### Table 2). light loading heavy loading

1996

"... In PAGE 13: ....3. Light-Loading and Heavy-Loading Approximations We also develop other approximations in Sections 5{7 based on asymptotics as s ! 1 with held xed, with either lt; 1 (light loading) or gt; 1 (heavy loading). These approximations are shown in Table2 . These formulas show that the workload factors wI and wS behave di erently:... In PAGE 14: ...Moreover, these formulas also serve as simple approximations. Since we already have reduced the G/GI/s/0 case to the M/M/s/0 case in (17) and (18), we primarily use the formulas in Table2 as convenient simple approximations for the canonical (M/M/s/0) workload factors (obtained by letting c2 a = 1 in Table 2). light loading heavy loading... In PAGE 14: ... Table2 . Approximation formulas for the workload factor w 2 of the estimators in (1), (2) and (4) for the G/M/s/0 model in light and heavy loading.... In PAGE 14: ... Approximation formulas for the workload factor w 2 of the estimators in (1), (2) and (4) for the G/M/s/0 model in light and heavy loading. With the exception of the light-loading simple-estimator formula, the formulas in Table2 are all in terms of the three variables s; and c2 a. (Given s, is equivalent to or .... In PAGE 14: .... (Given s, is equivalent to or .) The light-loading simple-estimator formula can be put in the same form by exploiting (15), which yields wS(s; ; c2 a) (1 + c2 a)3=2 2 p e 2=(1+c2 a) in light loading : (19) (Let w(s; ; c2 a) w(s; ; c2 a; 1; (c2 a + 1)=2).) Table2 is especially important for providing theoretical support for the simple-estimator work- load factor, because the di usion limit in Section 4 only applies directly to the indirect estimator. Note that formula (19) approximately satis es the general functional form (17) with S( ) = 2 1 ( ) = q 2= 2e 2=2 : (20) Similarly, the heavy-tra c indirect-estimator workload factor approximation in Table 2 can be expressed approximately as wI(s; ; c2 a) (1 + c2 a)3 2 4 4 ; (21) which is approximately of the form (17) with I( ) = 4 4 : (22)... In PAGE 14: ...) Table 2 is especially important for providing theoretical support for the simple-estimator work- load factor, because the di usion limit in Section 4 only applies directly to the indirect estimator. Note that formula (19) approximately satis es the general functional form (17) with S( ) = 2 1 ( ) = q 2= 2e 2=2 : (20) Similarly, the heavy-tra c indirect-estimator workload factor approximation in Table2 can be expressed approximately as wI(s; ; c2 a) (1 + c2 a)3 2 4 4 ; (21) which is approximately of the form (17) with I( ) = 4 4 : (22)... In PAGE 15: ...a + 1 here. Turning to the heavy-loading formula for the simple estimator and the light-loading formula for the indirect estimator, we note that these formulas are consistent with (17), because S( =pz) ! 2 as ! +1 and I( =pz) ! 2 as ! 1 for the M/M/s/0 model (corresponding to s ! 1 with xed as in Table2 ); see Table 1. Approximations (19){(22) reveal the essential form of the workload factors in light and heavy loading, but these formulas are not very accurate, e.... In PAGE 20: ... 5. The Simple Estimator In this section we derive the approximation formulas for wS in Table2 . The simple estimator is ^ BS(t) = L(t)= t, as in (2).... In PAGE 23: ...Table2 . From Table 1, we see that in the M/M/s/0 case with s = 400 the approximation wI 2:00 from (43) performs excellently for suitably large.... In PAGE 24: ... From Law (1975), the exact formula for the asymptotic variance in the M/GI/1 model is 2 Q = 4 2(1 )4 (c2 s + 1)3 + 5 3 2 m3 3m2 + 2 ! (c2 s + 1)2 (1 )3 + (1 + ) m3 3m2 + 3 2 m4 12m2 (c2 s + 1) (1 )2 ; (52) where mk is the kth moment of the service-time distribution, m1 = 1, c2 s = m2 m2 1 and the arrival rate is lt; 1. For the approximation here (with 1 lt; 1), we combine (18) and (22) of Whitt (1989) to obtain 2 m(M=G=1; = 1) 1(1 + 1)(c2 a + 1)3 4(1 1)4 ; (53) so that wI 2 I (1 + )(c2 a + 1)3 4s2( 1)4 for gt; 1; (54) as in Table2 . It is signi cant that (53) is exact for the M/M/1 model, but the role of c2 a in (53), and thus (54), is problematic, being based on heuristic heavy-tra c analysis in (17) of Whitt (1989).... In PAGE 38: ... Thus, wS serves as the approximation for wN. The canonical workload factors appearing in (17) and (18) can be calculated exactly by the methods of Section 3; approximations are given in Table2 . The approximations show that ^ BI(t) tends to be more e cient than the other estimators in heavy loading, but less e cient in light loading.... ..."

