Research

Research Resources

 

1. Supplementary resources for journal article: Bartimote-Aufflick, K. & Thomson, P. C. (2011). The analysis of ordinal time-series data via a transition (Markov) model. Journal of Applied Statistics, 38(9), 1883-1897.

Download R software

R code for ordinal repeated measures analysis

R code for adjustment of standard errors

Data description: The data format for this article appears as shown in the table below with five column variables. Tree ranges 1 – 100, Treatment is either Phosphonate or Control, Time ranges 1 – 36, Rating at the current time point (0, 1, 2, 3, 4), and LagRating which is the rating at the previous time point.

Tree Treatment Time Rating LagRating
1 Control 1 0 NA
2 Phosphonate 1 0 NA
3 Control 1 0 NA
4 Phosphonate 1 2 NA
5 Control 1 1 NA
1 Control 2 0 0
2 Phosphonate 2 1 0
3 Control 2 2 0
4 Phosphonate 2 3 2
5 Control 2 0 1
100 Phosphonate 36 0 0

2. Review of the appropriateness of technique in analysis of ordinal data (e.g. scores, ratings) in journal articles where the research relates to some aspect of science in agriculture, animal science, horticultural science, or land and water science. (Incorporates an example of Ordinal Logistic Regression.)

This work formed Chapter 2 of my research masters in statistics, a 2-year full-time degree that I completed part-time in the years 2003-2006. It reviews the appropriateness (from a statistical perspective) of analysis of ordinal data (e.g. scores, ratings) in journal articles where the research relates to some aspect of science in agriculture, animal science, horticultural science, or land and water science. Each article has been rated, and these ratings provide a data set with which to demonstrate the ordinal logistic regression technique.

Link to S-PLUS command code NB. This code should also work in R.
Data for Ordinal Logistic Regression