Research is Not Random

Research is Not Random

Many undergraduate students feel it when doing summer research or beginning their graduate studies; they just don’t know how to express it. Many students are uncertain of the procedure and methodology to follow in order to perform efficient research.

Numerous students may mistake presumptions for actual facts or intuition for experimental steps. This is probably because undergraduate students in Engineering rarely have the opportunity to be exposed to the fundamentals of scientific methodology. Thus, I really believe that a one credit course focusing on the do’s and don’t’s of research would be more than welcomed by undergraduates. What is really needed is a course that would not focus on a specific project, but rather on how research is done. The course could address simple, yet easily forgotten concepts such as detailed planning before experiments, finding sources of error and accounting for noise, recording several data points and following the proper method for analysis.

This may sound like another theoretical course, but it need not be! This optional undergraduate level course could focus on addressing questions most commonly raised by newly admitted graduate students. The course could make use of an existing project and reproduce it while keeping the attention on the “box” itself rather than the content. After partaking in several research projects, I have come up with some tips that are essential to obtain experimental results.

What are the steps to follow when planning an experiment?
It is essential to know beforehand each step of the experiment and predict the potential outcome, i.e. hypothesis. In order to be efficient, a student should know what to expect from performing a certain experiment. It could happen that the hypothesis is completely contradictory to the results but then the outcome can be compared with the initial idea and new experiments can be designed from it. Here, it is important to note that reading papers from the field can be extremely helpful, especially to predict experiment outcomes and even to formulate new experiment ideas. Also, a large part of research involves around optimizing concepts that have already been developed, therefore all the information a student needs is usually only few clicks away!

What data should be recorded and in what format should it be organized?
This question is crucial, especially when working on an experimentally-oriented project. A fact that is usually forgotten is to take pictures of experimental set ups or of any part of a developing device or prototype. These are strong proofs of the system’s functioning and it also gives greater insight into the actual project. As for data recording, almost all data must be recorded. Discrimination of data should be avoided. For certain types of research, it is critical to re-conduct a given experiment multiple times to ensure its validity and reproducibility. Recorded data can be organized either using tables for large amount of data points and multiple variables or by simply writing them in a notebook for simpler experiments! This may sound obvious, but it could be confusing to record data, especially when conducted several experiments at once. Also, if some details have not been recorded, then it could require a student to perform the entire experiment all over again.

How to analyse the results? Is a given graph or a picture appropriate?
Getting interesting looking results is fun, but what’s next? Analyzing data is basically asserting some pre-defined conclusion or formulating a new one. This is the most important part since this is where a student explains the significance of his results and its impact toward the final outcome of the project. Graphs are very useful but they can contain a large amount of information. So it is important to indicate errors by using indices or error bars. The same is true for pictures. They can be very beautiful at times, but it is important to ensure that they prove a point. Most importantly, scale bars or any other indication of size must be included. When analysing data, appropriate formulas must be used depending on the type of experiment. This is slightly complicated but it becomes intuition with practice and reading

Like what you see?

We release awesome content every Wednesday.
Stay updated; signup to our mailing list here:

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>