### Table 2. Classification error matrix of field-assigned vegetation class and satellite-derived vegetation classification in the Great Fish River basin. Figures in brackets are results obtained from spectral classification methods alone (prior to the incorporation of MSDI).

### Table 1. Assessor agreement for absolute judgments.

"... In PAGE 5: ... We chose to look at the latter in an attempt to average out differences in expertise, prior knowledge, or interpretation of the query. Agreement for absolute judgments is shown in Table1 . Each cell (J1, J2) is the probability that one assessor would say J2 (column) given that another said J1 (row).... ..."

### Table 3. Prior Specifications for u and s Parameters

in runoff

"... In PAGE 5: ... We employed conjugate priors for the u and s parameters arising in the autoregression models (5). First, we assume that conditional on the s parameters, the uij are mutually indepen- dent and have normal (Gaussian) distributions with means (E(uij) and variances tijsi 2, where the prior means E(uij) are listed in Table3 . The fixed constants tij are listed in Table 2.... In PAGE 5: ...d.(si 2) of the si 2 are given in Table3 . These quantities are given by the relations [e.... In PAGE 5: ...1, then var (si 2) 5 10(E(si 2))2. Also, the implied prior means and standard deviations of the uij are given in Table3 . The marginal prior variances of the uij are given by var (uij) 5 tijE(si 2).... In PAGE 7: ... We found no overwhelming evidence based on these diagnostics that the simulation did not reach reasonable equilibrium. In Table 5 we summarize the estimated posterior quantities; these may be compared with the prior inputs described in Table3 and 4. Note that the posterior means for all u regression coefficients are positive, in agreement with intu- ition.... ..."

### Table 13 shows the computed coefficients and statistics using the data of Table 3, for three cases: (1) assuming that each class is equally likely a priori, i.e. pi = 0:25; 8i (left column); (2) with unequal priors, in this case f0:1, 0:4, 0:1, 0:4g, i.e. with the highest expectation for the classes that are not mapped over the largest areas (center column); and (3) with unequal priors, in this case f0:4, 0:1, 0:4, 0:1g, i.e. with the highest expectations closely matching the largest mapped areas (right column).

"... In PAGE 29: ...Actual proportions f0:3742, 0:1104, 0:3865, 0:1288g Table13 : Tau statistics for the example confusion matrix With equal priors (left column), 0 2 is substantially lower than the chance agreement 2 computed for ^ k. This leads to a value of that is significantly higher than ^ k.... In PAGE 29: ... If the prior probabilities are quite different from the actual proportions (center column), tau is even closer to the overall accuracy; this is because the map is not what was expected, so agreement is a pleasant surprise. In this case, as shown in the rightmost column of Table13 , there is a smaller chance agreement, the index of agreement is closer to the overall accuracy, and the variability of the estimate is lower. The map is providing a large increase in information over the prior state of knowledge.... ..."

### Table 3. Posterior quantities based on 50000 Gibbs cycles under the noninformative prior.

"... In PAGE 13: ... We note that the prior distributions for #0E 0 ;#0E 1 and #0E 2 in the noninformative case are improper but satisfy the conditions of the theorem. A simple comparison of the results based on the two priors can be made by comparing the summary statistics in Table 2 with those in Table3 . We note that there is general agreement in the marginal posterior distributions of #28#12#12; #16#16; #0E#0E;#1A#1A; #0E 0 #29.... ..."

### Table 8 Number of Samples for Each Cement to be Tested in Compression Prior to Exposure (7 day), After 28 Days of Sulfate Exposure, and After 60 Days of Sulfate Exposure.

1999

"... In PAGE 10: ...able 7: Description of Curing Regimen for Cements Evaluated...............................................53 Table8 : Number of Samples for Each Cement to be Tested in Compression Prior to Exposure (7 day), After 28 Days of Sulfate Exposure, and After 60 Days of Sulfate Exposure.... In PAGE 64: ... However, compression tests will be repeated after 60 days of immersion for further verification. Table8 describes the test plan and the number of samples to be tested at each interval. The number of samples tested at each interval depends upon the agreement of the... ..."

### Table 1: Comparison between the re nement methods. n denotes the number of partici- pating agents and L is the number of possible agreements of the negotiation.

1995

"... In PAGE 12: ....3.1 Preliminaries In order to evaluate the bene ts of the learning agents, we compare their performance with two other non-learning approaches: communication and ignorance. Table1 shows the projected performance (in terms of the number of messages) for the three methods. The communication method requires the exchanges of full information about preferences prior to negotiation.... ..."

Cited by 1

### Table 2: Licensing agreements

"... In PAGE 4: ... Table2 shows that many libraries in CEE countries have been exposed to licensing. Licensing consortia are becoming more and more common in CEE countries.... ..."

### Table 4. Priors by Country

2007

"... In PAGE 38: ... Table4 . Priors by Country (Cont.... In PAGE 39: ... Table4 . Priors by Country (Cont.... ..."

### Table 3: Schedule of Experimental Sessions

"... In PAGE 4: ... They had read a simplified explanation of the study prior to their agreement to attend, and were given a more detailed explanation by videotape on arrival. Thus began the day, which proceeded as shown in the schedule shown in Table3 . The session was conducted by an experimenter who was blind to the hypotheses under test, although she was aware that different lighting systems were in use in various experimental sessions.... In PAGE 4: ... At the conclusion of the final questionnaire, participants were debriefed by videotape and provided with written information about the purpose of the experiment. The following sections describe the tasks grouped according to the outcome categories to which they relate (refer to Table3 for the sequence of presentation). Demographics and individual differences.... ..."