Research Methods for Politics PDF

Preview

Summary

Research Methods for PoliticsWeek 1Core ReadingWilliam L. Neuman (2011). Social Research Methods: Qualitative and QuantitativeApproaches. Seventh edition. Boston: Pearson, Chapter 1 on “Why do research?” (pp. 1–24). Social science research does not produce absolute truths. P Alternatives to social...

Description

Research Methods for Politics Week 1 Core Reading William L. Neuman (2011). Social Research Methods: Qualitative and Quantitative Approaches. Seventh edition. Boston: Pearson, Chapter 1 on “Why do research?” (pp. 1– 24).  

Social science research does not produce absolute truths. P2 Alternatives to social research: o Personal experience – a powerful influence – “I’ll believe it when I see it”. Leads us to misconception and misjudgment eg. Women continuing to buy face creams that don’t work due to fear. P3-4



Risks associated with relying on personal experience:

    

Over-generalization Selective observation – only allowing ourselves to observe some evidence and ignoring other facts. Premature closure – believing we have the answer to a question so no longer looking for evidence. Halo Effect – holding something as true because it is affiliated with a person we respect. False consensus – over-estimating the extent to which our own view matches the view of other people. Pp.5-6

o Opinions based on popular and media messages – following the masses and basing opinions on media and popular opinion. ‘Problem promoters’ – misleading reports etc. pp. 6-7 o Knowledge subordinated to ideological beliefs and values pp. 7-8

Lecture 1 – Introduction 

What is social science research? o o o



Social science is research without a predetermined position. We may have expectations, but we should not already know the answer to our question – should be open to all kinds of evidence. Should follow standard procedures, which allows you to communicate what you did to readers/other people – reduces risk of misjudgement/bias etc. (Newuman 2011, p5).

Research Questions o o

Research starts with a research question – not just with a topic. Often (but not necessarily), this is an explanatory (causal) question, for example:

• • •

o o o



Formulating a good question is often the most challenging part of the project. A research question is not the same as an essay question – a research question is normally narrower and more precise, as research often requires narrow questions. Typically, the question refers to something that is ‘puzzling’, or to a ‘problem’ in need of a solution.

Puzzles o

Hancké (2010, p. 234): Much social science is about resolving puzzles • •



o



Puzzles raise a question that should have been answered by existing theories but was not and which may, therefore, change our understanding

For instance: o



Why did we observe …? How can we account for …? These questions involve causality.

A single fact – e.g., an event that is not in line with what we should expect on the basis of existing theory A paradox – e.g., in two similar cases, where we would expect similar things to happen on the basis of theory, something the opposite happened

RQ: How can we account of this?

Examples of ‘puzzle’-based research questions o

Algeria & Tunisia •

o

2 2. Algeria and Tunisia Algeria and Tunisia are very similar countries in many respects that are relevant for political stability • Yet, while Tunisia experienced a revolution in 2011, Algeria did not • How can we account for this difference?

Arend Lijphart (1968) – The Politics of Accommodation  

 

According to pluralist theory, culturally heterogeneous and segmented countries are politically unstable. However, if we look at the Netherlands in the period 1920-1970, we find a country that was culturally heterogeneous and extremely segmented, but also extremely politically stable! How can we account for this? (Lijphart’s explanation: culturally heterogeneous and segmented countries can be stable if there is inclusive co-operation among political elites)

Problems • Sometimes social science research is designed to shed light on practical problems • Policy-makers have many unanswered questions that social science research can help to answer

• Problem-based research may in some cases then generate further ‘puzzles’

Week 2 Core Readings John Gerring (2012). “Mere description.” British Journal of Political Science 42 (4): 721– 746.   









