My long-time friend and colleague Mike Sharples told me about the recently released Innovating Pedagogy 2019 report from The Open University (U.K.). It’s the seventh in an annual series initiated by Mike. Each report aims to describe a number of…
Introduction to the New Statistics
The online community and blog for ITNS
Month: January 2019
Sizing up behavioral neuroscience – a meta-analysis of the fear-conditioning literature
Inadequate sample sizes are kryptonite to good science–they produce waste, spurious results, and inflated effect sizes. Doing science with an inadequate sample is worse than doing nothing. In the neurosciences, large-scale surveys of the literature show that inadequate sample sizes…
Sadly, Dichotomous Thinking Persists in HCI Research
A few words about the latest from Pierre Dragicevic. He’s an HCI researcher in Paris who totally gets the need for the new statistics. I’ve written about his work before, here and here. Now, with colleague Lonni Besançon, he reports…
New Stats blog: NeuRA Ahead of the Open Science Curve https://thenewstatistics.com/itns/2019/12/07/neura-ahead-of-the-open-science-curve/
flashbulb memory: @dstephenlindsay was visiting, and we went to Geoff Cummings’s talk. As it all sunk in, we slid down our chairs in horror. The next day, I starting fixing things in my lab—but Steve started fixing all of psychological science. https://journals.sagepub.com/doi/full/10.1177/0956797619893653
Bootstrap methods are not a license to use small sample sizes:
“When sample sizes are too small, no amount of bootstrapping can lead to reliable inferences because the skewness and the tails of certain distributions cannot be captured properly.”
https://psyarxiv.com/kxarf/
New Stats blog: I Join an RCT: A View From the Other Side https://thenewstatistics.com/itns/2019/11/27/i-join-an-rct-a-view-from-the-other-side/
Back in April the British Neuroscience Association adopted reforms to enhance research credibility (@robustgar):
https://journals.sagepub.com/doi/full/10.1177/2398212819844167
So impressed with this statement.
So envious of the BNA--way ahead of @SfNJournals on these issues (though good steps at eNeuro).
Today I'm thankful for any graduate program whose LoR portal requires only a letter upload and none of those rating scales.
Its only p-hacking if it comes from the pihâque region of France. Otherwise it's just sparkling malpractice
New Stats blog: Replications: How Should We Analyze the Results? https://thenewstatistics.com/itns/2019/11/22/replications-how-should-we-analyze-the-results/
We examined close to 1 million confidence intervals in PubMed abstracts and found a huge and unbelievable number of intervals that were just below the magic 0.05 statistically significant threshold https://bmjopen.bmj.com/content/9/11/e032506.full, with @Jonathan_D_Wren
Top
Recent Comments