You are on page 1of 5

Individual Project: Checkpoint 5

Introduction:
The hospital data set includes information about hospital costs, Mortality rating, and facility types. The
data is part of an Administration-wide effort to increase the availability and accessibility of information on
quality, utilization, and costs for effective and informed decision-making. Use the data set that you
generated for Checkpoint 1 to complete the rest of the checkpoints.

o Facility Types - Kind of organization operating the hospital: one of Government, Private,
Proprietary, Church, or Unknown
o Mortality Rating - Overall rating between 1 and 3 stars, with 3 stars being the highest rating; 1
represents the worst rating.
o Procedure Heart Attack Cost - Average cost of care for heart attacks

Upload the following documents:


1) This document with your answers.

Research Question
Primary question:
Is the rating of the hospital mortality affect the cost of the heart attack procedure?
Secondary question:
Does the hospital type a possible confounding variable for the cost of a heart attack procedure?

Conclusions and Limitations


Based on the information from the other checkpoints, answer the questions below:

1) Primary question: Is the rating of the hospital mortality affect the cost of the heart attack procedure?

a) Is there an association between hospital mortality rating and the procedure of Heart Attack
Procedure cost?
(Hint: Fitted Line Plot of the data, with 𝑥 as the explanatory variable and 𝑦 as the response
variable from checkpoint 3).
Attached Chart/graph, and answer the question.

- No, if we look at the graph at first sight we can tell that there isn’t a visible difference among
the hospital mortality rating and the heart attack procedure cost.
b) How strong is the relation between these factors?
(Hint: what is the coefficient of correlation and is it a strong or weak correlation, reference to ‘part a’)
- In order to find how strong the correlation between these two is, we would take the square root of
“R-Sq” which in this case is 0.6 and gives us an R value of 0.77. Because of what we can see visually,
and the R value we can tell that the relation among the heart attack procedure cost and the
hospital mortality rating is very weak.

c) Is there evidence you found in part a to support your claim (hint: From your previous hypothesis
testing, what is the name of the test static, interpret the p-value and decision? State the
conclusion in the context of the problem)?
Attached Chart/graph and Minitab output, and answer the question.
Analysis of Variance
Adj
Source DF Adj SS MS F-Value P-Value
Hospital Mortality 2 6179118 3089559 1.16 0.314
Rating
Error 297 790129625 2660369

Total 299 796308743

Means
Hospital
Mortality
Rating N Mean StDev 95% CI
1 100 23866 1729 (23545, 24187)
2 100 23866 1577 (23545, 24187)
3 100 24171 1583 (23850, 24492)
Pooled StDev = 1631.06

- Above is the results for an ANOVA test that was conducted to either support or deny the claim
made in the above question. From the ANOVA test we received a P-value of 0.314. If a P-value
is lower than 0.05, it means that the results we received were statistically significant and we
can suggest that the Heart attack procedure cost and the hospital mortality rating have a
correlation between one and the other. In this case, because our P-value was greater than
0.05, we do not reject the null hypothesis as there is NOT significally statistical data that
proves that there is a relationship between both.

d) Given your conclusion in part c, are you at risk of making a Type I error or a Type II error? Provide
your answer in the context of the problem.
- I am at risk of commiting a Type II error because the null hypothesis was rejected. A Type II
error means that we are at risk of a false negative, in the context of this problem we are at risk
that there IS a correlation between the Heart Attack Procedures and the Hospital Mortality
Rate.
-
e) What is the conclusion of your study for the primary question?
Is the rating of the hospital mortality affect the cost of the heart attack procedure?
(Hint: Summarize your findings from the analysis above)
- No, the rating of the hospital mortality does NOT affect the cost of the heart attack procedure.

2) Secondary question: Does the hospital type a possible lurking variable for the cost of a heart attack
procedure?

a) Is there an association between the facility types and the procedure of Heart Attack Procedure
cost?
(Hint: Relationship plot between the facility types and the procedure cost of heart attacks
from checkpoint 3).
Attached Chart/graph, and answer the question.

- Based on the scatter plot, although there seems to be a slight pattern within the proprietary
type, I do NOT think that there will be statistically significant proof to show that one is
associated with the other.

b) Is there evidence you found in part a to support your claim (hint: From your previous
hypothesis testing, what is the name of the test static, interpret the p-value and decision?
State the conclusion in the context of the problem.
Attached Chart/graph and Minitab output, and answer the question.

1. Paste the Minitab output (ANOVA table & Boxplot) below.


2. Analysis of Variance
Source DF Adj SS Adj MS F-Value P-Value

Facility Type 3 8828591 2942864 1.11 0.347


Error 296 787480152 2660406

Total 299 796308743

Means
Facility Type N Mean StDev 95% CI

Church 27 23986 1809 (23368, 24603)

Government 55 23805 1808 (23372, 24237)

Private 171 23912 1578 (23667, 24158)

Proprietary 47 24349 1494 (23881, 24817)

Pooled StDev = 1631.08

- In order to prove whether there was a correlation between the Facility type and the heart
attack procedure cost we performed an ANOVA test. We received a P-value of 0.347, which is
greater than 0.05. Because the probability that these results were due to random chance is
37%, we did not find statistically significant proof to reject the null hypothesis. Therefore, due
to the high chance of these results happening by chance, we fail to reject the null hypothesis.

c) Given your conclusion in part b, are you at risk of making a Type I error or a Type II error?
Provide your answer in the context of the problem.
- I am at risk of commiting a Type II error because the null hypothesis was rejected. A Type II
error means that we are at risk of a false negative, in the context of this problem we are at risk
that there IS a correlation between the Heart Attack Procedures and the Facility Type.

d) What is the conclusion of your study on the secondary question?


Does the hospital type a possible lurking variable for the cost of a heart attack procedure?

(Hint: Summarize your findings from the analysis above)


- No, the hospital type is not a confounding variable as we didn’t find statistically significant
evidence to prove that there is a correlation between both variables. The hospital type does
not affect the cost of a heart attack procedure.

3) Suggest additional two factors that could be possible lurking variables for the procedure of Heart
Attack Procedure cost for this study.
- I think one additional lurking variable can be demographics, the prices in different areas of not
only the United States but internationally may vary and there could be an effect on the cost
based on the location. Aside from that, the complexity of the procedure could be another one
as according to the study linked below it has a great influence on the time and resources that
the procedure takes, which is directly correlated with the cost.
- Reference Link: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9390661/

You might also like