Cited by 23

### Table 8: Summary of Significant Lighting Quality and Environmental Satisfaction Effects

"... In PAGE 8: ... Thus, MANOVA of the satisfaction data included seven variables: REF[Noise]; REF[Ventilation]; REF[Furniture]; REF[Washrooms]; REF[Lighting], Environmental Satisfaction, and Lighting Quality. Table8 , the summary of these results, shows an LPDA x DLQB interaction effect and an LPDB main effect. These effects are small, explaining 1-2% of the variance.... ..."

### Table 4. Photometric data, the complete table available electronically. The column called vari lists the results of the time series analysis: pv refers to periodic and iv to irregular variables (see also the comments in the text). is the standard deviation of the light curves from which outliers were rejected (see text). ptp is the estimated peak-to-peak variation in the light curves.

"... In PAGE 4: ... Using these plate solutions the final J2000 coordinates of our sources were cal- culated with the task CCTRAN. These coordinates are listed in Table4 . The RMS of the deviations between the fit and the coordinates of the reference stars is typically 0: 001.... In PAGE 6: ... To take into account these magnitude and colour changes we determined the IC magnitude and the (V IC) and (RC IC) colours at di erent epochs. The colours and magnitudes we list in Table4 are average values of these di erent measurements. The dates of the employed measurements are listed in Table 1.... In PAGE 8: ... Filter Limiting Magnitude 20 50 V 22:1 21:1 RC 21:5 20:4 IC 20:7 19:6 sure that they do not appear twice in the final catalogue. For stars with IC 16:0 mag the colours and magnitudes listed in Table4 were taken from the 500sec exposures. The mea- surements for stars with 12:5 mag IC 16:0 and with 10 mag IC 12:5 mag were taken from the 50 esc and 5 sec exposures, respectively.... In PAGE 8: ...laccomio et al. (2000). The reason therefor could be di erent transmission curves of the used filters and a di erent average colour of the employed flux standards. A small part of the final photometric catalogue is shown in Table4 and the complete table that contains all 10554 stars is available electronically. In Table 4 we also list the (RC H ) colour of the stars.... In PAGE 8: ... A small part of the final photometric catalogue is shown in Table 4 and the complete table that contains all 10554 stars is available electronically. In Table4 we also list the (RC H ) colour of the stars. Since no H standard star measurements were available only the instrumental colour is given.... In PAGE 11: ... In total we found 543 periodic variable stars. These stars are marked in Table4 with a pv in the column vari . For 520 (95:8%) of these periodic variables the periods determined with the two di erent periodogram techniques agree within the es- timated errors.... In PAGE 11: ... Out of the stars which we rejected after our visual inspec- tion another 136 were classified as possible periodic variables. These stars are marked with a pv? in the column vari of Table4 but are not listed in Table 6, because they are not used for any further analysis (see Paper II). 4.... In PAGE 13: ... In total we found 484 irregular variables out of the 5927 stars we analysed. The irregular variables are listed in column vari of Table4 as iv. The spatial positions of the irregular variables are shown in Fig.... In PAGE 16: ... In order to account for those statistical outliers in the light curves we have equated the ptp variation for each star with the magnitude di erence between the third highest and the third lowest data point. Both quantities and ptp variation are given in Table4 for each star. We find that the ptp vari- ation is typically three times the standard deviation but it can vary between two and five times the standard deviation in some extreme cases.... ..."

### Table 1: De nitions and uses of the variables in Fig.1 De nitions Uses

1999

"... In PAGE 11: ... guard(t) or action(t) references x). For example Table1 shows the de nitions and uses of the variables money and light in Fig. 1.... ..."

Cited by 23

### Table 7: Satisfaction grades for C2 under the Light load of the treatment booking system

"... In PAGE 7: ...98 1.00 Table7 shows the satisfaction grades for breach length (criterion C2) under the Light load. JIT produces between 43% to 48% higher satisfaction grades for palliative patients than ASAP, leading to less variability than under the Normal load.... ..."

### Table 4: Results of tting the time delay (T ), variable ratio (R) and excess constant component (Co) to minimize the PRHQ statistic for the 8 GHz and 15 GHz light curves. The error bars (95% con dence) are determined from jackknife samples. The last line of the table shows the averaged result from the two light curves.

