### Table 1. Software failure regression results.

2005

"... In PAGE 6: ... Consequently, they re ect differences in data reporting and collection. The positive coef cient in Table1 for the Svc variable may ap- pear to be counterintuitive because having a service agree- ment should help reduce error rate, i.e.... In PAGE 6: ....1.2 Predicting software failures To have results applicable in practice we predict software failures for a new release based on our tted using previ- ous releases (in this case one previous release). We re t- ted the model in Table1 to perform the prediction. We used only the most signi cant predictors that had p-values below 0:01 as shown in Table 2.... ..."

Cited by 3

### Table 2: Posterior Estimates Using the Power Prior for the Logistic Regression Model

"... In PAGE 16: ... From Tables 1 and 2, we can see that the posterior estimates of fl using the power prior with a0 = 0:415 are fairly close to those obtained under the hierarchical model. From Table2 , we also see that the posterior estimates are slightly difierent for the difierent values of a0. In particular, when more weight is given to the historical data, age and treatment become more \signiflcant quot;, that is, their 95% Highest Posterior Density (HPD) intervals do not include 0.... ..."

### Table 1. Outcome and Cost Estimates Used in the Decision Tree*

"... In PAGE 2: ...alization. Decision-analytic software (DATA 3.5; Tree- age, Williamstown, Mass) was used for all analyses. LIKELIHOOD OF CLINICAL EVENTS The probabilities and ranges of clinical events used in the decision model are shown in Table1 . The baseline risk of bacteriuria over the hospital stay in the control group (patients not receiving systemic antimicrobial agents and given short-term standard noncoated catheters) was sta- tistically pooled from several prospective studies.... In PAGE 2: ... OUTCOME ASSESSMENT AND SENSITIVITY ANALYSES The outcomes estimated using the decision-analytic model for 1000 hospitalized patients requiring short-term cath- eterization were the incidence of symptomatic UTI and bac- teremia and direct medical costs. To determine the effects of specific inputs on the results, we performed a series of 1-way sensitivity analyses using the ranges in Table1 . No- tably, the low estimate for the incidence of symptomatic UTI was 8% and that for the incidence of bacteremia was 0%.... In PAGE 3: ...The 1-way sensitivity analyses revealed that the silver- coated catheter strategy remained dominant through- out the ranges evaluated in Table1 . We determined threshold values for the following variables since they were particularly influential in the 1-way sensitivity analyses: baseline incidence of bacteriuria in the control group, probability of developing symptomatic UTI after bacte- riuria, efficacy of silver-coated catheters, and cost of a sil- ver-coated catheter.... ..."

### Table 5. Regression models and their explanative power Project Number of principal

"... In PAGE 6: ... We thus obtained a predictor that would take a new entity (or more precisely, the values of its metrics) and come up with a failure estimate. The regression models built using all the data for each project are characterized in Table5 . For each project, we present the R2 value which is the ratio of the regression sum of squares to the total sum of squares.... In PAGE 6: ... The F-ratio is to test the null hypothesis that all regression coefficients are zero at statistically significant levels. How does one interpret the data in Table5 ? Let us focus straight away on the R2 values of the regression models. The R2 values indicate that our principal components explain between 57.... In PAGE 7: ... However, for any top n predicted modules getting extra effort, one would never see more than one module not deserving that effort, and never more than one of the top n actual modules missed. All in all, both the R2 values in Table5 and the sensitivity of the predictions in Table 6 confirm our hypothesis H3 for all five projects, illustrated by the examples in Figure 2. In practice, this means that within a project, the past failure history of a project can successfully predict the likelihood of post-release defects for new existing entities; therefore, the predictors can also be used after a change to estimate the likelihood of failure.... ..."

### Table 5: Regressions of Matching Probabilities (U.S. Academic Market for New PhD Economists) Model (10) Model (11) Model (12) Job Market in January 2000

"... In PAGE 20: ...Table5 gives the estimation results of models (10), (11), and (12), in the same format as that of Table 2. It is clear that the regressions based on model (12) have the best t, followed by model (11).... In PAGE 21: ... These two elds have large numbers of openings while there are no candidates labeled as \any eld quot; or \general economics. quot; Estimation results not reported here (they are available upon request) show virtually identical parameters estimates as well as the goodness-of- t R2 for the results given in Table5 . Thus the fact that there are zero candidates in the thick elds \AF quot; and \A1 quot; does not a ect our estimation results nor the conclusions derived from them.... ..."

### Table 5. Parameter estimations for the logistic regression model

"... In PAGE 9: ... A large i indicates an important modality, a small i indicates a modality that does not contribute very much. Table5 shows the values for the parameters i obtained for the three modalities in our application. These values have been calculated on the training data using the SPSS software package32.... ..."

### Table 2 Hierarchical Regression Results

2002

"... In PAGE 8: ...1. Hierarchical Regression Results We report the results of the hierarchical regressions of the diffusion parameters in Column I of Table2 . The estimated effects of the various explanatory variables on the three BDM parameters are mostly consistent with the expected effects hypothesized in the previous sec- tion.... In PAGE 9: ...nd 0.2% respectively.8 The finding is of potential im- portance in assessing the attractiveness of emerging markets such as China, where such macroenviron- mental characteristics are undergoing rapid change. As noted in Table2 , our study is the first to present such empirical insights on penetration potential in the marketing literature. With respect to the estimated effects of the various variables on the coefficients of external and internal influence (which determine the speed of diffusion), we find a strong negative result for illiteracy level as expected.... In PAGE 11: ... Finally, since consumer and business products may have different diffusion patterns, we estimated the model by restricting the analysis to only consumer products (VCR players, camcorder, microwave, and CD players) and dropping cellular phones and fax machines, which are used by both consumers and businesses. The estimates are reported in column II of Table2 . There were hardly any differences in the estimates of penetration level.... ..."

Cited by 5

### Table 5. Parameter estimations for the logistic regression model

"... In PAGE 9: ... A large a15 a13 indicates an important modality, a small a15 a13 indicates a modality that does not contribute very much. Table5 shows the values for the parameters a15 a13 obtained for the three modalities in our application. These values have been calculated on the training data using the SPSS software packagea64a69a70 .... ..."

### Table 4: Parameter estimations for the logistic regression model

"... In PAGE 5: ... A large i indi- cates an important modality, a small i indicates a modality that does not contribute very much. Table4 shows the val- ues for the parameters i obtained for the three modalities in our application. These values have been calculated on the training data using the SPSS software package [12].... ..."

### Table 2. Source data for alternative retrofit actions.

"... In PAGE 28: ... The selection can be justified due to the ability of this method to take into account the ecological rucksack of a product. The MIPS method determines the total mass of abiotic and biotic materials, water and air (see: Table2 ) that are consumed during the entire life cycle of a product. ... In PAGE 29: ...Table2 . Examples of items in the various categories of natural resources (Sinivuori amp;Saari, 2006).... In PAGE 61: ... Specific hot water demands for different building types are summarized in Table 2. Table2 . Specific hot water demands.... In PAGE 66: ...061 0.061 Table2 . Energy consumption of the building estimated by the HOT2000 software.... In PAGE 67: ...15 U-value, windows, W m-2K-1 1 U-value, doors, W m-2K-1 0.5 Table2 . Reference energy prices (January, 2006)1.... ..."