POLS 210 Lecture Notes - Lecture 5: Institute For Operations Research And The Management Sciences, Gun Control
Document Summary
Independent variables are always in the columns, the dependent variables are always in the rows. Interpret by comparing across columns along dv value. Use li(cid:374)go like (cid:862)the results support the hypothesis(cid:863) do(cid:374)"t say true or pro(cid:448)e. Obviously, they are not the same number, but is the number different enough (sometimes there is error). Interested in mean values of dv for different iv values. We"re (cid:374)e(cid:448)er goi(cid:374)g to get the sa(cid:373)e (cid:374)u(cid:373)(cid:271)ers so (cid:449)e (cid:449)a(cid:374)t to talk a(cid:271)out the size of these differences. We have to be aware of the underlying scale of measurement. Ca(cid:374)"t just look at the (cid:374)u(cid:373)(cid:271)ers (cid:271)e(cid:272)ause (cid:449)e ha(cid:448)e to understand the scale. We do(cid:374)"t see the (cid:272)ausality i(cid:374) the ta(cid:271)les (cid:271)e(cid:272)ause (cid:449)e (cid:272)a(cid:374)"t see (cid:272)ausality. Causality is i(cid:374) your theory, it"s (cid:374)ot i(cid:374) reality for you to o(cid:271)ser(cid:448)e. The problem with this is we can find results that fit our hypothesis perfectly but are explained by something else entirely.