# My first blog post was a guest post

August 30, 2016

The title says it all. I wrote this piece about Publication Viability Analysis pondering about a pattern that I observed while looking at Hungarian ecologists publication output through time using the Web of Science database (the original post is in Hungarian).

I fit Ricker growth model to observed publication numbers and a non-stationary 2-phase model was the best fit. An intrinsic growth rate of 0.39 with carrying capacity K = 14 publications per year was characteristic to the pre-democracy (Soviet occupation era) phase (1978–1997). The democratic era (1998–2012) showed growth rate of 0.21 with a much higher K = 100 (variance went from 0.44 to 0.03).

The motivation for the analysis and post was that the number of publications has been persistently around K = 100 for 5 years. So I looked at the correlation between the number of PhD students and the number of publications, using different lag times between 0 and 10 years. The correlation was highest for a 7-year lag. This more or less indicates a cohort of PhD students (myself included) who started PhD around 2000.

This cohort represents the grandchildren of the post-WWII boom, births have declined after this cohort – therefore less PhD students to produce papers. If this is true, it also means that it takes considerable time for PhD students to reach peak publication productivity. So I proposed to shorten the lag to 3–5 years to boost publication numbers. Win for the students and win for Hungary.

Why am I bringing this old post up 3 years later? Because I wanted to see how well my estimates have held up from a short-term forecasting standpoint. Well, here are the results of a recent query with identical settings (ADDRESS=HUNGARY; CATEGORIES=ECOLOGY) for the years 2013–2015:

Year Count
2013 87
2014 114
2015 108

The figure shows the two phases used in modeling (grey and gold), and the forecast (tomato) with horizontal lines for carrying capacity. I wish these numbers have improved, but it is what it is. You can’t argue with science. In case you want to argue, just leave a comment!

#### Fitting removal models with the detect R package

In a paper recently published in the Condor, titled Evaluating time-removal models for estimating availability of boreal birds during point-count surveys: sample size requirements and model complexity, we assessed different ways of controlling for point-count duration in bird counts using data from the Boreal Avian Modelling Project. As the title indicates, the paper describes a cost-benefit analysis to make recommendations about when to use different types of the removal model. The paper is open access, so feel free to read the whole paper here.