calculations with your data, determine your experimental uncertainties, and show
mathematical fits on relevant plots.
When you show your calculations, you must explain what you are doing in words and embed
the equations into your sentences. You must also explain why you are performing a particular
calculation.
If you used a plot or graph to help you analyze your data, you must show it and describe it in
this section. You may have used either LoggerPro or Excel to fit a line to your data, and the
results of that fit are considered “analysis”, rather than “raw data”, because you made a
conscious decision to model your data in a certain way. If several plots of the same type are
used (for example, for multiple trials of an experiment) then show one plot as an example.
The final thing that belongs in the data analysis section is your error analysis, i.e., calculations of
your experimental uncertainties. You should have listed all measurement uncertainties in your
methods section, and in your data analysis section you propagate these errors into
uncertainties in your final result. You should conclude the error analysis section by calculating
your percent error, and (if applicable) a percent difference from the known value of the
quantity you are trying to measure. A detailed document explaining how to calculate your
errors is available, entitled “Understanding Experimental Error”. This document is available on
Canvas in the Lab folder.
3.4 Discussion
The typical length of the discussion section is 1 – 2 pages. This is the section where you
interpret your results, answer your initial scientific question, and discuss everything else
relevant to understanding your experiment. You may want to consider the following questions
as you write your discussion section:
• Is your data sufficient to answer your scientific question?
• Are there any other interpretations of your data?
• What sources of bias were you unable to remove from your data?
You also need to discuss the sources of uncertainty in your data. Regardless of whether you got
a very good result, or whether your result has something obviously wrong with it, there are
always sources of error in your experiment. When considering sources of error, you may refer
back to the handout from lab class about different types of experimental error. When writing
this section, you may want to consider the following questions:
• Are your uncertainties small enough to answer your scientific question?
• Of the several different sources of error in your experiment, which do you think is the
largest or most important?
• How might future experiments minimize these sources of uncertainty?
• Is your experiment biased in any way? If so, how would you improve this in future
experiments?