source("readraw_func.R")
##install.packages("httr")
library(httr)
# 先將收集的受試者檔案上傳到Rawdata資料夾
# OSFID改成個別raw file的網址,根據實際raw file增加
rawurls <- c(
"https://osf.io/tkp63/?action=download",
"https://osf.io/q4bge/?action=download",
"https://osf.io/5vqrd/?action=download",
"https://osf.io/qnx7y/?action=download",
"https://osf.io/5ghfn/?action=download",
"https://osf.io/c2df3/?action=download",
"https://osf.io/uvh45/?action=download",
"https://osf.io/8m3e9/?action=download",
"https://osf.io/x2gfy/?action=download",
"https://osf.io/a782s/?action=download",
"https://osf.io/d279h/?action=download",
"https://osf.io/v98bq/?action=download",
"https://osf.io/ryn2w/?action=download",
"https://osf.io/4utzs/?action=download",
"https://osf.io/j4z29/?action=download",
"https://osf.io/9hb2g/?action=download",
"https://osf.io/hdeqn/?action=download",
"https://osf.io/e37rk/?action=download",
"https://osf.io/fc7qh/?action=download",
"https://osf.io/re2m9/?action=download"
)
# 下載後的檔案名稱
filenames <- paste0("gazeraw", paste0(sprintf( "%02d", 1:length(rawurls)), ".csv"))
#filenames <- paste0("subject-", paste0(1:5, ".csv"))
# 執行完畢檢查是否已下載到本機資料夾
for(i in 1:length(rawurls)){
GET( rawurls[i], write_disk(paste0("RAW/", filenames[i]), overwrite = TRUE) )
}
# 讀入原始資料到data frame
DF <- data.frame() # 空白data frame
for(i in 1:length(filenames)){
DF <- readsubj(paste0("RAW/",filenames[i]), DF, 16)
} # 根據實際實驗設定,更改第三個argument
colnames(DF) <- c("RT", "ACC", "ID", "SOA", "Gaze", "Target_Congruency", "Response_Congruency")
#檢查data frame是否符合codebook
summary(DF)
## RT ACC ID SOA Gaze
## Min. :347.5 Min. :0.4375 gazeraw01: 16 100 :160 L:160
## 1st Qu.:480.3 1st Qu.:0.9375 gazeraw02: 16 1000:160 R:160
## Median :566.0 Median :1.0000 gazeraw03: 16
## Mean :569.3 Mean :0.9398 gazeraw04: 16
## 3rd Qu.:630.1 3rd Qu.:1.0000 gazeraw05: 16
## Max. :959.6 Max. :1.0000 gazeraw06: 16
## (Other) :224
## Target_Congruency Response_Congruency
## TC:160 RC:160
## TI:160 RI:160
##
##
##
##
##
print(read.csv("codebook_Tidy.csv"))
## Varibles Values
## 1 RT integers
## 2 ACC proportions
## 3 ID Char
## 4 SOA Char
## 5 Gaze Char
## 6 Target_Congruency Char
## 7 Response_Congruency Char
## Descrptions
## 1 Average correct RT
## 2 Propotions of accuracy
## 3 Subject ID
## 4 Levels of SOA: 100; 1000
## 5 Levels of gaze directions: L(Left); R(Right)
## 6 Levels of Target Congruency: TC(Congruent);TI(Incongruent)
## 7 Levels of Response Congruency: RC(Congruent);RI(Incongruent)
#回顧文獻的統計結果
x <- factor(rep(c("cued", "neutral", "uncued"),4))
SOA <- factor(rep(c("105 ms","300 ms", "600 ms", "1,005 ms"),each=3))
resp <- c(504, 509, 508, 481, 500, 481, 478, 488, 488, 470, 482, 483)
interaction.plot(x, SOA, resp, type = "b", xlab = "Cue Validity", ylab = "RT (ms)", main = "Identification", ylim=c(460, 520), lty = 1, pch = 1:4)
#總體描述統計
Two.Mean <- with(data = DF, tapply(RT, paste(SOA, Target_Congruency), mean) )
Two.Resp.Mean <- with(data = DF, tapply(RT, paste(SOA, Response_Congruency), mean) )
Four.Mean <- with(data = DF, tapply(RT, paste(SOA, paste(Gaze, paste(Target_Congruency, Response_Congruency))), mean) )
#二因子ANOVA
summary( aov(RT ~ SOA*Target_Congruency + Error( ID/(SOA*Target_Congruency) ), data = DF) )
##
## Error: ID
## Df Sum Sq Mean Sq F value Pr(>F)
## Residuals 19 3325800 175042
##
## Error: ID:SOA
## Df Sum Sq Mean Sq F value Pr(>F)
## SOA 1 61700 61700 16.91 0.000593 ***
## Residuals 19 69332 3649
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Error: ID:Target_Congruency
## Df Sum Sq Mean Sq F value Pr(>F)
## Target_Congruency 1 15784 15784 10.09 0.00497 **
## Residuals 19 29731 1565
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Error: ID:SOA:Target_Congruency
