GRE Computer Practice Test Book

This book is provided FREE with test registration by the Graduate Record Examinations Board. G R A D U A T E R E C O R D E X A M I N A T I O N S ® Computer Science Test Practice Book This practice book contains one actual full-length GRE Computer Science Test test-taking strategies Become familiar with test structure and content test instructions and answering procedures Compare your practice test results with the performance of those who took the test at a GRE administration. Visit GRE Online at www.gre.org Listening. Learning. Leading.Copyright © 2004 by Educational Testing Service. All rights reserved. EDUCATIONAL TESTING SERVICE, ETS, the ETS logos, GRADUATE RECORD EXAMINATIONS, and GRE are registered trademarks of Educational Testing Service. Note to Test Takers: Keep this practice book until you receive your score report. The book contains important information about content specifications and scoring.3 COMPUTER SCIENCE TEST PRACTICE BOOK Table of Contents Purpose of the GRE Subject Tests........................ 3 Development of the Subject Tests ....................... 3 Content of the Computer Science Test ............... 4 Preparing for a Subject Test ................................. 5 Test-Taking Strategies .......................................... 6 What Your Scores Mean ...................................... 6 Practice GRE Computer Science Test ................. 9 Scoring Your Subject Test ..................................47 Evaluating Your Performance.............................50 Answer Sheet .....................................................51 Purpose of the GRE Subject Tests The GRE Subject Tests are designed to help graduate school admission committees and fellowship sponsors assess the qualifications of applicants in specific fields of study. The tests also provide you with an assessment of your own qualifications. Scores on the tests are intended to indicate knowledge of the subject matter emphasized in many undergraduate programs as preparation for graduate study. Because past achievement is usually a good indicator of future performance, the scores are helpful in predicting success in graduate study. Because the tests are standardized, the test scores permit comparison of students from different institutions with different undergraduate programs. For some Subject Tests, subscores are provided in addition to the total score; these subscores indicate the strengths and weaknesses of your preparation, and they may help you plan future studies. The GRE Board recommends that scores on the Subject Tests be considered in conjunction with other relevant information about applicants. Because numeroou factors influence success in graduate school, reliance on a single measure to predict success is not advisable. Other indicators of competence typically include undergraduate transcripts showing courses taken and grades earned, letters of recommendation, the GRE Writing Assessment score, and GRE General Test scores. For information about the appropriate use of GRE scores, write to GRE Program, Educational Testing Service, Mail Stop 57-L, Princeton, NJ 08541, or visit our Web site at www.gre.org/codelst.html. Development of the Subject Tests Each new edition of a Subject Test is developed by a committee of examiners composed of professors in the subject who are on undergraduate and graduate facultiie in different types of institutions and in different regions of the United States and Canada. In selecting members for each committee, the GRE Program seeks the advice of the appropriate professional associations in the subject. The content and scope of each test are specified and reviewed periodically by the committee of examiners. Test questions are written by the committee and by other faculty who are also subject-matter specialists and by subject-matter specialists at ETS. All questions proposed for the test are reviewed by the committee and revised as necessary. The accepted questions are assembled into a test in accordance with the content specifications developed by the committee to ensure adequate coverage of the various aspects of the field and, at the same time, to prevent overemphasis on any single topic. The entire test is then reviewed and approved by the committee.4 COMPUTER SCIENCE TEST PRACTICE BOOK Subject-matter and measurement specialists on the ETS staff assist the committee, providing information and advice about methods of test construction and helping to prepare the questions and assemble the test. In addition, each test question is reviewed to eliminate language, symbols, or content considered potentially offensive, inappropriate for major subgroups of the testtakkin population, or likely to perpetuate any negative attitude that may be conveyed to these subgroups. The test as a whole is also reviewed to ensure that the test questions, where applicable, include an appropriate balance of people in different groups and different roles. Because of the diversity of undergraduate curricula, it is not possible for a single test to cover all the material you may have studied. The examiners, therefoore select questions that test the basic knowledge and skills most important for successful graduate study in the particular field. The committee keeps the test up-to-date by regularly developing new editions and revising existing editions. In this way, the test content changes steadily but gradually, much like most curriccula In addition, curriculum surveys are conducted periodically to ensure that the content of a test reflects what is currently being taught in the undergradduat curriculum. After a new edition of a Subject Test is first administeered examinees’ responses to each test question are analyzed in a variety of ways to determine whether each question functioned as expected. These analyses may reveal that a question is ambiguous, requires knowledge beyond the scope of the test, or is inappropriiat for the total group or a particular subgroup of examinees taking the test. Answers to such questions are not used in computing scores. Following this analysis, the new test edition is equated to an existing test edition. In the equating process, statistical methods are used to assess the difficulty of the new test. Then scores are adjusted so that examinees who took a difficult edition of the test are not penalized, and examinees who took an easier edition of the test do not have an advantage. Variatiion in the number of questions in the different editions of the test are also taken into account in this process. Scores on the Subject Tests are reported as threediigi scaled scores with the third digit always zero. The maximum possible range for all Subject Test total scores is from 200 to 990. The actual range of scores for a particular Subject Test, however, may be smaller. The maximum possible range of Subject Test subscores is 20 to 99; however, the actual range of subscores for any test or test edition may be smaller than 20 to 99. Subject Test score interpretive information is provided in Interpreting Your GRE Scores, which you will receive with your GRE score report, and on the GRE Web site at www.gre.org/codelst.html. Content of the Computer Science Subject Test The test consists of about 70 multiple-choice questions, some of which are grouped in sets and based on such materials as diagrams, graphs, and program fragments. The approximate distribution of questions in each edition of the test according to content categories is indicated by the following outline. The percentages given are approximate; actual percentages will vary slightly from one edition of the test to another. I. SOFTWARE SYSTEMS AND METHODOLOGY — 40% A. Data organization 1. Data types 2. Data structures