POLI 210 Lecture Notes - Lecture 5: Gateway Drug Theory, Spurious Relationship, Causal Inference

• School: McGill University
• Department: Political Science
• Course: POLI 210
• Professor: Dietlind Stolle

Document Summary

Empirical methods: experimental approach, field, survey, natural, lab, etc, non-experimental (observational approach, case study, medium-n, large-n statistical. Selection effects are a common cause of spurious relationships. The higher a (cid:272)ou(cid:374)try"s litera(cid:272)y rate, the (cid:373)ore de(cid:373)o(cid:272)rati(cid:272) it will (cid:271)e. Freedom of the press would be an indicator of how democratic a country is, not a cause of how literate the population is or how democratic the country is. There is no plausible source of spuriousness since nothing that could be measured in a survey would (cid:862)(cid:272)ause(cid:863) a perso(cid:374)"s gender. The researcher observes values of the dependent and independent variables for the samples subjects (cases), without anything being done to the subjects (such as imposing a treatment). A resear(cid:272)her (cid:272)o(cid:374)du(cid:272)ts a (cid:862)true(cid:863) e(cid:454)peri(cid:373)e(cid:374)t (cid:271)(cid:455) assig(cid:374)i(cid:374)g su(cid:271)je(cid:272)ts randomly to certain experimental conditions and then observing outcomes (dv). The experimental conditions are called treatments and represent the iv of the study.