# Day 74: Effortless Error Bars

For the cart and spring lab (see Day 73), it is tough for students to pull with a steady force so the carts accelerate at a constant rate. Accelerations may vary wildly from one trial to the next as students start to get a feel for how to do it properly. Good data collection in this case will require multiple trials for each pulling force. However, rather than averaging the accelerations and plotting the average acceleration for each force, I asked the students to graph all their data instead. This produces a vertical smear of points, which act sort of like error bars. I like the visual of the smear of uncertainty and I think it helps students with drawing the line of best fit as the aim to get the line through all the smears.

??

(If I recall correctly, I picked this technique up from the Den of Inquiry books. The Den of Inquiry books are about modeling physics experiments, which range in difficulty from basic on up to AP Physics level.)

Why is the best-fit line so bad? I guess it might have the right slope, but shouldn’t a real line of best fit go through the centers of your error blobs, rather than just nicking the topmost point in each collection?

Ah, well that is what I had to work with. Groups either graphed data by hand on graph paper or graphed the data using LoggerPro. For whiteboard discussions, I just ask the groups to sketch their graph on the boards. This was the only group that actually graphed points on their whiteboards, so I thought it would make a cool picture. But, I guess that doesn’t excuse their sloppy BFL.

This is a great idea. Very easy for kids to see that there is variation in their measurements, but that the relationship is still linear. Nice!