Paper 3
Key Terms
error - the difference between measured and the true value
uncertainty - the interval which the true value can lie in
accuracy - how close the measured/calculated value to the true value
precision - how consistence the values are, obtained from repeat measurements
Why result might be precise but not accurate?
there might be systematic error in the measuring instruments
so values are consistent when taken repeats, but not close to the actual value
Uncertainty
Calculating percentage uncertainty
if given uncertainty, just use this divide by the measured value
if given resolution, divide by 2 to get the uncertainty
for set of data and need to use the mean, percentage uncertainty = 100*(½ range/ mean)
Measuring instruments (resolution), high resolution so has low uncertainty
micrometer - 0.01mm
digital/vernier calliper - 0.01/0.1mm
if ask which one to use, almost always say micrometer
Why is there uncertainty?
the way measurements is made
reaction time
parallax error
reduce uncertainty for diffraction grating experiment
measure from nth order on one side to nth order the other side
use a larger grating to screen distance
use vernier calliper to record distance
use a grating with more lines per mm
Criticise
Criticising result table
data not recorded to the same d.p/s.f.
data should be recorded to more d.p/s.f.
not enough data recorded
repeats and mean not taken
the readings of independent variables are not in the same interval
position of the mass holder is not recorded for oscillating spring experiment
When asked to criticise a graph of log, mention
whether the graph is supposed to follow the observed trend
usually yes, if no, could be because of passing through/not passing through origin
measuring instruments not in high resolutions, so there might be systematic error, affecting the data recorded
see if there’s extrapolation of graph
usually more data points are needed to give a valid conclusion
graph where line supposed to go through the origin, if doesn’t there’s systematic error(e.g. zero error, parallax error); if line of best fits does not cover all/most points, then there’s random error
Improve Experiments
how to improve experiment of oscillating springs also mention
ruler may move during oscillations
move ruler closer to the spring
it is difficult to measure from the same position of the spring each time of the spring, so measure the position after adding mass and when mass is removed
In experiment of oscillating springs, use multiple oscillations and taking a mean, increasing time so reduces uncertainty
use set square to ensure spring is parallel, use fiducial marker to determine the equilibrium position
how to ensure accurate measurements?
measure from different places/orientations
repeats then take a mean reduces random error also always mention reduce percentage uncertainty
when taking measurements, if object very thin, take stacks of them and calculate a mean, this reduces percentage uncertainty
relationship between volume/length vs temperature
remember to stir water
parallax error when reading length/temperature
digital thermometer
let tube reach thermal equilibrium before taking reading because temperature of air otherwise may not be the same as temperature of the water because time is taken for the glass tube to reach the same temperature
Using other apparatus instead of standard ones
Using a voltmeter instead of oscilloscope
a.c. voltmeter allows reading to be taken directly
Gives a better resolution than measuring trace height on an oscilloscope
but voltmeter may draw current and affect the circuit it connects to
accuracy would depends on the calibration of the voltmeter
Using video instead of stopwatch
can use slow motion, so use when time is very short, can view motion more slowly
can record motion, so easy to judge when an oscillation is complete
why use a data logger with light gates instead of stopwatch?
Stopwatch - reaction time
so stopwatch has an uncertainty bigger than data logger
but repeats timings/taking a mean using a stopwatch can reduce uncertainty
light gate is difficult to set up in an oscillating spring system
datalogger used when need to record for a long period of time
higher sample rate
Other things to be aware of
when explain experiments involving recording sets of data, need to mention at least 5 sets
When using equations to explain, remember to define symbols, e.g. phi = work function, or just state in words, and in question with rearrange to use log and plot graph, always remember to write explicitly compare with y=mx +c
always always always mention reduce random error when taking mean!!!