Decision Support Systems MCQs (Set-4)

Difficulty: Hard

When outcome probabilities are credible, which measure averages financial results?

A Maximum observed payoff
B Unweighted score total
C Lowest possible loss
D Expected monetary value

Which measure accounts for a decision maker’s risk preference?

A Expected utility
B Median cycle time
C Simple payoff maximum
D Raw data count

If one option is no worse in every state and better in one, what condition is shown?

A Deadlock
B Dominance
C Sensitivity
D Data drift

Which analysis identifies the point where benefits cover costs?

A Scenario planning
B Data integration
C Break-even analysis
D Role-based access

How does robust optimization address uncertain inputs?

A Removes every constraint
B Maximizes a single uncertain input
C Assumes one forecast is exact
D Finds solutions resilient across cases

When a constraint’s shadow price is zero, what may this imply?

A Extra capacity adds no value
B The model has no variables
C The objective has no value
D All solutions are infeasible

Why use integer programming for indivisible decisions?

A Model only continuous percentages
B Require whole-number choices
C Ignore resource constraints
D Remove the objective function

What does Monte Carlo output help estimate?

A Only the largest input
B A fixed process sequence
C Outcome distribution under uncertainty
D One exact guaranteed answer

How does Bayesian updating use observed evidence?

A Replace probabilities with rankings
B Discard all prior information
C Guarantee a causal link
D Revise posterior beliefs

Demand shifts and forecast errors grow over time. What model risk may be present?

A Explores plausible conditions
B Data compression
C Index fragmentation
D Role conflict

When a recommendation changes customer behavior and its own input data, what risk emerges?

A Data normalization
B Feedback effects
C Linear feasibility
D Perfect calibration

Why can automation bias undermine DSS review?

A Data becomes more complete
B Models always show uncertainty
C Users over-trust automated advice
D Reviewers see every source

What does human-in-the-loop review add?

A Automatic approval of every result
B Removal of accountability
C Guaranteed unbiased choices
D Contextual judgment and oversight

When rankings shift after small weight changes, what should analysts test?

A Weight sensitivity
B Data masking
C Password strength
D Index coverage

Which feature separates a scenario from a probability forecast?

A It requires exact likelihoods
B Explores plausible conditions
C It always predicts one outcome
D It ranks options by utility

Which record supports tracing transformed data through a DSS?

A Model screen theme
B User profile
C Data lineage
D Decision deadline

What is a key tradeoff of anonymous GDSS input?

A Latency versus encryption
B Access versus normalization
C Accuracy versus storage
D Candor versus accountability

How can explainable AI help reviewers assess advice?

A Show influential factors
B Force a single action
C Hide model inputs
D Remove counterexamples

Which question does a counterfactual explanation answer?

A How keys identify database rows
B Effect of changing one input
C A database key definition
D When a backup was created

Why evaluate a model under several data conditions?

A Eliminate all bias
B Increase the number of users
C Reveal model uncertainty
D Guarantee a stable future

Which practice prevents test data leaking into training?

A Copy target values into inputs
B Reuse all labels
C Tune on final test cases
D Separate evaluation data

What should happen when critical decision data is incomplete?

A Assess sensitivity to missingness
B Treat blanks as zero always
C Hide the missing fields
D Assume all records are random

Why preserve the option to delay an irreversible investment?

A Remove all uncertainty
B Retain real-option value
C Lock in every forecast
D Maximize sunk costs

When should decision accountability be documented?

A Only for discarded alternatives
B Only after a failure
C Before implementing the chosen action
D Never for model advice

How should a DSS recommendation be treated when assumptions fail?

A Proceed without review
B Delete the source data
C Ignore changed conditions
D Reassess the model and choice