### Table 2. Feature selection for the breast cancer data: 51 features used by Hendenfalk et al. (2001)

2004

"... In PAGE 7: ... There are no misclassifications by Models I and II and two misclassi- fications (17th and 18th sample) by Model III. We compare our CV results with other popular classification algorithms in Table2 . All other methods have used 51 genes.... ..."

Cited by 10

### Table 2: Regressions Using the Log of Educational Attainment

1996

"... In PAGE 22: ... In fact, this is exactly what a model based on the Mincerian micro- foundations suggests, as shown in the previous section.16 Table2 illustrates the puzzle by including the logarithm of average edu- cational attainment in the regression. The speci cation is rst estimated in levels for 1990 and 1960, treating both years as steady state observations.... ..."

Cited by 1

### Table 2: A (2; 2) robust threshold scheme.

"... In PAGE 4: ... The bound on the size of the shares for di erent models of c-compact (k; n) threshold schemes with cheaters provided by Theorem 15 and Corollaries 16 and 20 of [5] are not correct since their derivation is based on the erroneous bound (2). For instance, the (2; 2) robust threshold scheme depicted in Table2 violates the bound provided by Theorem 15 of [5]. Using the corrected version of De Soete apos;s bound given by Theorem 2.... ..."

### Table 3: Overdispersed Multinomial Regression Model for 2000 Election Results, Florida Counties, Robust Parameter Estimates (FEC Data)

"... In PAGE 12: ...9 Each regressor xij3 is the rst principal component of the residuals produced as before. Tanh estimates for the parameters of the FEC model, along with sandwich standard errors, appear in Table3 . As happened with our base model de ned in (12), the estimates are precise only for the coe cients of the linear predictors for Gore and Bush.... In PAGE 12: ... Moreover, as measured by the estimated LQD dispersion parameters, the FEC model ts the data signi cantly worse than the base model: ^ = 7:01 for the FEC model versus ^ = 5:39 for the base model. *** Table3 about here *** 9For Gore and Bush we use the contributions given to any Democratic or Republican presidential candidate in the 2000 election cycle|i.e.... ..."

### Table 1: Comparison of the L1 norm of the error

"... In PAGE 4: ...Table 1: Comparison of the L1 norm of the error (a) Low error (b) High error Figure 3: Example tetrahedra relative to the u = 0:5 isosurface Table1 compares the L1 norm of the error for the model hyperbolic problem for an isotropic h- re nement scheme and for a hr-re nement scheme using the node movement algorithm (4) prior to each h-re nement step. In this case two Jacobi sweeps were performed on the mesh with a xed solution, then the solution was recomputed.... In PAGE 4: ... The results indicate that while the node movement scheme can help to reduce the error in the problem the improvement is not great in this case, although a lower error was obtained on each grid at the end of the r-re nement stage. Table1 (b) shows that the results are improved by choosing a more aligned initial mesh (in the sense of Iliescu [8]) as expected. There are various parameters in the adaptivity algorithm that can be tuned, such as the number of sweeps and the number of times the solution is recomputed during r-re nement, and improved results can be obtained by repeated experimentation.... In PAGE 7: ... The scheme leads to low solution errors quite rapidly, but the convergence stalls after the nal entry in the table. However the errors are much lower than those produced by the interpolation error scheme in Table1 for the same resolution, and are also superior to h-re nement based on the functional value. Figure 6 shows two tetrahedra with low functional value near the layer on the rst grid after optimisation.... ..."

Cited by 1

### Table 1: Overdispersed Multinomial Regression Model for 2000 Election Results, Florida Counties, Robust Parameter Estimates (Base Model)

"... In PAGE 10: ...ources for the vote and demographic data are as documented in Wand et al. (2001). Because xij2 varies over categories, the principal components also vary over categories. Tanh estimates for the parameters of (12), along with sandwich standard errors, appear in Table1 . Estimates are precise only for the components of the linear predictors for Gore and Bush.... In PAGE 10: ... Estimates are precise only for the components of the linear predictors for Gore and Bush. *** Table1 about here *** Figure 1 plots the studentized residuals for Buchanan, Nader, Gore and Bush against the expected vote proportion for each candidate in each county. For all four candidates the residuals 8As in Wand et al.... ..."

### Table 1 Determinants of Household Computer Ownership Descriptive Statistics and Model results Logistic Regression Model Results Descriptive Statistics

2006

"... In PAGE 18: ... Due to limited observations in some ethnic categories, only the categories of White, African American, White Hispanic, and Other or Mixed Races are used. Income is reported in a set of ranges in this data set (see Table1 ). Based upon the discussion above, we would expect higher income and higher educational attainment to be associated with greater computer ownership and would expect minority households to have lower computer ownership levels than white households.... In PAGE 18: ... Based upon the diffusion literature and past studies of computer use, we expect the presence of students to positively influence computer ownership and the presence of students who use a computer at school to cause a still higher rate of computer ownership. Summary statistics for the set of data are presented in the last three columns of Table1 . For dichotomous and categorical variables the frequency of each category and percent of the sample it represents are shown.... In PAGE 20: ...7 percent correct classifications with the model. Review of the model results in Table1 indicates that virtually all of the variables in the model were of the expected sign and were statistically significant. Of the variables hypothesized to affect computer ownership, only the presence of students without any computer use at school had an unexpected impact, but this coefficient was not statistically significant.... In PAGE 42: ... The results of our analyses are displayed in the following tables. Table1 : Results of the analysis on the image criteria. Criteria Comply with criteria Do not Comply No data Image size is 600*400 pixels or smaller 78% 13% 9% Use no more than 3 images per page 14% 84% 2% For some pages, it was impossible to determine the size of the largest image since Macromedia Flash was used.... ..."

### Table 5 Robustness checks

2004

"... In PAGE 16: ....3. Robustness of the static model Having taken as our starting point the specification of Cutler and Gruber (1996), we now examine the robustness of our results to changes in specification. Table5 presents these specification checks, beginning with a model including only eligibility and ending with our preferred specification that we will use for the remainder of the paper. From the first three columns, it is clear that while there is a negative relationship between eligibility and private coverage in the regression with no controls, this estimate of the effect of eligibility on private coverage is biased downward by the omission of age and year effects.... ..."

### Table 2: Summary of results for cross country regressions

"... In PAGE 7: ... We use the 1993 version of this classification, which classifies countries into five groups: (i) Low income economics (LIC) with a 1991 per capita GDP level of $635 or less, (ii) Lower middle income economies (MICL) with a 1991 per capita GDP level of $636-$2,555, (iii) Upper middle income economies (MICU) with a 1991 per capita GDP level of $2,556-$7,910, and High income economies with a 1991 per capita GDP level of $7,911 which are divided into (iv) OECD members (HICOECD), and (v) others (HICMISC). A summary of the results for these cross-country regressions are summarized in Table2 , while the detailed regression results are listed in Table A.2.... In PAGE 7: ... Convergence for all countries in the sample is rejected. As can be seen from Table2 , we find a lot of convergence when we consider the levels of the social indicators and much less when we consider their achievements. 3.... ..."