEPING THEM IN FULL VIEW OF THE PROFESSION, THE VALIDITY AND GENERALIZABILITY OF EXPERIMENTS CAN BE SIGNIFICANTLY IMPROVED. NOTES AND REFERENCES ARE PROVIDED. (MRK)