### Table 3 Execution times in microseconds for TARP operations (Average of 100 measure- ments).

2005

"... In PAGE 19: ... Finally, TARP gains significant performance improvements by amortizing the cost of ticket generation. Table3 summarizes the micro-benchmarks. The experimental environment was more controlled than that of the macro-benchmarks, therefore, even with 100 runs, a small standard deviation was achieved.... ..."

### Table 7.1. Confirmatory factor analysis for the Finn study (factor loadings below 0.35 not shown)

### Table 9 The Effect of CON Restrictions on the Number of Finn and Firm Diver:ific:ation

"... In PAGE 105: ... Rather, they are only for a subset of the firms supplying home health services in an area. DIFFERENCES IN MARKET STRUCTURE Based on some simple regression analyses reported in Table9 , it appears that the number of firms providing home See pp. 32-34 for a more complete discussion of the - quot; economic quot; theory of these regulations.... In PAGE 107: ...Table9 --Continued PDE = 3.... ..."

### Table 8. Implied Lag Valoes from Finn Demand Estimates Assuming Partial Demand Adjustment

1995

"... In PAGE 59: ... In all cases, F tests reject the joint hypothesis that all coefficients in a PDL structure are zero at a high confidence level. Table8 and figures 5 and 6 present the individual lag coefficients implied by the PDL estimates. While the lag structure differs considerably between the second and third order specification, the implied long-run elasticities are quite similar.... ..."

### Table 1. Coder efficiency comparison between an em- bedded coder (Tarp), non-embedded coder (ML), and non- embedded coder conditioned on the quantization step-sizes (MLjQ).

2005

"... In PAGE 2: ... When cre- ating compressed images at threshold, this maneuver (non- embedding + conditioning) can save over half the rate spent on coding the side information. Table1 compares the rate required to code the quantized data with (1) an embedded bit-plane coder, (2) a non-embedded coder (similar to the coder above, which codes the significance map without any conditioning) and (3) the proposed coder. An embedded Tarp-filter based coder is chosen as a basis for a comparison with the state-of-the-art, and the coder without conditioning is used to accurately illustrate the effect of the condition- ing.... ..."

Cited by 1

### Table 24: Neuroimaging volume file formats. ECAT and Interfile information kindly provided by Frdric Schoe- nahl. The author (Finn) is not sure that the information about the DICOM format is correct.

### Table (2) Hansen and Tarp Specification (1) (2) (3) (4) (5) Average annual per capita growth lagged -0.088 -0.028 -0.100 0.004 -0.119

### Table 1: The rst column gives the transformation index, while the second the base transformations. The remaining columns gives the constants required to make the normalized transformation scale and location invariant. These constants are provided for each of the three dependent variables considered in our dataset. The last two transformations were proposed by Finn (1998) and Hanssens and Weitz (1980)

"... In PAGE 16: ... In e ect, these constants normalize the base transformations considered to make them scale and lo- cation invariant{ a property that eases the qualitative interpretation of the regression estimates. Table1 presents the base transformations that we consider, which include those of Finn (1988) and Hanssens and Weitz (1980). The other seven are of the form ?1(y ), which are probit transformations with various values for the skewing parameter .... In PAGE 16: ... As implemented, our methodology may lead to di erent transformations for the noted, associated and read-most scores. |{ Table1 About Here.|{... ..."