Lack of scientific openness goes hand in hand with lack of basic statistical ability …

Regular readers will remember the long running saga concerning the Lang et al team’s bulge-disk decompositions (described in a previous post) in which I found that none of the key coauthors would even reply to my emails pointing out an issue of (potentially) major concern for their analysis: in brief, the need for an inclination vs. B/T diagnostic plot when attempting to convince the community of the robustness of their fitted profiles at high B/T against contamination from disk-less spheroids.  Well, it was then not a surprise to me to see that in the latest contribution from their Shaw prize winning team leader, Prof Genzel, on astro ph there is an apparent inability to deal with the statistical analysis of binomial data at even the most basic level.  All the usual astronomers mistakes are there—quoting confidence intervals with no mention of the significance to which they refer or the method used to compute them, which does not seem to match up with any standard approach (e.g. Bayesian, Normal approximation, Agresti-Coull)—plus no use of even basic Generalized Linear Modelling tools (cf. logistic/probit regression) to explore the significance of competing predictors (SSFR, bulge mass, total mass).

This entry was posted in Astronomy, Astrostatistics, Bad Science. Bookmark the permalink.

Leave a Reply

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

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

Google+ photo

You are commenting using your Google+ 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 )

Connecting to %s