Like last year, we started Honors Physics with Measurement Uncertainty activities. Based on last year’s experience, last fall’s Illinois Science Education Conference, and this summer’s QuarkNet workshop “Beyond Human Error,” we made some minor modifications.
With the popularity of the LHC’s five sigma result, there was more of a context in which to introduce the concept of measurement uncertainty. I mentioned how calculus and Monte Carlo techniques could be used, but we stuck with the Crank-Three-Times method for this algebra-based class.
What was really missing in last year’s activities was how to estimate the measurement uncertainty when performing computer-based experiments. There are so many factors that contribute much more significantly to the measurement uncertainty than the computer-based measurement devices. David Bonner presented on “Learning Physics Through Experiments: Significance of Students’ Interpretation of Error” at the Illinois Science Education Conference last fall. One great idea I took away from his session was a simple and effective approach to addressing this challenge where students perform many trials to establish a range of values from which the measurement uncertainty is determined.
We rewrote the fifth station to introduce students to this method. Rather than using stopwatches, we setup two daisy-chained photo gates connected to a LabQuest 2 to measurement the elapsed time as a cart travels from the first gate to the second. The uncertainty of the LabQuest 2 is insignificant compared to other factors that affect the motion of the cart. Students performed ten trials and determined the measurement uncertainty from the range of values that they measured. We will use this technique throughout the year to estimate the measurement uncertainty.
Question about “crank three times”…why not “crank two times”, once to find the max and once to find the min? What is the benefit of the “middle” calculation? If they want to express this using value +/- uncertainty they can calculate those from the max and min. Thanks!
The +/- uncertainty may not be evenly distributed and, therefore, the middle calculation is needed to calculate the nominal value. At least that is my understanding. I learned of the Crank-Three-Times method in this document: http://www.av8n.com/physics/uncertainty.htm