Tags: stats
2026
2025
2024
- The Ghost of p-values Past
- Type safety in R
- Why do anova(type="marginal") and anova(type="III") yield different results on lmer() models?
- Time for correlations
- Start with the programming language and statistical approach used by your community
- Back of the envelope calculations: pulp mill
- Exposing rather than hiding complexity
- Potential material for teaching
- Superindex and subindex in ggpairs axes labels
- greenR: Green spaces in R
- duckplyr: dplyr + DuckDB
- I haven't done an animation in R in ages
2023
- What Would Akaike Do?
- Some love for Base R. Part 4
- Anyone using other than RStudio?
- Infrequent doesn't disprove
- Sense-checking data
- Some love for Base R. Part 3
- Some love for Base R. Part 2
- Some love for Base R. Part 1
- Not a contribution to science
- The data may not contain the answer
- Flotsam 15: inference
2022
2021
2019
2018
2017
- Collecting results of the New Zealand General Elections
- Functions with multiple results in tidyverse
- Turtles all the way down
- Old dog and the tidyverse
2016
2014
- Mucking around with maps, schools and ethnicity in NZ
- Back of the envelope look at school decile changes
- Comment on Sustainability and innovation in staple crop production in the US Midwest
- Sometimes I feel (some) need for speed
- Less wordy R
- R as a second language
- Teaching linear models
2013
- Statistics unplugged
- Using Processing and R together (in OS X)
- Excel, fanaticism and R
- Flotsam 13: early July links
- My take on the USA versus Western Europe comparison of GM corn
- GM-fed pigs, chance and how research works
- Ordinal logistic GM pigs
- Analyzing a simple experiment with heterogeneous variances using asreml, MCMCglmm and SAS
- Subsetting data
2012
- An R wish list for 2013
- My R year
- Matrix Algebra Useful for Statistics
- When R, or any other language, is not enough
- Multisite, multivariate genetic analysis: simulation and analysis
- More sense of random effects
- Overlay of design matrices in genetic analysis
- A word of caution: the sample may have an effect
- Some regressions on school data
- Updating and expanding New Zealand school data
- New Zealand school performance: beyond the headlines
- New Zealand School data
- (Unsurprisingly) users default to the defaults
- Suicide statistics and the Christchurch earthquake
- m x n matrix with randomly assigned 0/1
- Mid-August flotsam
- INLA: Bayes goes to Norway
- Careless comparison bites back (again)
- Early August flotsam
- Split-plot 2: let's throw in some spatial effects
- Split-plot 1: How does a linear mixed model look like?
- Review: “Forest Analytics with R: an introduction”
- R’s increasing popularity. Should we care?
- Bivariate linear mixed models using ASReml-R with multiple cores
- Teaching code, production code, benchmarks and new languages
- R, Julia and genome wide selection
- If you have to use circles…
- Revisiting homicide rates
- Oracle’s strange understanding of R users
- Early-February flotsam
- Rstudio and asreml working together in a mac
- Mid-January flotsam: teaching edition
- R is a language
- Doing Bayesian Data Analysis now in JAGS
- Plotting earthquake data
2011
- An R wish list for 2012
- First impressions of Doing Bayesian Data Analysis
- R pitfall #3: friggin’ factors
- Tall big data, wide big data
- R, academia and the democratization of statistics
- On the (statistical) road, workshops and R
- If you are writing a book on Bayesian statistics
- No one would ever conceive
- Do we need to deal with ‘big data’ in R?
- Surviving a binomial mixed model
- On “true” models
- Coming out of the (Bayesian) closet: multivariate version
- Coming out of the (Bayesian) closet
- Teaching with R: the tools
- Multivariate linear mixed models: livin’ la vida loca
- Covariance structures
- Longitudinal analysis: autocorrelation makes a difference
- Teaching with R: the switch
- Spatial correlation in designed experiments
- Large applications of linear mixed models
- Linear mixed models in R
- Maximum likelihood
- Simulating data following a given covariance structure
- Upgrading R (and packages)
- On R versus SAS
- Linear regression with correlated data
- R pitfall #1: check data structure
- A shoebox for data analysis
- Python code to simulate the Monty Hall problem