### Table 2. Estimates and tests of nuisance parameters

"... In PAGE 7: ... As expected, we can reject the no-overdispersion hypothesis p2 quot;0 for both sexes (cf. Table2 ). There is signi cant and positive serial correlation in the number of days on sickness insurance for females while not for the males.... ..."

### Table 3 Econometric Estimates of the

in and

"... In PAGE 13: ... The percent of village land that villagers would desire to transact through rental contracts in period t ( * it R ) take place AFTER any administrative reallocations and thus responds to residual opportunities for mutually beneficial exchange that are created by intra-village dispersion in the contemporaneous distributions of endowments and off- farm, wage labor opportunity, where the latter is denoted w it s . 6 Table3 presents descriptive statistical indicators of these dispersion measures, as well as measures of the amounts of land reallocated by both administrative and market- 6 In one of the few studies that focuses on private rental transactions in China, Carter and Yao (1998) use standard household models to show that households become more likely to undertake land rental transactions as their land :labor factor endowment ratio becomes more extreme and as the household enjoys specific factors and skills that encourage it to specialize in either agricultural or non-agricultural activities. The intra-village dispersion measures hypothesized here to influence the activeness of the village land market capture in a simple way the forces identified by Carter and ... In PAGE 19: ... 3.3 Results Table3 presents the parameter estimates th at result from maximizing (3-10) for two slightly different model specifications. In addition to these SML estimates... In PAGE 22: ....e. , latent factors that increase land rentals decrease land that is administratively reallocated). As can be seen in Table3 , the SML-1 estimates of endowments, quotas and property rights now carry the expected sign, but they are statistically insignificant. The estimates shown as the SML Panel-2 estimates in Table 3 result from a slight specification change.... In PAGE 22: ... As can be seen in Table 3, the SML-1 estimates of endowments, quotas and property rights now carry the expected sign, but they are statistically insignificant. The estimates shown as the SML Panel-2 estimates in Table3 result from a slight specification change. As noted above, the intra-village coefficient of variation in off- farm earnings changed only slight between 1988 and 1993 even though the magnitude of off-farm earnings more than tripled over this period.... In PAGE 22: ... Arguably, the intra-village standard deviation in off-farm earnings is the better measure of the potential gains from exchange in the land rental market. As can be seen in Table3 , using this latter measure has surprisingly little effect on the market rental equation. However, it has a major effect on the administrative reallocation estimates (note that this change in variable effects the administrative reallocation equation only through its impact on the estimate of the latent village effect).... ..."

### Table 2: Model I, simulated powers, based on estimated nuisance parameters

"... In PAGE 13: ...f is it dominated by the POI. This pattern can be found for all values of . Obviously (and naturally) all tests lose power as increases. Table2 reports the powers of the same tests, but estimates of the nuisance parameter replace the true values. The estimates were obtained under H0 by the EM algorithm.... ..."

### Table 1. Summary of nuisance parameters and model parameters used in the data descriptive equations for the Thomson scattering data. Symbol Nuisance parameter description Quantity Unit

### Table 6: Model II, simulated powers, based on estimated nuisance parameters

"... In PAGE 14: ... For larger values of F -tests with large g perform better. Table6 shows the same results. Here and G2;(22) have been estimated under H1, all other parameters under H0.... ..."

### Table 1: Model I, simulated powers, based on known nuisance parameters

### Table 5: Model II, simulated powers, based on known nuisance parameters

"... In PAGE 14: ... The results for model II show a similar pattern. Table5 clearly shows that the power rises as the correlation coe cient =2 increases. For small values of the LBI performs better than the corresponding F -tests, but it is clearly dominated by the POI test.... ..."

### Table 1: The Impact of NuisanceWeight on the Utility of a Node Node Name Utilities for NuisanceWeight = 0 = 2 = 5

1997

"... In PAGE 19: ... At present, the value for NuisanceWeight in radar is set at 2. Table1 contains the utilities of the nodes in the inference tree in Figure 6 for the three values of NuisanceWeight modeling the above described behaviours of an information provider. For each value, the utility of the chosen node is underlined.... ..."

Cited by 6