Thread by @RSeriesa

Hello World! Today we launched our brand new @RSeriesa Twitter account!🥳

We also welcome @Fcorowe who joins our editorial board

Our *AIM*: share all the great work published in our journal 👉🏼 bit.ly/3sfKdiK & all about #stats! Interested? Give us a follow🙏

First🧵👇🏽

We’ll celebrate the launch of this account this week by highlighting the top 10 most cited journals published in JRSSA in 2020 🎉

*DISCLAMER* in reality they will be 13 given that a few of them have the same number of citations = same rank position. A bonus for you!

2/n

We start with the first 3 in reverse order from 10th to 1st

First up! A paper by Fadlalla Elfadaly & Paul Garthwaite: doi.org/10.1111/rssa.1…

*AIM* Establish a prior distribution to represent expert opinion in a multinomial model setting

3/n

*CHALLENGE* Find meaningful assessment tasks to determine parameter values in the prior distribution, & develop theory to determining these values

*METHOD* Sampling model from a multinomial logistic regression & a multivariate normal dist. to represent expert opinion

4/n

*RESULTS* A method which address the 👆🏽 challenges & interactive easy‐to‐use software which is freely available!

5/n

Kim et al. (2020) assessing the health effects of power plant emissions through ambient air quality doi.org/10.1111/rssa.1…

*AIM* Offer a novel statistical evaluation of key causal relationships between emissions, ambient pollution and health

6/n

*GAP* Systematic evidence of health impact assessments of emissions from coal power plants exist; yet assumed causal relationships between emissions, ambient pollution and health have remained empirically unverified

*METHOD* Bayesian methods and causal mediation analysis

7/n

*RESULTS* A unified epidemiological methodological framework integrating causal inference with direct measurements of coal emissions, pollution transport, ambient pollution and human health

8/n

Galbraith et al (2020) quantifying the association between discrete event time series with applications to digital forensics
doi.org/10.1111/rssa.1…

*AIM* Quantify associations between two discrete events exhibit temporal clustering

9/n

*METHODS* A non-parametric approach is used to investigate various score functions to quantify association, inc. marked point processes and summary statistics of inter-event times

10/n

*RESULTS* Two techniques are proposed:
1] a pop.-based approach to calculating score-based likelihood ratios when a sample from a relevant
population is avail.

2] a resampling approach to computing coincidental match prob. when only a single pair of event series is avail.

11/n

Stay tuned - next 3 articles TOMORROW! and more! 🙌🏽

END