### Table 1. Mean Results for Each Measure and Subject Group. Task Expert Novice Novice

1998

"... In PAGE 3: ... RESULTS AND DISCUSSION The PARADISE evaluation framework [7] posits that system performance can be modelled by determining the contributions of task success measures and a range of cost measures to user satisfaction. Table1 shows mean results on each task for each user group (Expert, Novice - Tutorial, Novice - No Tutorial), for the task success measures of perceived completion and kappa, for the cost measures of elapsed time, mean recognition score, user turns, ASR rejections, ASR time outs, cancellations, barge- ins and help messages, and for cumulative satisfaction. A two- way ANOVA for the mixed design, with task as the between- groups factor and user expertise as the within-group factor, was performed for each measure.... ..."

Cited by 26

### Table 1. Mean Results for Each Measure and Subject Group. Task Expert Novice Novice

1998

"... In PAGE 3: ... RESULTS AND DISCUSSION The PARADISE evaluation framework [7] posits that system performance can be modelled by determining the contributions of task success measures and a range of cost measures to user satisfaction. Table1 shows mean results on each task for each user group (Expert, Novice - Tutorial, Novice - No Tutorial), for the task success measures of perceived completion and kappa, for the cost measures of elapsed time, mean recognition score, user turns, ASR rejections, ASR time outs, cancellations, barge- ins and help messages, and for cumulative satisfaction. A two- way ANOVA for the mixed design, with task as the between- groups factor and user expertise as the within-group factor, was performed for each measure.... ..."

Cited by 26

### Table 4: Standard ML vs tw-constrained ML training

1994

"... In PAGE 16: ...ased on a similar reasoning. We call it t-constraint. tw-constraint The tw-constrained ML training is similar to the standard ML training, ex- cept that the probabilities p#28t=w#29 are not changed at the end of an iteration. The results in Table4 show the number of tagging errors when the model is trained with the standard or tw-constrained ML training. They show that the tw-constrained ML training still degrades the RF training, but not as... ..."

Cited by 163

### Table 5: Standard ML vs constrained ML training

1994

"... In PAGE 17: ...o a biclass model, i.e. a model where p#28t i =w 1 t 1 :::w i,1 t i,1 #29=h#28t i =t i,1 #29 The initial model is estimated by relative frequency on the whole training data and Viterbi tagging is used. As in the previous experiment, the results in Table5 show the number of tagging errors when the model is trained with the standard or t-constrained ML training. They show that the t-constrained ML training still degrades the RF training, but not as quickly as the standard ML.... ..."

Cited by 163

### Table 4 Standard ML vs. tw-constrained ML training.

### Table 3 Tests to reach criterion for test sets

"... In PAGE 11: ... One measure of learning facility is how quickly par- ticipants passed the criterion of a perfect score on one test. As seen in Table3 , the mean number of tests to reach criterion was smaller for experts than for novices on all four test sets. The difference between experts and novices was not significant for any one test set alone, but was significant when all four sets were considered together, t(90) 2.... ..."

### Table 1 Means and standard deviations of frequency of goal preference among novice and veteran teachers of humanities and sciences

"... In PAGE 8: ...XM quot;1.93) goals. Thus, it is apparent from an ex- amination of the interviews of the entire sample that these teachers expressed a de quot;nite preference for academic goals for their students over social and personal goals. However, examination of Table1 and Fig. 1 re- veals that the veteran humanities teachers expressed unique goal preferences di!ering from novice teachers of humanities and from both novice and experienced science teachers.... ..."

### Table 2 - Total time (in seconds) to type three strings for novices and experienced users (standard deviation in parentheses)

1993

"... In PAGE 6: ... 3 Results 3.1 Typing Rates Mean time to complete all three strings with each keyboard size appears, with standard deviations, in Table2 . Equivalent Words Per Minute rates (assuming 5 characters per word) appear in Table 3 and Figure 3.... ..."

Cited by 17