EXPERIMENTAL DATA
lab Session
Outline
Objective
Accuracy & Precision
Error Analysis
OBJECTIVE
Performing the experiment and collecting data is only the
beginning of the process of completing an experiment in
science
Understanding the results of any given experiment.
Presenting those results in a clear concise manner completes
the experiment by saying "This is what I did and this is what I
learned."
QUALITY
The quality of the experiment is considered through two important
aspects:
When you repeat the measurement
how close would new results be to the old,
how reproducible is the measurement? Scientists refer to this as the
precision of the measurement.
TYPES OF ERRORS
Errors may arise from three sources:
Gross blunders (very clear) in instruments. This type must be
eliminated by repairing the instrument or readjusting it.
A tape reading of 38.23m may be recorded as 38.32m in
the field book, The thermometer may be misread
Measurement may be made between the wrong
pegs..
Reading a tape incorrectly
Incorrectly marking
Wrong or inappropriate use of instruments,
Forgetting the reading or observations,
Unnecessary approximation.
Gross blunders
In practice there are variety of ways that can be employed to
reduced gross errors (Mikhail, 1974)
Taking multiple reading and checking for reasonable consistency
Careful checking of both pointing and recording
Using simple and quick technique for verification
Applying logic and common sense
Checking and verifying the performance of equipment, particularly
those with automatic readout
Repeating the experiment with perhaps slightly different technique
Increasing redundancy of the observation used in a model
TYPES OF ERRORS
Random errors: may be caused by personal or human errors, electronic fluctuations,
and influence of friction. Random errors follow a certain statistical distribution ((but not
always)).
Human errors (imperfection in vision and other senses especially touch) Negligible errors in
instruments Changing atmospheric conditions
Fixed errors: cause repeated readings to be in error by almost the same amount for
known or unknown reasons. They are called Systematic or Bias errors. Fixed errors may
be calculated. Thus, the readings can be corrected (correction factor). The measuring
tape expands in length with increase in temperature. It in neither possible to always
conduct the survey on the temperature the tape is designed for nor it is feasible to
invent a tape that is compatible in all temperatures.
Example: Measure temp. of hot gases in a furnace while the sensing element is exposed to
the flame radiation.
Example: A long thermometer used for measuring gas temp. in a tank. Heat losses from the
stem affect the reading.
Calibration:
Calibration is a comparison between the instrument readings and
a known standard and thus reduces the errors and improves the
accuracy.
The comparison is made with:
Primary Standard: chosen by Bureau of standards.
Secondary Standard: another instrument with better accuracy.
Known input source.
Frequency of calibration
Usually, instruments are factory-calibrated. However, frequent
Calibrations are required.
The frequency of calibration depends on :
The instrument itself
The environment of usage (materials, electronics, mechanical
linkages, dust, humidity, high temp., magnetic fields,..)
Transient measurement:
The transient effect is exponential as
Data analysis
Data analysis means understanding what your results mean. When
analyzing your data, try to think through the physical processes which are
occurring. Write your train of thought down. Ultimately, the goal is for you
to understand physics and the world a bit better. Your understanding of
your results probably occurs in stages, with each stage being a refinement
and possibly more mathematical than the previous stage
Data analysis should NOT be delayed until all of the data is recorded. Take
a low point, a high point and maybe a middle point, and do a quick
analysis and plot. This will help one avoid the problem of spending an
entire class collecting bad data because of a mistake in experimental
procedure or an equipment failure.
Error Analysis
Error Propagation