Experimental Skills

1
Experimental Skills

• In Stage 2 Chemistry
• Dr Simon Pyke
• RACI Conference December 2003

2
In this presentation

• Try to clarify some of the key ideas in the
  skills section Uncertainties (Errors)
• Illustrate some of these ideas with experimental
  data
• Go through the curriculum statement

3
Reasons for change

• Relationship between accuracy and precision.
• Precision
• Confusion about
• Precision of a single measurement
• Precision of a set of measurements
• Accuracy
• True, agreed or nominal value

4
Precision and accuracy
Key Ideas
Intended Outcome

• The accuracy of the result of an experiment is an
  indication of how close the result is to the true
  value, which is dependent on how well systematic
  errors are controlled.

• State which result of two or more experiments is
  more accurate, given the true value.

5
Precision and accuracy
Key Ideas
Intended Outcome

• Measurements are more precise when there is less
  scatter in the results, which is dependent on how
  well random errors are controlled.

• Determine which of two or more sets of
  measurements is more precise.

6
Precision and accuracy
Key Ideas
Intended Outcome

• The resolution of a measuring instrument is the
  smallest increment measurable by the measuring
  instrument.

• Select an instrument of appropriate resolution
  for a measurement.

7
Precision and Accuracy
Key Ideas
Intended Outcome

• The number of significant figures for a
  measurement is determined by the reproducibility
  of the measurement and the resolution of the
  measuring instrument.

• Record and use measurements to an appropriate
  number of significant figures

8
Accuracy

• The accuracy of a measurement (or a series of
  measurements) indicates its relation to the true
  value.
• The true value is the nominal, agreed or
  accepted value.
• Accuracy is effected significantly by systematic
  errors.

9
Precision

• One measurement
• Resolution of the measuring instrument
• True value not possible
• Several measurements
• Scatter of the data points

10
Precision

• The precision of a series of measurements is a
  measure of the agreement among the repetitive
  determinations
• Precision is associated with the random errors of
  the measurement process
• Quantified by statistical means e.g. standard
  deviation or range. It is about the spread of the
  measurements about the mean value

11
Precision

• High precision means low uncertainty in the
  measured value
• It is a measure of how well the result has been
  determined without reference to its true value
• Measure of the reproducibility of the result

12
Relationship Between Precision Accuracy
High precisionlow accuracy
Low precisionhigh accuracy (fluke)
High precisionhigh accuracy
Low precisionlow accuracy
13
Errors

• Random
• Scatter
• which influences precision
• Systematic
• Calibration of the instrument
• Accuracy in relation to true value

14
Random and Systematic Errors
Key Ideas
Intended Outcome

• Measurements are affected by random and/or
  systematic errors when measured values differ
  from the true value.

• Identify sources of random and/or systematic
  errors in an experiment.

15
Random and Systematic Errors
Key Ideas
Intended Outcome

• Increasing the number of samples minimizes the
  effects of random errors and increases the
  reliability of the data.

• Explain the importance of increasing the number
  of samples in an experiment.

16
Random and Systematic Errors
Key Ideas
Intended Outcome

• Repeating an experiment using fresh equipment and
  materials is a means of identifying systematic
  errors and verifying results.

• Explain the importance of repeating an
  experiment.

17
Precision and accuracy again
High precisionlow accuracy
Low precisionhigh accuracy (fluke)
High precisionhigh accuracy
Low precisionlow accuracy
18
Measuring the Boiling Point of Water
True value 100 degrees Celsius
19
Measurement of Speed of Light
20
Resolution

• The resolution of an instrument is the smallest
  increment measurable.
• E.g. 0.5 mm for metre rule, 0.01 seconds for
  stopwatch reading to nearest 100th of a second.
• The resolution of the measuring instrument can
  affect the precision of measurements but random
  errors also affect the precision.

21
Significant Figures
Key Ideas
Intended Outcome

• The number of significant figures for a
  measurement is determined by the reproducibility
  of the measurement and the resolution of the
  measuring instrument.

• Record and use measurements to an appropriate
  number of significant figures

22
Example Experiment
Time a cylinder rolling down an incline
Distance
23
Example data
The resolution of the stopwatch is 0.01 s but the
precision of the data does not match this.
24
Graph of time v. distance (1)
25
Relation of Scatter to Precision

• The scatter of the measured points about the line
  of best fit gives an indication of the precision
  of the experiment

26
Graph of time v. distance (2)
27
Presentation
Key Ideas
Intended Outcome

• Relationships between variables in an experiment
  can be shown by a line of best fit.

• Draw a line of best fit through a series of
  points on a graph such that the plotted points
  are scattered evenly above and below the line.

28
Interpretation and Evaluation
Key Ideas
Intended Outcome

• The scatter of the points above and below the
  line of best fit is probably due to random errors.

• Use the scatter in the graphs of data from
  similar experiments to compare the random errors
  in the experiments.

29
What are you really measuring?

• Real data sets of course do have errors
  associated with them.
• Experimental design should hopefully (!) minimise
  systematic error.
• Repetition and good experimental skills should
  minimise random errors.

30
The difficulty with real data

• A competitive protein binding assay

EC50 160 ?M 330 ?M 950 ?M
31
The true value problem

• Obtaining data with high precision is usually not
  the problem
• How do you know if your data is accurate ?
• Conceptually easy, but often technically
  difficult!

32
Use a different method!

• A direct protein binding assay

33
References

• Volker Thomsen, Precision and the Terminology of
  measurement. (The Physics Teacher Vol 35, Jan
  1997)
• PR Bevington DK Robinson, Data Reduction and
  Error Analysis.(McGraw Hill)
• Les Kirkup, Experimental Methods (Wiley)