Analyzing the results
After you complete your experiments, you have a very large task ahead of you. Analysis of data and interpretation of results is just as important as the experimental data that have been acquired.
When acquiring data, it is important to record all of your results. It is not always feasible to write all of your results in your notebook, and these records will need to stay with the lab. Use forms for tabulating results if necessary.
For computer output you should:
- record the location of these results by directory and file name
- create hard copies of any computer based results (such as simulation output or spreadsheet data)
- backup all of your softcopy data, preferably to optical storage media (such as CDs, DVDs, mini-drives and even online storage)
- organize and document your system of storage
As you review your results, be sure to avoid any personal bias in your analysis. Most research involving experimental work will need to have completed an uncertainty analysis to determine how certain your results are and where the most errors occurred. This will often require usage of statistical packages and knowledge of statistical analysis. Your mentor will typically want data to be analyzed according to a particular procedure. Similar to being objective when designing and conducting your research, you need to be objective in analyzing your results. Negative or unexpected results are often more important than having everything turn out as you expected. Much can be learned from mistakes!
Dealing with unexpected problems
One of the greatest challenges of research is dealing with unexpected problems. The best method to identify potential sources of error is to work backwards from the results, analyzing the inputs and outputs of each stage of the experiment. First, investigate how you interpreted the results. If this is where your mistake was made, finding this out now will require much less time and cost to make a correction compared to repeating the entire experiment again. You should also brainstorm with peers. This is an excellent way to jointly solve problems and should be done on a regular basis.
If you interpreted the results correctly, you should investigate the experiment that produced the results. Are there elements that can be improved? Are you using the correct equipment, and is it capable of performing the required task? Was there a step that you missed? Did you not account for all of the variables?
If everything to this point has checked out, it would be worth repeating the experiment, following the same exact steps as you recorded in your notebook. If the results are repeated, the next step will be to review the design of your experiment. Have you interpreted everything from your hypothesis correctly? Bring in colleagues to review your work and get their input. There is a chance that you have overlooked something, and a fresh pair of eyes may often see a problem that you have not been able to identify.
This cycle of refining your research and repeating experiments may occur many times. There have even been times that, after several attempts, a project may not yield successful results. When this occurs, the most you can do is learn from the experience and improve your abilities for future research. When research is successful, however, there are several ways in which you can present your hard work to others. This is discussed in more detail in Module 5.