and implementation techniques B. Program control and structure 1. Iteration and recursion 2. Procedures, functions, methods, and exception handlers 3. Concurrency, communication, and synchronization C. Programming languages and notation 1. Constructs for data organization and program control 2. Scope, binding, and parameter passing 3. Expression evaluation5 COMPUTER SCIENCE TEST PRACTICE BOOK D. Software engineering 1. Formal specifications and assertions 2. Verification techniques 3. Software development models, patterns, and tools E. Systems 1. Compilers, interpreters, and run-time systems 2. Operating systems, including resource management and protection/security 3. Networking, Internet, and distributed systems 4. Databases 5. System analysis and development tools II. COMPUTER ORGANIZATION AND ARCHITECTURE — 15% A. Digital logic design 1. Implementation of combinational and sequential circuits 2. Optimization and analysis B. Processors and control units 1. Instruction sets 2. Computer arithmetic and number representation 3. Register and ALU organization 4. Data paths and control sequencing C. Memories and their hierarchies 1. Performance, implementation, and management 2. Cache, main, and secondary storage 3. Virtual memory, paging, and segmentation D. Networking and communications 1. Interconnect structures (e.g., buses, switches, routers) 2. I/O systems and protocols 3. Synchronization E. High-performance architectures 1. Pipelining superscalar and out-of-order execution processors 2. Parallel and distributed architectures III. THEORY AND MATHEMATICAL BACKGROUND — 40% A. Algorithms and complexity 1. Exact and asymptotic analysis of specific algorithms 2. Algorithmic design techniques (e.g., greedy, dynamic programming, divide and conquer) 3. Upper and lower bounds on the complexity of specific problems 4. Computational complexity, including NPcompleetenes B. Automata and language theory 1. Models of computation (finite automata, Turing machines) 2. Formal languages and grammars (regular and context-free) 3. Decidability C. Discrete structures 1. Mathematical logic 2. Elementary combinatorics and graph theory 3. Discrete probability, recurrence relations, and number theory IV. Other Topics — 5% Example areas include numerical analysis, artificial intelligence, computer graphics, cryptography, security, and social issues. Note: Students are assumed to have a mathematical background in the areas of calculus and linear algebra as applied to computer science. Preparing for a Subject Test GRE Subject Test questions are designed to measure skills and knowledge gained over a long period of time. Although you might increase your scores to some extent through preparation a few weeks or months before you take the test, last-minute cramming is unlikely to be of further help. The following informatiio may be helpful. A general review of your college courses is probably the best preparation for the test. Howevver the test covers a broad range of subject matter, and no one is expected to be familiar with the content of every question. ▪6 COMPUTER SCIENCE TEST PRACTICE BOOK Use this practice book to become familiar with the types of questions in the GRE Computer Science Test, paying special attention to the directions. If you thoroughly understand the directions before you take the test, you will have more time during the test to focus on the questiion themselves. Test-Taking Strategies The questions in the practice test in this book illustrrat the types of multiple-choice questions in the test. When you take the test, you will mark your answers on a separate machine-scorable answer sheet. Total testiin time is two hours and fifty minutes; there are no separately timed sections. Following are some general test-taking strategies you may want to consider. Read the test directions carefully, and work as rapidly as you can without being careless. For each question, choose the best answer from the available options. All questions are of equal value; do not waste time pondering individual questions you find extremely difficult or unfamiliar. You may want to work through the test quite rapidly, first answering only the questions about which you feel confident, then going back and answering questions that require more thought, and concluding with the most difficult questions if there is time. If you decide to change an answer, make sure you completely erase it and fill in the oval corresponndin to your desired answer. Questions for which you mark no answer or more than one answer are not counted in scoring. As a correction for haphazard guessing, onefouurt of the number of questions you answer incorrectly is subtracted from the number of questions you answer correctly. It is improbable that mere guessing will improve your score significantly; it may even lower your score. If, however, you are not certain of the correct answer but have some knowledge of the question and are able to eliminate one or more of the answer choices, your chance of getting the right answer is improved, and it may be to your advantaag to answer the question. Record all answers on your answer sheet. Answers recorded in your test book will not be counted. Do not wait until the last five minutes of a testing session to record answers on your answer sheet. What Your Scores Mean Your raw score–that is, the number of questions you answered correctly minus one-fourth of the number you answered incorrectly–is converted to the scaled score that is reported. This conversion ensures that a scaled score reported for any edition of a Subject Test is comparable to the same scaled score earned on any other edition of the same test. Thus, equal scaled scores on a particular Subject Test indicate essentially equal levels of performance regardless of the test edition taken. Test scores should be compared only with other scores on the same Subject Test. (For example, a 680 on the Computer Science Test is not equivalent to a 680 on the Mathematics Test.) Before taking the test, you may find it useful to know approximately what raw scores would be required to obtain a certain scaled score. Several factors influennc the conversion of your raw score to your scaled score, such as the difficulty of the test edition and the number of test questions included in the computation of your raw score. Based on recent editions of the Computer Science Test, the table on the next page gives the range of raw scores associated with selected scaled scores for three different test editions. (Note that when the number of scored questions for a given test is greater than the number of possible scaled scores, it is likely that two or more raw scores will convert to the same scaled score.) The three test editions in the table that follows were selected to reflect varying degrees of difficulty. Examinees should note that future test editions may be somewhat more or less difficult than the test editions illustrated in the table.7 COMPUTER SCIENCE TEST PRACTICE BOOK Range of Raw Scores* Needed to Earn Selected Scaled Scores on Three Computer Science Test Editions That Differ in Difficulty Raw Scores Scaled Score Form A Form B Form C Number of Questions Used to Compute Raw Score *Raw Score = Number of correct answers minus one-fourth the number of incorrect answers, rounded to the nearest integer. Examples of Ways to Earn a Scaled Score of 700 on the Edition Labeled as “Form A” Number of Questions Questions Questions Questions Used Answered Answered Not to Compute Raw Score Correctly Incorrectly Answered Raw Score For a particular test edition, there are many ways to earn the same raw score. For example, on the edition listed above as “Form A,” a raw score of 31 would earn a scaled score of 700. Below are a few of the possible ways in which a scaled score of 700 could be earned on that edition. 