3 ms·
At least in online contexts, this often means 'there might be an effect, but it is smaller than our experiment could have predicted. Let's keep adding samples.'
by huac 10y ago
At least in online contexts, this often means 'there might be an effect, but it is smaller than our experiment could have predicted. Let's keep adding samples.'
- angry_octet 10y ago'... until we get the answer we're expecting'. Just expecting an effect is a bias towards outliers. http://doingbayesiandataanalysis.blogspot.com.au/2013/11/optional-stopping-in-data-collection-p.html?m=1 http://doingbayesiandataanalysis.blogspot.com.au/2013/11/opt...