CAREER COLUMN

What not to do in graduate school

Six limiting maxims PhD students should avoid.
Buddini Karawdeniya is a postdoctoral fellow at the Kim lab in the Lyle School of Engineering at Southern Methodist University, Dallas, Texas.

Search for this author in:

Adapted from Getty

During my time as a graduate student researching analytical sensors in the Dwyer laboratory at The University of Rhode Island in South Kingstown, I made a lot of mistakes — some of which matured into valuable lessons. If you are already in graduate school, or have decided to start, here are six things I recommend you do not do.

Compare yourself with others

I’ve met many scientists who spiral into stress and disappointment because they compare themselves unfavourably with others. Every research field, project and graduate student is unique. In some fields, it can take years to find a breakthrough worth publishing; in others, it’s easier to publish frequently. I became worried by the third year of my PhD, when it seemed as if it was taking me longer than others to publish my research project. It took me almost six years to complete my PhD, but my hard work paid off when I published a piece on my flagship project in Nature Communications, alongside almost a dozen other publications and two patent applications from various other projects. Instead of looking at what others are doing, learn to be introspective. Grow from your mistakes, and find more efficient and effective work tactics.

Blindly trust your data

I have learnt to be suspicious of my data. Consider what could go wrong when obtaining them — if something seems weird or wrong in some way, it probably is. I was once designing a sensor that would detect minuscule amounts of chemicals with a laser. One day, thrillingly, the signal looked fabulous: the laser power was turned all the way to the highest setting instead of my usual setting; and the higher the laser power, the higher the signal. Although it looked great, it turned out that the equipment was enormously overestimating the sensing performance and was therefore producing useless data. Be aware of issues such as sample contamination, labelling errors or faulty instrument calibrations. Just because you yourself obtained the data, do not blindly trust them.

Suffer alone

Graduate school is no easy gig. Failure is regular and often stings. Seek help and advice from those with more experience, such as senior colleagues, postdocs or your adviser. I’ve struggled with problems that I could see no way out of. For example, when I switched suppliers for a naturally derived polysaccharide, I witnessed unexpected results. I was lost after days of measurements: how could the same polysaccharide give me different signals? When I approached my adviser, he suggested a technique I hadn’t considered, and it helped me to uncover a difference in the composition of the polysaccharides that was behind the inconsistent signals.

Believe that more work is always better

We all get excited about where our latest data are taking us, and it’s not uncommon to work a few night shifts or weekends to get all of the data you need. Once, a colleague and I were unable to determine whether our nanosensor was working correctly, even after we had desperately tried many instrumental methods. In despair, we temporarily abandoned the mission and took a break. Rest brought clarity: after our break, we had the energy we needed to find an alternative solution to our problem. When working on something non-stop for long periods of time, we can develop tunnel vision or burn out. Take a break and start afresh when necessary.

Grow your records organically

Like most graduate students, I kept records in lab notebooks. This works for a few days, but not for an entire PhD project (especially when you are expected to handle multiple projects simultaneously). Keep clear records and maintain Excel sheets or Word documents with experimental logs. Build an appendix with links to the folders in which you’ve saved data, analysis and figures that you’re about to publish. By the time I completed my PhD, I had 19 lab notebooks, and I would have been doomed without the help of Excel files, appendices and indexes. Find ways to protect and keep track of your data from the beginning — don’t wait for a haystack to form around your needle.

Get stuck after one failure

It’s not easy to ditch a theory or an experiment you worked hard on. But knowing when to change direction is pivotal to success during a PhD. The first project I worked on as a graduate student was unsuccessful, despite 18 months of devotion. After this perceived failure, I was starting to feel disappointed. One day after a great discussion with my adviser, I moved on to a different project. After I worked out how to ‘fail fast’, I was able to dedicate my time to projects that brought me great success and eventually led to several interesting publications, research awards and fellowships.

Nature 572, 553-554 (2019)

doi: 10.1038/d41586-019-02255-7

This is an article from the Nature Careers Community, a place for Nature readers to share their professional experiences and advice. Guest posts are encouraged. You can get in touch with the editor at naturecareerseditor@nature.com.

Nature Briefing

An essential round-up of science news, opinion and analysis, delivered to your inbox every weekday.