Exam 1 Qualitative Questions

List the three ingredients of a simulation model identified in the introductory video for this topic and give an example of each type within the context of inventory management.
Click the card to flip 👆
1 / 20
Terms in this set (20)
List the three ingredients of a simulation model identified in the introductory video for this topic and give an example of each type within the context of inventory management.
Policy variables - order quantity, reorder point, safety stock
Random variables - demand, lead time
Outcome variables - holding costs, shortage/stockout costs, inventory turnover.
What is the general purpose of simulation as described in the introductory video for this topic?
The general purpose of simulation is to build an abstract representation of a real system that you can experiment with in a computer setting. This allows you to assess the effects of policy variables on outcomes in light of random variables on the computer, rather than engage in costly experimentation with the real system itself.
What is multicollinearity and what are its adverse effects? (Be specific)
Multicollinearity occurs when there is a high level of correlation between two or more of the independent variables in a regression model. The problem can cause a distortion of the sign and magnitude of the regression slope coefficients and (because of inflated coefficient standard errors) problems with their significance tests.
The key parameters of moving average and exponential smoothing forecasting models are n and α , respectively. Describe how the relative degree of stability/responsiveness of forecasts are affected by the settings of n and α. (Be specific).
As n increases, the stability of the forecast increases because you hare averaging over more data points. As n decreases, the forecast becomes more responsive to recent changes in demand because of the averaging over fewer demand points. As alpha decreases toward zero, forecasts are more stable because more weight is given to the previous smoothed value and less to the most recent demand. As alpha increases toward one, the forecast become much more responsive to recent demand.
Forecasting
The art and science of predicting future events
Elements of a Good Forecast
Timely - (enough time to support the decision)
Accurate - (necessary degree of accuracy)
Reliable - (performs consistently over time)
Meaningful Units - (product volume, $, etc.)
In Writing (accountability and clarity)
Easy to Understand and Use (simplicity is a virtue)
Cost Effective (cost-benefit tradeoffs)
Delphi method
Panel of experts identified, surveyed, collected and analyzed, returned to panel of experts. Several iterations of this occur until opinions converge
Strengths of Delphi Method
Broad range of expertise / opinion
Useful when data-driven models can't be used
Allows for freedom in responses because anonymous process
Weakness of Delphi Method
Time consuming and costly
Pooled ignorance
Limited accountability
Pooled ignorance
Nature of the iterative process leads to convergence of opinion on the middle ground.
RegressionExplanatory: determine which independent variable is the key driver of the dependent variable Predictive: create forecastsTotal Sum of Squares (TSS)sum of (y - ybar)squaredError Sum of Squares (ESS)sum of (y-yhat)squared. Minimizing ESS maximizes RSS and R SquaredRegression Sum of Squares = TSS-ESSTSS-ESSR squared (coefficient of determination)RSS/TSS. The proportion of variation in y that is explained by the independent variables (x's) R Square value of 1 = 100% perfectly correlatedIssues in RegressionAssumptions Multicollinearity Range of predictors OutliersAssumptions in testing significance of regression models:1. Normality of error terms 2. Homogeneity of variance of error termsMulticollinearityProblem that occurs when two or more of the independent variables are highly correlated with one another. Can result in improper signs for slope coefficients and significance testing problems.Range of predictorsOne should be careful about using a regression model to make predictions for independent variable values that are far outside the range of such values used to establish the regression equations. Example - if patient age used as independent variable to predict satisfaction, and you age range in independent variable is 40-70, the predictive value may not hold for patients aged 17.OutliersCan have strong effects on the fit of a regression model. They must be carefully considered. Might want to remove outliers when drawing a regression line.