31 31 0 39 70 31 34 12 24 70 31 38 28 4 70 800 47 41-42 37-38 700 31 27 22-23 600 16-17 15 11 500 3-4 3 0 70 69 678 COMPUTER SCIENCE TEST PRACTICE BOOK Practice Test To become familiar with how the administration will be conducted at the test center, first remove the answer sheet (pages 51 and 52). Then go to the back cover of the test book (page 46) and follow the instructions for completing the identification areas of the answer sheet. When you are ready to begin the test, note the time and begin marking your answers on the answer sheet.Copyright © 2002, 2003 by Educational Testing Service. All rights reserved. GRE, GRADUATE RECORD EXAMINATIONS, ETS, EDUCATIONAL TESTING SERVICE and the ETS logos are registered trademarks of Educational Testing Service. FORM GR0329 THIS TEST BOOK MUST NOT BE TAKEN FROM THE ROOM. GRADUATE RECORD EXAMINATIONS® Do not break the seal until you are told to do so. The contents of this test are confidential. Disclosure or reproduction of any portion of it is prohibited. COMPUTER SCIENCE TEST 29 9Unauthorized copying or reuse of any part of this page is illegal. COMPUTER SCIENCE TEST Time—170 minutes 70 Questions Notation, Conventions, and Definitions: In this test a reading knowledge of modern programming languages is assumed. The following notational conventions and definitions are used. 1. All numbers are assumed to be written in decimal notation unless otherwise indicated. 2. x denotes the greatest integer that is less than or equal to x. 3. x denotes the least integer that is greater than or equal to x. 4. g n O f n ( ) ( ( )) = denotes “g n ( ) has order at most f n ( )” and means that there exist positive constants C and N such that g n Cf n ( ) ( ) £ for all n N > . g n f n ( ) ( ( )) = W denotes “g n ( ) has order at least f n ( )” and for this test means that there exist positive constants C and N such that g n Cf n ( ) ( ) ≥ for all n N > . g n f n ( ) ( ( )) = Q denotes “g n ( ) has the same order as f n ( )” and means that there exist positive constants C C 1 2 , , and N such that C f n g n C f n 1 2 ( ) ( ) ( ) £ £ for alln N > . 5. $ denotes “there exists.” " denotes “for all.” Æ denotes “implies;” … is also used to denote “implies;” and fiis also used to denote “implies.” ÿ denotes “not”; “A” is also used as meaning “ÿA.” ⁄ denotes “inclusive or”; + also denote “inclusive or,” e.g., P Q + can denote “P Q or .” ≈ denotes “exclusive or.” Ÿ denotes “and”; also, juxtaposition of statements can denote “and,” e.g., PQ can denote “P and Q.” A boolean formula is satisfiable if it is true under some assignment of boolean values for its variables. A boolean formula is a tautology (or is valid) if it is true under all assignments of boolean values for its variables. 6. ∆ denotes the empty set. If A and B denote sets, then: A B » is the set of all elements that are in A or in B or in both; A B « is the set of all elements that are in both A and B; A is the set of all elements not in A that are in some specified universal set. A is the cardinality of A. 7. In a string expression, if S and T denote strings or sets of strings, then: An empty string is denoted by e or by L; ST denotes the concatenation of S and T ; S T + denotes S T » or { , } S T , depending on context; S n denotes SS S n . . . ; factors '() * S* denotes e + + + + S S S 2 3 . . . ; S + denotes S S S + + + 2 3 . . . . 10Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 8. In a grammar: a b Æ represents a production in the grammar. a b fi means b can be derived from a by the application of exactly one production. a b * fi means b can be derived from a by the application of zero or more productions. Unless otherwise specified, (i) symbols appearing on the left-hand side of productions are nonterminal symbols; the remaining symbols are terminal symbols; (ii) the leftmost symbol of the first production is the start symbol; and (iii) the start symbol is permitted to appear on the right-hand side of productions. 9. In a logic diagram: 10. represents a D-type flip-flop, which stores the value of its D input when clocked. 11. Binary tree traversal is defined recursively as follows. Preorder: visit the root, traverse the left subtree, traverse the right subtree Inorder: traverse the left subtree, visit the root, traverse the right subtree Postorder: traverse the left subtree, traverse the right subtree, visit the root 12. In a finite automaton diagram, states are represented by circles, with final (or accepting) states indicated by two concentric circles. The start state is labeled “Start”. An arc from state s to state t labeled a indicates a transition from s to t on input a. A label a b indicates that this transition produces an output b. A label a a ak1 2 , , . . . , indicates that the transition is made on any of the inputs a a ak1 2 , , . . . , . 13. For a program segment S and predicates P and Q, to say that the triple P S Q { } { } is partially correct means that if P is true before initiation of S, then Q is true upon termination of S. To say that P S Q { } { } is totally correct means that it is partially correct and S terminates for all inputs for which P is true. Given that P S Q { } { } is partially correct, a precondition is any assertion that implies P, and a postcondition is any assertion that is implied by Q. 14. A loop invariant for a while statement while B do S is an assertion that is true each time the guard B is evaluated during execution of the while statement. 11Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. Directions: Each of the questions or incomplete statements below is followed by five suggested answers or completioons Select the one that is the best of the choices offered and then mark the corresponding space on the answer sheet. 1. Suppose that a certain software product has a mean time between failures of 10,000 hours and has a mean time to repair of 20 hours. If the product is used by 100 customers, what is its availability? (A) 80% (B) 90% (C) 98% (D) 99.8% (E) 100% 2. The object-oriented paradigm includes which of the following properties? I. Encapsulation II. Inheritance III. Recursion (A) I only (B) II only (C) I and II only (D) II and III only (E) I, II, and III 12 Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 3. Which of the following algorithms has running time ( ) 2 n Q in the worst case but ( ) log n n Q on average? (A) Bubblesort (B) Mergesort (C) Heapsort (D) Quicksort (E) Tournament sort 4. Which of the following is the name of the data structure in a compiler that is responsible for managing information about variables and their attributes? (A) Abstract Syntax Tree (AST) (B) Attribute Grammar (C) Symbol Table (D) Semantic Stack (E) Parse Table 13Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 5. Consider the following pseudocode. x := 1; i := 1; while (x £ 1000) begin x := x 2 ; i := i + 1; end; What is the value of i at the end of the pseudocode? (A) 4 (B) 5 (C) 6 (D) 7 (E) 8 6. Suppose that ( ) , P x y means “x is a parent of y” and ( ) M x means “x is male”. If ( ) , F v w equals ( ) ( ) ( ) ( ) ( ) ( ) , , , , M v x y P x y P x v y v P y w Ÿ $ $ Ÿ Ÿ π Ÿ what is the meaning of the expression ( ) , ? F v w (A) v is a brother of . w (B) v is a nephew of . w (C) v is an uncle of . w (D) v is a grandfather of . w (E) v is a male cousin of . w 14 Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. Questions 7-8 are based on binary tree T shown below. 