  

‘Description’ is “employed as a euphemism for failed, or not yet proven, causal inference.” P721 “Studies that do not engage causal or predictive questions, or do not do so successfully, are judged ‘merely’ descriptive.” P721 “A descriptive question describes some aspect of the world” and “aims to answer what questions […] about a phenomenon or a set of phenomena. Descriptive arguments are about what is/was.” P722 “causal arguments attempt to answer why questions. Specifically, they assert that one or more factors generate change in some outcome, or generated change on some particular outcome.” P723 Whilst ‘descriptive’ is sometimes taken as being synonymous with ‘non-evaluative’ or ‘empirical’, Gerring refutes this meaning, and asserts that description can be normative. P723 “the terms causal and descriptive should be understood as forms of argumentation, not (or not secondarily) as characterizations of the sort of evidence available for causal inference.” P725 Taxonomy of descriptive arguments (p.725) shown below.

Accounts – “any analysis of an event or set of events with no explicit attempt to generalize beyond the specific of a particular case.” P725 Indicator – “aims to describe one feature (i.e., one dimension) of a population based on the empirical manifestation of some phenomenon.” P725 Associations – “Descriptive arguments about multidimensional components of a phenomenon, or the properties of various units” p726

 

 



Syntheses – “multidimensional arguments [which] attempt to group together diverse dimensions into distinct categories”, where “there is only one category”. P727 Typology – “where multiple categories are defined” and “the goal is to sort phenomena into discrete categories that are mutually exclusive ad exhaustive on the basis of a uniform categorization principle or principles.” P727 “In political science, causal studies are more common, and descriptive studies less common, than in other social sciences or the natural sciences.” P732 The importance of description is “thought to derive from its role in causal explanation. Political scientists describe in order to explain.”, but Gerring thinks that it should be pursued independently in some instances because if we only describe in order to explain then “we will know less about the world (descriptively) and what we know will be less precise, less reliable and perhaps subject to systematic bias” P733 If the mission is descriptive, and “no single causal theory guides the research” then the evidence collected will be broader and apply to many both causal and descriptive questions p. 734

Seminar Readings Renske Doorenspleet (2000). “Reassessing the three waves of democratization.” World Politics 52 (3): 384–406. 

Largely descriptive – see highlighted article. Some explanation (talks about the causal relationship between democratization (as the dependent variable) and decolonization in Africa (independent variable). Seminar

    

What is the difference between description and explanation? How can ‘mere description’ be used in the study of politics? What types of description can we distinguish? How can typologies be used in the study of politics? Is description useful in studying political science? Lecture 2 – Description and Explanation



Variables: o o o

o o o

A variable is a property that can take different values or characteristics or attributes (and it may change). Variables we are often interested in are age, gender, religion, democracy, corruption. Some of these are individual variables, whilst others are aggregate level variables. Eg. Corruption by definition cannot be the property of an individual as it relates to relationships between individuals. If something doesn’t change then it is a constant. In causal arguments, we find independent and dependent variables (IV and DV; X and Y). The dependent variable is the thing we are interested in and trying to explain.

o o o o 

Generalisation o o o o



A process of applying information collected about a phenomenon or a group to a set of phenomena or a larger group with similar characteristics. Generalisation typically results in loss of information, but this is necessary for most forms of analysis , and indeed for most forms of communication. ‘Generalisation’ does not carry negative connotations in political science as it does colloquially – different people are comfortable with different degrees of generalisation. Every time we use language we make decisions about how much information we are happy to lose (pen example). We all generalise constantly.

Inference o o o o o



Independent variables are “not problematical” – they are “taken as simply given” and “presumed to cause or determine a dependent variable” (Babbie 2004, p. 20). Explanans – the variable that explains. Dependent variable – “assumed to depend on or be caused by another” (idem). There are usually several independent variables and one dependent variable.

We generalise because we want to make inferences using the facts that we do know about the things that we don’t know. “the process of using the facts we know to learn about facts we do not know” (King, Keohane and Verba 1994, p. 46). Descriptive inference – an inference about what the world is. Causal inference – an inference about why something happens. In academic research, we aim to make inferences which are as accurate as possible.

