Deductive vs Inductive RCT Exam 2 Review Packet o Deductive is one that is claimed or intended to be valid o Inductive are not given with the intention of being valid o Deductive Standard validity o Inductive Standard Strong reasons o Adding premises to a valid argument cannot make it invalid but adding to an inductive argument can make it worse o Valid deductive arg are indefeasible as no new info can make them invalid o Strong inductive arg are defeasible as new info can undercut the strength of the argument and the support that the premises give for the conclusion o Deductive logical evaluation o Inductive support based Statistical Generalizations then draws a conclusion about the population as a whole Occur when one uses a sample portion of a population and o Sample of the population was surveyed and results are extrapolated and whatever was true from it is said to be true of the whole o You can tell when a sample size isn t large enough If its not it s a hasty argument generalization fallacy o Important to determine if the sample is biased If you test at only 1 location Fallacy of Biased Sampling o Ask Are the premises acceptable Is the sample too small Is the sample biased Are results affected by other sources of bias Statistical Application population we draw a conclusion concerning a member or subset of that population Opposite of Generalization From information concerning a o 97 of Repub in cali voted for McCain Marvin is a republican from Cali Reference Class above example Therefore marvin voted for McCain o When evaluation the strength of stat applications the percentage of F s that have the feature G is important As the figure approaches 100 the argument gains strength Can also get strong Stat applications when the figure approaches 0 o When the percentages are in the middle range such stat applications are weak F of a Stat Application is the reference class Ex republican party in o Should choose a reference class in a way that brings all relevant evidence to bear on the subject o One way in determining which reference class is stronger is to combine them What of Republicans from California who were relatives of Obama voted for mccain o To be successful such reasoning must take place within a broader framework that helps determine which features are significant and which features are not Sufficient and Necessary Conditions The Sufficient Condition Test SCT The Necessary Condition Test NCT o Any candidate that is absent when G is present is eliminated as a possible o Feature F is a sufficient condition for feature G if an only if anything that has feature F also has feature G o Feature F is a necessary condition for feature G if and only if anything that lacks feature F also lacks feature G o Ex Mercury is a sufficient condition for being a metal but its not necessary because there are other metals o Any candidate that is present when G is absent is eliminated as a possible sufficient condition of G Case1 Case 2 Case 3 A A A B B B Bcz G is present in Case 1 it can be ignored Bcz B and C are present when G is absent so it cannot be sufficient D isn t ruled out Ruling out a candidate is deductive and indefeasible D D D G G G C C C necessary condition of G Case 1 Case 2 Case 3 A A A B B B C C C D D D G G G A B D ruled out bcz they are absent when G is present C isn t ruled out as a candidate but need more testing o Watch for odd in which candidate is always present therefore and cannot A A A A o Beware of cases which candidate is always absent Case 1 Case 2 Case 3 Case 4 B B B B A fails SCT in Case 2 and 3 but passes NCT B Passes SCT but fails NCT in Case 1 and 4 C is eliminated by both Only D is not eliminated by either test so it is the only candidate for D D D D G G G G C C C C being both a necessary and sufficient condition of G o Besides looking for diversity in candidates we should look for diversity in o To test rigorously it involves seeking out cases in which failing the test is target feature G a live possibility Rigorous Testing fail NCT o SCT Cand present and Target Absent o NCT Cand absent and Target Present Applying these Tests to Find Causes o Normality our statements of sufficient and necessary conditions must be contextualized to normal situations o To determine which cases are relevant we take certain background assumptions for granted o Sometimes beliefs are common or specialized knowledge o Begin with a defeasible background assumption were looking for one cause that is causing one illness then apply NCT and SCT o Difference between saying something is a casual factor as opposed to a cause Permanent features of the context as casual factors and of changes that occur before the effect as the cause Doesn t always hold we would say the cause of contracting legionaires disease was the weakened immune sys of those who contracted it o Whatever survives after NCT SCT Testing may be called a casual condition or factor if they fit in well with our system of other casual generalizations o Some of these casual conditions will be called causes if they play a key role in our casual investigations Concomitant Variation absent o The SCT NCT require that sometimes some feature is present and sometimes is o Sometimes a feature is always present but just to a greater or less degree o Can use Concomitant Variation to see whether some feature is significantly responsible for some other feature o Ex Acid Rain Some Acid Rain is always present in the atmosphere o If A varies directly in proportion to B so that when A increases B increases and When A Decreases B Decreases A B Positive Correlation o If A varies directly in proportion to B so that when A decreases B increases A B negatively correlated o Difficult to determine whether there is causation where there is just correlation o When a correlation A is cause of B B is cause of A Some 3rd thing is the cause of both A B Correlation accidental o Have to be somewhat related NOT unrelated Inference to the Best Explanation o Once we settle on a cause we can argue that the cause has more explanatory value for some phenomenon than other causes o If the hypothesis fits well with our framework of beliefs and allows us to explain something better than any other hypothesis then we have a strong inductive arg o The Best Explanation is No competing hypothesis fits well with background beliefs and other facts o Ex 1 Observation There are bits of paper on the floor 2 Explanation The hypo that Samus tore up some printer …