reliable is such a subjective word. but generally, wikipedia is accurate. it's not perfect or fool proof, but it's accurate enough if you are a researcher. if you need cold hard stats, you have to rely on publications and journals gyud. for example, if you are conducting a social psychology experiment, you need to consult social psychology journals (i used to read a lot of them sa san carlos library) and use them as reference. of course, i apil gyud nimo sila sa biblio nimo.