7. Which of the following represents a postorder traversal of T ? (A) P Q U W X V Y Z (B) U Q X W P V Z Y (C) U X W Q Z Y V P (D) U X Z Q W Y V P (E) X Z U W Y Q V P 8. If T is a binary search tree with the smaller elements in the left subtree, which of the following nodes contains the fourth smallest element in T ? (A) Q (B) V (C) W (D) X (E) Z 15Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 9. Which of the following statements about Ethernets is typically FALSE? (A) Ethernets use circuit switching to send messages. (B) Ethernets use buses with multiple masters. (C) Ethernet protocols use a collision-detection method to ensure that messages are transmitted properly. (D) Networks connected by Ethernets are limited in length to a few hundred meters. (E) Packets sent on Ethernets are limited in size. 10. A k-ary tree is a tree in which every node has at most k children. In a k-ary tree with n nodes and height h, which of the following is an upper bound for the maximum number of leaves as a function of h, k, and n ? (A) logk n (B) logk h (C) logk n n (D) h k (E) k h 11. Which of the following is (are) true about virtual memory systems that use pages? I. The virtual address space can be larger than the amount of physical memory. II. Programs must be resident in main memory throughout their execution. III. Pages correspond to semantic characteristics of the program. (A) I only (B) II only (C) I and II (D) I and III (E) II and III 16Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 12. Let ( ) T n be defined by ( )1 7 T = and ( ) ( ) 1 3 T n n T n + = + for all integers 1. n ≥ Which of the following represents the order of growth of ( ) T n as a function of n ? (A) ( ) n Q (B) ( ) log n n Q (C) ( ) 2 n Q (D) ( ) 2 log n n Q (E) ( ) 2n Q 13. One approach to handling fuzzy logic data might be to design a computer using ternary (base-3) logic so that data could be stored as “true,” “false,” and “unknown.” If each ternary logic element is called a flit, how many flits are required to represent at least 256 different values? (A) 4 (B) 5 (C) 6 (D) 7 (E) 8 14. Hash tables can contribute to an efficient average-case solution for all of the problems described below EXCEPT: (A) Counting distinct values: Given a set of n keys, determine the number of distinct key values. (B) Dynamic dictionary: Support the operations of insert, delete, and search in a dictionary. (C) Range search: Given values a and b, find all the records whose key value is in the range a b , . (D) Symbol table lookup: Given a program identifier, find its type and address. (E) Finding intersections: Given two sets of keys, find all key values in common to both sets. 17Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 15. An invariant for the loop below is “z* = k n x b and ≥ k 0”. x := b; k := n; z := 1; while (k ≠ 0) { if odd(k) then z := z*x; x := x*x; k := Í ˙ Î ˚ k/2 ; } When the loop terminates, which of the following must be true? (A) = n x b (B) = n z b (C) = n b x (D) = n b z (E) π k 0 16. In the Internet Protocol (IP) suite of protocols, which of the following best describes the purpose of the Address Resolution Protocol? (A) To translate Web addresses to host names (B) To determine the IP address of a given host name (C) To determine the hardware address of a given host name (D) To determine the hardware address of a given IP address (E) To determine the appropriate route for a datagram 17. A certain pipelined RISC machine has 8 general-purpose registers R0, R1, . . . , R7 and supports the following operations. ADD Rs1, Rs2, Rd Add Rs1 to Rs2 and put the sum in Rd MUL Rs1, Rs2, Rd Multiply Rs1 by Rs2 and put the product in Rd An operation normally takes one cycle; however, an operation takes two cycles if it produces a result required by the immediately following operation in an operation sequence. Consider the expression AB ABC BC + + , where variables A, B, C are located in registers R0, R1, R2. If the contents of these three registers must not be modified, what is the minimum number of clock cycles required for an operation sequence that computes the value of AB ABC BC + + ? (A) 5 (B) 6 (C) 7 (D) 8 (E) 9 18Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 18. Below is a precedence graph for a set of tasks to be executed on a parallel processing system S. Efficiency is defined as the ratio between the speedup and the number of processors. (The speedup is defined as the ratio of the time taken to perform a set of tasks on a single processor to the time taken to perform the same set of tasks on a parallel processor.) System S has four processors (CPU’s). If each of the tasks T T 1 8 , . . . , takes the same time, what is the efficiency of this precedence graph on S ? (A) 25% (B) 33 13 % (C) 50% (D) 100% (E) 125% 19. Let ( ) , G V E = be a finite directed acyclic graph with 0. E > Which of the following must be true? I. G has a vertex with no incoming edge. II. G has a vertex with no outgoing edge. III. G has an isolated vertex, that is, one with neither an incoming edge nor an outgoing edge. (A) I only (B) II only (C) III only (D) I and II only (E) I, II, and III 19Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. Questions 20-21 are based on the following information. Array A contains 256 elements of 4 bytes each. Its first element is stored at physical address 4,096. Array B contains 512 elements of 4 bytes each. Its first element is stored at physical address 8,192. Assume that only arrays A and B can be cached in an initially empty, physically addressed, physically tagged, direct-mapped, 2K-byte cache with an 8-byte block size. The following loop is then executed. for (i = 0; i < 256; i++) A[i] = A[i] + B[2*i]; 20. During the execution of the loop, how many bytes will be written to memory if the cache has a write-through policy? (A) 0 (B) 256 (C) 1,024 (D) 2,048 (E) 4,096 21. During the execution of the loop, how many bytes will be written to memory if the cache has a write-back policy? (A) 0 (B) 256 (C) 1,024 (D) 2,000 (E) 4,000 20 Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 22. According to the IEEE standard, a 32-bit, single-precision, floating-point number N is defined to be N = (−1)S × 1.F × 2E−127 where S is the sign bit, F the fractional mantissa, and E the biased exponent. A floating-point number is stored as S E F : : , where S, E, and F are stored in 1 bit, 8 bits, and 23 bits, respectively. What is the decimal value of the floating-point number C1E00000 (hexadecimal notation) ? (A) 26 (B) −15 (C) -26 (D) -28 (E) −59 23. Which of the following predicate calculus formulas must be true under all interpretations? I. ( ) ( ) ( ) ( ) ( ) ( ) xP x xQ x x P x Q x " ⁄" Æ" ⁄ II. ( ) ( ) ( ) ( ) ( ) ( ) x P x Q x xP x xQ x " ⁄ Æ " ⁄" III. ( ) ( ) ( ) ( ) ( ) ( ) xP x xQ x x P x Q x $ ⁄$ Æ$ ⁄ (A) I only (B) III only (C) I and II (D) I and III (E) II and III 21Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. Questions 24-25 are based on the following procedures. procedure mystery a : integer; b : integer; procedure enigma(x,y) begin y = y + b; x = b + x; b = x + b; a = y; end enigma; begin a = 2; b = 7; enigma(a,b); write(a); write(b); end mystery; The output of procedure mystery depends on the parameter-passing method used. 24. Suppose that all parameters are passed by value. Which of the following values are output when procedure mystery is called? a b (A) 2 7 (B) 2 9 (C) 9 14 (D) 14 16 (E) 30 30 25. Suppose that all parameters are passed by reference. Which of the following values are output when procedure mystery is called? a b (A) 2 7 (B) 2 9 (C) 9 14 (D) 14 16 (E) 30 30 22Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 26. Let A and B be two sets of words (strings) from S*, for some alphabet of symbols S. Suppose that B is a subset of A. Which of the following statements must always be true of A and B ? I. If A is finite, then B is finite. II. If A is regular, then B is regular. III. If A is context-free, then B is context-free. (A) I only (B) II only (C) III