Lesson 5 of 12 · 20 min
Measuring g from experimental data
Determine the acceleration due to gravity from drop data by linearising, plotting with error bars and using minimum and maximum lines of best fit to find the gradient's uncertainty.
What you will be able to do
- Process repeated trials into a mean with an absolute uncertainty, and carry the uncertainty into a squared quantity.
- Linearise by plotting against , and find g from the gradient.
- Use error bars and minimum and maximum lines of best fit to find the uncertainty of a gradient and intercept.
- Compare an experimental value with the accepted value using percentage error.
The data book gives as the mean acceleration due to gravity on Earth. (It ranges from about at the equator to at the poles.) You can measure it yourself with a ball, a metre rule and a timer. The physics is one equation; the skill is in handling the data well enough to trust the answer.
The experiment. Release a ball from rest and time its fall through a measured height , using light gates, an electromagnet release with a timer, or slow-motion video. Repeat three times at each of at least five heights. The height is the independent variable (you choose it); the time is the dependent variable (you measure it), and it carries most of the uncertainty.
| Drop height (m) | Mean time (s) | () |
|---|---|---|
| 0.20 | 0.205 ± 0.005 | 0.0420 ± 0.002 |
| 0.40 | 0.287 ± 0.005 | 0.0824 ± 0.003 |
| 0.60 | 0.354 ± 0.006 | 0.1253 ± 0.004 |
| 0.80 | 0.407 ± 0.006 | 0.1656 ± 0.005 |
| 1.00 | 0.457 ± 0.007 | 0.2088 ± 0.006 |
Worked example
Processing repeated timing trials
A ball is dropped from rest through 0.60 m between two light gates. Three trials give fall times of 0.348 s, 0.355 s and 0.360 s.
(a) Calculate the mean time and its absolute uncertainty.
(b) Calculate and its absolute uncertainty, for plotting with an error bar.
Step 1: Mean of the trials
.
Step 2: Absolute uncertainty of the mean
.
This is the formula in the data book. Half the range covers the spread of the trials either side of the mean.
Step 3: Percentage uncertainty in t
.
Step 4: Square the mean and double the percentage uncertainty
. For a power, multiply the percentage uncertainty by the power: . So .
Squaring multiplies by itself, and for multiplication percentage uncertainties add: .
Try it yourself first, then check the working one step at a time.
Answer
(a) (b) , so the error bar runs from about to .
Method
Mean = sum of trials ÷ number of trials.
Absolute uncertainty = ± half the range.
For a squared quantity, double the percentage uncertainty, then convert back to an absolute uncertainty.
Watch out for
Squaring the absolute uncertainty (), which makes the error bar far too small.
Quoting the mean to more decimal places than the measurements.
Three trials at a drop height of 1.00 m give times of 0.450 s, 0.462 s and 0.459 s. What is the absolute uncertainty of the mean time, in seconds?
Why linearise? A ball dropped from rest obeys . A graph of against is a curve, and it is hard to read a constant off a curve or to see whether a point is out of line. Rearrange the equation so that it has the form of a straight line, :
Plot against . If the physics is right the points lie on a straight line through the origin, with gradient , so .
The least-squares line has gradient about and an intercept very close to zero, so . Drag a point to see how one bad trial pulls the line and changes g.
Key idea
How to linearise
- Write the physics equation linking the variables.
- Rearrange it into , putting the measured (dependent) quantity, or a function of it, on the left.
- Identify what and represent physically. Here and .
- Plot, draw the line of best fit, measure the gradient, and solve for the constant.
Error bars and the uncertainty of the gradient. Plot each point with a vertical error bar showing its uncertainty. Draw the line of best fit, then the maximum (steepest) and minimum (least steep) lines that still pass through every error bar. These usually join opposite corners of the first and last bars, but check the middle bars too. Then
The same lines give an uncertainty range for the intercept. Here the intercept should be zero; if its whole range excludes zero, suspect a systematic error, such as a timer that starts late.
Worked example
Determining g from a linearised graph
Students drop a ball from rest through heights from 0.20 m to 1.00 m and plot against with vertical error bars. The line of best fit has gradient . The steepest and least steep lines that pass through every error bar have gradients and .
(a) Show that the gradient of the graph equals .
(b) Determine and its uncertainty.
(c) Calculate the percentage error compared with the accepted value of , and comment.
against , with the line of best fit (solid) and the maximum and minimum lines (dashed).
Step 1: (a) Rearrange the equation of motion
Released from rest, , so becomes . Rearranging for the plotted quantity:
This has the form with and , so the gradient is .
Plotting t² rather than t turns the curve into a straight line through the origin, whose gradient you can measure.
Step 2: (b) Best value of g
. Keep the extra digits until the uncertainty is known.
Step 3: Uncertainty in the gradient
, a percentage uncertainty of .
Step 4: Uncertainty in g
Dividing 2 by keeps the same percentage uncertainty: . Rounding the uncertainty to one significant figure, .
Check with the extreme lines: and ; half their difference is .
Step 5: (c) Percentage error and comment
. The accepted value lies inside the range to , so the result agrees with it within uncertainty. The minimum and maximum lines also give intercepts between about and , a range that includes zero, so there is no evidence of a systematic timing error.
Try it yourself first, then check the working one step at a time.
Answer
(b) (c) percentage error ; the accepted value lies within the uncertainty range.
Method
Rearrange the physics equation into y = mx + c with the plotted variables.
Match the gradient (and intercept) to the physical constants.
Gradient uncertainty = ± half the difference of the max and min gradients.
Carry the percentage uncertainty through to g.
Compare with the accepted value using percentage error.
Watch out for
Writing . The gradient is , not , because is on the vertical axis.
Drawing the max and min lines through the first and last points only, without checking the middle error bars.
A trolley's velocity–time data are plotted with error bars. The line of best fit has gradient ; the maximum and minimum lines have gradients and . What is the absolute uncertainty in the trolley's acceleration?
A second method. Film the falling ball against a scale, or use a ticker timer, and find its velocity at several times. A velocity–time graph of free fall is a straight line whose gradient is itself, with the same minimum and maximum line treatment for its uncertainty.
An object falls from rest. Which graph of the measured data should be a straight line through the origin?
Common mistake
Read what the gradient really is
With on the vertical axis the gradient is , so . If you plot against instead, the gradient is and . Always write the rearranged equation next to your graph before you calculate.
Note
Uncertainty and error are different
Uncertainty (± a range) describes the spread of your measurements: precision. Percentage error compares your result with the accepted value: accuracy.
A result can be precise but inaccurate if there is a systematic error.
A group measures . Calculate the percentage error compared with , to two significant figures.
Summary
Mean ± half the range for repeated trials; double the percentage uncertainty when you square a quantity.
Linearise by rearranging into : for free fall from rest, , so plot against and use .
Maximum and minimum lines pass through every error bar; .
An intercept range that excludes the expected value points to a systematic error.
Uncertainty measures precision; percentage error against the accepted value measures accuracy.
Finished the lesson?
Try the 4 quick checks above and the lesson completes itself, or mark it complete now.