The solution seems quite simple, but I can't find the code that works. I just try to run the following SAS code in RStudio :
proc ttest data=jump_data;
var JFM_BJH;
class clc_sex;
where ageg=2;
run;
ageg = age groups and this categorical variable has 5 levels.
Thank you very much in advance
This link maybe useful
https://www.programmingr.com/statistics/t-test/
# t test in R example (Hypothesis testing in R)
Shipment
Simpler without tidyverse:
t.test(JFM_BJH ~ clc_sex, data = jump_data, subset = ageg==2, var.equal = TRUE)
If there are several age groups it might be better to explicitely include ageg in the model, like
model
At the risk of starting a war, let me say that compared to the SAS and R code that has been posted above, I find the corresponding Stata code refreshingly simple and very readable. YMMV. :-)
ttest JFM_BJH if ageg==2, by(clc_sex)
To get the t-test for each level of ageg separately, you would do this:
by ageg, sort : ttest JFM_BJH, by(clc_sex)
But as Jochen Wilhelm noted, it very likely makes far more sense to estimate a model that includes sex, age group and the interaction between them. Again, in Stata, that might look like this:
regress JFM_BJH clc_sex##ageg
Apologies for introducing Stata code into a thread where the OP asked for help with how to do something using R. I know how irritating it can be when someone asks, "How do I do X using Package A?", and the response is, "Use Package B". And I am not suggesting that Joel Blanchard should use Stata rather than R. As noted above, I was just struck by how much simpler its code is for these relatively simple tasks. To repeat what I said above, YMMV!
Donning my flame-retardant suit now. ;-)
Bruce Weaver ,
vive la diversité :)
R is made for sophisticated analyses, that's why such simple tasks look unneccesarily complicated ;)
t test in R:
t.test(x, y, alternative = "two.sided", var.equal = FALSE)
for example:
> x = rnorm(20, 0,1) # sample from standard normal distribution
> y = rnorm(20,0,1) # sample from standard normal distribution
> t.test(x, y, alternative = "two.sided", var.equal = FALSE)
Welch Two Sample t-test
data: x and y
t = -0.70095, df = 37.241, p-value = 0.4877
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
-0.8470274 0.4115379
sample estimates:
mean of x mean of y
-0.2168400345 0.0009047068
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA