### Table 2 Correct Classification Percentages Over Three Successive Echoes for Two Models of the Generalization Result

"... In PAGE 11: ... A total of 300 simulated subjects were run, for 1,200 observa- tions in the generalize condition and 1,200 in the control. Correct response percentages on the three successive echoes are shown in the top half of Table2 . For Echo 1, classification accuracy does not differ reliably between the generalize and control conditions (z = 1.... In PAGE 12: ... HINTZMAN generalize condition (an average of 19/22). The cumulative effect of this difference increases across successive echoes, as Table2 shows. There is another way in which traces might interact, and it deserves some attention because it is suggested by the resonance metaphor that was introduced earlier.... In PAGE 12: ... Again there were three classification attempts follow- ing each probe: one using the echo produced by the probe itself, one using the echo from the ensuing trace interaction described by Equation 5, and a third obtained by applying the interaction procedure yet again. Correct classification percentages are shown in the bottom half of Table2 . Data from this resonance model followed the same pattern as those from the conversion model: the generalize and control conditions did not differ reli- ably on Echo 1 but did on Echo 2 (z = 2.... ..."

### Table 3: Accuracy comparison of online stacked graphical learning for sequential partitioning. We evaluated two local models: MaxEnt and MBW. We compared to a competitive graphical model - conditional random fields. The standard stacked model used two-fold-cross-validation predictions. The online stacked graphical model is based on MBW. We used 1 level of stacking.

### Table 1. Test models

2005

"... In PAGE 3: ... The bifurcation diagram shows hysteresis. 3 EXAMPLES We have tested the discovery tool on a variety of example models that are listed in Table1 . Figure 1 shows the progress of the fitness in a typical optimization run for locating a Hopf bifurcation.... ..."

### Table 4 Comparison of Models with Weighting Based on PAYROLLa

2001

"... In PAGE 12: ... Thus, the logarithmic transformation did not remove the heteroscedasticity. Table4 shows the results of fitting several mod- els, using PAYROLL to weight subject-specific variances. Because the logarithmic transforma- tion did not capture all of the heteroscedasticity, we fit the models using both PP and LnPP as response variables.... In PAGE 12: ... Plots of standardized residu- als from these models versus PAYROLL, not dis- played here, indicate that this weighting captures the heteroscedasticity. Among the models with LnPP as the response, Table4 shows that the model with YEAR as an additional random effects component, but not fixed effects, provides the best fit to the data. This model can be written as LnPPit 5 a1i 1 a2iYEARt 1 b1 1 eit~PAYROLLit!1/2, where {eit} is an i.... In PAGE 12: ...here {eit} is an i.i.d. sequence of noise terms. Among the models with PP as the response vari- able, Table4 shows that the model with no addi- tional random or fixed components is the pre- ferred choice. This model is PPit 5 a1i 1 b1 1 eit~PAYROLLit!1/2.... ..."

Cited by 6

### Table 5 ARMD trial

2007

"... In PAGE 20: ... This finding is mildly confirmed by the random-intercept model, when the data as observed are used. The results for the random-effects models are given in Table5 . We observe the usual relationship between the marginal parameters of Table 4 and their random-effects counterparts.... ..."

