Balancing broomsticks to a t!
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Physics. The day in general physics: instruments, methods and results across the field.
Kai: We're looking at the paper: "Balancing broomsticks to a t!".
Mira: Here's the gist of the paper. Balancing broomsticks to a t!
Kai: The paper introduces an experiment using balancing metersticks to illustrate and refine the concept of 't'' in rotational mechanics, which guides students in comparing experimental results and iterating future measurements The core of the physics lab experience is structured around this t' paradigm, where the distance between values divided by their combined uncertainty informs students about their measurements and next steps The t' Framework and Interpretation The t' paradigm forms the core of our physics lab experience Results of measurements are compared to either theoretical predictions or other sets of measurements, and the distance between the values is divided by their combined uncertainty This value, labelled t’, gives the students information about their measurements and helps them determine what their next steps should be in iterating future experiments Ideally the students cycle through several upgrades of experiments If the t’ value is less than one, then the two measurements are indistinguishable given their uncertainties If t’ is greater than 3, then one is clearly larger than the other Values between 1 and 3 indicate inconclusive data As top-heavy broomsticks should be easier to balance, it is expected that after several rounds of improving experiments, t’ should be greater than 3 when comparing balancing times in the two orientations Experimental Setup and Variables Students are instructed to tape 200g of mass at the 67cm point on wooden metersticks, which act as stand-ins for broomsticks Secretly, one group is advised to place their mass at 80cm instead The instructor demonstrates balancing the meterstick on their hand, and students measure how long they can balance the meterstick (fig.1) This requires students to calculate means and standard errors for their data Groups can go outside of the lab room to either outdoors or a hallway for safety when moving around to keep the meter stick balanced Refining Experimental Procedures Once students have found their t’ values they critique their experiments and look for refinements Common issues that arise include training effects where humans become better at tasks with practice, so due to training effects later attempts may result in longer balance times Alternating the two orientations between measurements would help mitigate this effect The suggestion to refine an experiment often results in taking more data, as more trials should reduce the standard error The instructor should point out that the standard error shrinks as the square root of the number of trials, so diminishing marginal returns sets in quickly Addressing Data Variability The balancing times in this lab can have wide variability, which leads to the question of outliers and rejection of data This provides for the opportunity to discuss Chauvenet’s criterion, which is not often a first-day topic With the use of tables or a z-score calculator students can determine the probability of the result, and consider rejecting the outlier if the product of the probability and the number of trials is less than ½ Students are told to include results both with and without the outlier for transparency The asymmetry of the weighted stick could be increased, either by adding more mass or moving the mass farther from the midpoint When students suggest this, the 80cm group is revealed, and their t’ result is shared out It is typically above 3, indicating that the balancing times in the top-heavy orientation clearly longer than those for bottomheavy Groups often have multiple members act as stickbalancers, and using only one person is a good control variable, as balancing skill varies Using multiple people is a teachable moment if it is clearly indicated in a group’s data If the data are summed across all of the group members, then the standard error will be larger, as it accounts for variation due to both luck and the possibly different skill levels of the participants As the standard error is in the denominator of the t’ ratio, this artificially depresses the t’ value when compared to the t’ calculated for each individual in the combined measurement References The paper cites several sources regarding balancing sticks and reaction times Robert H.
Mira: Romer, “Balancing Sticks and Reaction Times Revisited,” Phys.
Kai: Teach., 43, 68-69 (2005) https://doi.org/10.1119/1.1855735 Leonid Minkin and Anthony C.
Title and authors: Kai: We're looking at "Balancing broomsticks to a t!" and the authors, David Marasco from Foothill College, who frame this as a classic question in rotational mechanics about whether it's easier to balance something heavy-side up or heavy-side down.
Mira: Essentially, the paper uses this specific balancing experiment as a mechanism to introduce and review that 't' framework so students understand how to compare any two sets of measurements against each other using that ratio.
Lev: The core idea they stress is that the distance between those values divided by their combined uncertainty is the critical piece of information you need to know what your next experimental step should be.
Kai: They outline clear thresholds for this 't' value: if it’s less than one, the two measurements are indistinguishable given their uncertainties, and if it’s greater than three, one measurement is clearly larger than the other.
Mira: The paper also sets a clear expectation that after you cycle through several rounds of improving your experiments, especially with top-heavy sticks, you should end up with a 't' value greater than three when comparing the balancing times in both orientations.
