I am working on MHA-FP5017 Data Analysis and my project is called, "Predicting an outcome using regression models." I've submitted my work and have rec'd a response from my instructor that I've missed an important piece. He says, "You need to interpret the effect size of the regression coefficient. For example, what is the meaning of the regression coefficient of 107 for the age variable? Below is what I've submitted, and my excel. Any advice would be greatly appreciated. I'm lost...and this is the last time I can resubmit. I'm a solid A student typically, but this class is foreign to me.

Introduction

ABC Medical Center is attempting to determine the reimbursement amount necessary to operate the facility for the following year. Costs of caring for the patient can vary, and the majority of the dollars come directly from charges eligible for reimbursement from the medical coverage of each patient. In healthcare and most other businesses, the interpretation of statistical data is crucial to making sound business decisions. With statistical modeling, regression models combine data to make essential variables, formulate and then analyze a model appropriate for the situation. (Casson & Farmer, 2014) With multiple regression and analysis, we can predict reimbursement amounts for the following year. For this task, the dependent variable is the amount of money necessary for the following year. The four independent variables are hospital cost in dollars, patient age, count of patient risk factors, and patient satisfaction scores.

Statistical Significance

A p-value (above .05) justifies that the null hypothesis does not significantly impact overall costs. In our study, the p-value is .000 for the age category and 0.02 for patient risk. Both of these numbers indicate that these variables are significant statistically. Age and risk then play a factor in the cost of care every time. The category of satisfaction has a p-value of 0.149; this means that the patient's satisfaction does not significantly impact the cost of care. Correlation is a relation that exists between mathematical or statistical variables which tend to vary, be associated, or occur together. Interpretation of correlation coefficients differs among scientific research areas and there no absolute rules for the interpretation of their strength. (Akoglu, 2018).

The value or R-squared explains how close the date is to the regression line or equation for the data. A higher R-squared value reveals an increase in the variability of the dependent variable (costs, for our purposes). The combination n independent variables make independent variables. When looking at the data, we see that the Rsquared=0.11. The hospital will need to make a decision on reimbursement expectations using a variance of 11%. The recognition that the analysis of the variables is significant statistically, except for patient satisfaction. We can begin to determine what patients will pay for a hospital stay using age and risk factors by plugging the numbers into a linear regression equation.

Calculating reimbursement costs based on each patient from the previous year, we will use the following equation:

Equation: y = 6652.176 + 107.036*(age) + 153.557*(risk) -9.1958(satisfaction)

The dataset attached can be used to reference the example below:

Example 1: Row 10 Predicted Reimbursement Cost y = 6,652.176 + 107.036*(82) + 153.557*(9) - 9.1958(66) y = 6,652.176 + 8,776.952 + 1,382.013 -606.922 y = 16,811.141 - 606.922 y = 16,204.219

Example 2: Row 30 Predicted Reimbursement Cost y = 6,652.176 + 107.036*(70) + 153.557*(9) - 9.1958(71) 5 y = 6,652.176 + 7,492.52 + 1,382.013 - 652.9018 y = 15,526.709 - 652.9018 y = 14,873.807

Example 3: Row 132 Predicted Reimbursement Cost y = 6,652.176 + 107.036*(69) + 153.557*(9) - 9.1958(25) y = 6,652.176 + 7,385.484 + 1,382.013 - 229.895 y = 15,419.673 - 229.895 y = 15,189.778

By averaging the three examples, we can determine the predicted cost of care for patients over 65 is $ 15,422.60. The average from the previous year came in at $14.906.51. The numbers for the upcoming year are higher, and therefore, the facility may see changes in the reimbursement costs. The administration will need to have its team closely monitor increases and decreases and implement plans to reduce risks to the budget. Leadership directly working in the facility can brainstorm for ways to increase overall revenue and decrease operation costs. In contrast, THOUGH not a massive factor in the project, strategies for rising satisfaction rates should also be a concern for myriad other reasons. Conclusion In this exercise and most others concerning reimbursement, perhaps it is wiser not to include satisfaction rates as they have little bearing on the outcome. Making future decisions to keep an organization viable and in line with long-term goals may be necessary. All healthcare facilities face regulatory adjustments to reimbursement rates that may go in either direction. Reimbursement is ever-changing and requires constant monitoring.

SUMMARY OUTPUT
Regression Statistics
Multiple R	0.33626289
R Square	0.11307273
Adjusted R Square	0.09837228
Standard Error	2482.42864
Observations	185
ANOVA
	df	SS	MS	F	Significance F
Regression	3	142200787.6	47400262.5	7.6917861	7.2555E-05
Residual	181	1115403803	6162451.95
Total	184	1257604590
	Coefficients	Standard Error	t Stat	P-value	Lower 95%	Upper 95%	Lower 95.0%	Upper 95.0%
Intercept	6652.17624	2096.817927	3.17251019	0.0017757	2514.82518	10789.5273	2514.825184	10789.5273
age	107.035898	28.91090095	3.70226783	0.0002835	49.9901507	164.0816463	49.99015068	164.0816463
risk	153.55707	66.68461014	2.30273626	0.0224312	21.9778617	285.1362779	21.97786172	285.1362779
satisfaction	-9.1946899	6.358071506	-1.4461445	0.1498661	-21.740163	3.350783708	-21.74016342	3.350783708