Research Questions o o

o o o o

o

We should not yet know the answer to our research question, though we may have expectations (which will form our hypotheses). The question needs to be researchable. Some questions are unknowable (eg. What would have happened if Britain did not vote for Brexit?), and some ask for answers that lie in the future (eg. What will be the effect of Brexit on the economy). Time and resources are limited. Some questions are unethical eg. Is torture an effective way of obtaining information from terrorists? The question needs to be interesting and relevant – they need scientific relevance, social importance, timeliness etc. The question should be original. What is originality? Making sure you don’t ask the exact same question that someone has already addressed, or making sure you don’t give the same answer. Types of research questions: o

o

Univariate descriptive questions – describe a situation, a trend, a phenomenon, or different manifestations of a phenomenon. What is happening or how it is happening. Eg. ‘What percentage of the electorate are expected to vote for the Conservatives in the next election?’, ‘How often do Kings students use Facebook?’. Relational descriptive questions – what is the relationship (association) between variables A and B? No argument about causality. Eg. ‘Is there a relationship between age and voting behavior in the Scottish independence referendum?’.

o



Explanatory or causal questions – whether one or more variables (independent variables) affects or causes a certain outcome (dependent variable). An interest in some ‘cause-effect’ relationship. This is the most demanding type of question – requires a description of the variables and a demonstration that they are related. Eg. ‘Does ethnic diversity drive down generalised trust?’.

Types of Description o o o o o

Description of a single event without an attempt to generalise (accounts). Description of one specific feature of a population based on the empirical manifestation of the phenomenon (a single variable or indicator). Description of the relation between a phenomenon and the properties of various units, or a phenomenon and time (association – see relational questions). Description of a set of attributes associated with a ‘common theme’ (synthesis). Descriptive typologies. These can be simple or multidimensional. They can be descriptive or explanatory.

Week 3 Core Readings Stephen van Evera (1997). Guide to Methods for Students of Political Science. Ithaca, NY: Cornell University Press, Chapter 1 on “Hypotheses, laws, and theories: A user’s guide” (pp. 7–48). 

What is a theory? o

o o o





 

“Theories are general statements that describe and explain the causes or effects of classes of phenomena. They are composed of causal laws or hypotheses, explanations, and antecedent conditions.” Pp. 7-8 A causal law or causal hypothesis with explanation is a theory. P. 9 “The proposal “A  B” is the theory’s prime hypothesis, while the proposals that “A  q”, “q  r” and “r  B” are its explanatory hypotheses.” P. 13 “A “theory” that cannot be arrow-diagrammed is not a theory and needs reframing to become a theory.” Pp. 14-15

Hypothesis – “A conjectured relationship between two phenomena. Like laws, hypotheses can be of two types: causal (“I summarise that A causes B”) and noncausal (“I summarise that A and B are caused by C; hence A and B are correlated but neither causes the other”).” P. 9 Antecedent conditions – “a phenomenon whose presence activates or magnifies the action of a causal law or hypothesis. Without it causation operates more weakly (“A causes some B if C is absent, more B if C is present”)” Pp. 9-10 “We can restate an antecedent condition as a causal law or hypothesis (“C causes B if A is present, otherwise not”” p. 10 Intervening variable – “A variable framing intervening phenomenon included in a causal theory’s explanation. Intervening phenomena are caused by the IV and cause the DV. In the theory “Sunshine causes photosynthesis, causing grass to grow”, photosynthesis is the intervening variable.” P. 11



 

Condition variable – “A variable framing an antecedent condition. The values of condition variables govern the size of the impact that IVs or IntVs have on DVs and other IntVs. In the hypothesis “Sunshine makes grass grow, but only if we also get some rainfall,” the amount of rainfall is a condition variable.” P.11 Prime hypothesis – “The overarching hypothesis that frames the relationship between a theory’s independent and dependent variables.” P.11 What is a good theory? o o o

