Georgia Tech ISYE 6644 Simulation and Modeling for
Engineering and Science: Practice Test 1 (Units 1-7) -
Questions with Correct Answers
100 questions
1. A 'what-if' question for a hospital emergency department simulation could be:
A. How would adding one triage nurse change average patient wait?
B. What is the exact wait of the next patient?
C. Which random seed is best?
D. What color should entities be?
Correct Answer: A. How would adding one triage nurse change average patient wait?
Rationale: Scenario analysis.
2. For an M/M/1 queue, the proportion of time the server is busy equals:
A. 1 - rho
B. rho^2
C. rho
D. 1/rho
Correct Answer: C. rho
Rationale: Utilization.
3. A deterministic model is one that:
A. Always uses exponential interarrivals
B. Contains no random components
C. Requires many replications
D. Has random service times
Correct Answer: B. Contains no random components
Rationale: Same inputs give same outputs.
4. A static simulation is one in which:
A. Time advances event by event
B. The system is always a queue
C. Time plays no role, such as Monte Carlo integration
D. Entities move through stations
Correct Answer: C. Time plays no role, such as Monte Carlo integration
Rationale: Monte Carlo.
5. The simulation clock:
A. Measures computer CPU time
B. Keeps track of the current simulated time
C. Counts random numbers used
D. Stores entity attributes
Correct Answer: B. Keeps track of the current simulated time
Rationale: Advances to next event time.
,6. Validation of a simulation model asks:
A. Does the code compile?
B. Is the random number generator fast?
C. Does the model accurately represent the real system (built the right model)?
D. Is the clock correct?
Correct Answer: C. Does the model accurately represent the real system (built the right
model)?
Rationale: Compare to real system.
7. A steady-state (non-terminating) simulation studies:
A. Long-run behavior of a system that does not naturally end
B. Only the first hour of operation
C. Systems with no randomness
D. Terminating events only
Correct Answer: A. Long-run behavior of a system that does not naturally end
Rationale: Initialization bias issues.
8. Which famous early use of Monte Carlo methods is associated with Los Alamos?
A. Inventory control in 1990
B. Arena development
C. Airline yield management
D. Neutron diffusion calculations in the 1940s
Correct Answer: D. Neutron diffusion calculations in the 1940s
Rationale: Ulam, von Neumann.
9. The future event list (event calendar) contains:
A. Past events already completed only
B. All random numbers ever generated
C. Scheduled events ordered by their time of occurrence
D. Only entities in queue
Correct Answer: C. Scheduled events ordered by their time of occurrence
Rationale: Drives next-event time advance.
10. A 'model' in simulation is best described as:
A. An exact copy of the real system
B. A simplified representation of a system used to study its behavior
C. A random number generator
D. A statistical test
Correct Answer: B. A simplified representation of a system used to study its behavior
Rationale: Abstraction.
11. For an M/M/1 queue, the expected time in system W equals:
A. lambda/(mu - lambda)
B. 1/(mu - lambda)
C. mu/(mu - lambda)
D. 1/lambda
Correct Answer: B. 1/(mu - lambda)
Rationale: W = L/lambda.
, 12. Simulation is often preferred over analytical models when:
A. A simple formula gives the exact answer
B. The system is too complex for tractable closed-form analysis
C. There is no randomness and one equation suffices
D. No data or logic about the system exists
Correct Answer: B. The system is too complex for tractable closed-form analysis
Rationale: Complex, stochastic systems.
13. Which is NOT a typical step in a simulation study?
A. Problem formulation
B. Data collection
C. Experimental design and output analysis
D. Guaranteeing the answer without data
Correct Answer: D. Guaranteeing the answer without data
Rationale: Studies need data and analysis.
14. A resource is:
A. An arriving customer
B. A random variate
C. An output statistic
D. Something entities compete for, such as a server, machine or clerk
Correct Answer: D. Something entities compete for, such as a server, machine or clerk
Rationale: Has capacity.
15. Initialization bias (warm-up) occurs because:
A. Initial conditions, such as an empty system, are not representative of steady state
B. Random numbers are biased
C. The clock starts at 1
D. Entities have no attributes
Correct Answer: A. Initial conditions, such as an empty system, are not representative of
steady state
Rationale: Delete warm-up data.
16. The derivative of e^(2x) is:
A. e^(2x)
B. 2e^(2x)
C. e^(2x)/2
D. 2x e^(2x-1)
Correct Answer: B. 2e^(2x)
Rationale: Chain rule.
17. The integral of 2x from 0 to 3 is:
A. 6
B. 3
C. 18
D. 9
Correct Answer: D. 9
Rationale: x^2 from 0 to 3.
Engineering and Science: Practice Test 1 (Units 1-7) -
Questions with Correct Answers
100 questions
1. A 'what-if' question for a hospital emergency department simulation could be:
A. How would adding one triage nurse change average patient wait?
B. What is the exact wait of the next patient?
C. Which random seed is best?
D. What color should entities be?
Correct Answer: A. How would adding one triage nurse change average patient wait?
Rationale: Scenario analysis.
2. For an M/M/1 queue, the proportion of time the server is busy equals:
A. 1 - rho
B. rho^2
C. rho
D. 1/rho
Correct Answer: C. rho
Rationale: Utilization.
3. A deterministic model is one that:
A. Always uses exponential interarrivals
B. Contains no random components
C. Requires many replications
D. Has random service times
Correct Answer: B. Contains no random components
Rationale: Same inputs give same outputs.
4. A static simulation is one in which:
A. Time advances event by event
B. The system is always a queue
C. Time plays no role, such as Monte Carlo integration
D. Entities move through stations
Correct Answer: C. Time plays no role, such as Monte Carlo integration
Rationale: Monte Carlo.
5. The simulation clock:
A. Measures computer CPU time
B. Keeps track of the current simulated time
C. Counts random numbers used
D. Stores entity attributes
Correct Answer: B. Keeps track of the current simulated time
Rationale: Advances to next event time.
,6. Validation of a simulation model asks:
A. Does the code compile?
B. Is the random number generator fast?
C. Does the model accurately represent the real system (built the right model)?
D. Is the clock correct?
Correct Answer: C. Does the model accurately represent the real system (built the right
model)?
Rationale: Compare to real system.
7. A steady-state (non-terminating) simulation studies:
A. Long-run behavior of a system that does not naturally end
B. Only the first hour of operation
C. Systems with no randomness
D. Terminating events only
Correct Answer: A. Long-run behavior of a system that does not naturally end
Rationale: Initialization bias issues.
8. Which famous early use of Monte Carlo methods is associated with Los Alamos?
A. Inventory control in 1990
B. Arena development
C. Airline yield management
D. Neutron diffusion calculations in the 1940s
Correct Answer: D. Neutron diffusion calculations in the 1940s
Rationale: Ulam, von Neumann.
9. The future event list (event calendar) contains:
A. Past events already completed only
B. All random numbers ever generated
C. Scheduled events ordered by their time of occurrence
D. Only entities in queue
Correct Answer: C. Scheduled events ordered by their time of occurrence
Rationale: Drives next-event time advance.
10. A 'model' in simulation is best described as:
A. An exact copy of the real system
B. A simplified representation of a system used to study its behavior
C. A random number generator
D. A statistical test
Correct Answer: B. A simplified representation of a system used to study its behavior
Rationale: Abstraction.
11. For an M/M/1 queue, the expected time in system W equals:
A. lambda/(mu - lambda)
B. 1/(mu - lambda)
C. mu/(mu - lambda)
D. 1/lambda
Correct Answer: B. 1/(mu - lambda)
Rationale: W = L/lambda.
, 12. Simulation is often preferred over analytical models when:
A. A simple formula gives the exact answer
B. The system is too complex for tractable closed-form analysis
C. There is no randomness and one equation suffices
D. No data or logic about the system exists
Correct Answer: B. The system is too complex for tractable closed-form analysis
Rationale: Complex, stochastic systems.
13. Which is NOT a typical step in a simulation study?
A. Problem formulation
B. Data collection
C. Experimental design and output analysis
D. Guaranteeing the answer without data
Correct Answer: D. Guaranteeing the answer without data
Rationale: Studies need data and analysis.
14. A resource is:
A. An arriving customer
B. A random variate
C. An output statistic
D. Something entities compete for, such as a server, machine or clerk
Correct Answer: D. Something entities compete for, such as a server, machine or clerk
Rationale: Has capacity.
15. Initialization bias (warm-up) occurs because:
A. Initial conditions, such as an empty system, are not representative of steady state
B. Random numbers are biased
C. The clock starts at 1
D. Entities have no attributes
Correct Answer: A. Initial conditions, such as an empty system, are not representative of
steady state
Rationale: Delete warm-up data.
16. The derivative of e^(2x) is:
A. e^(2x)
B. 2e^(2x)
C. e^(2x)/2
D. 2x e^(2x-1)
Correct Answer: B. 2e^(2x)
Rationale: Chain rule.
17. The integral of 2x from 0 to 3 is:
A. 6
B. 3
C. 18
D. 9
Correct Answer: D. 9
Rationale: x^2 from 0 to 3.
