Midterm 2

Question 1

What do descriptive statistics do (2)? What are the limits (3)?

Answer 1

Helps us describe / classify data

Can be used as part of a larger argument

(cannot identify causes, be an argument by itself, or observe trends)

Question 2

What is hypothesis testing?

Answer 2

Testing the effect of “x” on “y”

Question 3

What does rejecting the hypothesis mean (language)?

Answer 3

Testing our hypothesis against the null: If there is no effect of “x” on “y,” then the relationship will be a straight line

Question 4

What is Y = β0 + β1x + ε

Answer 4

β0 is the intercept (b)
Β1 is the “m”
ε is the error term

Question 5

What are the different tests? When are they used?

Answer 5

Z-score

T-test: used for an “x” with two values, but also ratios/intervals

F-test: used for an “x” with discrete values (categories)

Chi-squared test: used when both “x” and “y” have discrete values (categories)

Pearson correlation: used for an “x” with multiple values – checks against t-test

Question 6

What is the standard deviation divided by sample size?

Answer 6

Standard error

Question 7

What is the standard error?

Answer 7

the average difference between the predicted value and the actual value of the relationship

Question 8

What is the problem of causality?

Answer 8

If we compare treated vs. non-treated individuals, we are not going to get the causal effect right because individuals self-select into treatment

Maybe other variable is the reason for observing the outcome and not X.

Question 9

What are causal inference problems (4)?

Answer 9

Reverse causation (endogeneity): think A causes B, but B causes A

Selection effects

Confounding variable: A and B are caused by C and there is no causal relationship between A and B

Spurious correlation without an underlying relationship

Question 10

What are varieties of research design (2)?

Answer 10

Experimental approach: survey experiments, lab experiments

Observational studies: case studies, small-n comparisons, large-n observational studies

Question 11

What are benefits/drawbacks of experimental approach?

Answer 11

Benefits: Other factors can be excluded; Random assignment prevents self-selection

Drawbacks: hard to convince everyone about external validity of results

Question 12

What are benefits/drawbacks of observational studies?

Answer 12

Benefits: real world data on political phenomena/behaviours, not manipulated/invented

Drawbacks: causal inference problems are hard to overcome

Question 13

What is a case study? Author?

Answer 13

John Gerring: the intensive study of a single case where the purpose is to shed light on a larger class of cases (a population).

Question 13

What are the characteristics of case studies (6)?

Answer 13

Central to qualitative methods.

Can be both descriptive and explanatory.

explanatory: goal of the case study is to explain an outcome through an in-depth study of the mechanisms or processes that lead to that outcome.

Style of analysis tends to be narrative.

May or may not be representative of the population.

Case studies may be theory generating or theory testing

Question 14

How do we choose case studies?

Answer 14

They are interesting and in need of explanation: why did the USSR collapse? How do we compare?

Caution from Geddes: don’t select on the dependent variable! Hard to know what is the population for the dependent variable.

Question 15

What is field research (4)?

Answer 15

Traveling to the site where the political processes you’re researching are taking place to collect data.

Different methods (surverys, archival data,…)

Globalization has reduced need for fieldwork partially.

Local vs. overseas researchers; positionality; power structure of the academy. North Americans researching “overseas,” not relying on local researchers. Are foreign researchers conducting field research in US/Canada?

Local researchers are engaging in field research

Question 16

Why do we conduct field research (5)?

Answer 16

Some data is not available remotely

Some questions require locally obtained data
Knowing the context of the data helps evaluate it

Diversified data helps scientific progress, especially when we aim to produce widely generalizable theories

Experiments run in one country cannot always be replicated in other countries (context)

Comparative studies require we check whether the concepts/data we have is comparable enough across the cases

Question 17

How to conduct field research (6)?

Answer 17

Preparation:
Obtain a REB certificate,

Decide what data you would be looking for

Prepare some contacts, leads

In the field:
Keep a daily research diary with impressions / relevant thoughts about how what you observed relates to hypothesis. You will not remember these things later.

Be prepared to re-evaluate and change direction either for pragmatic or for theoretical reasons

Be safe; consider the ethical ramifications

Question 18

What are challenges during field research (3)?

Answer 18

Ethical challenges

Immersing yourself in the field does NOT guarantee that you will get the right data

Immersing yourself in the field does NOT guarantee that you will apply the right theories to the data and interpret things correctly

Question 19

What is ontology? What are the paradigms?

