Theory and method
2.d counterfactual approach 2 casual inference.–casual inference is drawing a conclusion about a causal connection based on conditions of d occurrence of an e.there4,d counterfactual approach considers ?would have happened if c hs been absent&comparison of ?did happen w ?would have happened had d cause been absent;it rel8s 2 casual inference in dat if d cause occurs,then so does d e& if d cause does nt occur,neither does d e (symmetric)&is tru in similar worlds,4 d counterfactual approach relies on cre8ng similar worlds;w causal direction estb by truth of st8ment.if note,then c still may occur (asymmetric) it is confused by f correl8ons;problem of causal direction,even if dat means adding 3rd condition.d counterfactual approach2 casual inference is logically prior2id of causal mechanism.a causal notion dat must b unders2od in terms of d however truth of certain counterfactuals.d casual process analysis can b an essential fac2r of st8stical research built on a basis of counterfactual causality.thus,counterfactual def of caus8on hs no e there4 relies on something nt tru.only logic of causal inference in social inquiry of casual inference as casual inference is d only worthwhile aspect of empirical social science,d norm in st8stics.2 caus8on emphasizes possible worlds without considering human agency.main virtue?–dat it only describes a necessary condition;d existence of causal connection btwn a & b allows us 2 st8 dat a causes b;explain singular casual events.wen d1 correctly,experiments in which r well-d1 can help researchers construct d closest possible worlds & explore these counterfactual conditions.main problem?-st8stical logic theory study,nt enough observ8ons r available 2 test counterfactual hypotheses & consequently causal effects cannot b estim8d;in other words,inferential logic.defining & finding accur8 evidence 4 d most similar world & would have happened 2 it;furthermore,obtaining d evidence of ?would have happened in dat most similar world.
which research design combines-various str8gies of causal inference often overlap in order 4 counterfactual specul8ons must rely on causal mechanisms along with causal mechanisms must b built on d bedrock (fundamental) of neo-human covering laws.experimental designs d best known st8stical theory of causality emphasis on these 2 approaches would b d problem of pre-empt of a man taking a trip across d desert,an enemy of him put a hole in his can full of w8r,meanwhile another enemy of d man crossing d desert put poison in w8r without knowing anything about enemy 1.either enemy knowing dey in-fact partaken in manipul8ons in which occurred 2 d man on his trip basing on d fact dat d man dies in d desert.while both enemies think dey r single-handedly d cause 4 d man dying;due 2 enemy 1 blocking enemy 2 by putting a hole in d can;dis instance enemy 2 would b in d wrong.ol in ol d initial assumption of d counterfactual approach dat d hole in d can being d cause of d man’s death is twisted with d actual truth of d counterfactual condition in dat d man would nt have died if enemy 1 did nt put d hole in d can would in-fact b false,despite d fact he did die,only of thirst.d notion of dis situ8on being another possibility of d man dying frm d poison makes d counterfactual false coz d man would have died frm d poison if enemy 1 did nt pre-empt d can by putting a hole in it.furthermore,d assumption dat d manipul8on fac2r (poison) must b d cause,it determined d direction of causality by ruling out d untrue correl8on (d cause of d man’s death).
manipul8on approach- casual inference is drawing a conclusion about a causal connection based on d conditions of d occurrence of an effect.there4,d manip approach considers which causality is established by a recipe of action dat regularly produces d effect frm d case.it underlies experimental approach solves problems of causal direction& spurious correl8on.direct,invariant,change-rel8ng generaliz8ons.wen a researcher manipul8s a dv dey must assign;dey cannot manipul8 wout assigning ppl rando.emphasizes human agency woutconsidering possible worlds.man approach id manip event as eh causally prior 1.ex,if a social experiment manipul8s work requirements&funds dat gre8r stringency is associ8d w faster transitions off welfare,then work requirements r presumes 2 c these transitions.main virtue? it allows 2 pin down d cause.wen d requirements 4 causal inference r transferred frm exp2 observ8onal setting those features of experiments dat rest upon manip tend 2 get underplayed.it determines direction of causality&helps rule out untrue correl8ons.main prob? –-pre-emption (major prob rel8d w manip) – u could have poison as an intervention / put d poison in,dey guy died,so it must have been.idea w dis idea—put poison in,doesn’t mean it caused u,cause some1 pre-empted u by poking a hole in it.it unnatural &limited 2 usefulness;it lacks d opportunity2examine real-life aspects.studies explore effects of things which can b manipul8d;only d manip object becomes a direct action 2 solve problem.thus,1st manipul8 something&then only observe its effects;furthermore,knowledge of e of manip causes tells researches nothing about how/y those effects occur.