Testing your hypothesis

Flow Chart - Design

Now that you have created a sound hypothesis, the next step is to test it.  This requires careful planning.  How can you account for all potential variables?  Do you have the proper instruments to measure your work?  Will any of your instruments require calibration?  What output do you need to collect?  Are there other variables or output that you will want to collect data for that you may not need immediately, but may use for other studies?  These all need to be considered before you begin your experiment.

It is extremely important to be objective in conducting an experiment.  As part of developing your hypothesis, you have predicted the results you expect to see.  There can often be pressures, such as deadlines for papers or reports, which can influence your judgment.  Reporting incorrect results is unethical, however, and can have dire consequences on not only your career, but also those with whom and for whom you are working.  Results contrary to what you expected are not a bad thing, and can often be quite beneficial.  They force you to consider everything from the original hypothesis to the laboratory technique and methods of data analysis to explain the outcome.

Another important aspect of research is recording everything you do.  Use your research notebook by describing your setup and every step of the experiment.  Write your notes so that you will know exactly what you did and what occurred throughout the experiment.  If there are any problems, or you want to improve your procedure, this will allow you to diagnose issues by having the information organized in one place.  As noted earlier, it is important that someone else is able to replicate your work.

Watch out for any indication of errors in your results and anomalies in your data (such as extreme values and results) that are inconsistent with established theories, and be sure to record them.  These deviations can include:

  • personal errors from subjective interpretation
  • mistakes from incorrectly recording of data
  • instrument errors from uncertainty in the instrumentation or mis-calibration
  • nonsystematic errors which arise from uncontrolled (and unmeasured) variables

The ability to detect these errors is a desirable trait to have as a professional in any field as it reflects your breadth of knowledge and expertise.  It is a skill that requires a strong understanding of the concepts related to your research and can take a long time to acquire. Progressing slowly is normal for beginners, so keep working on it and you will definitely reap the benefits!

Previous | Next

Return to Index