MA 16 - Analysis: Linear Programming and Regression Analysis Flashcards
Most important issue in estimating a cost function
Determining if a cause-and-effect relationship exists between the level of an activity and costs related to that level of activity. If there is, this is a cost driver.
What creates economic plausability?
A cause-and-effect relationship, not just correlation
5 steps to estimating a cost function
- Choose dependent variable (y axis) - this is cost to be predicted.
- Identify independent variable (x axis) - cost driver.
- Collect data on both.
- Plot data.
- Estimate the cost function.
linear cost function formula
y = a + bX
y is predicted cost (dependant variable)
a = fixed costs
b = activity level of cost driver
X = cost driver
SAME AS TC = FC + VC x X
Then it is compared with Y (actual cost)
What does goodness of fit measure?
How close y matches Y (predictive cost vs actual cost)
what is r2?
Goodness of fit - the closer to 1 the more positive correlation there is.
-1 = negative correlation
0 = no correlation
2 further steps for regression analysis
Use regression analysis in Excel to provide a summary output
Analyze results based upon economic plausability, goodness of fit and statistical sig
What are you looking for when analyzing the results?
Economic plausability - does this make sense?
Goodness of fit - r2 - closer to 1 the greater the accuracy
Significance F - should be less than 0.05 - shows whether the results are statistically significant
P-value - should be less than 0.05 - shows stat sig of independent variables.
What does linear programming do?
Offer a technique to determine an optimal outome.
Steps for linear programming
- determine variables
- identify decision to be made
- define constraints
- define objective
- use solver to fill in the cells
