avova one way vs multiway Hello I'm running anova with R and I am wondering what the differce is between multiway anova and single anova. I know single anova gives a different answer than multiway anova but I dont know how to quantify multiway anova.
Could someone explain in words the difference between:


*

*base~col1

*base~col1+col2

*base~col1*col2


http://www.gardenersown.co.uk/Education/Lectures/R/anova.htm#anova_1
How do I interpret the results of col1+col2 and col1*col2
because col2 is giving a P(>F) of  0.473 on single anova but P(>F) 4.21e-07 in multiway anova. col2 does correlate well with base having only a 2% coloration for 711 individuals. According to http://vassarstats.net/rsig.html 2% coloration isn't significant. 
Thank you
edit:
i know my col1 and col2 have some interactions is that causing the low multiple anova p value?
i am using 
aov.ex2 <- aov(base~col1+col2,data=data1) 
summary(aov.ex2)
aov.ex2 <- aov(base~col1,data=data1) 
summary(aov.ex2)
aov.ex2 <- aov(base~col1,data=data1) 
summary(aov.ex2)

and the results are not simular
 A: Simply, single ANOVA tests between-groups for a single independent variable (col1) for a response variable (base). Multiway ANOVA tests between-groups for two or more independent variables (col1 and col2). The choice to use one or the other depends on your experimental design. 
base~col1 is testing a single "main" effect
base~col1+col2 would be testing two "main" effects only
base~col1*col2 would be testing two "main" effects along with the "interaction" effect between independent variables.
Some more info about two-way ANOVA
Edit: Given your updated post:
I suspect you made a mistake in the code you gave as example, but here are some explanations:
aov.ex2 <- aov(base~col1+col2,data=data1) 
summary(aov.ex2)

This would test for only main effects of col1 and main effects of col2. In other words, are there any differences between groups for col1 or col2?
aov.ex2 <- aov(base~col1,data=data1) 
summary(aov.ex2)

This would test for only the main effect of col1. In other words, are there any differences between groups for col1?
It seems like your basic problem is that with the inclusion of a second independent variable you are getting different results. This post may be of relevance.
