How do I do a t-test only on data that fulfills a certain criterion?

Question

I'm looking at whether the intake rate of a monkey eating leaves changes from the first quarter of time spent in a tree to the last quarter.

I've analysed this generally, but now I want to stratify by food type, to see if the food type leads to a significant change.

Here is a sample of my data:

Intake.Q1	Intake.Q4	FOOD.TYPE
6	6	M
19.5	5	M
5.1	2.5	M
4.8	2.1	Y
10.6	5.4	Y
6	2.5	Y

This is the code I used to generally do a t-test on all data, regardless of food type. t.test(R_DATA$Intake.Q1,R_DATA$Intake.Q4)

What would I add to this code in order to stratify based on 'M' and 'Y', so I can see whether there was a significant change in intake rate between Q1 and Q4 when eating Mature leaves, and the same for Young leaves?

Answer 1

The t.test is for comparing only 2 conditions A vs B. Your problem is A vs B vs C.

So one option is to perform 2 separate t.test and compare Q1 to Q4 for both Food types.

The other option is to perform a ANOVA on the data. An for this it is easier to rearrange the data.

R_DATA <- structure(list(Intake.Q1 = c(6, 19.5, 5.1, 4.8, 10.6, 6), 
                         Intake.Q2 = c(6, 5, 2.5, 2.1, 5.4, 2.5), 
                         Food.Type = c("M", "M", "M", "Y", "Y", "Y")), 
                    class = "data.frame", row.names = c(NA, -6L))

library(tidyr)
df_long <-pivot_longer(R_DATA, cols= -Food.Type,  names_to= "Intake")

results <-aov(value ~ Intake * Food.Type, data=df_long)

summary(results)

             Df Sum Sq Mean Sq F value Pr(>F)
Intake            1  67.69   67.69   3.345  0.105
Food.Type         1  13.44   13.44   0.664  0.439
Intake:Food.Type  1   2.71    2.71   0.134  0.724
Residuals         8 161.87   20.23