Gaussian versus count-data hurdle models: cigarette consumption by women in the US

AbstractCounts data with excessive zeros are frequently encountered in practice. For example, the number of health services visits often includes many zeros representing the patients with no utilization during a follow-up time. A common feature of this type of data is that the count measure tends to have excessive zero beyond a common count distribution can accommodate, such as Poisson or negative binomial. Zero-inflated or hurdle models are often used to fit such data. Despite the increasing popularity of ZI and hurdle models, there is still a lack of investigation of the fundamental differences between these two types of models. In this article, we reviewed the zero-inflated and hurdle models and highlighted their differences in terms of their data generating processes. We also conducted simulation studies to evaluate the performances of both types of models. The final choice of regression model should be made after a careful assessment of goodness of fit and should be tailored to a particular data in question.

Download Full-text

P66 Estimating inequalities in moderate-vigorous physical activity among adolescents in England and the US using hurdle models

10.1136/jech-2020-ssmabstracts.159 ◽

2020 ◽

Author(s):

S Scholes ◽

JS Mindell

Keyword(s):

Physical Activity ◽

Vigorous Physical Activity ◽

Hurdle Models ◽

The Us

Download Full-text

Using Count Data Models to Predict Epiphytic Bryophyte Recruitment in Schima superba Gardn. et Champ. Plantations in Urban Forests

Forests ◽

10.3390/f11020174 ◽

2020 ◽

Vol 11 (2) ◽

pp. 174

Author(s):

Dexian Zhao ◽

Zhenkai Sun ◽

Cheng Wang ◽

Zezhou Hao ◽

Baoqiang Sun ◽

...

Keyword(s):

Count Data ◽

Human Disturbance ◽

Negative Binomial ◽

Urban Environments ◽

Tree Planting ◽

Count Data Models ◽

Urban Tree ◽

Schima Superba ◽

Hurdle Models ◽

Positive Effects

Epiphytic bryophytes are known to perform essential ecosystem functions, but their sensitivity to environmental quality and change makes their survival and development vulnerable to global changes, especially habitat loss in urban environments. Fortunately, extensive urban tree planting programs worldwide have had a positive effect on the colonization and development of epiphytic bryophytes. However, how epiphytic bryophytes occur and grow on planted trees remain poorly known, especially in urban environments. In the present study, we surveyed the distribution of epiphytic bryophytes on tree trunks in a Schima superba Gardn. et Champ. urban plantation and then developed count data models, including tree characteristics, stand characteristics, human disturbance, terrain factors, and microclimate to predict the drivers on epiphytic bryophyte recruitment. Different counting models (Poisson, Negative binomial, Zero-inflated Poisson, Zero-inflated negative binomial, Hurdle-Poisson, Hurdle-negative binomial) were compared for a data analysis to account for the zero-inflated data structure. Our results show that (i) the shaded side and base of tree trunks were the preferred locations for bryophytes to colonize in urban plantations, (ii) both hurdle models performed well in modeling epiphytic bryophyte recruitment, and (iii) both hurdle models showed that the tree height, diameter at breast height (DBH), leaf area index (LAI), and altitude (ALT) promoted the occurrence of epiphytic bryophytes, but the height under branch and interference intensity of human activities opposed the occurrence of epiphytic bryophytes. Specifically, DBH and LAI had positive effects on the species richness recruitment count; similarly, DBH and ALT had positive effects on the abundance recruitment count, but slope had a negative effect. To promote the occurrence and growth of epiphytic bryophytes in urban tree planting programs, we suggest that managers regulate suitable habitats by cultivating and protecting large trees, promoting canopy closure, and controlling human disturbance.

Download Full-text

EXTENSIONS OF HURDLE MODELS FOR OVERDISPERSED COUNT DATA

Health Economics ◽

10.1002/hec.2892 ◽

2012 ◽

Vol 22 (11) ◽

pp. 1398-1404 ◽

Cited By ~ 3

Author(s):

Helmut Farbmacher

Keyword(s):

Count Data ◽

Hurdle Models

Download Full-text

Zero-Inflated Poisson Model with Group Data

Advanced Materials Research ◽

10.4028/www.scientific.net/amr.569.627 ◽

2012 ◽

Vol 569 ◽

pp. 627-631

Author(s):

Jun Yang ◽

Xin Zhang

Keyword(s):

Parameter Estimation ◽

Maximum Likelihood ◽

Model Selection ◽

Count Data ◽

Maximum Likelihood Estimate ◽

Poisson Model ◽

Cigarette Consumption ◽

Chi Square ◽

Group Data ◽

Chi Square Test

The Zero-inflated Poisson model has been widely used in many fields for count data with excessive zeroes. In fact, group data are often collected for many count data, such as cigarette consumption. In order to solve the problem, Zero-inflated Poisson model with group data is investigated in this paper. Parameter estimation is given by the maximum likelihood estimate, model selection is discussed by the Chi-square test, and one real example is given for application in the end.

