### Table 1: The Dependent Variables and Their Hazard Rates

"... In PAGE 11: ... After the event occurs, no more state-years are observed for the state (thus limiting my analysis to the risk set, the set of states which are at risk of event occurrence at each point in time). Table1 summarizes the event histories of these four dependent variables. This is a discrete-time event history analysis which, for simplicity without... In PAGE 16: ... These variables also pick up exogenous effects influencing all 51 public utility commissions and legislatures, such as FERC Orders 888 and 889 in 1996, the growth in the national economy during the study period, and the evolution of the national policy debate over electricity sector restructuring. I include dichotomous (0,1) time variables for 1994, 1995, 1996, 1997, 1998, and 1999 and treat 1993 as the baseline, so that constant term in the models subsumes the single event recorded in 1993 (see Table1 ). As Figure 2 shows, the cumulative event time paths follow the expected S-shaped pattern for innovations, although those for legislation and implementation have not yet begun their deflection.... ..."

### Table 4: Statistics on experiments with 1000 random variable orders.

"... In PAGE 5: ... Experiments on both the schemes are done, using each one of them as initial variable order for the corresponding instance. Results are reported in Table4 . The results for the last two instances are not re- ported as all of the variable orders resulted in huge memory requirement and had to be aborted.... ..."

### Table 2: VAR{order selection Lag order Order criterion:

"... In PAGE 6: ... The results for the BIC and AIC criterion up to the truncation order pmax = 8 are given below. The criteria reach its minimum for p = 5 and p = 6, respectively (see Table2 ). To be on the save side p = 6 is selected.... ..."

### Table 9 Ordered probit regressions for government relief for terrorism lossesa

"... In PAGE 19: ...eliefs, increasing from 48.6 percent for the below-average risk category to 62.8 percent for the above-average risk category. As with the regression results for terrorism risk beliefs, the probit regressions for the two terrorism-aid questions reported in Table9 show how far fewer significant effects than the natural hazard regressions. This difference arises in part because unlike natural hazards, which are geographically concentrated in well-known areas, terrorism risks are poorly un- derstood.... ..."

### Table 2 Proportional hazard model of retention

"... In PAGE 29: ... For example, students from schools with higher ratings will tend to have more friends or classmates who are also attending college, and, as a result, may view leaving college before completion to return home as less desirable than other students. In an attempt to capture both the effect that good schools have on college grades and these other types of effects, we estimate the duration model from Table2 for the Kentucky subsample and include the school quality measure. Because the estimated effects of the majority of the variables are very similar to those in Table 2, the entire set of results is not included.... In PAGE 29: ... In an attempt to capture both the effect that good schools have on college grades and these other types of effects, we estimate the duration model from Table 2 for the Kentucky subsample and include the school quality measure. Because the estimated effects of the majority of the variables are very similar to those in Table2 , the entire set of results is not included. However, the coefficient and t-statistic associated with the school quality variable was found to be -.... In PAGE 32: ... 42 The family income variable in the NELS-88 is categorical with classes defined as in Table 8. To estimate the analog of column 2 of Table2 (in which family income is categorical), we define the low income variable to include all income classes below and including the (11440,17160] class, we define the high income variable to include the income classes (40040,57200] and (57200,85800], and we define the middle income variable to include all other income classes. To estimate the analog of column 1 of Table 2 (in which family income enters as a continuous variable), because the exact income value is not observed, we first fit a lognormal distribution to the categorical data (using both the individuals from Table 8 and the individuals that were removed from the final sample due to higher family incomes).... In PAGE 32: ... To estimate the analog of column 2 of Table 2 (in which family income is categorical), we define the low income variable to include all income classes below and including the (11440,17160] class, we define the high income variable to include the income classes (40040,57200] and (57200,85800], and we define the middle income variable to include all other income classes. To estimate the analog of column 1 of Table2 (in which family income enters as a continuous variable), because the exact income value is not observed, we first fit a lognormal distribution to the categorical data (using both the individuals from Table 8 and the individuals that were removed from the final sample due to higher family incomes). We then compute the likelihood contribution for person i by 30 relative to the tuition explanation), we estimate the duration model of attrition using data from the NELS-88.... In PAGE 32: ... 41 The descriptive statistics for these individuals is shown in Table 8. The analogs to column 1 and column 2 of Table2 are shown in Table 9. 42... In PAGE 33: ...amily income of $5,000. For the NELS-88 sample, these numbers are .71 and .64 respectively. 31 estimated effects of family income are very similar to those in Table2 , despite the fact that the students in the NELS-88 are attending tuition charging institutions. 43 Thus, although one should be very careful about the conclusions that can be drawn from this comparison, the exercise suggests that family environment factors may be the driving force in determining the strong relationship between family income and educational outcomes.... In PAGE 42: ...958 * represents t statistic greater than two. Sample size is smaller than the sample size in Table2 because the 172 students who did not stay in school long... ..."

### Table 4: TLB Results for tlb hit0 BP 1BNtlb hit1 BP 0

2002

"... In PAGE 9: ... When way 0 entry matches, the address is com- puted from the data stored in way 0 and tlb hit0 BP 1 (and vice verse). Table4 shows results based upon the assertion for tlb hit0 BP 1 and tlb hit1 BP 0. We emphasize that for symbolic simula- tion alone, if no manual effort is involved to adjust the variable ordering, the simulation would abort when more than 19 sym- bolic bits are used.... ..."

Cited by 1

### Table 2. Mixed Model on Mean Fixation Duration

2004

"... In PAGE 4: ...377 Second Page .353 Table2 shows the significant effects of independent variables on mean fixation duration1. The analysis reveals that the gender of a subject, the viewing order of a web page, and the interaction between page order and site type all significantly contribute to the differences in mean fixation durations.... ..."

Cited by 2

### Table 2. Logit Stepwise Modela

2000

"... In PAGE 13: ...verall classification and 34.0 percent Type II errors (i.e. misclassification of capital inadequate banks to be adequate).17 From the results of logit analysis in Table2 , 16 out of 48 variables were retained in the one-year prior logit model, and only six variables are significant at the 5 percent level or better. Out of these 16 variables, 13 were significant in the univariate t-tests using the same 1988 data (Table 1) -- thus, both the univariate and multivariate statistical analyses seem to identify the same significant predictors for the most part.... ..."

### Table 3. Descriptive statistics for variables used in the Ordered Logistic Regression

2006

"... In PAGE 14: ... Descriptive statistics of dependent variables after clustering Table 2. Descriptive statistics of independent variables after clustering Table3 . Descriptive statistics for variables used in the Ordered Logistic Regression Table 4.... ..."

Cited by 1