1 2
y1 48 44
y2 38 39
y3 49 56
y4 3 4
y5 55 28
y6 99 101
y7 121 120
y8 2 6
1) Given this descriptive statistics where 1 and 2 are the outcome (Y = 1 or 2 ) and Y1 - Y8 are the variables, I want to perform independent t-test using unequal variance. Y4 and Y8 are binary variables, and I need to perform chi-square. I want the results from these tests as my third column to see which variable is a driving factor of the group distinction (Y = 1 or 2). How would I be able to do this in R?
2) If the outcome changes to three categories (Y = 1, 2, and 3), how can I perform ANOVA for continuous variables and chi-square for Y4 and Y8 in R?
First of all, you shouldn't mix the binary variables with the rest of the measurements. I will start by separating the input dataframe in two dataframes.
df2 <- df1[c(4, 8), ]
df3 <- df1[-c(4, 8), ]
Now the tests. The t.test will need the data in long format, see this question for other ways of reshaping the dataset.
chisq.test(df2)
long <- reshape2::melt(df3)
t.test(value ~ variable, long)
Data in dput format.
df1 <-
structure(list(`1` = c(48L, 38L, 49L, 3L, 55L,
99L, 121L, 2L), `2` = c(44L, 39L, 56L, 4L, 28L,
101L, 120L, 6L)), class = "data.frame",
row.names = c("y1", "y2", "y3", "y4", "y5",
"y6", "y7", "y8"))