Tips for designing experiments

One guarantee of being a scientist is that you perform many experiments in your career that do not work. I always warn trainees just joining my lab that this will happen and that they should expect it. This is an issue that I discuss in particular with undergraduate students because often the only lab experience they have is through undergrad labs run as part of a course and those experiments are designed to work and have already undergone extensive troubleshooting. I make sure to tell my students that they will run many failed experiments and that this is a normal part of doing science and is a cornerstone of the scientific method. I tell them that so far in my career I’ve only had a few experiments that worked out perfectly the first time and that a failed experiment can happen for reasons other than their abilities or talent for doing science.

That being said, there are things that you can do to decrease the chances that an experiment will fail right out of the gate. I offer some tips below:

1) The first thing that I suggest to students is that they do extensive reading of the literature and established protocols related to their experiment before starting to design it. I’ve heard the phase “one hour in the library can save you one month in the lab” and I absolutely believe it! It’s really important to understand the rationale behind a particular protocol and the nuts and bolts of why you are doing each step. In the days of commercial kits I think that many people forget this crucial step and it often causes issues later.

2) Make sure that you are including all reasonably possible positive and negative controls as part of your experiment. From talking with several of my colleagues recently it has become clear that many undergraduate and graduate students have not had explicit training in how to determine what the appropriate controls should be for an experiment or are simply not including them. By including controls in your experiment you allow yourself the capability of narrowing down where problems cropped up in your experiments. When an experiment fails, this step can save you a massive amount of time when it comes to troubleshooting and determining what went wrong. The presence of control and experimental groups also ensures that you will be able to conduct statistical analyses of your data in an attempt to demonstrate whether your results are significant.

3) Write up an extremely detailed step by step protocol for your experiment. Try to think about what might go wrong and where key steps are in the protocol. Attempt to troubleshoot the experiment before you even do it. The plans for your experiment should be written in your lab notebook and not on paper towels, scrap pieces of paper, etc. This will ensure that your experiment will be reproducible and will help you to identify potential issues before you get rolling. If someone in the lab has done the experiment or protocol before, go and talk to them. They may have tips or tricks that are not explicitly written down that are valuable. Write out the protocol in your own words with as much detail as you can. I tell my students that if they needed to perform the experiment without thinking about the steps, the protocol should be detailed enough that they could do this.

4) Ensure that all of the materials and reagents that you need for your experiment are available and ready to go before you start the experiment. There is nothing worse than getting part way through a long protocol only to realize that you’ve run out of Tris buffer and have to order more in from the supplier. Complete any prior steps that are needed before starting the experiment. Do you need to culture cells, wrangle critters, grow plants, etc.?

5) Conduct a small trial run of your experiment. Starting things off with a pilot experiment allows you to save money, time, and can allow you to discover problems with the design of your experiment before you fully commit large amounts of resources to it.

Designing good experiments is an art form that requires years of practice in order to get better at it. I am still working on designing the perfect experiment, but I have certainly improved this skill by extensive practice over the years. Due to the challenge of designing an effective experiment it is truly amazing when an experiment works beautifully on the first attempt. This is one of the eureka moments that scientists live for!


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s