INTRODUCTION TO ERRORS AND EXPERIMENTAL PRACTICES

A Guide to Sound Scientific Methodology

GROUP : 28
INTRODUCTION TO RESEARECH DIBYENDU NASKAR.
AND ENTREPRENEURSHIP ROLL NO- A25MM09002
DATE: 11/09/2025 PHD SMMME
1
 The Importance of Errors
• Why We Estimate Errors
• Measurements are never exact. We must
indicate the precision or reliability.
• An error estimate shows how close a result is
likely to be to the true value, like f=(256±2)
mm.
• Without error estimates, we cannot draw
significant conclusions from experimental
results.
• Example: Is a difference in resistance from
200.025Ω to 200.034Ω significant? It
depends entirely on the error margin.
• The purpose of an experiment often
determines the acceptable error.
AI GENERATED IMAGE 2
 Types of Errors (Random Errors)
• Vary and are equally likely to be
positive or negative.
• Cause a spread of readings around
the true value.
• Examples: Errors in
starting/stopping a stopclock,
irregularities in motion.
• Can be estimated by statistical
methods and reduced by repeating
measurements.

AI GENERATED IMAGE 3
Types of Errors (Systematic Errors)
• Constant throughout a set of
readings.
• Cause readings to be consistently
displaced from the true value.
• Examples: Inaccurate apparatus,
ignored correction factors (e.g.,
thermal EMFs).
• Dangerous: Repeated
measurements don't reveal them.

AI GENERATED IMAGE 4
 Accuracy vs. Precision
•Accuracy: How close a measurement is to
the true value. A result is accurate if it has little
systematic error.
•Precision: How close repeated
measurements are to each other. A result is
precise if the random error is small.
•Why it matters: A result can be very
precise (low random error) but highly
inaccurate due to a hidden
systematic error. This is why systematic
errors can be more dangerous.

AI GENERATED IMAGE 5
 Systematic Errors - A Cautionary Tale

The Millikan's Oil-Drop Experiment

• Millikan's measurement of the elementary
charge (e) was affected by a systematic error.
• The value he used for the viscosity of air was too
low.
• Resulting value: e=(1.591±0.002)×10−19 C
• Present value:
e=(1.602 176 46±0.000 000 06)×10−19 C
• This single systematic error affected the values of
several other atomic constants for decades.

AI GENERATED IMAGE 6
 The Preliminary Experiment
• The Trial Run
• A crucial step for any real experiment.
• Purpose:
• To "learn" the experiment's techniques and
routines.
• To check that all apparatus is working properly.
• To find a suitable range for each variable.
• To estimate initial errors, which helps plan the
main experiment.
• This step is about planning and is more valuable
than a lot of theoretical pre-work.
• It should be flexible and modified as the
experiment progresses.
AI GENERATED IMAGE 7
 Checking the Obvious
• "Horse-Sense" in the Lab
• Before starting, always check the basics:
• Is the apparatus mechanically firm?
(Three points define a plane!)
• Is the apparatus level?
• Are all optical surfaces clean?
• Are optical components properly
aligned?
• These are simple, common-sense checks
that can prevent major issues.
• A warning on cleaning: Use a soft camel-
hair brush for expensive, coated lenses,
not a cloth.
8
AI GENERATED IMAGE
 Personal Errors
• The Experimenter as an Instrument
• Recognize that you, the experimenter, are also a source of error.
• Common personal errors:
• Wishful errors: Biasing results
towards an expected outcome.
• Unconscious repetition: Repeating
a previous reading, even if it was a
mistake.
• Best practices to reduce these:
• Ensure physical and mental comfort.
• Arrange apparatus and instruments
for easy access and reading.
• Maintain good lighting and ventilation.
• Work out results as you go to catch
mistakes early. AI GENERATED IMAGE 9
 The Importance of Repetition
• Repeat, Repeat, Repeat... with a Purpose
• Repeating measurements of a single
quantity helps:
• Avoid mistakes in reading and
recording.
• Provide a means of estimating random
error.
• A common mistake: A student gets two
wildly different readings (e.g., 56∘30′ and
60∘12′).
• The solution: Do not guess which is right
or average them. The only sensible action
is to make more measurements until the
results start to make sense.
AI GENERATED IMAGE 10
 Working Out Results
• Don't Wait Until the End
• It's bad practice to wait until the
end of an experiment to work out
results.
• Working out results as you go
along allows you to:
• Catch mistakes while the
process is fresh in your mind.
• Discover if a change needs to
be made to the apparatus.
• Determine the next steps in
the experiment.

AI GENERATED IMAGE 11
Conclusion

• Errors are an integral part of science, not a failure.

• Distinguish between systematic and random errors and take steps to eliminate
the former.
• Begin every experiment with a well-thought-out preliminary run.
• Always check the obvious before you start making measurements.
• Be aware of your own personal biases and errors.
• Work out results as you go to ensure a smooth, sensible, and successful
experiment.

12
Thank You

13