1 Data management

Load data

dta2 <- read.table("C:/Users/ASUS/Desktop/data/physical_activity.txt", h=T)
head(dta2)
##   bmi count
## 1  22     0
## 2  31     0
## 3  29     0
## 4  33     0
## 5  18     0
## 6  25     0
str(dta2)
## 'data.frame':    357 obs. of  2 variables:
##  $ bmi  : int  22 31 29 33 18 25 28 30 29 16 ...
##  $ count: int  0 0 0 0 0 0 0 0 0 0 ...

2 Analysis & output

  • I am interested to know whether the subject’s BMI can predict their
    engagement in physical activity (whether the heavier the sujects (higher
    BMI)are, the less likely they will engage in physical activity) ?
  • Poisson regression is used to predict a dependent variable that consists
    of “count data”
  • 結果顯示受試者的BMI越高，確實與越減少激烈運動次數有顯著關連。
names(dta2) <- c("BMI","Physical")
m1<-glm(Physical~BMI, family = poisson, data=dta2)
summary(m1)
## 
## Call:
## glm(formula = Physical ~ BMI, family = poisson, data = dta2)
## 
## Deviance Residuals: 
##     Min       1Q   Median       3Q      Max  
## -1.8404  -1.1270  -0.9570  -0.8127   4.8654  
## 
## Coefficients:
##             Estimate Std. Error z value Pr(>|z|)    
## (Intercept)  2.27048    0.43383   5.234 1.66e-07 ***
## BMI         -0.10898    0.01729  -6.302 2.94e-10 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## (Dispersion parameter for poisson family taken to be 1)
## 
##     Null deviance: 795.57  on 356  degrees of freedom
## Residual deviance: 756.44  on 355  degrees of freedom
## AIC: 946.99
## 
## Number of Fisher Scoring iterations: 6