Regression analysis of variates observed on (0, 1): percentages, proportions and fractions

By R. Kieschnick, Bruce D. McCullough. 2003

521 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/07564413e6f4808e2b86fdf478ad040ab4ef1e76

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Statistics and Probability

Citations in the corpus

Beta Regression in R (2010, background): “Future research and implementation shall focus on the situation where the data contain zeros and/or ones (see e.g., Kieschnick and McCullough 2003).”
Inflated beta distributions (2007, background): “,t 3(y),t 4(y)) with t 1(y) = y1l{0,1}(y), t 2(y) = 1l{0,1}(y),t 3(y) = log(y) if y∈ (0,1) and 0 if y∈ {0,1} and t 4(y) = log(1 − y) if y∈ (0,1) and 0 if y∈ {0,1}.Note that the BEINF density function (10) can be written as exp{η⊤T(y)−B∗(η)}h(y), (12) where B∗(η) = log{1 +exp[η 2 +M(η 1) −B(η 3,η 4)]”
Analysing continuous proportions in ecology and evolution: A practical introduction to beta and Dirichlet regression (2019, methodology): “…approaches to model continuous proportions with two categories are as follows: to apply ordinary least squares techniques without transformations (Kieschnick & McCullough, 2003); or apply the arcsine transformation (Sokahl & Rohlf, 1995) or logit transformation followed by nonlinear least…”
Wealth Shocks, Unemployment Shocks and Consumption in the Wake of the Great Recession (2015, unknown): “The quasi ML approach proposed by PW has been found to perform very well in estimation problems involving fractional variables (Kieschnick and McCullough, 2003) and requires no additional assumptions about other features of the data generating process (e.”
Alternative Estimating and Testing Empirical Strategies for Fractional Regression Models (2011, unknown): “Although this setup corresponds to a very simple case, it is probably the most usual approach in applied work; see for example Cook, Kieschnick and McCullough (2008) and Ramalho and Silva (2009). In order to generate samples of fractional data with a given proportion of zero outcomes, we used two di”
Mixed and Mixture Regression Models for Continuous Bounded Responses Using the Beta Distribution (2012, methodology): “Several authors (Kieschnick & McCullogh, 2003; Paolino, 2001; Smithson & Verkuilen, 2006) have presented arguments that traditional methods such as regression-linearizing and variance-stabilizing transformations or robust estimators are not always appropriate for modeling this kind of limited…”
The Kumaraswamy distribution: median-dispersion re-parameterizations for regression modeling and simulation-based estimation (2012, methodology): “…that have been advanced for using the Beta distribution in regression models in a range of contexts and disciplines (Ferrari and Cribari-Neto 2004; Kieschnick and McCullough 2003; Paolino 2001; Smithson and Verkuilen 2006) apply unchanged if we substitute the Kumaraswamy for the Beta distribution.”
On beta regression residuals (2008, background): “Kieschnick and McCullough 3 identify different modelling strategies that are commonly used in such applications.”
Some New Facts about the Unit-Rayleigh Distribution with Applications (2020, background): “The unit-distributions are of particular interest because of the following argues: (i) any rv X with bounded support of the form (0, a), with a > 0, can be rescaled on (0, 1) as Y = X/a, and thus Y follows a certain unit-distribution, (ii) the unit-distributions allow us to define general and simple”
A better lemon squeezer? Maximum-likelihood regression with beta-distributed dependent variables. (2006, result): “Kieschnick and McCullogh (2003) compared it with their beta regression model on two data-sets and found the beta model slightly outperformed it both times in terms of goodness of fit.”
Regression analysis of proportions in finance with self selection (2008, methodology): “Of the various econometric specifications that Kieschnick and McCullough (2003) test, they fail to reject the applicability of either a regression model based upon the beta distribution that they propose or the quasi-likelihood model proposed by Papke and Wooldridge (1996).”