STEM Subject Lab/Packs/NM-LAB-07
New this season

Race Data & Graphs.

We time every run and we never trust a single number. Six runs, a mean, a range, and one suspicious result that has to be argued about. Then the split times from the timing gates become a distance–time graph, and the gradient tells you how fast the car was going along the lane.

Open the simulation Jump to paper sheet

01Learning objectives

01

Calculate mean, median and range for a set of race times and choose which one to report.

02

Identify an outlier, justify excluding it, and recalculate.

03

Plot a distance–time graph from timing-gate splits and read speed from the gradient.

04

Explain why you need repeated runs before you can call a design change an improvement.

02From the Nightmare car

Every test day produces a sheet like the one in Fig. 7. Six runs of the same car, the same batch of cartridges, the same launcher. Five of them agree to within a few hundredths of a second. One does not. That run had a slow launch trigger and we all saw it happen. The question is what to do with it. Our rule is that a result is only left out if there is a physical reason written on the sheet, and the mean of what is left is what we compare against the next design change. The timing gates every 5 m give a second picture: split times, which become a distance–time graph.

Test day: six runs and four gates Run sheet, configuration B Run 11.212 s Run 21.184 s Run 31.239 s Run 41.196 s Run 51.620 sslow trigger Run 61.203 s mean of 6 = 1.276 s mean of 5 = 1.207 s (reported) range of 5 = 0.055 s, median 1.203 s Distance against time from the gates, run 2 t (s)x (m) 00.40.81.205101520 Δt = 0.23 s Δx = 5 m gradient = 5 / 0.23 ≈ 22 m/s Steeper means faster. The curve bends early because the car is still speeding up.

Fig. 7 Left: six runs of the same set-up. Run 5 is left out for a written physical reason and the mean of the other five is reported. Right: the timing-gate splits plotted as distance against time. The gradient of each segment is the average speed over that 5 m.

03Core idea

Three summaries, one decision

mean = Σx / n   ·   median = middle value when sorted   ·   range = largest − smallest

The mean uses every value, so one bad run drags it along. The median ignores the extremes, so it is a good check on the mean. The range tells you how repeatable the car is. A small range is worth as much as a fast mean, because a car that is only sometimes fast loses knockout races.

Outliers

A value a long way from the rest is an outlier. We flag anything more than about 10% away from the median. Flagging is not the same as dropping. A run is only dropped when there is a physical reason, such as a slow trigger, the car touching the lane wall, or a cartridge misfire. Otherwise it stays in and the mean is honest.

Gradient is speed

speed over a segment = Δx / Δt This is the gradient of the distance–time graph between two gates.

A straight line means a steady speed. A line that curves upwards means the car is speeding up, and a line that flattens means it is slowing down. Our runs curve hard in the first 5 m, while the cartridge is pushing, and are nearly straight after that. That is exactly what LAB-05 predicts.

i

Why six runs? One run tells you what happened once. Three runs give you a rough mean. Six runs let you see the range and spot a bad launch without guessing. More than ten and you have used a lot of cartridges for a small gain in confidence.

04Timing-gate analyser

Type in your own run times or use ours. The analyser works out the summaries, flags a suspected outlier, and lets you leave it out to see the effect. Below that, edit the split times to rebuild the distance–time graph and read each segment’s gradient.

A. Six runs

RunTime (s)Speed (20 m ÷ t)Flag
sMean
sMedian
sRange
sBest run

B. Splits from the gates

Gates at 0, 5, 10, 15 and 20 m. Enter the time the car crossed each gate.

Gate (m)Time (s)Segment speed (m/s)

Segment speed = Δx / Δt, the gradient between that gate and the one before it.

05Worked example

Six runs: 1.212, 1.184, 1.239, 1.196, 1.620, 1.203 s.

Sorted: 1.184, 1.196, 1.203, 1.212, 1.239, 1.620

Median = (1.203 + 1.212) / 2 = 1.2075 s. The 1.620 s run is 34% above the median, so flag it.

Mean of 6 = 7.654 / 6 = 1.276 s. Mean of 5, with run 5 left out for the slow trigger, = 6.034 / 5 = 1.207 s

Range of 5 = 1.239 − 1.184 = 0.055 s

Report: 1.207 s, the mean of five runs, with one run left out for a slow launch trigger. The run that was left out still stays on the sheet.

Gradient between the 10 m and 15 m gates: t = 0.620 s and 0.850 s.

speed = Δx / Δt = 5 / (0.850 − 0.620) = 5 / 0.230 = 21.7 m/s

Compare that with the first segment: 5 / 0.390 = 12.8 m/s. The car’s average speed more than doubles between the first and third segments. The graph curves, then straightens out.

06Student questions Teacher mode is off. Answers are hidden.

Q1 · Summarise[3]

Five runs: 1.221, 1.198, 1.205, 1.231, 1.190 s. Find the mean, median and range.

Answer key

Sorted: 1.190, 1.198, 1.205, 1.221, 1.231

Mean = 6.045 / 5 = 1.209 s · Median = 1.205 s · Range = 0.041 s

Q2 · Judge[3]

A seventh run records 1.05 s, the fastest ever. The launcher operator says nothing unusual happened. Should it be left out? What should the team do next?

Answer key

No. There is no physical reason to leave it out, so it stays in the data (1). But one fast run is not evidence of a faster car. The team should repeat the run, at least three more times, to see whether it happens again (1). If it never happens again, it stays on the sheet as an unexplained result and the median is reported next to the mean (1).

Q3 · Gradient[3]

The car crosses the 15 m gate at 0.850 s and the 20 m gate at 1.078 s. Find the average speed over the last segment. Is the car still accelerating? How do you know?

Answer key

speed = 5 / (1.078 − 0.850) = 5 / 0.228 = 21.9 m/s

Only just. It was 21.7 m/s on the previous segment and is 21.9 m/s now (1). The graph is almost straight, so the car has stopped gaining speed. The thrust ended long ago and drag is balancing what is left (1).

Q4 · Explain[2]

Why does Team Nightmare report the mean of several runs rather than the best run when comparing two body designs?

Answer key

The best run is partly luck, from the launch and the cartridge. The mean averages that out, so the comparison is about the car and not about the day (1). A design is only called faster if its mean beats the other mean by more than the range of either set (1).