Our reporter Monya Baker runs through some of the statistical tools she found when writing her latest story.
As I reported in a Nature feature published this week, I found more online courses that were being developed than were actually in place. Resources to help scientists do more robust research are set to expand quickly. For example, the National Institute of General Medical Sciences has a competitive program that awards funds to institutions to enhance graduate student training; of 15 such supplements awarded in 2015, a dozen involved data analysis, statistics, or experimental rigor. You can find more here, and that is only a fraction of what is available. Some courses are still being developed and piloted to select students; others are being offered only to those in a particular department or training grant. If you find one that interests you, it can’t hurt to ask.
The NIH has funded a series of online modules which will appear at the NIH Training Clearing House as they are ready. Those from the Society of Neuroscience are already there. The site also lists meetings (some with webcasts) on specific techniques for everything from cell culture to structural biology.
One quick recommended reading is Ten Simple Rules for Effective Statistical Practice in PLoS Computational Biology. The authors of that work have published many other general articles as well. Life scientists may also want to spend some time with Nature’s Statistics for Biologists. From here, you can find the “Points of Significance” column, short series explaining practical approaches and common misconceptions. David Vaux has an Annual Reviews article on basic statistics for cell biologists. (subscription required).
Several universities—too many to list—offer free statistical consulting and collaboration. You may want to read these thoughts from Karl Broman about what it feels like to be a statistical consultant before reaching out. The Center for Open Science also offers free statistical and methodological consulting.
I hope to cover tools to learn coding and big data analysis in a future article, so this post focuses only on basic statistics, experimental plans, and protocols. Even with this focus this post can only offer the slimmest sampling of resources available. Please chime in on what resources you think useful in the comments below. The best suggestions will come from you.
Suggested posts
Transferable skills: What are scientists good at (other than science?)
Finding job satisfaction as a data scientist
Big data jobs are out there – are you ready?
Report this comment
Also, here are more online tools to help plan experiments.
We profiled the Experimental Design Assistant previously. https://www.nature.com/news/web-tool-aims-to-reduce-flaws-in-animal-studies-1.19459
Here are some other tools.
The g-power program: https://www.gpower.hhu.de/en.html
The JASP program for Bayesian analysis. https://jasp-stats.org
A program to estimate effects subject to publication bias. https://rvanaert.shinyapps.io/p-uniform/
Hat tip to Jelte Wicherts
Report this comment
Here is the link for a residential course in Advanced Methods for Reproducibility
https://www.bristol.ac.uk/expsych/events/reproducibility2017/
Report this comment
Here is Jeff Leek’s guide for how to share data with a statistician
https://github.com/jtleek/datasharing
Report this comment
Great article on analysis, thanks !
Léo from https://www.nature.com/protocolexchange/labgroups/290999