Lev: That outcome sounds like what we'd expect if the physics is working as predicted and your measurement technique has stabilized enough to overcome initial setup noise.
Kai: The experimental setup involves taping two hundred grams of mass at sixty-seven centimeters on wooden meter sticks and measuring how long you can balance the stick in two different ways, which they call top-heavy versus bottom-heavy.
Mira: And they add a slight twist by suggesting that one group might place their mass at eighty centimeters instead of the sixty-seven centimeter point to see how that changes the outcome.
Lev: That’s a classic way to introduce an asymmetry into the system, which is fundamental when you're studying rotational dynamics, forcing you to account for where the center of mass is actually located.
Kai: They also discuss how training effects can skew those balance times over repeated attempts, suggesting that alternating the two orientations between measurements would help mitigate that effect.
Mira: That speaks directly to the practical difficulty of doing these kinds of experiments where human skill naturally improves with practice, so you need a systematic way to control for that.
Kai: So they summarize it as a quick and fun implementation that introduces the 't' paradigm, which then guides students in comparing results and figuring out how to iterate their measurements.
The paper's summary: Mira: Now moving into what they suggest as improvements for this lab experience, the paper really shifts away from just doing more of the same experiment right away.
Kai: They propose that instead of immediately running another round of upgraded experiments, the class should focus on critiquing their original implementation and sharing their thinking about what issues you encountered during that first attempt.
Lev: I think that's a much more valuable learning step than just trying to get a longer balance time on the second try, because you need to understand *why* it was shorter or longer first.
Kai: They specifically call out training effects as a common issue and suggest alternating orientations between measurements as a way to mitigate that effect from later attempts.
Mira: That’s interesting because it suggests an active process of refinement rather than just passive data collection after the initial run is done.
Lev: And they also bring up using only one person for balancing as a good control variable, since individual balancing skill varies quite a bit in a group, so you want to isolate that effect.
Kai: They suggest that if multiple people are involved, it becomes a teachable moment if the data clearly shows it's important to separate the skill variation from the stick's properties.
Mira: And they talk about how adding more trials can reduce standard error, but they warn you that this improvement comes with diminishing returns quite fast as you get more data.
The paper's improvements: Kai: So to wrap up on "Balancing broomsticks to a t!", the paper highlights that this exercise is really about establishing a robust framework for how students should approach comparing different measurement results.
Mira: It’s about understanding that the distance between those measurements, divided by their combined uncertainty is the key metric for judging how good your comparison is.
Lev: And they show you that if you stick to this 't' paradigm and keep iterating, you build confidence in your experimental process over time.
Kai: The implication for us is that we need to make sure students aren't just running experiments blindly but are using these statistical tools to guide their next steps thoughtfully.
Mira: And it seems like the paper really encourages a cycle of critique before you jump into more complex experimental setups, which saves time and confusion in the long run.
Lev: For me, it’s about recognizing that initial results are often just noise until you apply these iterative checks, which is something we do constantly when we're debugging real systems.
Kai: So they leave us with this concrete example of how a simple physical setup can introduce a deep concept in rotational mechanics and error analysis.
Mira: And it’s a great reminder that even seemingly simple tasks have layers of statistical thinking underneath them that are important for any physics lab.
Lev: That's the point, you learn how to manage uncertainty before you try to run something that actually matters on real hardware.
Conclusion: Kai: So we looked at "Balancing broomsticks to a t!", which is basically using meter sticks to figure out how to compare measurements when you have uncertainty involved in rotational mechanics.
Mira: Right, so they use the balancing experiment to define that 't' value, which is the distance between two values divided by their combined uncertainty.
Lev: I mean, it’s a way for students to see how measurement error affects their comparison of theoretical predictions versus experimental results.
Kai: Exactly. They set up two hundred grams of mass on a meter stick and tape it at sixty-seven centimeters, comparing top-heavy versus bottom-heavy balance times.
Mira: And they introduce the idea that if you keep running these experiments, you should get a 't' value greater than three for top-heavy sticks, which tells you one orientation is clearly longer.
Lev: From my side, it’s about how much noise we need to filter out before we can say one measurement is actually better than the other on real hardware.
Kai: They also talked about how training effects mess things up later on, so they suggest alternating the orientations between measurements to try and fix that bias.
