THE MOCK-WITNESS PARADIGM APPROACH CORRECTED THE FAILINGS OF THE FORMER 'CHECKLIST APPROACH' TO ELIMINATING BIAS IN LINEUPS. THE CHECKLIST APPROACH ENTAILED SPECULATING ON ALL POSSIBLE BIASES IN A LINEUP (WHETHER SEVERAL DIFFERENT RACES, SKIN TONES, HEIGHTS, BUILDS, ETC., WERE REPRESENTED) AND THEN TRYING TO CONTROL FOR THEM. THE PROBLEM WAS THE APPROACH COULD NOT DEAL WITH THE FACT THAT VARYING FACTORS HAVE MULTIDIMENSIONAL AND UNKNOWN WEIGHTS. THE MOCK-PARADIGM APPROACH DEFINED A BIASED LINEUP AS 'ONE WHERE A PERSON WHO WAS NOT A WITNESS TO THE CRIME IS MORE LIKELY TO PICK THE SUSPECT OUT OF THE LINEUP THAN WE WOULD EXPECT BY CHANCE.' CHANCE WAS THEN SET AT A CERTAIN NUMBER IN PROPORTION TO THE NUMBER OF PERSONS IN THE LINEUP. THE MOCK-PARADIGM APPROACH, HOWEVER, HAD TWO DRAWBACKS. ITS INFLEXIBLE FORMULA FOR DERIVING THE VALUE OF CHANCE WAS INSENSITIVE TO THE AUDITION OF BAD FOILS OF A LINEUP AND IT COULD NOT LABEL 'SHOW UPS' (LINEUPS OF ONLY ONE PERSON, THAT PERSON BEING THE SUSPECT). A SOLUTION WOULD BE USING THE PROBABILITY OF MOCK WITNESSES CHOOSING THE SUSPECT AS THE SUMMARY STATISTIC RATHER THAN TESTING THE PROBABILITIES AGAINST A NULL HYPOTHESIS. THE PROBABILITY WOULD BE TRANSFORMED INTO AN EQUATION WHICH REFLECTS THE FUNCTIONAL SIZE OF A LINEUP; I.E., THE NUMBER OF LINEUP MEMBERS RESEMBLING THE CRIMINAL. THIS FUNCTIONAL SIZE ESTIMATE IS INVARIANT AND BOTH APPROPRIATELY INSENSITIVE TO ADDITION OF BAD FOILS IN THE LINEUP AND SENSITIVE TO SHOW UPS. THE FUNCTIONAL SIZE CONCEPT WOULD WORK BEST WITH A SAMPLE OF ENOUGH WITNESSES TO PRODUCE A 95 PERCENT CONFIDENCE LEVEL OF PLUS OR MINUS .5. WHEN TESTED IN THE CASE OF U.S. V. MILLS, THE PARADIGM INDICATED A TOTAL OF 60 COLLEGE STUDENTS (MOCK WITNESSES) WHO CORRECTLY IDENTIFIED THE SUSPECT FROM A 6-PERSON LINEUP. THE FUNCTIONAL SIZE CONCEPT CAN AUGMENT THE CRIMINAL JUSTICE PROCESS EITHER AS A POST HOC CALCULATION OF THE FUNCTIONAL SIZE OF LINEUP FOR EITHER THE DEFENSE OR PROSECUTION, OR AS AN AID TO POLICE IN LINEUP CONSTRUCTION BY PROVIDING KNOWLEDGE OF POTENTIAL BIASES ON THE PART OF POLICE INVESTIGATORS. THIS TECHNIQUE COULD ALSO BE USED FOR EYEWITNESS-IDENTIFICATION RESEARCH AS WELL AS FOR SCREENING OUT FALSE-IDENTIFICATION WITNESSES. NOTES AND TWENTY-FOUR REFERENCES ARE APPENDED.