Theory and Argument Preface How people think theoretically how people generate theories and the process of developing theory The world is composed of about 50 variables Arguing exchanges reasons Theory what it is to think theoretically Theories are arguments Theories and Arguments They summarize evidence induction guaranteed by explanation o Generalize from particular observations to particular conclusions o Statistics induction Sample sizes give us an inductive idea They generate hypotheses and other implications deduction o Deduce further consequences o Philosophy of science theories in the natural or social disciplines and what they have in mind is frankly a somewhat unrealistic vision o Not the whole the picture o Deduction word substitution o Certain conclusion They begin by collecting clues into a first suspicion or hypothesis abduction does not result in a firm conclusion o Process of looking at a bunch of observations and imagining something that would account for them o That thing you will imagine is the hypothesis that you will test o Example Police detective all of the clues would be accounted for if she did it Are you in the will Did he break up with you or did you break up with him o Reasoning to the best conclusion o Best shot to getting it right o Suggestive evidence o Works from clues matter in these respects Argument as Method The type of methodology quantitative critical qualitative doesn t Which methodology you use does not matter All of them argue in a particular way Use methodology to test theory Actually method is argument Design 1 o Frighten the students o Test the students o Conclude that terror improves performance Design 2 o Be nice to the students o Test students o Notice second group scored higher o Conclude that being nice is better teaching Argument Method one shot case study Second design two group no pretest experiment is better than first For reasons arguments that you can find in a design book o Comparison vs intuitive sense control comparable groups etc o These reasons have been worked out over the decades i e arguments have been settled Examples of argumentative conclusions in social science o Results shouldn t depend on the investigator o Data and procedures must be public o Comparisons are required o Things that could happen 5 by chance aren t real P 05 o Need 70 reliability in measurement Formal research procedures are nothing more than a codification of the principles of critical thought Follow instructions and avoid fatal criticisms Need theory and data o Theory offers premises o Data can test and extend the premises o Conclusions are new theory Example o Theory says A causes B o Data say A causes B but only for males o Conclusion need a theory that includes A and B and also distinguishes male from female experience of A and B Arguing toward Theory Theory and data should be of high quality Argue for a theory Theory s quality this whole class o Truth beauty and justice in evaluating a theory What are high quality data Data o Authentic reliable and valid Theory has to be accurate ethics and social justice quite relevant o Even recognizing the phenomena is not simple operationalization can be controversial e g relational happiness Scale Duration Observer s judgment o Investigator interpretive or distanced e g data from respondents point of view or quasi objective Communication Models Models and theories are the same things model is a template for the process of reaching a theory Simplified representation of the world Four Models o Individual Choice o Exchange o Adaptation o Diffusion Includes important elements Necessary skills and techniques to identify a model o An ability to abstract from reality to a model o A facility at derivation within an abstract model o A competence at evaluation a model o A familiarity with some common models Man is capable of producing more complex behavior than he is capable of understanding Speculative Models o Trying to find out different theoretical explanations of what you observe o You have more than one model so you have a bigger picture What processes lead to what you observe most important thing in our communication theories Process o Observe the fact simply state what we observe o Speculate the process that results in your observation o Deduce implications from process o Test validity of implications refine your model You may deduce many explanations Whether the implications are true collect data and statistical test so then the model can be generalized Arguing Toward Theory Preface Theory and data should be of high quality Theory s quality this whole class Truth beauty and justice Data Authentic reliable and valid Even recognizing the phenomena is not simple operationalization can be controversial e g relational happiness Scale Duration Observers judgment Investigator interpretative or distanced e g data from respondents point of view or quasi objective o Individual makes judgment Distanced or involved o Survey distanced Getting involved nuance data but raises the possibility of changing the data because of the observer s presence The better the data the more you trust the theory test The bottom line is the theory is the theory good or bad Need to be modified Good Arguments May Not Be Conclusive Why is social science tentative o If H is true Data will show X o The Data show X o So H is true o Theories change data gets reinterpreted o Base syllogism that a social scientist works with on a day to day basis o If the hypothesis is right I am going to find X But this is invalid affirming the consequent o If P then Q if you re rick then you re happy o Q you re happy o So P you re rich o Formal fallacy o Represents day to day thinking Argument as Method So we try to falsify Popper not verify Logical Positivism The more falsification efforts something survives the more confidence we have in it This is a community argument you will get some stubborn people Intro to Models in the Social Sciences Ch 1 What We Are Up To o Pleasures of thinking about human behavior o Book is about invention of conjectures o Imagination speculation with some discipline Model Theory etc A simplified picture of a part of the real world Will have some veridical features but not all So can have several models of the same thing Willing to accept as synonyms theory paradigm hypothesis idea How much can I learn about the whole phenomenon by just understanding one thing one part Looking at a relationship like a balance sheet economic model