tag:blogger.com,1999:blog-6870909117486903221.post7904983679870955323..comments2023-04-30T07:03:04.307-07:00Comments on RCID 803: Week 12: Experiment and Quant. DescriptionJ. Ridouthttp://www.blogger.com/profile/14942536451496395729noreply@blogger.comBlogger1125tag:blogger.com,1999:blog-6870909117486903221.post-4855760484744255062009-03-30T13:02:00.000-07:002009-03-30T13:02:00.000-07:00As I reviewed my notes, the reading, and also the ...As I reviewed my notes, the reading, and also the assorted posts...I hit a snag. Simply put, is a quasi-or full experimental design more powerful? As best I can gather, the answer is "it depends." If incorrect, I'd like your thoughts. But if we can control the variable at issue through randomization, we can more effectively use an experimental design. Not all variables are so kind, however. Your comments regarding correlation versus causation reminded me of this point...if properly designed, neither can "prove" causation, merely strongly suggest it. But does one come closer than the other?Glen Southergillhttps://www.blogger.com/profile/02598743614464187339noreply@blogger.com