Mira: And they touched on data variability too, suggesting using Chauvenet’s criterion to decide if you should toss an outlier based on its probability of being wrong.
Lev: That statistical filtering is crucial; you can collect a million trials, but if your statistical test isn't right, the resulting comparison is still garbage.
Kai: So what this means for us in the quantum world is that before we build anything, we need to have a solid way to compare our measured parameters against what theory says they should be.
Mira: It’s about establishing that baseline confidence level from the very start of the experimental process.
Lev: Because when you're building error correction codes for real hardware, those initial comparisons are how you figure out where your overhead will really land in terms of actual performance versus just theoretical limits.
Kai: That sounds like a good way to frame it before we move onto something more complex, like those papers on transfer learning or turbulence models.
Mira: Absolutely. It’s about making sure our foundational measurement skills are sound before we try to apply them to huge simulations or complex physical systems.
Lev: Before you can predict how a model will behave in a turbulent flow, you need to know how well you can measure the basic inputs reliably first.
Kai: So that’s it for "Balancing broomsticks to a t!". It’s all about using simple tools to build up that critical thinking cycle for experimental science.
Foothill College
physics.ed-ph
Submitted: 2026-11-02
Updated: 2026-11-02
Comments: 3 pages, 1 figure
Journal ref: Phys. Teach. 64, 434-435 (2026)
DOI: 10.1119/5.0322733
License: http://creativecommons.org/licenses/by-nc-sa/4.0/
Importance score: 74/100
The gist: Balancing broomsticks to a t! The paper introduces an experiment using balancing metersticks to illustrate and refine the concept of 't'' in rotational mechanics…
Key concepts
- t' Framework
- The t' framework compares two measurements (like theoretical vs. experimental) by dividing the distance between them by their combined uncertainty. If t' < 1, measurements are indistinguishable; if t' > 3, one measurement is clearly larger than the other.
- Iterative Improvement Cycle
- Instead of immediately trying a new setup after a measurement, students should first critique their original experiment and propose specific refinements based on observed issues. This focuses their thinking on improving the current implementation before moving to an upgraded experiment.
- Training Effects Mitigation
- To prevent human skill improvement from skewing results over multiple trials, students should alternate the two orientations of the meterstick between measurements. This helps mitigate training effects by ensuring consistent performance across different experimental conditions.
- Chauvenet's Criterion
- This criterion is used to assess data variability and identify outliers. By calculating the probability associated with a result, students can reject an outlier if the product of that probability and the number of trials is less than 0.5.
Terminology
Summary
Balancing broomsticks to a t! p. 0 Balancing broomsticks to a t! David Marasco (Foothill College)
A classic question from rotational mechanics asks “Is it easier to balance a broomstick
heavy-side up or heavy-side down?Where the paper says it
The t' Framework and Interpretation
The t' paradigm forms the core of our physics lab experience Results of measurements are compared to either theoretical predictions or other sets of measurements, and the distance between the values is divided by their combined uncertainty This value, labelled t’, gives the students information about their measurements and helps them determine what their next steps should be in iterating future experiments Ideally the students cycle through several upgrades of experiments If the t’ value is less than one, then the two measurements are indistinguishable given their uncertainties If t’ is greater than 3, then one is clearly larger than the other Values between 1 and 3 indicate inconclusive data As top-heavy broomsticks should be easier to balance, it is expected that after several rounds of improving experiments, t’ should be greater than 3 when comparing balancing times in the two orientations
Experimental Setup and Variables
Students are instructed to tape 200g of mass at the 67cm point on wooden metersticks, which act as stand-ins for broomsticks Secretly, one group is advised to place their mass at 80cm instead The instructor demonstrates balancing the meterstick on their hand, and students measure how long they can balance the meterstick (fig.1) This requires students to calculate means and standard errors for their data Groups can go outside of the lab room to either outdoors or a hallway for safety when moving around to keep the meter stick balanced
Refining Experimental Procedures
Once students have found their t’ values they critique their experiments and look for refinements Common issues that arise include training effects where humans become better at tasks with practice, so due to training effects later attempts may result in longer balance times Alternating the two orientations between measurements would help mitigate this effect The suggestion to refine an experiment often results in taking more data, as more trials should reduce the standard error The instructor should point out that the standard error shrinks as the square root of the number of trials, so diminishing marginal returns sets in quickly
