Science, Technology and the Environment (SC NATS 1840A) –
Thursday, September 13, 2012
READINGS CHAPTER 19.5, 19.1, 26, 28 (DUE NEXT WEEK)
Statistical significance & scientific (un)certainty:
The outcome of any measurement is characterised by
Accuracy: How close the outcome is to the ‘true’ value.
Precision: The degree to which repeated measurements yield the same value.
Uncertainty in a measurement descries the range within which the ‘true’ value most likely lies.
(Example: Podium width= 55.35 cm (estimate) + 0.05 cm + 14% uncertainty
Average weight 144.2 pounds + 20 pounds + 8.6%
Average weight 144.6 pounds + 12 pounds
Lies between 132.2 pounds to 156.2 pounds
Controlled experiment with N1 subjects and N2 controls (or measurements). Averages differ and
have some uncertainty.
Is the difference real? Or just the result of random chance?
Male vs. Female weight
Tuesday: 19 Males, average = 164 pounds + 38 pounds (23%)
39 Females, average = 133 pounds + 23 pounds (17%)
Is the difference real? To be statistically significant the difference In the averages must be greater than the error ranges
Random fluctuations might mimic a real response in a controlled experiment. Statistical significance
Conducting experiment with many subjects and controls.
Repeating several times.
An experiment should be describable in sufficient detail to allow any independent replication to yield
the same o