POL SCI 3 Study Guide - Midterm Guide: Interrupted Time Series, Cut-Point, Blind Experiment
Document Summary
Pilot your induced state treatment and provide evidence for a manipulation check, gather evidence on the state you are manipulating, on how treatment and control differ on the state you plan to induce and possible confounding states. Subjects (6) (1) demand effects (a) subjects figure out hypothesis and alter behavior, either help or harm researcher by providing supportive/ unsupportive evidence (b) sol: deception, controversial and requires debriefing (i) mild deception - mask purpose (ii) Lying - problematic because false info prevent adjustment to the correct belief (iii) deception where subjects pay huge cost to part of the experiment - time and effort (c) sol: make study blind! Observational studies (1) non qe - sgc, ogpp (2) qe. Quasi experimental design: improve iv over non qe by researcher exerting control on (1) when we gather data on y and/or (2) on whom data is gathered and/or (3) when x occurs. Iv scale: its > ics/ mits > mics > ets.