To quantitatively throw out a measurement you must perform a

To quantitatively \"throw out\" a measurement, you must perform a statistical test on your data called a Q-Test. The formula for the Q-test is

where gap = (questionable value ? next closest number) and range = (largest number ? smallest number).

Q_calculated is then compared to Q_table. If Q_calculated > Q_table then the questionable datum can be discarded.

What is the gap of these values (questionable value ? next closest value)? (Take the absolute value if necessary to make this number positive.)

Q (90% confidence)	number of measurements
0.76	4
0.64	5
0.56	6

Solution

Results 1 42.45 42.67 42.21 41.98 43.55 Results 2 42.45 42.67 42.21 41.98 44.21 for result 1 gap 0.88 range 1.57 Q calc 0.56 Q table(n=5) 0.64 for result 2 gap 1.54 range 2.23 Q cal 0.69 Q table(n=5) 0.64 it is possible to reject the 44.21 value in Result 2 since Qcalc > Qtable at the 90% confidence