### Table 7. Comparisons of gambles with

in Economics

2002

"... In PAGE 6: ...hart 1. Parameter stability of previous coefficient ............................................27 Table7 .... In PAGE 6: ...able 6. Average valuations by decision frame....................................................42 Table7 .... In PAGE 7: ...able 6. Average valuations for gains and losses.................................................73 Table7 : Description of variables in regression models.... In PAGE 37: ... Table7 . Logit results: Models including dealership and manager dummies The dependent variable is PURCHASE, which equals 1 if a contract was purchased.... In PAGE 52: ... Of course, both gambles have an expected value of $5. Table7 compares gambles with similar ranges for the pooled data and for the context and abstract treatments. These gambles are over possible gains.... In PAGE 84: ... The independent variables used in this analysis are presented in Table 7. Table7 : Description of variables in regression models EV Expected value of gamble COV Coefficient of variation of gamble Order = 1 if investment decisions completed first = 0 if insurance decisions completed first Range measure of ambiguity in the gamble (ex. A gamble of a 0-20% chance at $50 has a range of EVs of $0-10.... ..."

### Table 6: P-values of two-sided pairwise t tests between two levels of linkage disequilib- rium at the true trait locus. The signiflcance level is 0:05=15 = 0:003. Asterisk * indicates signiflcance after the Bonferroni correction.

2007

"... In PAGE 26: ... We carried out pairwise two-sided t-tests to compare lod scores between any two LD levels at the true trait locus, adjusting the p-values using the Bonferroni correction. Table6 lists the p-values of these pairwise comparisons.... ..."

### Table 6--Mean scores for relative performance of attributes for 3 endemic cooking cultivars by elevation

2004

### Table 1: A performance summary of the described methods. Starred and plus-marked results are highly significant against the noOV-noLM and noOV-LM systems, respectively, according to the two-sided t-test (D4 AK BCBMBCBCBE). Italicized and boldface results are significant against noOV-noLM and noOV-LM according to the sign test (D4 AK BCBMBCBH).

2003

"... In PAGE 6: ... 4.5 Experimental Results In Table1 , we show results for each of the exper- imental conditions described above. The scores in Table 1 are evaluated using only one reference trans- lation, which yields lower scores than the multiple reference translations used in DARPA evaluations.... In PAGE 6: ...5 Experimental Results In Table 1, we show results for each of the exper- imental conditions described above. The scores in Table1 are evaluated using only one reference trans- lation, which yields lower scores than the multiple reference translations used in DARPA evaluations. The highest performing system is highlighted in the table for both BLEU and NIST.... ..."

Cited by 9

### Table 1: A performance summary of the described methods. Starred and plus-marked results are highly significant against the noOV-noLM and noOV-LM systems, respectively, according to the two-sided t-test (D4 AK BCBMBCBCBE). Italicized and boldface results are significant against noOV-noLM and noOV-LM according to the sign test (D4 AK BCBMBCBH).

"... In PAGE 6: ... 4.5 Experimental Results In Table1 , we show results for each of the exper- imental conditions described above. The scores in Table 1 are evaluated using only one reference trans- lation, which yields lower scores than the multiple reference translations used in DARPA evaluations.... In PAGE 6: ...5 Experimental Results In Table 1, we show results for each of the exper- imental conditions described above. The scores in Table1 are evaluated using only one reference trans- lation, which yields lower scores than the multiple reference translations used in DARPA evaluations. The highest performing system is highlighted in the table for both BLEU and NIST.... ..."

### Table 7 shows per-topic comparisons of our submitted runs with the medians in their category for the measures reported by NIST: Average Precision, Precison@10, Precision@20 and Precision@30, respectively. In each comparison, we show the number of topics on which the run scored higher than the median, lower than the median and tied with the median (Higher-Lower-Tied). Differences statistically significant at the 1% level by the two-sided Wilcoxon signed rank test are marked with two asterisks (**), and differences just significant at the 5% level are marked with a single asterisk (*):

2002

"... In PAGE 9: ... Table7 : Per-topic comparison of Submitted runs with Medians The per-topic comparisons show a lot of ties in the early precision scores, particularly P@10, because of the small number of documents considered. Still, in each measure, the difference of the hum01tlx and hum01tl runs with the medians is statistically significant at the 1% level by the two-sided Wilcoxon signed rank test (the calculation of the significance level discards the ties, following [5]).... ..."

### Table 2: Mean zero-one test error scores for the BCI experiments. The mean was taken over ten single error scores. The p-value for a two-sided sign test against the SVM error scores are given in brackets.

"... In PAGE 7: ... We exclusively used a linear kernel. Table2 shows the mean zero-one loss for DATA I and DATA II and the constructed Universa. On the DATA I dataset, there is no improvement in the error rates for the subjects FS and JL com- pared to an SVM without Universum.... ..."

### Table 2: Mean zero-one test error scores for the BCI experiments. The mean was taken over ten single error scores. The p-value for a two-sided sign test against the SVM error scores are given in brackets.

"... In PAGE 7: ... We exclusively used a linear kernel. Table2 shows the mean zero-one loss for DATA I and DATA II and the constructed Universa. On the DATA I dataset, there is no improvement in the error rates for the subjects FS and JL com- pared to an SVM without Universum.... ..."

### Table 8: Two-sided matching to nearest neighbor

1999

"... In PAGE 7: ...From Table 7 and Table8 we see that two-sided nearest neighbor local recoding leads to stronger protection accompanied with larger average distance within pairs. The advantage of the two-sided nearest neighbor local recoding is that it withstands both the fishing... ..."

Cited by 1

### Table 2: Estimates of mean accuracy differences between all pairs of Comparison feature sets and Base feature sets. Statistical significance of the differences was tested using the two-sided t-test, as indicated: *=p lt; 0.05 and **=p lt; 0.01.

"... In PAGE 5: ... We trained models and evaluated test set accuracy for each of the feature sets described above. Mean accuracies for both cross-validation and random sampling are given in Table 1, with mean paired accuracy differences by random sampling in Table2 ; significance was measured by two-tailed t-test. For comparison, Table 1 also lists previous results from two previous studies.... In PAGE 5: ...0.2/90.1% accuracy) by combining appraisal group features (Attitude Type and Orientation) with the bag-of-words fea- tures (for coverage), demonstrating how appraisal analysis helps sentiment classification. This improvement is signifi- cant to a 99% confidence level (see Table2 ). Again we note that including Force does not seem to help.... ..."