o

o

o o o



“A good theory has large explanatory power. The theory’s independent variable has a large effect on a wide range of phenomena under a wide range of conditions.” P.17 The extent of explanatory power is determined by explanatory range, importance and applicability. Pp. 17-18 “Good theories elucidate by simplifying. Hence a good theory is parsimonious. It uses few variables simply arranged to explain its effects.” This unfortunately involves sacrificing explanatory power. P. 19 “A good theory is “satisfying”, that is, it satisfies our curiosity. A theory is unsatisfying if it leaves us wondering what causes the cause proposed by the theory.” “The further removed a cause stands from its proposed effect, the more satisfying the theory.” P. 19 “A good theory is clearly framed. Otherwise we cannot infer predictions from it, test it, or apply it to concrete situations.” “A clearly framed theory includes a full outline of the theory’s explanation. It does not leave us wondering how A causes B.” P. 19 “A good theory is in principle falsifiable. Data that would falsify the theory can be defined (although it may not now be available).” P. 20 “A good theory explains important phenomena: it answers questions that matter to the wider world, or it helps others answer such questions.” Pp. 20-21 “A good theory has prescriptive richness. It yields useful policy recommendations.” “A theory gains prescriptive richness by pointing to manipulable causes, since manipulable causes might be controlled by human action.” P. 21

How can theories be made? o o

o

“Some scholars use deduction, inferring explanation from more general, alreadyestablished causal laws.” Pp. 21-22 “Others make theories inductively: they look for relationships between phenomena; then they investigate to see if discovered relationships are causal; then they ask “of what more general causal law is this specific cause-effect process an example?”” p. 22 Aids to theory-making:

  



“We can examine “outlier” cases, that is, cases poorly explained by existing theories.” P. 22 “The “method of difference” and “method of agreement” (proposed by John Stuart Mill) can serve as aids to inductive theory-making.” P.23 “We can select cases with extreme high or low values on the study variable (SV) and explore them for phenomena associated with it. If the values on the study variable are very high (if the SV phenomenon is present in abundance), its causes and effects should also be present in unusual abundance, standing out against the case background.” P. 25 “We can select cases with extreme within-case variance in the value on the study variable and explore them for phenomena that covary with it. If values on the study variable vary sharply, its causes and effects should also vary sharply, standing out against the more static case background.” P.25



 

 

o

“Counterfactual analysis can aid inductive theorizing. The analyst examines history, trying to “predict” how events would have unfolded had a few elements of the story been changed, with a focus on varying conditions that seem important and/or manipulable.” P.25 “Counterfactual analysis helps us recognize theories, not make them.” P. 26 “Theories can often be inferred from policy debates.” “Theories inferred in this fashion are sure to have policy relevance, and they merit close attention for this reason” P. 26 “The insight of actors or observers who experienced the event one seeks to explain can be mined for hypotheses. Those who experience a case often observe important unrecorded data that is unavailable to later investigators. Hence they can suggest hypotheses that we could not infer from direct observation alone.” P. 26 “Large-n data sets can be explored for correlations between variables.” P. 27 “We can fashion theories by importing existing theories from one domain and adapting them to explain phenomena in another.” P. 27

How can theories be tested?

  

     

Experimentation – Observation – large-n – “A large number of cases – usually several dozen or more – is assembled and explored to see if variables covary as the theory predicts.” P.29 Observation – case study – “The analyst explores a small number of cases (as few as one) in detail, to see whether events unfold in the manner predicted and (if the subject involves human behavior) whether actors speak and act as the theory predicts.” P. 29 “A strong test is one whose outcome is unlikely to result from any factor except the operation or failure of the theory.” P. 31 Hoop tests – can kill a theory, but give little support if passed. P. 31 Smoking-gun test – “passage strongly corroborates the explanation, but a flunk infirms it very little.” pp. 31-32 Doubly-decisive tests – “passage strongly corroborates an explanation, a flunk kills it.” P. 32 Straw-in-the-wind test – “such tests can weigh in the total balance of evidence but are themselves indecisive.” P. 32 “Test as many of a theory’s hypotheses as possible” otherwise it is part tested. P. 35

Seminar Readings George Tsebelis (1999). “Veto players and law producti...