Statistical Analysis on mostly boolean values So I have a large dataset, and I was wondering what the best way to conduct statistical analysis of it is. I'm very green in terms of statistical methods, but I learn quickly. Basically, each item has a couple attributes, and each attribute has several possibilities. Each item has their specific attribute set-up in terms of booleans for each possibility of each attribute (i.e. 1 means it has that possible version of that attribute, 0 means it doesn't). For some of the attributes, each item can only have one possible value, and for some it can have multiple. I'm using R through RStudio to conduct my analysis. Any help would be appreciated!
Basically, I have a bunch of error messages generated by different customers. Each error has different attributes, which include server downtime, country of origin, the type of change, which specific item(s) caused the problem, and a couple others. I'm mostly trying to look for which applications/types of changes prompt downtime.
 A: If you are trying to predict downtime from attributes of messages, it sounds like you want some form of regression.
If downtime is a binary variable (yes/no) then you probably want logistic regression.
If downtime is continuous (e.g. in minutes or seconds) you probably want "regular" (ordinary least squares) regression.
If downtime is in some other format, please tell us.
Both these forms of regression (and others) are available in the R function glm. 
A: In R, your attributes are called Factors [ categorical variables, eg male/female or age range[20-30,30-40,..]] ... recommend you look the  term up
its probably best to leave them as that ( rather than what it sounds like you are doing is "manually" splitting them into a "1-of-N" representation. 
Using R's factor representations then its easy to look at interactions - in r you would do something like 
y~sex+age.range +sex * age.range and it is tehn easy to get R to spit out what are relevant interactions [ eg maybe some behaviours are specific to 20 something males?]