only (D) I and II only (E) I, II, and III 27. A CPU has an arithmetic unit that adds bytes and then sets its V, C, and Z flag bits as follows. The V-bit is set if arithmetic overflow occurs (in two’s complement arithmetic). The C-bit is set if a carry-out is generated from the most significant bit during an operation. The Z-bit is set if the result is zero. What are the values of the V, C, and Z flag bits after the 8-bit bytes 1100 1100 and 1000 1111 are added? V C Z (A) 0 0 0 (B) 1 1 0 (C) 1 1 1 (D) 0 0 1 (E) 0 1 0 23Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 28. Let k be an integer greater than 1. Which of the following represents the order of growth of the expression 1 n i i k = Â as a function of n ? (A) ( ) n k Q (B) ( ) log n n k Q (C) ( ) log nk n Q (D) ( ) 2kn k Q (E) ( ) 1 k n Q + 29. Mergesort works by splitting a list of n numbers in half, sorting each half recursively, and merging the two halves. Which of the following data structures will allow mergesort to work in ( ) log O n n time? I. A singly linked list II. A doubly linked list III. An array (A) None (B) III only (C) I and II only (D) II and III only (E) I, II, and III 24Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 30. Consider the following function. double power(double base, unsigned int exponent) { if (exponent == 0) return 1.0; else if (even(exponent)) return power(base*base, exponent/2); else return power(base*base, exponent/2)*base; } How many multiplications are executed as a result of the call power(5.0, 12) ? (Do not include divisions in this total.) (A) 12 (B) 9 (C) 8 (D) 6 (E) 5 31. Which of the following statements about datagrams sent by a node in a network using IPv4 protocol is (are) true? I. Datagrams at the source must be the size of the smallest maximum transmission unit (MTU) of all the links on a path to the destination. II. Datagrams may be fragmented during routing. III. Datagrams are reassembled only at the destination. (A) I only (B) II only (C) III only (D) I and III (E) II and III 25Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 32. Of the following problems concerning a given undirected graph G, which is currently known to be solvable in polynomial time? (A) Finding a longest simple cycle in G (B) Finding a shortest cycle in G (C) Finding ALL spanning trees of G (D) Finding a largest clique in G (E) Finding a node coloring of G (where adjacent nodes receive distinct colors) with the minimum number of colors 33. Two processors, M-5 and M-7, implement the same instruction set. Processor M-5 uses a 5-stage pipeline and a clock cycle of 10 nanoseconds. Processor M-7 uses a 7-stage pipeline and a clock cycle of 7.5 nanoseconds. Which of the following is (are) true? I. M-7’s pipeline has better maximum throughput than M-5’s pipeline. II. The latency of a single instruction is shorter on M-7’s pipeline than on M-5’s pipeline. III. Programs executing on M-7 will always run faster than programs executing on M-5. (A) I only (B) II only (C) I and III only (D) II and III only (E) I, II, and III 26Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 34. For the following code, the bias of each conditional branch in the code is labeled on the control flow graph to the right. For example, the Boolean expression if_condition evaluates to true on one-half of the executions of that expression. do { U; if (if_condition) { V; if (break_condition) break; } else W; X; } while (loop_condition); Y; What is the expected number of times that U executes? (A) 0.5 (B) 1 (C) 1.5 (D) 2 (E) More than 10 35. Consider the following grammar. S → ( S ) S → x Which of the following statements is (are) true? I. The grammar is ambiguous. II. The grammar is suitable for top-down parsing. III. The grammar is suitable for bottom-up parsing. (A) I only (B) II only (C) III only (D) II and III only (E) I, II, and III 27Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 36. A logic circuit has three input bits: 0 1 , , x x and 2, x where 0 x is the least significant bit and 2 x is the most significant bit. The output from the circuit is 1 when its input is any of the 3-bit numbers 1, 4, 5, or 6; otherwise, the output is 0. Which of the following expressions represents the output from this circuit? (A) 2 1 0 __ __ __ x x x + + (B) 2 0 2 1 __ __ x x x x + (C) 1 0 2 0 __ __ x x x x + (D) 2 1 0 2 1 __ __ x x x x x + (E) 2 1 0 __ x x x + 37. Which of the following problems can be solved by a standard greedy algorithm? I. Finding a minimum spanning tree in an undirected graph with positive-integer edge weights II. Finding a maximum clique in an undirected graph III. Finding a maximum flow from a source node to a sink node in a directed graph with positive-integer edge capacities (A) I only (B) II only (C) III only (D) I and II only (E) I, II, and III 28Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. Questions 38-39 are based on the following information. A certain randomized algorithm A is intended to determine whether a given positive-integer input n is prime by generating a random bit string r and, based on the values of n and r, by outputting either Yes (indicating that n is prime) or No (indicating that n is composite). Execution of algorithm A guarantees the following. (i) If n is prime, then the output of A is always Yes. (ii) If n is composite, then there is a probability p > 0 such that the output of A is No with probability p and is Yes with probability 1 -p. On an input m, algorithm A is executed k times (k > 0) and generates a random string ri on the ith execution, 1 £ £ i k, where r1, r2, . . . , rk are mutually independent. 38. If m is composite, what is the probability that in each of the k different executions the output of A is Yes ? (A) 1 (B) pk (C) 1 − p k a f (D) 1 − pk (E) 1 1 − −p k a f 39. Suppose that in each of the k different executions the output of A is No. What is the probability that m is composite? (A) 1 (B) pk (C) 1 − p k a f (D) 1 − pk (E) 1 1 − −p k a f 29Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 40. In systems with support for automatic memory management, a garbage collector typically has the responsibility for reclaiming allocated memory objects whose contents cannot affect any future legal computation. Such objects are identified by determining that they cannot be reached from a root set. Which of the following is NOT part of the root set in a typical garbage collector? (A) Actual parameters of the active procedures (B) Dynamically allocated objects on the heap (C) Global variables of the program (D) Local variables on the call stack (E) Values in machine registers 41. For a connected, undirected graph G V E = , , a f which of the following must be true? I. ∑ ∈ v V degree(v) is even. II. E V ≥ −1 III. G has at least one vertex with degree 1. (A) I only (B) II only (C) III only (D) I and II (E) II and III 30Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 42. Which of the following conditions can be expressed by a Boolean formula in the Boolean variables p p p p 1 2 3 4 , , , and the connectives ∧ ∨ , (without ¬ ) ? I. At least three of p p p p 1 2 3 4 , , , are true. II. Exactly three of p p p p 1 2 3 4 , , , are true. III. An even number of p p p p 1 2 3 4 , , , are true. (A) I only (B) II only (C) III only (D) I and III (E) II and III 43. Consider the collection of all undirected graphs with 10 nodes and 6 edges. Let M and m, respectively, be the maximum and minimum number of connected components in any graph in the collection. If a graph has no selflooop and there is at most one edge between any pair of nodes, which of the following is true? (A) M = 10, m = 10 (B) M = 10, m = 1 (C) M = 7, m = 4 (D) M = 6, m = 4 (E) M = 6, m = 3 31Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 