Inferential Statistics is not Inferential – sci five | University of Basel – Medium
"The earth is flat (p > 0.05).

I confess. Throughout my scientific life, I have used a method that I knew or felt was deeply flawed. What’s more, I admit to have taught — and I do still teach — this method to my students. I have a number of questionable excuses for that. For example, because the method has shaped a big part of science in the last century, I think students ought to know about it.

But I have increasingly come to believe that science was and is largely a story of success in spite of, and not because of, the use of this method. The method is called inferential statistics. Or more precisely, hypothesis testing.
The method I consider flawed and deleterious involves taking sample data, then applying some mathematical procedure, and taking the result of that procedure as showing whether or not a hypothesis about a larger population is correct."
statistics:bayesian  statistics:frequentist  confidence_intervals  to_teach  regression  statistics  philosophy_of_science  philosophy_of_statistics  statistics:error_based 
4 days ago
Twitter
Even with real data, it can be tricky to formulate a prior for complex data. With community d…
from twitter
9 days ago
Twitter
Okay, thanks. I can see that it could be really difficult to come up with a single value for uncertainty.…
from twitter
13 days ago
Twitter
This is great! I'm really excited to see that it includes depth data as well.

Is there any way to get i…
from twitter
13 days ago
Twitter
This looks fantastic! And for marine spatial folks out, it'll output depth as well
from twitter
13 days ago
Twitter
I'd like to see an option for what character encoding is used when creating the reference file; r…
from twitter
14 days ago
Twitter
Agreed! Also, I really need to make a mental note to…
from twitter
5 weeks ago
Twitter
Most of the time, the global smooth term is just a us…
from twitter
5 weeks ago
Twitter
I think there are cases when you could think of the g…
from twitter
5 weeks ago
Twitter
Basically, it will predict the global function (since…
from twitter
5 weeks ago
Twitter
You don't need to use predict differently, but you do…
from twitter
5 weeks ago
Twitter
RT : It's like someone reached into my brain, unscrambled all the random thoughts I've had on GAMs for the last eight ye…
from twitter
5 weeks ago
Twitter
RT : Our detection paper is out! This was a super fun one to write thanks to great collaborators:
from twitter
5 weeks ago
Twitter
If you just use the `predict` function for new data,…
from twitter
5 weeks ago
Twitter
Thanks! More curves would likely help; I wanted to av…
from twitter
5 weeks ago
Twitter
2/2 you could then calculate that measure of dissimilarity…
from twitter
5 weeks ago
Twitter
1/2 It's not something you can do out of the box, but it is…
from twitter
5 weeks ago
Twitter
That's a good catch; it should refer to the group-spe…
from twitter
5 weeks ago
Twitter
Thanks for all this feedback! (and apologies for the…
from twitter
5 weeks ago
Twitter
:) You'll just have to take my word on it I guess... If it m…
from twitter
5 weeks ago
Twitter
:) fair. First non-trivially palindromic then. Also the firs…
from twitter
5 weeks ago
Twitter
RT : Hi Twitter hive mind! I'm co-chair of a student conference held in Ottawa in spring. We're brainstorming potential…
from twitter
6 weeks ago
Twitter
Monica Granados () is in Ottawa, currently working on Open Science policy for Environment Can…
from twitter
6 weeks ago
Twitter
My best replacement for broken journal RSS feeds has been PubMed's adva…
from twitter
6 weeks ago
Untitled (https://onlinelibrary.wiley.com/action/cookieAbsent)
RT : Towards a multi‐trophic extension of metacommunity ecology - Guzman - - Ecology Letters - Wiley Online Library
from twitter
6 weeks ago
precisely: An R package for estimating sample size based on precision
"precisely is a study planning tool to calculate sample size based on precision rather than power. Power calculations are focused on whether or not an estimate will be statistically significant; calculations of precision are based on the same prinicples as power calculation but turn the focus to the width of the confidence interval.