Addressing Data Variability
The balancing times in this lab can have wide variability, which leads to the question of outliers and rejection of data This provides for the opportunity to discuss Chauvenet’s criterion, which is not often a first-day topic With the use of tables or a z-score calculator students can determine the probability of the result, and consider rejecting the outlier if the product of the probability and the number of trials is less than ½ Students are told to include results both with and without the outlier for transparency The asymmetry of the weighted stick could be increased, either by adding more mass or moving the mass farther from the midpoint When students suggest this, the 80cm group is revealed, and their t’ result is shared out It is typically above 3, indicating that the balancing times in the top-heavy orientation clearly longer than those for bottomheavy Groups often have multiple members act as stickbalancers, and using only one person is a good control variable, as balancing skill varies Using multiple people is a teachable moment if it is clearly indicated in a group’s data If the data are summed across all of the group members, then the standard error will be larger, as it accounts for variation due to both luck and the possibly different skill levels of the participants As the standard error is in the denominator of the t’ ratio, this artificially depresses the t’ value when compared to the t’ calculated for each individual in the combined measurement
References
The paper cites several sources regarding balancing sticks and reaction times Robert H. Romer, “Balancing Sticks and Reaction Times Revisited,” Phys. p. 2 References:
1. Romer, “Balancing Sticks and Reaction Times Revisited,” Phys. Romer, “Balancing Sticks and Reaction Times Revisited,” Phys. 43,
68-69 (2005) https://doi. 43,
68-69 (2005) https://doi. Leonid Minkin and Anthony C. Leonid Minkin and Anthony C. Zable, “From Balancing a Stick to Tightrope Walking:
How Added Mass Influences the Act of Balancing,” Phys. Zable, “From Balancing a Stick to Tightrope Walking:
How Added Mass Influences the Act of Balancing,” Phys. 60, 549-553 (2022)
https://doi. 60, 549-553 (2022)
https://doi. Holmes and D. Holmes and D. Bonn, “Quantitative Comparisons to Promote Inquiry in the
Introductory Physics Lab,” Phys. Bonn, “Quantitative Comparisons to Promote Inquiry in the
Introductory Physics Lab,” Phys. 53, 352–355 (2015)
https://doi. This
individual could balance the stick for minutes. Taylor, An Introduction to Error Analysis, 2nd ed. Taylor, An Introduction to Error Analysis, 2nd ed. (University Science Books,
Sausalito, 1997), p.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Incorporate a t’ framework review into all subsequent introductory mechanics lab courses to ensure students are familiar with
t’ concepts
before performing experiments, addressing the lack of familiarity for students from other institutions. -
Implement an iterative improvement cycle where students critique their original implementation and propose refinements rather than immediately performing an additional round of upgraded experiments, focusing on
sharing their thinking on issues in their original implementation.
-
Develop a mechanism to mitigate training effects by alternating the two orientations between measurements, as suggested by the paper to help
mitigate this effect
where humans become better at tasks with practice. -
Establish a clear protocol for handling data from multiple group members, ensuring that standard error calculations reflect individual skill variation rather than artificially depressing the t’ value when summing results.
-
Integrate Chauvenet's criterion into the data analysis workflow, allowing students to
determine the probability of the result
and consider rejecting outliers ifthe product of the probability and the number of trials is less than ½.
Abstract
A classic question from rotational mechanics asks, "Is it easier to balance a broomstick heavy side up or heavy side down?" Measuring the time a student can balance an asymmetrically weighted meter stick in two orientations is a straightforward exercise that can be used to introduce or review the t' framework, which guides lab students in comparisons of results and iterative improvement of their measurements. As a two-year college physics program, many of our students attend from other institutions, so while t' is taught in the introductory mechanics labs, there is no guarantee that a student in a following lab course will be familiar with the material. To address this, the t' concepts are reviewed in the first session of subsequent lab courses. The broomstick experiment is a quick and fun implementation.
Related papers
- Raytracing Black Holes with AI Accelerators
- FlowMLLab: An open-source framework for reproducible computational-fluid-dynamics and scientific-machine-learning experiments
- Students' Validation Criteria for AI-Generated Physics Solutions
- Thermal fluctuations and electronic readout using a moving-coil movement
- Predictive Modeling for High Impact Active Learning Classrooms
- A Reproducible Inertia Dynamometer for Vocational Schools: Design, Fabrication, and Pedagogy