### TABLE 7 Posterior Medians and 95% Credible Intervals for the Squared Logistic Model Parameters for the Muhlberger (1999) Political Motivation Data Set Parameter Mdn 95% Credible Interval

### Table 1: Posterior Means (95% credible intervals) for model param- eters for the disk-trace data.

2004

"... In PAGE 9: ...hows the posterior mean for each of the seasonal components v1, . . . , v48. Posterior means of parameters obtained for both the long-memory Tobit model and the model of Brockwell et al. (2003) are given in Table1 . The seasonal components, as well as... ..."

Cited by 1

### Table 2: Posterior Means (95% credible intervals) for model param- eters for the Edinburgh rainfall data.

2004

Cited by 1

### Table 2. Coverage probabilities of 95% credible intervals

"... In PAGE 4: ... After burn-in, 100 000 samples were collected for posterior inference. Table2 shows that the proportion of data sets for which the estimated 95% credible intervals contained the actual parameter values accorded well with the nominal probability. The absence of significant departures from expectation under a binomial distribution suggests convergence of the EF algorithm to the correct distribution.... ..."

### Table 1: Posterior means (95% credible regions) for the engaged parameters t after process- ing all data for the 12 new patients. Parameter

1997

Cited by 52

### Table 4. Parameter estimates and credible intervals for the D.persimilis and D.p.bogotana data set (Dpe/Dpb) under the EF and IM methods

### Table 1: Simulated Data - true values, posterior mean (PM) and 95% credibility interval for p =0.1 and n =10.000, n =1.000 and n = 100.

2004

"... In PAGE 11: ...roup (p =0.01) . The estimates of the parameters for each scenario are given in table 1below. Table1 shows the results for data with p =0.1.... In PAGE 13: ...other parameters the same as before, almost all parameters are close to their true values. The results are given in the second part of Table1 , and only in a few cases is the true value not included in the 95% credibility interval. However, in these exceptional cases, the true values are very close to it.... ..."

Cited by 1

### Table 1. Posterior mean and 95% credible intervals of model parameters.

"... In PAGE 10: ... Which model is preferred by the data? Can we determine which studies prefer a given model over some other speci ed model? 6. How do the results obtained here compare with those of Iyengar and Greenhouse (1988) and Patil and Taillie (1989)? Table1 contains the posterior estimates and 95% credible intervals for all parameters and models. All posterior plots are density estimates generated in S-Plus using a normal kernel.... In PAGE 10: ... All posterior plots are density estimates generated in S-Plus using a normal kernel. Note in Table1 that the posterior of for all four models di er substantially from zero. This re ects the assumption, built into the weight functions, that bias is present in the studies.... In PAGE 11: ...reater than zero (e.g., Morris and Normand, 1992). While Table1 shows that the 95% credible interval for 2 is greater than zero for all four models, one might argue that this must necessarily follow from the Inverse Gamma complete conditional form for 2. However, Figure 1 shows that the prior asymptote for 2 at zero ( ( 2) = ?2) has been updated by the data to produce a posterior mode which has been shifted in a positive direction away from zero.... In PAGE 11: ... If so, then the random e ects model is contra-indicated. Table1 and Figure 2 indicate, however, that no such signi cant shrinkage has occurred; for example, 95% credible intervals for 1 and 4 do not overlap for any of the models. This lends further support to the random e ects model.... ..."

### Table 3 Range of estimated posterior means of the spatial effects as well as estimated posterior means, medians and 95 % credible intervals given in brackets for the spatial hyperparameters in the considered models for the meningococcal disease data.

"... In PAGE 14: ...293 when spatial effects are included. The range of the estimated spatial effects in the NB model, see Table3 , is considerably smaller than in the Poisson model where unexplained heterogeneity in the data is captured by the spatial effects alone. However, part of the data variability in the NB model is explained by spatial effects as well rather than the parameter r alone.... In PAGE 15: ... Unobserved heterogeneity still present in the data after adjusting for covariates is captured better by the GP and NB model, whereas the assumptions of extra zeros is obviously not appro- priate for this data. When spatial effects are included to the ZIP model the estimated proportion of extra zeros drops even further, indicating that unexplained heterogeneity is picked up mostly by the spatial effects alone like in the Poisson model, the range of the estimated posterior means of the spatial effects in the ZIP and the Poisson model is almost the same, see Table3 . The map plot of the estimated posterior means of the spatial effects in the Poisson model, given in Figure 3 roughly represents the spatial pattern of the observed relative risk in each region yi ti , i = 1, .... ..."