Download data file: https://drive.google.com/open?id=0B6i_JzkQ2f6IYWxZSTdyRjRLWE

Question

Download data file:

https://drive.google.com/open?id=0B6i_JzkQ2f6IYWxZSTdyRjRLWE0

The systolic and diastolic blood pressure (mm Hg) was recorded for 30 patients with moderate essential hypertension (Table below) [1]. Half of the group was measured immediately after taking the drug captopril (Drug=2) while the other half received a placebo (Drugs = 1). The dataset b1d.dat is Using R software: (a) Fit a single regression model to predict Diastolic from Systolic BP, with separate slopes and intercepts for each Drug. (b) Determine if the slopes are different for each Drug. (c) Determine if the intercepts are different for each Drug.

Explanation / Answer

(a) Fitting OLS regression models (using excel data analysis):

Regression model for Drug-2:

For drug -1:

(b) Yes, slopes look different for both the drugs

however, we need to test it statistcially:

H0: No difference

Ha: yes difference

we use t-test

t-stat=PE/SE

PE= slope difference

SE=SQRT((SE1^2+SE2^2)/2)*SQRT(2/n)

(c) Intercept:

SUMMARY OUTPUT Regression Statistics Multiple R 0.664773 R Square 0.441923 Adjusted R Square 0.398994 Standard Error 8.118514 Observations 15 ANOVA df SS MS F Significance F Regression 1 678.4998 678.4998 10.2943 0.006855 Residual 13 856.8335 65.91027 Total 14 1535.333 Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Intercept 52.43849 18.78504 2.791503 0.015276 11.85589 93.02109 Systolic 0.338516 0.105507 3.208472 0.006855 0.110582 0.566451

Related Questions

Navigate

• Browse (All)
• Browse D
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.