ECON 2500 Lecture Notes - Lecture 11: List Of Statistical Packages, Statistical Inference, Sampling Distribution

• School: York University
• Department: Economics
• Course: ECON 2500
• Professor: Andrei Semenov

Document Summary

Lecture 11: the (cid:272)o(cid:374)ti(cid:374)ui(cid:374)g (cid:396)e(cid:448)olutio(cid:374) i(cid:374) (cid:272)o(cid:373)puti(cid:374)g is ha(cid:448)i(cid:374)g a d(cid:396)a(cid:373)ati(cid:272) i(cid:374) ue(cid:374)(cid:272)e o(cid:374) statisti(cid:272)s. the exploratory analysis of data is becoming easier as more graphs and calculations are automated. Normal distributions for data: while no data are exactly normal, the t procedures are useful in practice because they are (cid:396)o(cid:271)ust. No(cid:374)etheless, (cid:449)e (cid:272)a(cid:374)(cid:374)ot use t (cid:272)o(cid:374) de(cid:374)(cid:272)e i(cid:374)te(cid:396)(cid:448)als a(cid:374)d tests if the (cid:396)o(cid:271)ust(cid:374)ess, page (cid:1008)(cid:1007)(cid:1006) F test for equality of spread, page 474 are strongly skewed, unless our samples are quite large. The fundamental reasoning is still based o(cid:374) aski(cid:374)g, (cid:862)what (cid:449)ould happe(cid:374) if (cid:449)e applied this (cid:373)ethod (cid:373)a(cid:374)(cid:455) ti(cid:373)es? (cid:863) a(cid:374)s(cid:449)e(cid:396)s to this (cid:395)uestio(cid:374) a(cid:396)e still gi(cid:448)e(cid:374) (cid:271)(cid:455) (cid:272)o(cid:374) de(cid:374)(cid:272)e levels and p-values based on the sampling distributions of statistics. But their effectiveness and range of use are so great that they are rapidly becoming the preferred way to do statistical inference: this is already true in high-stakes situations such as legal cases and clinical trials.