dents also seem satisfied with the grading results ; their questions about our grading dropped significantly. We found, not surprisingly, that specifyin g what we expect from the students as they design and write their programs, as well as what we expect from ourselves when we grade those programs, has had a distinctly positive effect . The students can see and incorporate the concept s that improve their skills (and their grades), an d graders can uniformly assess how well they di d it. References . 1. Hamm, R. Wayne, Kenneth D . Henderson , Marilyn Repsher and Kathleen Timmer, " A Tool for Program Grading : The Jacksonville University Scale", SIGCSE Bulletin (15, 1) , 1983, pp . 248-252 . 2. Ledgard, Henry and John Tauer, Professional Software, Volume II: Programmin g Practice . Menlo Park : Addison-Wesley. will receive the award and give th e keynote presentation at the 1995 SIGCSE Annual Conference in Nashville , Tennessee . A Note on Physical Implementation s of Logical Entities Joseph S. Fuld a The Foundation for Economic Educatio n As early as the sophomore year, compute r science majors are introduced to th e concept of different physical realizations o f logical entities and that the choice o f physical implementation may affect th e degree to which the logical entity i s realized . In this note, we give a simple an d compelling example of this fundamenta l and critical idea . Among the marvels of modern technolog y are today's ATM's ; ATM's make bankin g easy, quick, and untied to banker's hours . Today, banks share information in vas t networks such as NYCE, CIRRUS, and MAC . Originally, these networks coul d handle cash withdrawals and balanc e inquiries, but neither deposits no r transfers between accounts . Deposits still require paperwork and verification an d one has to use one's own bank's ATM's t o carry out the transaction. Since in-branc h transfers involve filling out a withdrawa l order followed by a deposit slip, and sinc e deposits cannot be handled by th e networks, transfers were excluded fro m menus given on the networks . About a year after the ATM networks came int o use, someone realized that th e withdrawal-deposit realization of th e transfer transaction was a poo r implementation of the (logical ) transaction . With balances available fo r all accounts, transfers could also b e handled electronically . At my bank, a t least, in-branch transfers are stil l implemented by the withdrawal-deposi t model . But machine transfers--which unlike deposits involve no paperwork--ar e not: They are fully automated . Physica l implementation of logical entities matter . 7 1987 . 3. Miller, Nancy E. and Charles G . Peterson , "A Method for Evaluating Student Written Computer Programs in an Undergraduat e Computer Science Programming Languag e Course", SIGCSE Bulletin (12), 1980, pp . 9-17. 4. Olson, David M ., "The Reliability of Ana lytic and Holistic Methods in Rating Stu dents' Computer Programs", SIGCSE Bulletin (20, 1), 1988, pp . 293-298 . '**Educator Award Continued From Page 2* * G. Michael Schneide r Dept . of Mathematics and Compute r Scienc e Macalester Colleg e 1600 Grand Avenue St. Paul, MN 5510 5 (612) 696-645 8 SCHNEIDER@MACALSTR .EDU Please submit this information no late r than September 10, 1994 . The winner SIGCSE BULLETIN Vol . No . 3 Sept .

/lp/association-for-computing-machinery/a-note-on-physical-implementations-of-logical-entities-Kv8fSx94nf