It's a freaking bad day. You've spent countless hours on the Internet trying to figure out how multilevel modeling works, but the only things you can find are academic papers filled with jargon, obscure equations, and indecipherable lines of code. 'Why can't I understand anything about stats?!' you ask yourself. Well, you've got to cool it now! Learning multilevel modeling can be a real bear, and this paper is precisely made for you to get the hang of it as easily as possible. If you're here,

more »

... u probably already know that the general aim of multilevel modeling is to simultaneously analyze data at a lower level (usually participants) and at a higher level (usually clusters of participants). In other words, multilevel modeling enables one to disentangle the effects of lower-level variables (e.g., individual effects) from the effects of higher-level variables (e.g., contextual effects) and examine how lower-level and higher-level variables interact with one another (interactions involving variables at different levels are called ' cross-level interactions'). Let us give you an example. In early 2000, a New Zealand team of scientists conducted research involving approximately 700 cats from 200 households (i.e., on average, 3.5 cats per household; Allan et al., 2000) . The team treated the cats (level-1 units) as nested in households (level-2 units) and used multilevel modeling to disentangle the effects of level-1 cat variables (e.g., does the cat have long legs?) from the effects of level-2 household variables (e.g., is there a dog in the household?) in predicting cat obesity. They found that short-legged cats living in dog-free households tend to be chubbier. 1 After reading the present paper, you will be able to handle this kind of (feline) two-level hierarchical design. Our paper is divided into four parts: