Correlation R/Rstudio Example

Data: WOW_data

研究者想了解三個時間點的閱讀成績(reading_3、reading_2、reading_1)和三個時間點的學生自評師生衝突(CO_S_3、CO_S_2、CO_S_1)之間的關係

用R讀取剛剛的CSV檔，並將此資料命名為 corre_raw

corre_raw <- read.csv("D:/104/ML_R/WOW_data.csv",header=TRUE,sep=",")

將要分析的變項存成corre，並用head()看一下是否正確

corre <- corre_raw[c(12,17,22,15,21,26)]
head(corre)

##   reading_3 reading_2 reading_1   CO_S_3   CO_S_2   CO_S_1
## 1       492       478       472 1.000000 1.000000 1.000000
## 2       494       486       478 1.000000 1.000000 1.500000
## 3       523       505       500 1.666667 1.333333 1.833333
## 4       476       457       431 2.333333 1.833333 1.500000
## 5       515       502       500 1.833333 1.166667 1.000000
## 6       450       444       429 1.500000 2.166667 1.666667

載入套件並進行相關分析type=pearson、spearman (以皮爾森積差相關為例)

library(sandwich)
library(RcmdrMisc)
library(corrplot)

pearson_r <-rcorr.adjust(corre,type="pearson")
pearson_r

## 
##  Pearson correlations:
##           reading_3 reading_2 reading_1  CO_S_3  CO_S_2  CO_S_1
## reading_3    1.0000    0.9024    0.8709 -0.1574 -0.0853 -0.1589
## reading_2    0.9024    1.0000    0.8867 -0.1142 -0.1078 -0.1687
## reading_1    0.8709    0.8867    1.0000 -0.1139 -0.0763 -0.1403
## CO_S_3      -0.1574   -0.1142   -0.1139  1.0000  0.2931  0.3011
## CO_S_2      -0.0853   -0.1078   -0.0763  0.2931  1.0000  0.4577
## CO_S_1      -0.1589   -0.1687   -0.1403  0.3011  0.4577  1.0000
## 
##  Number of observations: 193 
## 
##  Pairwise two-sided p-values:
##           reading_3 reading_2 reading_1 CO_S_3 CO_S_2 CO_S_1
## reading_3           <.0001    <.0001    0.0288 0.2385 0.0273
## reading_2 <.0001              <.0001    0.1137 0.1356 0.0190
## reading_1 <.0001    <.0001              0.1149 0.2915 0.0516
## CO_S_3    0.0288    0.1137    0.1149           <.0001 <.0001
## CO_S_2    0.2385    0.1356    0.2915    <.0001        <.0001
## CO_S_1    0.0273    0.0190    0.0516    <.0001 <.0001       
## 
##  Adjusted p-values (Holm's method)
##           reading_3 reading_2 reading_1 CO_S_3 CO_S_2 CO_S_1
## reading_3           <.0001    <.0001    0.2185 0.5683 0.2185
## reading_2 <.0001              <.0001    0.5683 0.5683 0.1710
## reading_1 <.0001    <.0001              0.5683 0.5683 0.3094
## CO_S_3    0.2185    0.5683    0.5683           0.0004 0.0002
## CO_S_2    0.5683    0.5683    0.5683    0.0004        <.0001
## CO_S_1    0.2185    0.1710    0.3094    0.0002 <.0001

進行繪圖

corrplot(pearson_r$R$r ,method = "circle")

corrplot(pearson_r$R$r ,method = "ellipse")

corrplot.mixed(pearson_r$R$r ,lower = "number" , 
upper="ellipse")

date: “2016年1月23日,第一版”

author: “邱浩恩”