The ISA for Physics

1
The ISA for Physics

• What you need to revise

2
Know your experiment

• Know the type of equipment you used and why you
  chose it .
• Know the name of each piece of the equipment and
  what it does (measures - if applicable).
• Know the sensitivity of the equipment that
  should have been evident in the way you wrote
  results in the table.
• Know why you chose the number of readings to take.

3
Terminology

• Make sure you KNOW all of the words to describe
  variables etc.
• Remember it is the independent variable that you
  change by a regular amount (and usually plot on
  the x-axis)
• Remember it is the dependent variable that you
  measure as it changes in response to the changes
  in the independent variable (usually plotted on
  the y-axis).
• The control variables are kept constant to make
  the results valid.

4
Why are control variables kept constant?

• If that variable had changed as well as the
  independent variable the dependent variable would
  have responded to that change too making the
  experiment invalid. Do not simply say to make
  it a fair test if asked to explain why control
  variables are kept constant expand on the idea
  and say it would affect the dependent variable.

5

• If you are asked what you were trying to find
  out or the purpose of your investigation.
  they want to know what you decided to
  investigate.
• They want you to describe how you decided on
  varying your independent variable and observing
  the effect this has on the dependent variable
• They want a brief description of the fair test
  you carried out

6
Minimising errors

• A good experimenter checks the equipment is
  reliable before beginning she looks for zero
  error on instruments and calibration errors and
  uses instruments of a suitable precision.

7
Minimising errors

• Zero errors can sometimes be adjusted manually
  otherwise they can be noted and deducted from all
  readings. A column of actual readings should be
  recorded and then a column of readings corrected
  for zero error should be drawn.

8
Minimising errors

• Calibration errors are a bigger problem.
• They can mean that all of the results are out by
  a percentage because two fixed points have not
  been fixed correctly.
• You can spot drastic calibration errors by
  checking meters against each other but you
  dont know the whole batch may be wrong!
• If you were doing an important research project
  you would check the meters against standard
  resistors etc. to see whether they were properly
  calibrated.

9
Minimising errors

• Using instruments with sensitive scales (small
  divisions more significant figures) give
  results with more precision.
• This will not make your results more reliable or
  accurate (valid) just more precise.

10
Minimising errors

• Any possible causes of errors spotted while
  carrying out the experiment (sparking, heating,
  fluctuations in the meters) are noted.
• A repeat set of readings is taken to check the
  reliability of the first set any that differ
  widely on repeating must be checked again.

11
How can you tell if your readings are reliable?

• When repeated the results are virtually the same
  giving you an average that barely differs from
  either set.
• When plotted a smooth curve or straight line is
  obtained no obvious anomalies they make a
  pattern.

12
Is reliable the same as accurate?

• NO!!!
• If there is a problem with the accuracy of your
  measuring instruments and/or an error in how you
  are using the equipment you will get repeatable
  data that is wrong.

13
Conclusion

• State the findings in terms of how the
  independent variable affected the dependent one.
• Refer to the graph shape to describe the
  relationship and make a conclusion from that.
• Does it go through the origin?
• Is it the same in negative bias as forward bias?
• Does it indicate direct proportionality?
• If you can use numbers do so!

14
Evaluation

• Are your results valid for ALL components or only
  true for the one you did?
• In order to make a more general conclusion how
  many others would you have to test?
• Would you expect all of the results to be
  identical?
• What tolerance would you allow?