This package includes a Shiny app to help with calculations, which you can start with launch_precisely_app(). You can also find a live version at malcolmbarrett.shinyapps.io/precisely."
shiny  R_packages  statistics:experimental_design  experimnents  to_try  to_teach 
7 weeks ago
Twitter
Thanks Andrew! Hopefully our paths cross soon, so we can compare notes on the…
from twitter
7 weeks ago
Twitter
(2/2) and just released his course on GAMs, which looks like an amazing resource for learning (…
from twitter
7 weeks ago
Twitter
And just to signal boost some other great GAM work that just came out (1/2): just released his gratia pack…
from twitter
7 weeks ago
Twitter
As today is my Discovery Grant submission day, it feels like a good time to finally make the big announcement: I'l…
from twitter
7 weeks ago
Twitter
Have you ever wanted to fit the same nonlinear regression model for a bunch of different species (or sites, or indi…
from twitter
7 weeks ago
Twitter
This is also my first big experiment with an entirely open workflow. You can view the whole history of this project…
from twitter
7 weeks ago
Twitter
Just submitted a paper to PeerJ for review, just in time for the NSERC Discovery Grant deadline this week! This is…
from twitter
7 weeks ago
Zotero Blog » Blog Archive » Zotero Comes to Google Docs
Finally! is now available within

Now all Google Docs needs is better math support...
from twitter
7 weeks ago
How People Learn II Learners, Contexts, and Cultures
The second report from the National Academies of Science Engineering and Medicine, sumamarizing research into how people learn. Definitely going to be handy going ahead...
learning  pedagody  psychology  psychology:applied  to_teach  to_read  Books 
8 weeks ago
Do more with R: drag-and-drop ggplot
Drag-and-drop interface for ggplot2 in Rstudio. Definitely worth digging further in to!
ggplot2  R  R_packages  GUI  plotting  graphics 
8 weeks ago
Canadian Hydrographic Service Non-Navigational (NONNA-100) Bathymetric Data
Bathymetric data for all of Canada at 100 m spatial resolution (not sure about depth resolution yet).
marine_biology  oceanography  maps  GIS  to_try  spatial_ecology 
8 weeks ago
Twitter
Sounds really interesting! Is there any chance there'll be a recording available…
from twitter
10 weeks ago
Twitter
I might actually have to write this up as a paper at some point... as far as I can tell it's not yet been…
from twitter
12 weeks ago
Twitter
This should give similar CIs to the delta method in cases when a linear logi…
from twitter
12 weeks ago
From the bottom of the heap
I was inspired by 's great stats tutorials at , and by the number o…
from twitter
12 weeks ago
Twitter
I was inspired by 's great stats tutorials at , and by the number o…
from twitter
12 weeks ago
Twitter
I have a new help policy: if I'm asked a hard question at least twice, I'm going to turn it into an online…
from twitter
12 weeks ago
Twitter
RT : Seriously considering writing my book with aka R

Does anyone have any t…
from twitter
september 2018
Twitter
We've been writing a paper in Rmarkdown and I love it; so refreshing to mix code w/ writ…
from twitter
september 2018
Twitter
True; for 2800 data points, that is a pretty small number of outliers…
from twitter
september 2018
Twitter
Third, you might want to think about using something like the student…
from twitter
september 2018
Twitter
Second, given the high degree of concurvity between your predictors,…
from twitter
september 2018
Twitter
Given the description, I'd have a couple suggestions.

