EHS 280 Linear regression analysis
Article Quiz1 worksheet. Quiz Due in Dropbox 9/13/2024 at 11:59PM
First Name: LastName UMB ID#
1. Save on your laptop computer LinearReg-HTN12.xlsx sent via email and also posted in
Notes1 folder in weekly folders on Blackboard..
2. Upload the excel spreadsheet to StatCrunch. Log into www.statcrunch.com, use
email/password to sign-in (top right of landing page). Select Upload data
3. Data will open in StatCrunch ready to analyze.
4. Fit the following Simple linear regression models/equations and one multiple linear
regression model, i.e., Stat->Regression->Simple/Multiple
5. Copy screenshot of StatCrunch and paste in boxes on the left. Enter requested numbers
from the output in boxes on the right.
Models to be fitted using StatCrunch
A. Model (Equation) 1: Y=DiastolicBP, X=age B. Model (Equation) 2: Y=DiastolicBP, X=smokers
C. Model (Equation) 3: Y=DiastolicBP, X=sex D. Model(Equation) 4: Y=DiatolicBP, X=BMI
E. Model (Equation) 5: Y=DiastolicBP, X=age +sex +BMI
A. Model 1: Y=DiastolicBP, X=age (Completed Example)
StatCrunch Output
equation DiastolicBP=88.41-0.12*Age
R correlation coefficient=-0.20
R2 R-squared=0.03998 or 4%
Intercept 88.41
Slope -0.12
P-value* for slope, p=0.5337
significant (p<0.05)? no
P-value* For model, p=0.5337
significant (p<0.05)? no
* If p-value for slope is less than 0.05, we
say the slope is statistically significant
(i.e., good) at the 0.05 level, otherwise if
p>0.05 the slope is NOT significant (i.e.,
not good).
* If p-value for model is less than 0.05, we
say the model is statistically significant at
the 0.05 level, otherwise if p>0.05 the
model is NOT significant.
1
2
Using above example as template, run the following regression models and complete the
table on the right panel.
Model(Equation) 2: Y=DiatolicBP, X=smokers
Paster StatCrunch Output here
equation DiastolicBP=
R correlation coefficient=
R2 R-squared=
Intercept
Slope
P-value* for slope, p=
significant (p<0.05)?
P-value* For model, p=
significant (p<0.05)?
* If p-value for slope is less than 0.05, we
say the slope is statistically significant at
the 0.05 level, otherwise if p>0.05 the
slope is NOT significant.
* If p-value for model is less than 0.05, we
say the model is statistically significant at
the 0.05 level, otherwise if p>0.05 the
model is NOT significant.
Model(Equation) 3: Y=DiatolicBP, X=sex
Paster StatCrunch Output here
equation DiastolicBP=
R correlation coefficient=
R2 R-squared=
Intercept (a)
Slope (b)
P-value* for slope, p=
significant (p<0.05)?
P-value* For model, p=
significant (p<0.05)?
* If p-value for slope is less than 0.05, we
say the slope is statistically significant
(i.e., good) at the 0.05 level, otherwise if
p>0.05 the slope is NOT significant (i.e.,
not good).
* If p-value for model is less than 0.05, we
say the model is statistically significant at
the 0.05 level, otherwise if p>0.05 the
model is NOT significant.
3
Model(Equation) 4: Y=DiatolicBP, X=BMI
Paster StatCrunch Output here
equation DiastolicBP=
R correlation coefficient=
R2 R-squared=
Intercept(a)
Slope (b)
P-value* for slope, p=
significant (p<0.05)?
P-value* For model, p=
significant (p<0.05)?
* If p-value for slope is less than 0.05, we
say the slope is statistically significant
(i.e., good) at the 0.05 level, otherwise if
p>0.05 the slope is NOT significant (i.e.,
not good).
* If p-value for model is less than 0.05, we
say the model is statistically significant at
the 0.05 level, otherwise if p>0.05 the
model is NOT significant.
Model (Equation) 5: Y=DiastolicBP, X=age +sex +BMI [Note, this is a multiple linear
regression model
Paste StatCrunch Output here
MLR equation DiastolicBP=
R2 R-squared=
Intercept(a)
Slope (age) p-val=
Slope (sex) p-val=
Slope (bmi) p-val=
P-value* For model, p=
Significant?
* If p-value for model is less than 0.05, we
say the model is statistically significant at
the 0.05 level, otherwise if p>0.05 the
model is NOT significant.
Do not hesitate to ask if you are not sure about anything for this exercise.
4
5
