What’s so novel about hierarchical modelling?

Daily rant: I noticed this paper by Soiaporn et al (2013) on astro ph today, which has been accepted by Annals of Applied Statistics.  The last clause here really stuck in my craw because my own fine structure paper was last year rejected by AoAS on the basis that its topic was more suited for an astronomical journal and that the statistical methodology was not sufficiently novel (Bayesian model selection with Chopin & Robert’s nested importance sampling).  I naturally disagreed since this was the first use of NIS on a real scientific problem and I thought we demonstrated nicely its potential for prior-sensitivity analysis (which incidentally one of the referees then suggested we should think about doing — despite having already done this prominently in the submitted manuscript!!).  That’s all water under the bridge I suppose because the rejection spurned me on to find some nice results in semi-parametric model selection.  The main focus of my renewed butt-hurt is that the Soiaporn et al paper presents a topic of even more specific astronomical interest (magnetic deflection scales for cosmic rays) and does so using even more traditional techniques (it’s a hierarchical model explored with MCMC and its marginal likelihood computed with Chib’s estimator).  

I feel like Chalmers in that old Simpsons episode where Ned Flanders gets a laugh from telling the kids he’s going to put the “pal back in principal”, while Chalmers gets only awkward silence and a cough from adding that he’s going to put the “super back in superintendent”.  It’s the exact same joke … what gives, Leo?

Image

Advertisements
This entry was posted in Astrostatistics, Rants. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s