Falsification Labs Teaching experimental skills John Welch Cabrillo

‘Falsification Labs’ Teaching experimental skills. John Welch, Cabrillo College • Students want to get the ‘right answer’ in lab, but this isn’t really how science is done. • They ask: “Is my result close enough? ” – meaning ‘do I pass? ’ • Where do they get this attitude? Chemistry Lab? Previous school experience?

How do we train them to ‘think like scientists’? • The only ‘right answer’ is careful investigation that is well documented. • Results that are inconsistent with theory are valid and exciting, (once carefully checked). • Decisions on consistency need to be made based on statistical analysis.

November, 2005 paper in Physics Education

Lessons from experience: • Statistical uncertainty concepts are necessary. • Students buy the fake theories, even if they know the ‘right’ answer. • Students will think they did something wrong. • Standardizing the experimental conditions (initial height, angle, etc. ) allows for group collaboration.

“What did we do wrong? ” Steps in guiding them to discovery. 1. Is there really anything ‘wrong’? - Can we say the results are ‘close enough’? - Is theoretical value within the confidence interval agreed on by the class? (2 sigma, etc)

“What did we do wrong? ” Steps in guiding them to discovery. 2. Have them consider possible sources of error: - Did you double check all measurements? - How far off, and in which direction, would a measurement have to be to account for the discrepancy? Is that likely? - Were there any shaky assumptions? If we didn’t ignore friction, etc, how would results change?

“What did we do wrong? ” Steps in guiding them to discovery. 3. Check in with other researchers. - Write your results on the white board. - How does your data compare to those of other groups? - Is it likely that everyone made the same ‘mistake’? - Is there a flaw in the experimental design?

“What did we do wrong? ” Steps in guiding them to discovery. 4. What’s left to consider? - They’ll still say stuff like ‘human error’, etc. - Some will start doubting theory, and will notice the funny names. - Is it possible that theory might not be right? - After all your checking, would you be comfortable telling a TV or newspaper reporter that you think Floogle and Depew are wrong? (95% confidence) - Call ‘authors’ on the phone to discuss results.

Resources: Falsification Lab activities available at Eric Ayars’ or my websites (google ‘Falsification Labs’) http: //phys. csuchico. edu/web/ayars/falsification/ http: //www. cabrillo. edu/~jwelch/falsification/false. html Please invent your own labs or modify these and send them to Eric or John to add to archives.