Download Full-text

Asymmetric information and opportunism in insider share-pledging

Managerial Finance ◽

10.1108/mf-06-2020-0322 ◽

2021 ◽

Vol ahead-of-print (ahead-of-print) ◽

Author(s):

Michael R. Puleo ◽

Steven E. Kozlowski

Keyword(s):

Private Information ◽

Abnormal Returns ◽

Risk Assessments ◽

Corporate Policies ◽

Market Reactions ◽

Content Type ◽

Hurdle Models ◽

Cumulative Abnormal Returns ◽

Earnings Surprises ◽

The Us

PurposeAmid growing attention from investors, regulators and advisory firms in recent years, this study assesses whether managers exploit private information to time share-pledge transactions and extract personal benefits while avoiding unintended market scrutiny.Design/methodology/approachWe use hand-collected pledging data for a random sample of S&P 1500 firms to examine whether private information influences insider share-pledging activity using Heckman selection and two-part hurdle models of the pledge decision. We also conduct an event study analysis of announcement returns to measure market reactions to pledging news and determine whether share-pledge disclosures affect investor risk assessments.FindingsConsistent with insiders timing pledges prior to anticipated performance declines, both the likelihood and level of pledging increase significantly with negative earnings surprises. New share-pledges precede significant decreases in abnormal returns, and public announcement of new pledging corresponds with significant negative cumulative abnormal returns. The evidence suggests that insiders exploit private information to time pledges, and that investors update risk assessments and value estimates based on information conveyed by these transactions.Practical implicationsOur findings hold important implications for governance and regulation of pledged shares, indicating that permissive reporting requirements in the US facilitate informed pledging and may undermine incentive alignment between managers and shareholders. The analysis promotes transaction-specific disclosures and transparent corporate policies for insider share-pledging.Originality/valueOurs is among the first empirical analyses of share-pledging in US firms and the first to examine the role of private information in pledging decisions. We offer novel evidence on the opportunistic use of pledged shares and provide insight to predictors of share-pledging behavior.

Download Full-text

Zero-Inflated and Hurdle Models of Count Data with Extra Zeros: Examples from an HIV-Risk Reduction Intervention Trial

The American Journal of Drug and Alcohol Abuse ◽

10.3109/00952990.2011.597280 ◽

2011 ◽

Vol 37 (5) ◽

pp. 367-375 ◽

Cited By ~ 120

Author(s):

Mei-Chen Hu ◽

Martina Pavlicova ◽

Edward V. Nunes

Keyword(s):

Risk Reduction ◽

Hiv Risk ◽

Count Data ◽

Intervention Trial ◽

Hiv Risk Reduction ◽

Hurdle Models ◽

Risk Reduction Intervention

Download Full-text

Extended negative binomial hurdle models

Statistical Methods in Medical Research ◽

10.1177/0962280218766567 ◽

2018 ◽

Vol 28 (5) ◽

pp. 1540-1551

Author(s):

Maengseok Noh ◽

Youngjo Lee

Keyword(s):

Count Data ◽

Negative Binomial ◽

Random Effect ◽

Random Effect Model ◽

Real Data ◽

Hurdle Models ◽

Poisson Models ◽

Effect Model ◽

Zero Rate ◽

General Statistical

Poisson models are widely used for statistical inference on count data. However, zero-inflation or zero-deflation with either overdispersion or underdispersion could occur. Currently, there is no available model for count data, that allows excessive occurrence of zeros along with underdispersion in non-zero counts, even though there have been reported necessity of such models. Furthermore, given an excessive zero rate, we need a model that allows a larger degree of overdispersion than existing models. In this paper, we use a random-effect model to produce a general statistical model for accommodating such phenomenon occurring in real data analyses.

Download Full-text