44. Consider the following pseudocode, where n is a nonnegative integer. x = 0; i = 0; while i < n do x = x + 2i; i = i + 1; end Which of the following is a loop invariant for the while statement? (Note: a loop invariant for a while statement is an assertion that is true each time the guard is evaluated during the execution of the while statement.) (A) x i = − 2 1 and 0 i n ≤ < (B) x i = − + 2 1 1 and 0≤ < i n (C) x i = − 2 1 and 0≤ ≤ i n (D) x i = − + 2 1 1 and 0≤ ≤ i n (E) x > 0 and 1≤ < i n 45. In order to find a solution * x to the equation ( ) 0 f x = for a polynomial ( ) f x of degree 2 ≥ with derivative ( )* 0, f x π ¢ Newton’s method does iterations of the form ( ) ( ) 1 , t t t t f x x x f x + = -¢ starting with some initial value 0 * x x π sufficiently close to the desired solution * x to ensure convergence to *. x For fixed values of 0 x and *, x which of the following represents the order of growth of the minimal number of iterations required to compute * x to b bits of accuracy as a function of ? b (A) ( ) 1 O (B) ( ) loglog O b (C) ( ) log O b (D) ( ) O b (E) ( ) O b 32Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 46. Which of the following statements about a remote procedure call is true? (A) It is used to call procedures with addresses that are farther than 16 2 bytes away. (B) It cannot return a value. (C) It cannot pass parameters by reference. (D) It cannot call procedures implemented in a different language. (E) It is used to call procedures at an outer nesting level. 47. Let M be a single-tape, deterministic Turing machine with tape alphabet blank, , , 01 { } and let C denote the (possibly infinite) computation of M starting with a blank tape. The input to each problem below is M, together with a positive integer n. Which of the following problems is (are) decidable? I. The computation C lasts for at least n steps. II. The computation C lasts for at least n steps, and M prints a 1 at some point after the nth step. III. M scans at least n distinct tape squares during the computation C. (A) None (B) III only (C) I and II only (D) I and III only (E) I, II, and III 33Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 48. Which of the following is NOT a property of bitmap graphics? (A) Fast hardware exists to move blocks of pixels efficiently. (B) Realistic lighting and shading can be done. (C) All line segments can be displayed as straight. (D) Polygons can be filled with solid colors and textures. (E) The complexity of the image representation is independent of the image. 49. In a pipelined RISC computer where all arithmetic instructions have the same CPI (cycles per instruction), which of the following actions would improve the execution time of a typical program? I. Increasing the clock cycle rate II. Disallowing any forwarding in the pipeline III. Doubling the sizes of the instruction cache and the data cache without changing the clock cycle time (A) I only (B) II only (C) III only (D) I and II (E) I and III 34Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. Questions 50-51 refer to the following information. The All Pairs Shortest Path Problem can be specified as follows. Input: Directed graph G V E , , ( ) where V n = 1 2 , , . . . , k p Cost C i j , a fOE » • { } + for all i j V , , OE where C i j , a f= ∞ if and only if i j E , a foe Definition: D i j , a f is the length of the shortest path from i to j for all i j V , . OE If there is no path from i to j, then D i j , . a f = • If i j = , then D i j , . a f = 0 Problem: Find D i j , a f for all i j V , . OE The Floyd-Warshall algorithm gives a dynamic programming solution to the problem by defining an array A k i j , , a f for 0 £ £ k n and i j V , OE by the following condition. A k i j ( , , ) is the length of a shortest path from i to j such that all intermediate nodes on the path are in 1 2 , , . . . , k k p (where no intermediate nodes are allowed if k = 0 ). Then D i j A n i j , ,, . a f a f = The algorithm computes A k i j , , a f using a recurrence on k, where the initial step is given as follows. A i j Ci j 0, , , a f a f = for all i j V , OE such that i j π A i i ( , , ) 0 0 = for all i V OE 50. Which of the following is the general step in the recurrence, where 1 £ £ k n? (A) A k i j A i k A k j k , , min , , , , a f a f k p = ( ) + < ` ` ` (B) A k i j A k i A k j k , , min , , , , a f a f k p = -( )+ -< ` ` ` 1 1 (C) A k i j A k i j A k i k A k k j , , min , , , , , , , a f a f a f k p = --( )+ -1 1 1 (D) A k i j C i j A k i k A k k j , , min , , , , , , a f a f a f k p = ( ) + (E) A k i j C i j A k i k A k k j , , min , , , , , , a f a f a f k p = -( )+ -1 1 51. What is the running time of the Floyd-Warshall algorithm? (A) Q n ( ) (B) Q n2 d i (C) Q n3 d i (D) Qn n 3 log d i (E) Q n4 d i 35Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 52. A transaction schedule is serializable if its effect is equivalent to that of some serial schedule. Consider a bookkeeping operation consisting of two transactions — T1 and T2 — that are required to keep the sum A + B + C unchanged. Which of the following pairs of transactions will always result in a serializable schedule? I. T1 T2 Lock A; Lock B; A = A -10; B = B -20; Unlock A; Unlock B; B = B + 10; C = C + 20; II. T1 T2 A = A -10; Lock B; Lock B; B = B -20; B = B + 10; Unlock B; Unlock B; C = C + 20; III. T1 T2 Lock A; Lock A; A = A -10; B = B -20; Unlock A; Unlock A; B = B + 10; C = C + 20; (A) I only (B) II only (C) III only (D) I and II (E) II and III 36Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 53. Which of the following is NOT a reasonable justification for choosing to busy-wait on an asynchronous event? (A) The wait is expected to be short. (B) A busy-wait loop is easier to code than an interrupt handler. (C) There is no other work for the processor to do. (D) The task must meet some hard real-time deadlines. (E) The program executes on a time-sharing system. 54. The Singleton design pattern is used to guarantee that only a single instance of a class may be instantiated. Which of the following is (are) true of this design pattern? I. The Singleton class has a static factory method to provide its instance. II. The Singleton class can be a subclass of another class. III. The Singleton class has a private constructor. (A) I only (B) II only (C) III only (D) I and III only (E) I, II, and III 37Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 55. Assume that a target t is an integer value stored in some element of integer array x, which is sorted in nondecreasing order, and consider the following outline of a loop to search for t. while ( π x[h] t) { P; } If establishes the invariant “ £ < x[h] t x[k]”, which of the following definitions for P, taken individually, would ensure that the loop terminates with , x[h] t = assuming that t appears in the array? I. if x[h]< t then h := h + 1 II. h := h + 1 III. k := k -1 (A) I only (B) II only (C) III only (D) I and II (E) I and III 56. Assume that a debugger places a breakpoint at a load instruction at virtual address 0x77E81234 (hexadecimal notation) in a debugged process P. If the text segment of P begins at 0x77E80000 in P’s virtual address space and if the debugger has mapped this same text segment at 0x01000000 in its virtual address space, which of the following is the virtual address used by the debugger in its WRITE operation, along with a description of how the debugger has mapped the virtual memory page containing this address? (A) 0x01001234; page mapped with READ/WRITE access (B) 0x01001234; page mapped with COPY-ON-WRITE access (C) 0x76E81234; page mapped with READ/WRITE access (D) 0x76E81234; page mapped with COPY-ON-WRITE access (E) 0x77E81234; page mapped with READ/WRITE access 38Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 