### Table 3 Model Parameters for Far Echoes

### Table 2: Stated Preference Models

1997

"... In PAGE 9: ... 4.1 Stated Preference Models Table2 gives the results of our pure stated preference models. The first three columns replicate the MNL model in Brownstone and Train (1996) for the current version of the data.... In PAGE 10: ... 40% of sample have college education Constant for CNG 1 for compressed natural gas vehicle, zero otherwise Constant for methanol 1 for methanol vehicle, zero otherwise Used 1 if vehicle is used; zero otherwise. The last four columns of Table2 give the estimates for the best fitting Mixed Logit specification. The normally distributed random coefficients were determined by testing exclusion restrictions against a larger model with all coefficients allowed to vary.... In PAGE 11: ... They had components for Size, Luggage Space, Non-EV, and Non-CNG. The specification in the last columns of Table2 obtains a lower Log Likelihood value with the same degrees of freedom. However, the relatively large error component for the Price variable implies that the model will generate an implausible positive price effect approximately 30% of the time.... In PAGE 11: ... The normalized coefficients can then be interpreted as the average amount that a respondent with median income would be willing to pay for an additional unit of a particular attribute. For example, the results in Table2 indicate that the typical respondent would pay either $1000 or $492 (MNL or Mixed Logit) to purchase a Sports Utility Vehicle with all other attributes held constant. As in previous studies, these average willingness to pay figures are quite similar for MNL and Mixed Logit.... In PAGE 13: ... The first three columns of Table 3 give results for the best fitting MNL model using the variables described in Table 1. Generally the coefficients are larger in magnitude than the MNL estimates for the SP data given in Table2 . This indicates that the variance of the stochastic portion of the utility indices are lower for the RP data.... In PAGE 13: ... While these estimates show substantial variation across the sample, the standard deviations are not so large as to lead to a significant number of cases with positive price and cost effects. The normalized coefficients (computed as in Table2 ) show no significant differences in average effects between the models. The normalized coefficients for Used vehicles indicate that respondents are willing to purchase a used vehicle with identical attributes to a new one if the used vehicle costs $2200 less.... ..."

### Table 2: Stated Preference Models

"... In PAGE 9: ... 4.1 Stated Preference Models Table2 gives the results of our pure stated preference models. The first three columns replicate the MNL model in Brownstone and Train (1996) for the current version of the data.... In PAGE 10: ... 40% of sample have college education Constant for CNG 1 for compressed natural gas vehicle, zero otherwise Constant for methanol 1 for methanol vehicle, zero otherwise Used 1 if vehicle is used; zero otherwise. The last four columns of Table2 give the estimates for the best fitting Mixed Logit specification. The normally distributed random coefficients were determined by testing exclusion restrictions against a larger model with all coefficients allowed to vary.... In PAGE 11: ... They had components for Size, Luggage Space, Non-EV, and Non-CNG. The specification in the last columns of Table2 obtains a lower Log Likelihood value with the same degrees of freedom. However, the relatively large error component for the Price variable implies that the model will generate an implausible positive price effect approximately 30% of the time.... In PAGE 11: ... The normalized coefficients can then be interpreted as the average amount that a respondent with median income would be willing to pay for an additional unit of a particular attribute. For example, the results in Table2 indicate that the typical respondent would pay either $1000 or $492 (MNL or Mixed Logit) to purchase a Sports Utility Vehicle with all other attributes held constant. As in previous studies, these average willingness to pay figures are quite similar for MNL and Mixed Logit.... In PAGE 13: ... The first three columns of Table 3 give results for the best fitting MNL model using the variables described in Table 1. Generally the coefficients are larger in magnitude than the MNL estimates for the SP data given in Table2 . This indicates that the variance of the stochastic portion of the utility indices are lower for the RP data.... In PAGE 13: ... While these estimates show substantial variation across the sample, the standard deviations are not so large as to lead to a significant number of cases with positive price and cost effects. The normalized coefficients (computed as in Table2 ) show no significant differences in average effects between the models. The normalized coefficients for Used vehicles indicate that respondents are willing to purchase a used vehicle with identical attributes to a new one if the used vehicle costs $2200 less.... ..."

### Table 5: Hurricane Andrew Estimation Results for S amp;P Insurer Stock Price Return (Rt) Model

2005

"... In PAGE 11: ... Otherwise, the estimation was undertaken in the same fashion that was used for analyzing the effects of Floyd. Table5 presents the results obtained from estimating the synoptic lifecycle event study for Andrew. Comparing the results of Floyd to Andrew, the former of which ... In PAGE 25: ...Figure 3: Storm Map of Hurricane Andrew Notes: Points on the figure represent the path of the hurricane. Points on this figure correspond to the data presented in Table5 . The path can be differentiated by a variety of storm characteristics.... ..."