First: I'd us…
from twitter
september 2018
Twitter
They're supposed to reintroduce the separate git repo () but I don't…
from twitter
september 2018
Working Offline in Overleaf v2 - Help and FAQs - Overleaf
They're supposed to reintroduce the separate git repo () but I don't…
from twitter
september 2018
Twitter
...and obviously I mean **Wiener** process here. Which is a good…
from twitter
september 2018
Twitter
Although in this case, I was talking about the continuous time au…
from twitter
september 2018
Twitter
I think we do agree, although we're starting to hit the hard limi…
from twitter
september 2018
Twitter
Although there's no need to assume differentiable basis functions…
from twitter
september 2018
Twitter
You could obtain that in this case by setting k= # of distinct da…
from twitter
september 2018
Twitter
Although part of it is also the k=3 versus the default (k=10) tha…
from twitter
september 2018
Twitter
Great post! Ah, the fun of topological poles. Now I wonder how easy it'd be to recreate t…
from twitter
september 2018
Twitter
I love this! You can also really see the effect of switching fr…
from twitter
september 2018
Twitter
That makes sense... Everything starts to look Poisson (or eve…
from twitter
september 2018
Twitter
Hard to advise without seeing what the residuals look like... Although I gene…
from twitter
august 2018
Twitter
This would be so frustrating. Also, how is "incremental progress" a bad thing deserving rejection? All good science…
from twitter
august 2018
Twitter
Too many things to tell you about! If you're looking for board games: check out Pub Randolf on St. De…
from twitter
august 2018
Twitter
I know the feeling... I keep forgetting about, then delightfully rediscovering the starts_with(), ends_…
from twitter
august 2018
Twitter
Found on the way to work this morning: a rose gall! Looks like a flower, but actually there'll be a bunch of wasp e…
from twitter
august 2018
Twitter
I don't recall, but we got it from Void Gallery just off 19th Street. I know the guy who runs…
from twitter
august 2018
Twitter
Glad you're enjoying! Papers do better with a really long simmer, right? :)
from twitter
july 2018
Twitter
This is also a great example of Simpon's paradox ():
A higher % of Republicans…
from twitter
july 2018
Simpson's paradox - Wikipedia
This is also a great example of Simpon's paradox ():
A higher % of Republicans…
from twitter
july 2018
Twitter
That is an amazing amount of fish! Also: we should chat fish stats at some…
from twitter
july 2018
Twitter
I'd you're in St. John's on this lovely Saturday morning, come out to the Georgetown neighborhood yard sale! A…
from twitter
july 2018
« earlier      
advice advice:acedemic agriculture agroecosystems animal_biology animation anthropology apps art artificial_intelligence behaviour behaviour_ecology biodiversity blog_post books c++ causal_inference civil_rights classes_and_papers climate_change collaboration colour comics community_ecology computers conservation cooking cooking:baking cooking:breakfast cooking:desserts cooking:dinner cooking:indian cooking:italian cooking:methods cooking:pasta cooking:pressure_cooker cooking:sauces cooking:soup cooking:vegan cooking:vegetarian crime criticism culture culture:popular data_management databases dataset decision_making development developmental_biology dispersal dynamics ecology ecology:community ecology:landscape economic_growth economics economics:behavioural economics:macro economics:micro economics:natural_resource ecosystem_function education environmental_issues epidemiology evolution fantasy feminism fiction fiction:science_fiction finance fisheries fishing forests game_theory gaming gaussian_processes gender genetics geography ggplot2 gis git github gmo graduate_school graph_theory graphics gui health high_dimensional_methods high_performance_computing history history_of_culture history_of_military history_of_politics history_of_science history_of_technology horror housing humour humour:acedemic humour:geeky humour:technical image_processing inequality internet invasion_biology james_nicoll jobs justice landscape_ecology latex law learning lecture_notes library limnology linguistics literature machine_learning macroecology macroeconomics make management maps marine_biology markdown math:dynamical_systems math_and_stats media medicine microeconomics military movement_ecology movies nature networks news non_parametrics numeric_methods object_tracking open_science pedagody philosophy_of_science physics plants plotting political_science politics politics:environmental politics:gender politics:police politics:progressive politics:usa politics:workplace population_ecology presentation productivity_tool programming programming:best_practices programs_to_use project_management psychology psychology:applied psychology:congnitive python r r_function r_hadley r_packages race rcpp recipe regression remote_sensing reproducible_research review romance rstudio science_fiction science_in_public sexism shiny short_story simulation smoothing_and_penalization socialism sociology sociology:gender software_engineering spatial_ecology statistical_clustering statistical_computing statistical_methods statistical_software statistics statistics:additive_models statistics:bayesian statistics:ecological statistics:gams statistics:hierarchical statistics:movement statistics:multivariate statistics:networks statistics:regression statistics:spatial statistics:time_series stochastic_methods stochasticity teaching teaching:coding teaching:ecology teaching:statistics teaching:undergraduate technology textbook theoretical_ecology tidyverse time_series to_read to_teach to_try tracking trophic_cascades tropical_forests tumblr tutorial version_control video video_analysis visualization walleye web_applications wisconsin word_processors writing

Copy this bookmark:



description:


tags: