Results

**1 - 9**of**9**### Table 2 Size-Based Illusory Correlation: Learning History (Based on the Design and Procedure of McConnell et al., 1994, Experiment 2)

2003

"... In PAGE 10: ... In addition, increased memory for undesirable minority behaviors is driven by the competition property of the delta algorithm as described earlier. Table2 represents a simplified simulated learning history of a typical illusory correlation experiment as conducted by McConnell et al. (1994, Experiment 2).... In PAGE 11: ...e., likability ratings, frequency estimations, and group assignments), the group nodes were turned on and the resulting activation of the evaluative nodes was read off (denoted by a question mark, see bottom panel of Table2 ). As noted earlier, no additional external activation was provided to the evaluative nodes (or any other measurement node) because null activation is a neutral resting activation state that allows an unbiased assess- ment of the evaluative activation generated directly or indirectly by the group nodes.... In PAGE 11: ... In a group assignment task, behaviors are presented and participants have to indicate as fast as possible by which group member they were performed. To reflect this measure, each episodic node from different sets of behaviors (AH11001,AH11002,BH11001,BH11002) was activated one at a time (see bottom panel of Table2 ). This episodic activation spreads to the group nodes and so determines response times.... In PAGE 11: ...3 Results. The 18 statements succinctly listed in Table2 were processed by the network for 50 participants with different random orders. Figure 5 depicts the mean test activation for all simulated dependent measures, together with the observed likabil- ity and reaction time data from McConnell et al.... ..."

Cited by 7

### TABLE 3 Regression Model

in Public Utility Companies: Institutional Ownership And The Share Price Response To New Equity Issues

### Table 1: Timing information for the models used for the paper: 1. Dodecahedron, 2. Torus-I, 3. Open surface, 4. Torus-II. is 0.00 for all models. All timings are in sec- onds. Torus-II is the twice Doo-Sabin subdivided Torus-I.

"... In PAGE 6: ...priate result is obtained in the next time cycle. 7 Results Table1 presents a set of models used. The first column gives the number of vertices, edges and faces for the cor- responding model.... In PAGE 6: ... Torus-II is the twice Doo-Sabin subdivided Torus-I. Figure 3 presents the partial energy against T=Tmax for model 1 ( Table1 ). The partial energy plot diverges for T=Tmax greater than 1.... ..."

### Table 2. Comparison of macro elements for a given smoothness r. The best elements will have the lowest degree possible, will have stable dimensions, and will use the least number of degrees of freedom. We begin by examining the dimension of the superspline space S := fs 2 Sr d(4v) : s 2 C v(v)g de ned on a Powell-Sabin cell. By Lemma 3.2 of [6], dimS = v + 2 2

1999

"... In PAGE 17: ...1) Our macro-elements have two advantages over these macro-elements: they use lower degree polynomials, they use a smaller number of degrees of freedom, for r 5. Table2 shows a comparison of the macro-elements in (9.1) with our new macro elements for 1 r 10.... ..."

Cited by 10

### Table 1 gives the runtime statistics for the local deformation examples in Figure 16. The dolphin model contains 15,774 vertices. The control mesh contains 44 vertices, 48 faces for the Doo-Sabin subdivision surface, and 44 vertices, 84 faces for the Loop and Modified Butterfly subdivision surfaces since the control meshes are triangulated for the later two cases. The runtime data is collected from a PC with Pentium-IV 1.7GH CPU, 256M memory and Windows 2000 OS. The runtime for attaching the object to the subdivision surface is high in general. As it is computed only once, it will not influence the system response. We also can conclude from Table 1 that the runtime of three deformation methods are of the same order, they increase rapidly with the increase of subdivision depth. For the models with tens of thousand vertices, the deformation runtime can fulfill the realtime requirement under the subdivision depth 4. According to our experiments, the Doo-Sabin method is more efficient than the others.

"... In PAGE 14: ... Table1 : Runtime comparison among three subdivision methods, where Para-Time is the runtime for attaching the object to the subdivision surface, and Deform-Time is the runtime for computing the deformed object. Figures 17 and 18 are deformation examples using the primitive control meshes: sphere and cube.... ..."

Cited by 1

### (Table 3.9). The Texas Gulf Coast includes the seven bay systems including Sabine Lake and the Galveston, Matagorda, San Antonio, Aransas, Corpus Christi, Upper Laguna Madre, and Lower Laguna Madre bay systems. The commercial fishing industry is largest in the Trinity-San Jacinto estuary region. Some estuaries that have a relatively low catch attributable to their bays from both inshore and offshore fishing, have a sizeable commercial fishing industry when landings from all Gulf areas are considered (such as the Sabine- Neches estuary region in Table 3.9). Direct impacts of commercial fishing in the Texas Gulf Coast region were estimated as $205 million in the 1987 study (Fesenmaier et al., 1987), compared to $175 million in 1995, representing a decrease of about 15% in current dollars (Table 3.9). In order to compare the value of output from commercial fishing in real terms, direct impacts for 1987 and 1995 were deflated by the Producer Price Indices (PPI) for those years. In real dollars, direct impacts of commercial fishing in 1987 and 1995 were $199 and $141 million respectively, showing an even larger decrease of about 29% from 1987 to 1995.

### Table 1: Number of vertices for the triangulations successively re ned by the global triadic subdi- vision scheme, and by the vertex-centered local subdivision scheme with di erent -values.

"... In PAGE 17: ... The dimensions of the successively re ned Powell-Sabin spline spaces are 63, 102, 195, 273 and 450. In Table1 we compare the growth of the number of vertices for the global triadic scheme from [19] with our local scheme. Table 2 gives the minimal angle of the corresponding meshes, and the mean value of the minimal angle in the triangulations.... ..."

### Table 1: Convergence rates for several = ( 1; 2)T for uniform re nement (upper) and adaptive re nement (lower). 4. References 1 Bank, R.E., Dupont, T.F., Yserentant, H.: The hierarchical basis multigrid method. Numer. Math. 52, 427-458 (1988) 2 Bank, R.E., Gutsch, S.: Hierarchical basis for the convection-di usion equation on unstructured meshes. Ninth Interna- tional Symposium on Domain Decomposition Methods for Partial Di erential Equations (P. Bj rstad, M. Espedal and D. Keyes, eds.), J. Wiley and Sons, New York, (1996) to appear 3 Bank, R.E., Gutsch, S.: The generalized hierarchical basis two-level method for the convection-di usion equation on a regular grid. submitted to the Proceedings of the 5th European Multigrid Conference in Stuttgart (1996) 4 Bank, R.E., Xu, J.: The hierarchical basis multigrid method and incomplete LU decomposition. Seventh International Symposium on Domain Decomposition Methods for Partial Di erential Equations (D. Keyes and J. Xu, eds.), 163-173. AMS, Providence, Rhode Island (1994) 5 Reusken, A.A.: Approximate cyclic reduction preconditioning. Preprint RANA 97-02, Eindhoven University of Technology Addresses: Randolph E. Bank, Department of Mathematics, University of California at San Diego, USA Sabine Gutsch, Mathematical Seminar II, Christian-Albrechts-University Kiel, Germany

### Table 3: Sobolev and Holder smoothness for some schemes with symbol (43)

"... In PAGE 19: ...g., Table3 in Section 3.1) and Table 4 in Section 3.... In PAGE 25: ... Finally, let us check for sum rules of order k = 4, where ~ D(2;1)P (~ !j) = ? ~ D(1;2)P (~ !j) = 1 3(a ? 8b)(~ zj ? ~ z2j) ; ~ D(3;0)P (~ !j) = ~ D(0;3)P (~ !j) = 0 ; j = 0; 1; 2 : Thus, we obtain one more additional condition, a = 8b ; (46) which together with (44) and (45) implies that the only scheme in the considered class with sum rules of order k = 4 is given by a = 8=27 ; b = 1=27 ; c = 2=27 ; d = 5=9 : (47) Since ~ D(4;0)P (~ !j) = 12c + 2 3(a + 16b)(~ zj + ~ z2j) 6 = 0 ; the order k = 4 is exact. The computed values of stability indicators and smoothness exponents for the only scheme with kmax = 4, some schemes from the one-parameter family a = 1 3 ? b; c = 1 18 + b 2; d = 2 3 ? 3b (0 b 2 9) with kmax = 3, and two more schemes with kmax = 2 are shown in Table3 . All considered masks are non-negative, however, only the rst scheme satis es P (!) 0.... In PAGE 25: ... All considered masks are non-negative, however, only the rst scheme satis es P (!) 0. The stability indicators ~ stab correctly predict the schemes with unstable (although the numbers in the 4th and 5th row of Table3 are relatively small, the internally set error tolerances for StableD.m are such that values ~ stab 1e ? 8 safely indicate stability).... In PAGE 26: ... This is because the limiting surface is piecewise quartic C3 with respect to a Powell-Sabin split of the initial triangulation. The examples with kmax = 2 in Table3 have unstable , see the discussion at the end of Section 3.3.... In PAGE 26: ...3. It should be noted that most of the Holder exponents shown in the last column of Table3 are anticipated values based on our numerical evidence. The exception is the rst scheme, where one can prove P (!) 0 and use the above-mentioned characterization of the critical Holder exponent from [15].... In PAGE 27: ...Example 2 Example 3 Example 4 l sl 1( ) sl 1( ) sl 1( ) sl 1( ) sl 1( ) sl1( ) sl 1( ) sl 1( ) 1 3:3142 0:6658 2:6309 0:2584 3:0000 0:7780 3:0000 1:2244 2 3:3142 1:2107 2:6309 0:7154 3:0000 1:3311 2:7491 1:6549 3 3:3142 1:8107 2:6309 1:1944 3:0000 1:7356 2:9550 1:8659 4 3:3142 2:1759 2:6309 1:5319 3:0000 1:9665 2:7491 2:0358 5 3:3142 2:3990 2:6309 1:7384 3:0000 2:1661 2:9419 2:1555 6 3:3142 2:5522 2:6309 1:8946 3:0000 2:2756 2:7491 2:2280 7 3:3142 2:6608 2:6309 1:9825 3:0000 2:3858 2:9092 2:2938 8 3:3142 2:7427 2:6309 2:0766 3:0000 2:4484 2:7491 2:3416 9 3:3142 2:8061 2:6309 2:1209 3:0000 2:5176 2:8884 2:3841 10 3:3142 2:8570 2:6309 2:1855 3:0000 2:5566 2:7491 2:4200 11 3:3142 2:8985 2:6309 2:2105 3:0000 2:6040 2:8766 2:4484 Example 5 Example 6 Example 7 l sl 1( ) sl 1( ) sl 1( ) sl 1( ) sl 1( ) sl 1( ) 1 2:4531 0:7431 2:0000 0:0851 1:4321 0:2247 2 2:2849 1:3167 2:0000 0:5278 1:4321 0:6872 3 2:4423 1:5664 2:0000 0:8207 1:4321 0:9557 4 2:2849 1:7048 2:0000 1:0207 1:4321 1:1157 5 2:3568 1:8089 2:0000 1:1654 1:4321 1:1997 6 2:2849 1:8831 2:0000 1:2750 1:4321 1:2499 7 2:3438 1:9476 2:0000 1:3605 1:4321 1:2774 8 2:2849 1:9762 2:0000 1:4289 1:4321 1:3018 9 2:3231 2:0210 2:0000 1:4848 1:4321 1:3170 10 2:2849 2:0401 2:0000 1:5313 1:4321 1:3309 11 2:3226 2:0728 2:0000 1:5704 1:4321 1:3406 Table 4: Upper and lower Holder smoothness estimates for Table3 and l 11... In PAGE 34: ... It turns out (we leave this as an exercise for the interested reader) that we arrive at an approximate scheme of the type discussed in Section 3.2, with parameters a = d = 1=3 ; b = 0 ; c = 1=9 : (59) Unfortunately, the scalar for this set of parameters is also unstable (see Table3 ), and the maximal order of sum rules is only kmax = 2, i.... In PAGE 40: ... Setting KDiff to some positive integer allows to display information about partial derivatives (divided di erences) of re nable vectors; this option is not used below. We include pictures for all three examples of Table 1, some of Table3 (this also covers the schemes (V V; 1) and (V V; 2) from Table 5) and of Table 6. The graph for the interpolatory scheme (1) or, what is the same, (V V; 1) shown in Figure 8 exhibits a point singularity which was mentioned in the introduction.... In PAGE 41: ...Table3 , from [25] Figure 12: Graph of : Approximating scheme No. 1 in Table 3, same as (V V; 2) [5] C.... In PAGE 41: ...Figure 12: Graph of : Approximating scheme No. 1 in Table3 , same as (V V; 2) [5] C.... ..."

**1 - 9**of

**9**