negative binomial distribution

(redirected from Negative binomial)
Also found in: Encyclopedia.

negative binomial distribution

A distribution parameterised by a mean and an aggregation parameter that is large when aggregation is small; as it becomes larger, the negative binomial distribution approximates a Poisson distribution. Aggregated distributions are often well described empirically by the negative binomial distribution. For instance, macroparasites are typically aggregated in their host populations, such that most hosts harbour few or no parasites while a few have large parasite burdens.
References in periodicals archive ?
Bean, "Two-sample inference in highly dispersed negative binomial models," The American Statistician, 2011.
We used GLMs (binomial and negative binomial families) without interactions or random effects to examine the NYC-wide and Staten Island data.
Incidence rate ratios (IRR) are reported based on mixed-effects negative binomial modeling.
The Poisson and negative binomial models were compared, and the viability of their use in leaf count data of coffee seedlings was analysed.
where NA = number of sample units, C = permitted error, x = population mean, and kc = common parameter of aggregation of the negative binomial frequency distribution (0.9134) determined previously.
When analyzing the frequency of PEs in the frequent exacerbators, using the negative binomial model, clinically meaningful and statistically significant results in favor of Apulmiq were observed in ORBIT-4, with an estimated risk reduction of 49% for all PEs (RR = 0.51; 95% CI: 0.31-0.85; p = 0.0094).
Poisson and negative binomial (NB) regression analysis were performed to assess the hypothesis that the college of study and level of study are predictors of the proportion of student participants with PBL exposure.
The "Static Negative Binomial Model" section reviews briefly the negative binomial model and motivates the dynamic frailty framework.
The results of the negative binomial regression models predicting marijuana use (columns 1-4) and other illicit drug use (columns 5-8) are presented in Table 1.
The negative binomial model (2) estimated the number of FDI projects a county-cluster received.
Poisson and negative binomial regression models are designed to control for highly skewed distributions of dependent variables using maximum likelihood procedures for parameter estimation.