Results

**1 - 3**of**3**### Table 2: Example of a problem description and annotated source example

"... In PAGE 9: ...Table 2: Example of a problem description and annotated source example Annotated source examples were constructed by presenting a typical solution, a brief descrip- tion of why the code works, and an example of the output that the code produces when run with a particular input. See Table2 for a sample source example. Participants worked through two chapters of an introductory LISP textbook (Anderson, Corbett, amp; Reiser, 1987) containing approximately 26 total pages of material using BATBook, an electronic book and problem solving environment (Faries amp; Reiser, 1988).... ..."

### Table 1: Running time (seconds) for the veri cation of Fischer apos;s mutual exclusion protocol. f-reach stands for forward reachability, b-reach stands for backward reachability.

1999

"... In PAGE 13: ...i Controller i i i i i i i i Train i i i L-i =/=[] go i i y :=0 L:=[] appr L:=[i] appr go let i=hd(L) y :=0 y := 0 y gt;=10 occ2 free1 occ1 i fari free2 cross start wait i i x=0 x=0 y lt;=5 y lt;=20 y lt;=20 i i stop let i=tail(L) x:=0 L:=L::i y gt;=10 y :=0i i y :=0 L:=L-i x:=0 stop leave leave i appri leave near L-i=[] Figure 2: The bridge crossing system. Table1 shows the running time for verifying Fischer apos;s mutual exclusion protocol with 2, 3 and 4 processes.... In PAGE 13: ... Hence we list Uppaal apos;s performance in the column for forward reachability. Based on the data in Table1 , we can make the following observations: The tool Uppaal, which is based on di erence bound matrices, is faster than the other two tools for verifying the safety property using forward reachability. Backward reachability is much faster than forward reachability for verifying the safety prop- erty, across all tools.... ..."

Cited by 14

### Table 1: Minimum of EAAy( ; n0)=E(S) and EMAy( ; n0)=E(S) over 1 n0 N p and n0 that achieves these minima when p = 2.

2005

"... In PAGE 9: ... A similar procedure was used to nd a lower bound for EMA( ; n0). Table1 lists these bounds for 2 f0:05; 0:1g, for 10 N 50, and for p = 2; 3. The value of n0 achieving the corresponding minimum bound is also listed.... In PAGE 9: ...nd for p = 2; 3. The value of n0 achieving the corresponding minimum bound is also listed. For p = 2 test treatments, the Section 3 bounds are achieved by the EAA- or EMA-optimal design which always has the most balanced test treatment allocation. In particular, the optimal solutions when = 0:05 correspond to those in Table1 of Spurrier and Nizam (1990). For p = 3, although the designs most balanced in the test treatments for the listed values of n0 may not be the optimal designs, they are extremely ef cient.... In PAGE 9: ... (2004) while tables of bounds based on the lower bound (7) of Section 3, are available in Bortnick (1999) for 4 p 10. EAA- or EMA-optimal or ef cient designs can be constructed from the information given in Table1 as follows. For p = 2 and = 0:05 or 0.... In PAGE 9: ... If (N n0) is odd, then an optimal design has n1 = (N n0 1)=2 and n2 = (N n0 + 1)=2 observations, respectively, on the two test treatments. For example, from Table1 , an EAA(.05)-optimal design with N = 25 observations has n0 = 10; n1 = 7; n2 = 8, while the EMA(.... In PAGE 9: ...05)-optimal design has n0 = 9; n1 = n2 = 8 Similarly, for p = 3, to obtain a highly ef cient design, N n0 is divided as evenly as possible between n1; n2; n3. Thus, from Table1 , an EAA(.05)-ef cient design with N = 25 observations has n0 = 9 observations on the control, n1 = n2 = 5 and n3 = 6 observations, respectively, on the three test treatments, whereas an EAA(.... ..."

**1 - 3**of

**3**