Regression with an ordered categorical response

1989 ◽  
Vol 8 (7) ◽  
pp. 785-794 ◽  
Author(s):  
T. J. Hastie ◽  
J. L. Botha ◽  
C. M. Schnitzler
Keyword(s):  
Categorical Response ◽  
Ordered Categorical

Ordinal logistic regression analysis of flight task ratings

2006 ◽  
Vol 110 (1109) ◽  
pp. 447-456
Author(s):  
R. Bradley ◽  
W. M. Maclaren
Keyword(s):  
Logistic Regression ◽  
Ordinal Logistic Regression ◽  
Practical Significance ◽  
Response Variable ◽  
Explanatory Variables ◽  
Ordinal Logistic Regression Analysis ◽  
Categorical Response ◽  
The Relationship ◽  
Ordered Categorical ◽  
Categorical Scale

Abstract The relationship between a pilot workload rating for a simulated flight task in the proximity of an offshore platform helideck and three experimental factors – wind speed, wind direction and pilot is investigated. The statistical method employed is ordinal logistic regression, which allows the specifying and fitting of regression relationships between ordered categorical response variables and explanatory variables. The response variable in this context is a pilot’s rating of the workload induced by certain flight tasks, measured on an ordered categorical scale 1 to 10. Estimates of the effects of the explanatory variables are given and their practical significance discussed.

Fleece rot and body strike in Merino sheep. I. Evaluation of liability to fleece rot and body strike under experimental conditions

1988 ◽  
Vol 39 (5) ◽  
pp. 917 ◽  
Author(s):  
HW Raadsma ◽  
AR Gilmour ◽  
WJ Paxton
Keyword(s):  
Linear Models ◽  
Common Denominator ◽  
Control Groups ◽  
Experimental Conditions ◽  
Merino Sheep ◽  
Experimental Induction ◽  
Categorical Response ◽  
Ordered Categorical ◽  
Underlying Variable ◽  
Accurate Indicator

The conceptual model that variation in resistance of Merino sheep to fleece rot and body strike could be attributed to an underlying variable with a continuous normal distribution, termed 'liability', was examined in four experiments. A total of 640 young Merino sheep were exposed from 0 to 13 days to a simulated rainfall treatment to allow the development of fleece rot. The incidence of fleece rot ranged from 0% of sheep affected in unwetted control groups and increased up to 100% affected with increased exposure to wetting. Assuming binomial variation, the response in fleece rot and flystrike incidence was fitted on an underlying scale using generalized linear models. In the case of fleece rot severity, grades of exudative fleece rot were shown to fit an ordered categorical response model. The assessment, induction and liability of sheep to fleece rot were found to be highly repeatable in all three cases, indicating that liability to fleece rot could be accurately screened for using experimental induction. Fleece rot severity was found to be an accurate indicator of liability to body strike but not of other forms of flystrike. The possible mechanisms and factors contributing to the liability of sheep to fleece rot and body strike have been discussed, and moisture status of the fleece during the development of the two diseases was indicated as an essential common denominator.

Sign in / Sign up

Export Citation Format

Share Document