57. Company X shipped 5 computer chips, 1 of which was defective, and Company Y shipped 4 computer chips, 2 of which were defective. One computer chip is to be chosen uniformly at random from the 9 chips shipped by the companies. If the chosen chip is found to be defective, what is the probability that the chip came from Company Y ? (A) 29 (B) 49 (C) 12 (D) 23 (E) 57 58. An Euler circuit of an undirected graph is a circuit in which each edge of the graph appears exactly once. Which of the following undirected graphs must have an Euler circuit? I. A complete graph with 12 vertices II. A complete graph with 13 vertices III. A tree with 13 vertices (A) I only (B) II only (C) III only (D) I and II (E) I and III 59. Consider the following two languages. L x ab x a b L x abc x a b c 12 = OE{ } = OE{ } , * , , * has equally many ’s and ’s has equally many ’s, ’s, and ’s k p k p Which of the following is true about L1 and L2 ? (A) L1 and L2 are both regular. (B) L1 is regular, and L2 is context-free but not regular. (C) Neither L1 nor L2 is regular, but both are context-free. (D) L1 is context-free but not regular, and L2 is not context-free. (E) Neither L1 nor L2 is context-free. 39Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 60. Consider the following possible data structures for a set of n distinct integers. I. A min-heap II. An array of length n sorted in increasing order III. A balanced binary search tree For which of these data structures is the number of steps needed to find and remove the 7th largest element (log ) O n in the worst case? (A) I only (B) II only (C) I and II (D) I and III (E) II and III 61. Which of the following problems is (are) decidable? I. Given a (finite) string w, is w a prefix of the decimal expansion of p ? II. Given a program and an input, is the program’s output the decimal expansion of p ? III. Given a program that takes as input a prefix of the decimal expansion of p, is the program’s output always the same for every prefix? (A) I only (B) II only (C) III only (D) I and II only (E) I, II, and III 40Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 62. Which of the following problems would have polynomial time algorithms if it is assumed that P NP π ? I. Given a combinational circuit with n inputs and m outputs and n2 gates, where each gate is either AND, OR, or NOT, and given m values c cm 1, . . . , in 0 1 , , k p either find a string of n input values b bn 1, . . . , in 0 1 , k p that would produce c cm 1, . . . , as output or determine that c cm 1, . . . , is not a possible output of the circuit. II. Given an n by n matrix A with rational number entries, either find the exact inverse A−1 of A or determine that A−1 does not exist. (Assume that each rational number is expressed as a pair a/b of integers b π 0 a f, where a and b are expressed in binary notation.) III. Given a directed graph with nodes numbered 1, 2, . . . , n, and given positive integer weights assigned to the edges, either find the length of a shortest path from node 1 to node n or determine that no such path exists. (Here the length of a path is the sum of the lengths of the edge weights on the path.) (A) I only (B) II only (C) III only (D) I and II (E) II and III 63. Which of the following characteristics of a programming language is best specified using a context-free grammar? (A) Identifier length (B) Maximum level of nesting (C) Operator precedence (D) Type compatibility (E) Type conversion 41Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 64. Consider the following function. f(k) { x = 2; for i = 1 to k x = x * x; return x; } If n and k are positive integers, then the least value of k such that f(k) > n is approximately (A) 2 2 log (log ) n (B) 2 log n (C) n (D) 2 log n n (E) 2n 65. Let T be a depth-first search tree of a connected undirected graph G. For each vertex v of T, let pre v ( ) be the number of nodes visited up to and including v during a preorder traversal of T , and post v ( ) be the number of nodes visited up to and including v during a postorder traversal of T. The lowest common ancestor of vertices u and v in T is a vertex w of T such that w is an ancestor of both u and v, and no child of w is an ancestor of both u and v. Let u v , ( ) be an edge in G that is not in T, such that pre u pre v ( ) < ( ). Which of the following statements about u and v must be true? I. post u post v ( ) < ( ) II. u is an ancestor of v in T. III. If w is the lowest common ancestor of u and v in T, then w u = . (A) I only (B) II only (C) III only (D) I and II (E) II and III 42Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 66. Consider languages L and L1, each over the alphabet a b , , { } where L ww x L 1= OE contains some as a substring k p. Which of the following must be true about L and L1 ? I. If L is regular, then L1 is regular. II. If L is context-free, then L1 is context-free. III. If L is recursive, then L1 is recursive. (A) I only (B) III only (C) I and III only (D) II and III only (E) I, II, and III 67. For each nonnegative integer n, let Rn be the greatest possible number of regions into which the plane can be partitioned by n straight lines. For example, R R 0 1 1 2 = = and . Then Rn has order (A) Θ n a f (B) Θn n log a f (C) Θ n2 d i (D) Θ 2n d i (E) Θ n ! a f 43Unauthorized copying or reuse of any part of this page is illegal. GO ON TO THE NEXT PAGE. 68. Which of the following comes closest to being a perfectly secure encryption scheme? (A) The Caesar Cipher, a substitution cipher (B) DES (Data Encryption Standard), a symmetric-key algorithm (C) Enigma, a transposition cipher (D) One-time pad (E) RSA, a public-key algorithm 69. Suppose Q and R are languages. Assuming P NP π , which of the following implies that R is not in P ? (A) R is in NP. (B) Q is in NP and Q is polynomial-time reducible to R. (C) Q is in NP and R is polynomial-time reducible to Q. (D) Q is NP-complete and Q is polynomial-time reducible to R. (E) Q is NP-complete and R is polynomial-time reducible to Q. 44Unauthorized copying or reuse of any part of this page is illegal. 70. Let N be the set of all natural numbers. Which of the following sets are countable? I. The set of all functions from N to 0 1 , k p II. The set of all functions from 0 1 , k p to N III. The largest subset of N (A) None (B) I and II only (C) I and III only (D) II and III only (E) I, II, and III S T O P If you finish before time is called, you may check your work on this test. 45NOTE: To ensure prompt processing of test results, it is important that you fill in the blanks exactly as directed. ® A B C D A B C D A B C D A B C D EEEE A C D E B 0123456789 0123456789 0123456789 0123456789 0123456789 6. TITLE CODE PRINT: ___________________________________________________________________ (LAST) (FIRST) (MIDDLE) SIGN: ____________________________________________________________________ I Educational Testing Service Princeton, New Jersey 08541 DO NOT OPEN YOUR TEST BOOK UNTIL YOU ARE TOLD TO DO SO. Sample Answer Example: What city is the capital of France? (A) Rome (B) Paris (C) London (D) Cairo (E) Oslo CORRECT ANSWER PROPERLY MARKED IMPROPER MARKS GRADUATE RECORD EXAMINATIONS SUBJECT TEST B. The Subject Tests are intended to measure your achievement in a specialized field of study. Most of the questions are concerned with subject matter that is probably familiar to you, but some of the questions may refer to areas that you have not studied. Your score will be determined by subtracting one-fourth the number of incorrect answers from the number of correct answers. Questions for which you mark no answer or more than one answer are not counted in scoring. If you have some knowledge of a question and are able to rule out one or more of the