An introduction and tutorial is presented for regression and trend analyses when data are observed in the form of proportions. The basic binomial probability model is assumed. Emphasis is placed on quantal response analyses, useful when a concomitant predictor variable or other form of ordered measure is recorded with each proportion observation. These include regression methods and trend tests, residual analysis, and inverse regression/effective stimulus estimation.
ASJC Scopus subject areas
- Safety, Risk, Reliability and Quality
- Strategy and Management
- Management Science and Operations Research
- Industrial and Manufacturing Engineering