11.2.3 Reducing the Impact of Errors

• Random errors can be reduced by repeated trials and measurements
• Multiple measurements when averaged will reduce the impact of a random error on the average
  • The more readings you have the lower the possibility that a random error will skew the results
• If you spot a random error in a data table then you can omit it in the calculation of an average
• For example, in a titration you can leave out results that are not concordant when finding the average titre:

Calculating the average volume delivered in a titration should not include non-concordant volumes. Run 3 (and the Rough run) is omitted from the calculation of the average volume delivered

• Systematic errors cannot be reduced by repetition
• Systematic errors can only be reduced by changing the procedure and making sure you are using the instruments correctly
• If you cannot actively reduce systematic errors you must still try to identify them and comment on them in your evaluation

• A skill that is very important in data processing is the ability to discuss the impact of different types of errors on an experimental conclusion
• This is an integral part of the Internal Assessment, but it can also be examined in the written exam papers
• You should always evaluate random errors and systematic errors in an investigation
• This includes assessing the relative impact of errors, for example:
  • Whether a particular error has a major or minor effect on the final result
  • Which errors produce the largest impact on a final result

You should be able to state what the impact would be of not using a draught shield in a simple combustion calorimetry experiment

• Accuracy is how close you are to an accepted value
• Precision is a measure of how many decimal places you can express your results to
• Imagine you are shooting at a target: the following results show the difference between these concepts

Accuracy and precision in target shooting

• In practical chemistry terms, if you have a literature value for a final calculation, then it is very easy to compare how close you got to the literature value, in other words how accurate you were
  • For example in enthalpy of combustion experiments, did you get close to the Data Book value?

• Sometimes you can control precision by changing instrument
  • For example if you change from a two decimal place to three decimal place balance, you are making your measurements more precise

Answer

The correct option is C.

Random errors can be reduced by repetition. All the other procedures would only affect systematic errors.