"... In PAGE 13: ... We compute errors on the parameters by forming the 95% con dence intervals from the data of Figure 8, which have been rescaled by the necessary \expansion factor quot; (N ? 1)=pN (see Efron amp; Tibshirani 1993). Values from the di erent frequencies for T and R are combined as described above, and the results are presented in Table4 . We caution that since the light curve data points are not independent, the jackknife simulations are likely to... ..."

### Table 2: Identi ed parameters for the dynamic experiments with the lightly lled sample

1998

"... In PAGE 23: ... Table2 . Note the relatively good agreement across experiments (especially ; C1; a1; a2; a3.... In PAGE 23: ... For example, Figures 16-17 each depict two ts that produce similar responses. Figure 16 includes the responses for the 3 lb tip mass given by the previously identi ed C1 and gv from Table2 , and by gv( ) = e4:15 ? 22:77 2 + 4:23 3, C1 = e3:71: Figure 17 depicts the following responses for the 2 lb tip mass: one model uses the previous C1 and gv; and the other uses gv = e4:04 ? 23:63 2 + 4:07 3, C1 = e4:06. In our model the damping or resisting force is provided by the internal variable quot;1, so we expect that its contribution to the total force is alternately positive and negative, starting with being positive, i.... ..."

Cited by 1

### Table 1. Research Approaches and Relevant Variables to this Research

in Korea

"... In PAGE 2: ... Hence in this study, we make an attempt to take an integrative view of the development of trust. A significant body of knowledge from three research streams in trust, summarized in Table1 , sheds light on the antecedents of initial trust. Table 1 also includes variables relevant to this research, suggested by extant trust research.... In PAGE 2: ... A significant body of knowledge from three research streams in trust, summarized in Table 1, sheds light on the antecedents of initial trust. Table1 also includes variables relevant to this research, suggested by extant trust research. It is to be noted here that the antecedents of trust presented here are not exhaustive and are limited to those variables that were deemed to be of interest in this research.... ..."

### Table 4. Variability Parameters

"... In PAGE 4: ...RESULTS 3.1 Variability Pattern Figure 1 shows the light-curves for the X{rays, the UV con- tinuum at 1800 A and the CIV line (the lightcurves for Ly are very similar in shape to those of CIV, although with smaller amplitude; see Table4 ). The UV continuum light curve shows a few large events which typically last one to two years and have amplitudes of up to a factor of 5.... In PAGE 4: ... During the IUE observations, the UV continuum flux varied by a factor larger than 10 although the actual dy- namic range (Rmax) of the continuum variations is di cult to determine due to the fact that the lowest flux, in 1984, is de ned by a relatively noisy spectrum (SWP22624). The variability parameters (Fvar = = F)andRmax for line and continuum are given in Table4 . Since Rmax is rather un- certain because of the large error associated with the lowest brightness state as well as the sparse temporal coverage in the sampling, we give in Table 4 also Rmax (JD gt; 7600), which only considers the data taken after 1988.... In PAGE 4: ... The variability parameters (Fvar = = F)andRmax for line and continuum are given in Table 4. Since Rmax is rather un- certain because of the large error associated with the lowest brightness state as well as the sparse temporal coverage in the sampling, we give in Table4 also Rmax (JD gt; 7600), which only considers the data taken after 1988. This should give a better evaluation of the variability, since this part of the lightcurve is adequately sampled with respect to the variability time scale.... In PAGE 4: ... The measurement errors for constant flux (non-variable) source are assumed to be normally dis- tributed and we estimated the error in Fvar through a Monte-Carlo calculation for data sets which are consistent with the measurements. It is clear from Table4... In PAGE 6: ....Wamsteker et al. crease slightly with increasing continuum brightness, con- trary to the behaviour shown in Figure 4 (right). From Fig- ure 2 (middle) and Table4 , it is clear that this relation is in part driven by the larger variation in the CIV line as compared to Ly (by a factor of two). Regression analysis for the relation between Ly and CIV for all components shows that a linear relation between the two lines does not give a statistically signi cant description of their relation ( 2 red gt; 1:6).... In PAGE 6: ...nd 0.54 for 1988 and 1991 respectively (Inda et al.,1994). The steeper soft X{ray spectral slope for the observation of 1991 is certainly due to the large tted NH ( Table4 ). The spectral indices of optical to UV, UV to X{ray, soft X{ray, and ME X{Ray are all very similar for 3C390.... ..."

### Table 4. Regression variables.

1993

"... In PAGE 10: ... This analysis was performed for every survey carried out using the method with a 3-vehicle start lag. The overall results were as follows:- Table4 . The PCU values derived from synchronous regression analysis Minibuses were shown to have a bigger influence, in terms of space required, than light trucks.... ..."