## Df Sum Sq Mean Sq F value Pr(>F)
## SOA:Target_Congruency 1 8250 8250 11.49 0.00307 **
## Residuals 19 13640 718
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Error: Within
## Df Sum Sq Mean Sq F value Pr(>F)
## Residuals 240 658586 2744
#另一種二因子ANOVA
summary( aov(RT ~ SOA*Response_Congruency + Error( ID/(SOA*Response_Congruency) ), data = DF) )
##
## Error: ID
## Df Sum Sq Mean Sq F value Pr(>F)
## Residuals 19 3325800 175042
##
## Error: ID:SOA
## Df Sum Sq Mean Sq F value Pr(>F)
## SOA 1 61700 61700 16.91 0.000593 ***
## Residuals 19 69332 3649
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Error: ID:Response_Congruency
## Df Sum Sq Mean Sq F value Pr(>F)
## Response_Congruency 1 348 347.8 0.384 0.543
## Residuals 19 17209 905.7
##
## Error: ID:SOA:Response_Congruency
## Df Sum Sq Mean Sq F value Pr(>F)
## SOA:Response_Congruency 1 565 565.1 0.578 0.457
## Residuals 19 18591 978.5
##
## Error: Within
## Df Sum Sq Mean Sq F value Pr(>F)
## Residuals 240 689278 2872
#三因子ANOVA?
#summary( aov(RT ~ + Error( ID/() ), data = DF) )
summary( aov(RT ~SOA*Response_Congruency*Target_Congruency + Error( ID/(SOA*Response_Congruency*Target_Congruency) ), data = DF) )
##
## Error: ID
## Df Sum Sq Mean Sq F value Pr(>F)
## Residuals 19 3325800 175042
##
## Error: ID:SOA
## Df Sum Sq Mean Sq F value Pr(>F)
## SOA 1 61700 61700 16.91 0.000593 ***
## Residuals 19 69332 3649
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Error: ID:Response_Congruency
## Df Sum Sq Mean Sq F value Pr(>F)
## Response_Congruency 1 348 347.8 0.384 0.543
## Residuals 19 17209 905.7
##
## Error: ID:Target_Congruency
## Df Sum Sq Mean Sq F value Pr(>F)
## Target_Congruency 1 15784 15784 10.09 0.00497 **
## Residuals 19 29731 1565
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Error: ID:SOA:Response_Congruency
## Df Sum Sq Mean Sq F value Pr(>F)
## SOA:Response_Congruency 1 565 565.1 0.578 0.457
## Residuals 19 18591 978.5
##
## Error: ID:SOA:Target_Congruency
## Df Sum Sq Mean Sq F value Pr(>F)
## SOA:Target_Congruency 1 8250 8250 11.49 0.00307 **
## Residuals 19 13640 718
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Error: ID:Response_Congruency:Target_Congruency
## Df Sum Sq Mean Sq F value Pr(>F)
## Response_Congruency:Target_Congruency 1 7297 7297 1.5 0.236
## Residuals 19 92427 4865
##
## Error: ID:SOA:Response_Congruency:Target_Congruency
## Df Sum Sq Mean Sq F value Pr(>F)
## SOA:Response_Congruency:Target_Congruency 1 646 646.4 0.294 0.594
## Residuals 19 41836 2201.9
##
## Error: Within
## Df Sum Sq Mean Sq F value Pr(>F)
## Residuals 160 479667 2998
#成對t檢定
ID.100.Means <- matrix( with(data = subset(DF, SOA == 100), tapply(RT, paste(ID, paste(SOA, Target_Congruency) ), mean) ), ncol = 2, byrow =TRUE)
t.test(ID.100.Means[,1], ID.100.Means[,2], paired = TRUE, var.equal = TRUE) # SOA = 100ms
##
## Paired t-test
##
## data: ID.100.Means[, 1] and ID.100.Means[, 2]
## t = -0.6836, df = 19, p-value = 0.5025
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -15.805069 8.022721
## sample estimates:
## mean of the differences
## -3.891174
ID.1000.Means <- matrix( with(data = subset(DF, SOA == 1000), tapply(RT, paste(ID, paste(SOA, Target_Congruency) ), mean) ), ncol = 2, byrow =TRUE)
t.test(ID.1000.Means[,1], ID.1000.Means[,2], paired = TRUE, var.equal = TRUE, data = subset(DF, SOA == 1000) ) # SOA = 1,000 ms
##
## Paired t-test
##
## data: ID.1000.Means[, 1] and ID.1000.Means[, 2]
## t = -4.8729, df = 19, p-value = 0.0001057
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -34.59641 -13.80623
## sample estimates:
## mean of the differences
## -24.20132