Interesting post, food for thought. I have just started my PhD in grassland restoration and seeking for methods to use in my research particularly in microbial and plant communities. I agree with you that the simpler the best, of course with certain nuances. However, to be successful and publish own research data we are forced to use novel methods. Moreover, we reference last 10 years published papers and majority of them use new methods. Therefore, for proper citation need to follow how they do, specially for young researchers.
P.S. I like your articles and find them useful for myself.
Great post. When designing experiments we always ask "What else can we take away?" There's a point where the answer is 'nothing,' then we stop designing and start doing. Why? It's cheaper, more efficient, and generally easier. The same knowledge is generated.
I agree 100%. Just like our numerical models, our experimental methods should be as simple as possible to answer a question. I do like exploring, and when needed, developing new methods, and new surprises await, but that’s not an excuse to leave hypotheses behind.
We are currently developing a survey to assess soil health through a simple citizen science approach, mid-tier agronomist methods, and high-tier research methodologies. Our objective is to explore potential correlations among these approaches, and the anticipated findings are expected to be quite intriguing.
Interesting post, food for thought. I have just started my PhD in grassland restoration and seeking for methods to use in my research particularly in microbial and plant communities. I agree with you that the simpler the best, of course with certain nuances. However, to be successful and publish own research data we are forced to use novel methods. Moreover, we reference last 10 years published papers and majority of them use new methods. Therefore, for proper citation need to follow how they do, specially for young researchers.
P.S. I like your articles and find them useful for myself.
Thanks for writing!
Yes, this is exactly my point: that it creates this sort of increasing pressure in the practice of science, including in publishing.
Great post. When designing experiments we always ask "What else can we take away?" There's a point where the answer is 'nothing,' then we stop designing and start doing. Why? It's cheaper, more efficient, and generally easier. The same knowledge is generated.
Thanks for writing. That is really a great approach, love it!!
Very often there is the opposite tendency, like, what else can we cram into this experiment....
I agree 100%. Just like our numerical models, our experimental methods should be as simple as possible to answer a question. I do like exploring, and when needed, developing new methods, and new surprises await, but that’s not an excuse to leave hypotheses behind.
My favorite experiment is the Kerr et al (2002) microbial rock-paper-scissors. Elegant and simple. https://pubmed.ncbi.nlm.nih.gov/12110887/
Yes! I know this paper, really like it as well!
Agree that this same point also pertains to other aspects of research, not just experimental design.
We are currently developing a survey to assess soil health through a simple citizen science approach, mid-tier agronomist methods, and high-tier research methodologies. Our objective is to explore potential correlations among these approaches, and the anticipated findings are expected to be quite intriguing.
Very cool! Will be very interesting to see what you find.
I really like thinking on simplicity/ bare minimum. Something like "Occam's razor" for methods.
And in addition, two examples which might have been done 100 years ago:
- Decomposition study using tea bags
- Soil your Undies: measuring soil health with underwear
Yes, absolutely!
Soil your undies is brilliant on so many levels...
Thanks for sharing your view. Now, who will set the standards of bare minimum method? What do you think?
Thanks. We definitely don't need any standards. This is just food for thought in terms of experimental design.