answer choices as incorrect, your chances of selecting the correct answer are improved, and answering such questions will likely improve your score. It is unlikely that pure guessing will raise your score; it may lower your score. You are advised to use your time effectively and to work as rapidly as you can without losing accuracy. Do not spend too much time on questions that are too difficult for you. Go on to the other questions and come back to the difficult ones later if you can. YOU MUST INDICATE ALL YOUR ANSWERS ON THE SEPARATE ANSWER SHEET. No credit will be given for anything written in this examination book, but you may write in the book as much as you wish to work out your answers. After you have decided on your response to a question, fill in the corresponding oval on the answer sheet. BE SURE THAT EACH MARK IS DARK AND COMPLETELY FILLS THE OVAL. Mark only one answer to each question. No credit will be given for multiple answers. Erase all stray marks. If you change an answer, be sure that all previous marks are erased completely. Incomplete erasures may be read as intended answers. Do not be concerned that the answer sheet provides spaces for more answers than there are questions in the test. Copy the Test Name and Form Code in box 7 on your answer sheet. TEST NAME ___________________________________ FORM CODE _______________________ Copy this code in box 6 on your answer sheet. Then fill in the corresponding ovals exactly as shown. SUBJECT TEST A. Print and sign your full name in this box: 2 9 7 0 0 Computer Science GR0329 4647 COMPUTER SCIENCE TEST PRACTICE BOOK Scoring Your Subject Test Computer Science Test scores typically range from 540 to 860. The range for different editions of a given test may vary because different editions are not of precisely the same difficulty. The differences in ranges among different editions of a given test, however, usually are small. This should be taken into account, especially when comparing two very high scores. In general, differences between scores at the 99th percentile should be ignored. The score conversion table on page 49 shows the score range for this edition of the test only. The worksheet on page 48 lists the correct answers to the questions. Columns are provided for you to mark whether you chose the correct (C) answer or an incorrect (I) answer to each question. Draw a line across any question you omitted, because it is not counted in the scoring. At the bottom of the page, enter the total number correct and the total number incorrect. Divide the total incorrect by 4 and subtract the resulting number from the total correct. This is the adjustment made for guessing. Then round the result to the nearest whole number. This will give you your raw total score. Use the total score conversion table to find the scaled total score that corresponds to your raw total score. Example: Suppose you chose the correct answers to 38 questions and incorrect answers to 22. Dividing 22 by 4 yields 5.5. Subtracting 5.5 from 38 equals 32.5, which is rounded to 33. The raw score of 33 correspoond to a scaled score of 760.48 COMPUTER SCIENCE TEST PRACTICE BOOK Worksheet for the Computer Science Test, Form GR0329 Only Answer Key and Percentages* of Examinees Answering Each Question Correctly Correct (C) Incorrect (I) Total Score: C – I/4 = ____________ Scaled Score (SS) = ____________ QUESTION TOTAL Number Answer P + C I QUESTION TOTAL Number Answer P + C I 1 D 65 36 C 71 2 C 84 37 A 32 3 D 71 38 C 80 4 C 76 39 A 47 5 B 70 40 B 30 6 C 96 41 D 47 7 ** ** 42 A 66 8 ** ** 43 C 65 9 A 41 44 C 48 10 D 78 45 C 18 11 A 68 46 C 56 12 C 46 47 D 33 13 C 76 48 C 56 14 C 60 49 E 60 15 B 69 50 C 26 16 D 42 51 C 36 17 B 48 52 B 43 18 C 77 53 E 33 19 D 71 54 E 37 20 C 42 55 D 71 21 A 25 56 A 27 22 ** ** 57 D 68 23 D 41 58 B 39 24 D 70 59 D 38 25 E 72 60 E 24 26 A 35 61 A 53 27 B 63 62 E 36 28 A 62 63 C 52 29 E 27 64 A 45 30 ** ** 65 E 18 31 E 45 66 E 25 32 B 42 67 ** ** 33 A 55 68 D 40 34 D 34 69 D 35 35 D 49 70 D 23 * The P+ column indicates the percent of COMPUTER SCIENCE Test examinees who answered each question correctly; it is based on a sample of November 2003 examinees selected to represent all COMPUTER SCIENCE Test examinees tested between July 1, 2000 and June 30, 2003 ** Items 7, 8, 22, 30, and 67 were not scored when this form of the test was originally administered. 49 COMPUTER SCIENCE TEST PRACTICE BOOK Score Conversions and Percents Below* For GRE Computer Science Test, Form GR0329 Only TOTAL SCORE Raw Score Scaled Score % Raw Score Scaled Score % * Percent scoring below the scaled score is based on the performance of 10,579 examinees who took the COMPUTER SCIENCE Test between July 1, 2000 and June 30, 2003. 65 900 99 61-64 890 99 57-60 880 98 53-56 870 97 50-52 860 95 48-49 850 93 46-47 840 91 44-45 830 87 42-43 820 84 40-41 810 82 38-39 800 78 37 790 74 35-36 780 71 34 770 67 32-33 760 63 31 750 58 29-30 740 55 28 730 51 27 720 47 26 710 44 24-25 700 41 23 690 37 22 680 34 21 670 31 20 660 28 19 650 25 17-18 640 22 16 630 20 15 620 18 14 610 16 13 600 14 12 590 12 11 580 10 10 570 8 9 560 7 8 550 6 7 540 5 6 53 4 5 520 3 4 510 3 2-3 500 2 1 490 1 0 480 150 COMPUTER SCIENCE TEST PRACTICE BOOK Evaluating Your Performance Now that you have scored your test, you may wish to compare your performance with the performance of others who took this test. Both the worksheet on page 48 and the table on page 49 use performance data from GRE Computer Science Test examinees. The data in the worksheet on page 48 are based on the performance of a sample of the examinees who took this test in November 2003. This sample was selected to represent the total population of GRE Computer Science Test examinees tested between July 1, 2000 and June 30, 2003. The numbers in the column labeled “P+” on the worksheet indicate the percentages of examinees in this sample who answered each question correctly. You may use these numbers as a guide for evaluating your performance on each test question. The table on page 49 contains, for each scaled score, the percentage of examinees tested between July 1, 2000 and June 30, 2003 who received lower scores. Interpretive data based on the scores earned by examinnee tested in this three-year period will be used by admissions officers in the 2004-05 testing year. These percentages appear in the score conversion table in a column to the right of the scaled scores. For example, in the percentage column opposite the scaled score of 730 is the number 51. This means that 51 percent of the GRE Computer Science Test examinees tested between July 1, 2000 and June 30, 2003 scored lower than 730. To compare yourself with this population, look at the percentage next to the scaled score you earned on the practice test. Note: due to changes in the test-taking population, the percentile rank data changes over time. Percentile rank information is kept current on the GRE Web site and may be obtained by visiting the GRE Web at www.gre.org/codelst.html, or by contacting the GRE Program. It is important to realize that the conditions under which you tested yourself were not exactly the same as those you will encounter at a test center. It is impossible to predict how different test-taking conditions will affect test performance, and this is only one factor that may account for differences between your practice test scores and your actual test scores. By comparing your performannc on this practice test with the performance of other GRE Computer Science Test examinees, however, you will be able to determine your strengths and weaknesses and can then plan a program of study to prepare yourself for taking the GRE Computer Science Test under standard conditions.54721-007623 • U54E8 • Printed in U.S.A. 724474®