Physics 6 / Virtual Lab / Measurement and uncertainty

Making a measurement is really a two-step process. First, you have to correctly read the tool that you are using.
Second, you have to decide how much uncertainty to ascribe to that reading. The latter part can be complicated
because there are various sources of uncertainty that may creep into a measurement.

Let’s discuss reading the tool first. When it comes to measuring time interval and inertial mass, we shall use
familiar digital devices: stopwatches and scales. These are pretty straightforward, but one always has to
remember to zero these tools. Zeroing a stopwatch means setting the initial time to 0.00, and zeroing a scale
means setting it to 0.0 so that the tare weight is not included. As for measuring length, we often use an analog
ruler:

Once the tool has been read, one must estimate or calculate the uncertainty.

The tool contributes uncertainty because of the coarseness of its calibration or scale. For example, a standard
metric ruler has an uncertainty of about ±1 mm because that is the smallest interval printed on the ruler and the
human eye cannot generally differentiate tenths of a millimeter.

Sometimes the procedure, or act of measuring, introduces more uncertainty. Think about whether the nature of
the technique tends to make the reading of the tool more difficult. For example, it is much harder to read the
length of a wriggling baby than a stick which does not move. So even if the baby and stick turn out to have
equal readings, one might assign the baby's length a larger uncertainty.

Lastly, sometimes the quantity itself is inherently ambiguous so one should think about the range of values that
the measurement might lie in, rather than a specific value. For example, a baby's weight fluctuates in a matter of
hours due to its feeding and going to the bathroom.
1. For each of the following objects, estimate the greatest uncertainty in a single measurement of the quantity
with the tool and procedure indicated. Use the most convenient units. Explain where this uncertainty comes
from.

object tools procedure greatest uncertainty

-------------------------------------------------------------------------------------------------------------------------------
paper width ruler lay ruler along edge

person’s height meter sticks stack 2 meter sticks and look for
top of head

person’s weight scale weigh at noon

time for a ball to stopwatch start watch when ball leaves

fall from dinner table, stop when it hits floor
table to floor
2. If one repeatedly measures a quantity under identical conditions, then the uncertainty may be calculated
rather than estimated. Here is some data for the time that it takes a ball to fall 2 meters from an initial state of
rest: ​0.55, 0.66, 0.63, 0.74 s 

a) Calculate the best estimate by taking the arithmetic average of the data,

best estimate of ​Δt ​=

b) Calculate the uncertainty from the expression (high – low)/2, a prescription that we shall frequently
refer to as the “hi-lo method.”

uncertainty of ​Δt ​=

c) Explain why the hi-lo method makes sense. It often gives an overestimate of the uncertainty compared to
more statistically rigorous measures. Why do you think that is?

d) Percent uncertainty is the ratio of uncertainty to the best estimated value, expressed as a percentage. It is
helpful for quickly characterizing how precise your measurement is—a low percent uncertainty corresponding
to a more precise estimate.

percent uncertainty of ​Δt​ =

e) The average velocity of the ball may be computed from v​av​ ​= Δx/Δt.​ Assume that the displacement is exactly
Δx = ​2 m. Compute the best estimate for the average velocity.

best estimate of v​av​ =

f) Now devise a procedure to compute the uncertainty of the average velocity using the hi-lo method as a
guiding principle.

uncertainty of v​av​ =

g) Confirm that the percent uncertainty of the average velocity equals the percent uncertainty of the time
interval.

percent uncertainty of v​av​ =