4. Regression Coefficients
b0= 141.35
b1= 24.775
b2= 19.075
b3= 9.925
b4= 16.775
b5= -0.325
b6= 1.375
b7= -1.825
The Regression Coefficient is the constant b in the regression equation.
5. INTERACTIONS GRAPHS AND INTERPRETATION
Credit History Low High
Good (-) 30.8 56.4
Fair(+) 49.8 142
0
100
200
Low High
AxisTitle
Credit History
High Low Chart
Good (-) Fair(+)
Mild Interaction – Its better to have
Positive credit which is higher than
negative credit which is low.
6. Interactions graphs and interpretation
Mortgage Size Low High
<$500,000 30.8 69.4
>$500,000 41.8 142
Mortgage Size Low High
<$500,000 30.8 69.4
>$500,000 41.8 142
30.8
69.441.8
142
0
100
200
300
Low High
Mortgage
<$500,000 >$500,000
Mild Interaction – The mortgage
that is greater than $500,000 has
the highest approval times. It is
better to have the higher approval
time.
7. Interactions graphs and interpretation
Region Low High
Western 54.6 142
Eastern 30.8 120.6
0
100
200
300
Low High
Region
Western Eastern
Mild Interaction – The eastern region
has the highest approval time. It is
better to be in the eastern region for
the mortgage approval times.
8. INTERACTIONS GRAPHS AND INTERPRETATION
Interactions can be synergistic or antagonistic:
Synergistic interaction is positive.
Two variables involved produce an effect that is larger than would be predicted if the effects of
the two were additive.
Antagonistic Interaction is negative.
The effects of the two factors is smaller than would be predicted by the additive effects of the
two factors.
The easiest way to interpret interactions is to construct a plot of the averages of the four groups.
Researched by Hoerl and Snee (2012).
9. INTERACTIONS GRAPHS AND INTERPRETATION
Synergy
Stems from the idea that integration involves unity and wholeness. Through this unity
synergy can be achieved.
Synergy manifests itself through a positive interaction effect.
Antagonistic
The opposite of synergy
Exhibits negative returns.
Failure to achieve consistency.
Researched by Kolsarici, C., & Vakratsas, D. (2018).
10. ANALYSIS OF THE SAMPLE SIZE
Objective – Agreement must be obtained to ensure success.
Output variables Identify the output or measure the process performance.
Identify the levels of input to be studied.
Verify the available resources for the size of the experiment.
Time – the amount of time for the number of tests performed and when the results are needed.
Funds – How much money to spend depends on the amount spent on personnel and the
experimentation. No more than 20% should be spent on the first experiment.
.
11. ANALYSIS OF THE SAMPLE SIZE
Replication - replicating some or all of the experiment is important as it
increases the sensitivity of the experiment. By doing so assists us in
detecting smaller differences.
Randomization – Running a test in an experiment in a random order. This
guards against any unknown changes that may have occurred during the
conduct of the experiment.
The planning t the results f the test. Will the data be collected electronically or manually.
Researched by Hoerl and Snee (2012).
12. VARIABLES OF INTEREST TO MEASURE
AND STUDY
Annual Salary
Marital Status
Criminal Background Check
Debt to Earnings Ratio
13. INTERACTIONS BETWEEN DOE WITH 3
FACTOR EXPERIMENT
Factorial experiments enable us to identify interactions and is a three-factor experiment. The three possible
two factor interactions are (x1x2, x1x3, x2x3) and one three factor interaction (x1x2x3). Although they are rarely
important in real applications, a three factor interaction would mean that the interaction of x1 and x2 is
dependent on the level of x3 .
Researched by Hoerl and Snee (2012).
14. RECOMMENDATIONS AND
CONCLUSION OF THE DOE
Use the annual salary, marital status, background checks and debt to earnings ratio instead of
just the credit check, region and mortgage size.
Use other sample times for the approval rate because the interaction effect charts have mild
interaction instead of strong interaction.
Use different sample numbers because the effect numbers, which are the difference between
the avg(+) minus the avg(-), are below 20 and some are negative numbers.
15. REFERENCES
Singleton, C., Gilman, J., Rollit, J., Zhang, K., Parker, D. A., & Love, J. (2019). A design of experiments approach for the rapid
formulation of a chemically defined medium for metabolic profiling of industrially important microbes.
PLoS ONE, 14(6), 1–18. https://doi.org/10.1371/journal.pone.0218208
Hoerl, R., Snee, Ron. (2012) Statistical Thinking, Improving Business Performance. Hoboken, New Jersey: John Wiley & Sons, Inc.
Kolsarici, C., & Vakratsas, D. (2018). Synergistic, Antagonistic, and Asymmetric Media Interactions. Journal of Advertising,
47(3), 282–300. https://doi.org/10.1080/00913367.2018.1471757