Answer 19

what exists?

Positivists
Interpretivists

Question 20

What is epistemology? Paradigm views?

Answer 20

How can we know what exists?

Positivist: a real/true world exists out there to be discovered and researched

Interpretivists: a real world does not pre-exist independently of human activity or language

Question 21

What does interpretivism encompass?

Answer 21

All challenge the notion of raw data extracted by neutral methods designed to control the subjectivity of the scientific observer

Question 22

What is interpretivism (4)?

Answer 22

knowledge as historically situated and entangled in power relationships.

the world as socially made.

eschew the individualist assumptions that characterize rational-choice and behaviorist literature.

interested in language and other symbolic systems.

Question 23

What is epistemological reflexivity?

Answer 23

Being reflexive to the ways in which concepts and styles of reasoning, as well as scholarly commitments, are historically situated and enmeshed in power relationships.

Question 24

What is the pearson correlation

Answer 24

Line that defines the relationship of the variables (between 1 and -1)

Question 25

What is a focus group (4)?

Answer 25

Group of individuals is convened to discuss a set of questions centered on a particular topic to generate conversations uncovering individual opinions.

Helps reveal group consensus, where it exists, on the issue at hand.

The potential for data collection emerges from the ‘‘range of experiences and perspectives’’ that these focused conversations uncover.

Used in marketing to know how to advertise. In political science and sociology usually as part of a mixed method approach

Question 26

What is the Herfindahl-Hirschman index of religion?

Answer 26

The higher the number, the fewer religions; lower means more

Question 27

What is the unit of analysis in a focus group?

Answer 27

Individual unit of analysis

Group unit of analysis– consensus on the topic, what cleavages and how strong

Interactive unit of analysis– which alternative routes were dismissed quickly as obviously wrong

Question 28

What are advantages of focus groups (4)?

Answer 28

Individual unit of analysis is a tool for triangulation with other methods

The group unit of analysis can be a tool for a pre-test of a theory before using other methods

The interactive unit of analysis can be a tool for hypothesis generation as it produces unexpected, new observations/ideas

It’s cheaper than surveys and produces more data than individual interviews

Question 29

What are pitfalls of focus groups (4)?

Answer 29

Difficult moderation, sometimes dynamic just doesn’t work

Danger of groupthink developing and producing artificial consensus

Danger that the group dynamic leads individual participants to exaggerate, minimize, or withhold experiences (desirability bias)

Focus groups are not representative samples of the population (hard to know if their results are generalizable)

Question 30

What are best practices for focus group research (3)?

Answer 30

Formulate the main purpose of the focus group in the research design (whether/how to limit participant selection by some variable)

Decide which unit of analysis you would exploit primarily and how

Provide questions for other researchers to use for replication/to understand how the study was run

Question 31

What are interviews?

Answer 31

One-on-one conversation, semi-structured with some questions in mind, with an individual whose behaviour / beliefs are relevant to the phenomenon studied

Common methods tool in research on elites, whose individual agency usually has outsized influence on political processes.

Different research goals– pre-test, context useful for understanding statistical data, missing data, process-tracing.

Question 32

What are advantages of interviews (4)?

Answer 32

Useful addition and corrective to data that is obtained through formal channels

Useful to assess the interaction, regardless of the content that the interviewee provides

Useful tool in environments/situations where there are severe limitations on officially produced data

Useful as a tactic for obtaining other relevant contacts with individuals you didn’t know were important to your research question

Question 33

What are disadvantages of interviews (5)?

Answer 33

People won’t to reveal information they don’t want. Sometimes you get nothing useful beyond the official speeches.

Ethical challenges: may put a respondent at risk with your line of questioning /may obtain information which raises ethical dilemmas about whether to disclose it or not. Interviews for background research only vs. interviews for citation and attribution

REB process and informed consent requirements can be a serious liability for this method in certain contexts.

Occasionally dangerous to the researcher

Because it’s an intimate format where success depends on establishing rapport, might bias the results by interviewer developing sympathy/antipathy towards the subject interviewed.

Question 34

What are best practices for interviews (5)?

Answer 34

Ask open ended questions, not very specific ones

Ask questions about what is important/useful/preferred by the respondent to elicit their own opinion, rather than their reaction to your opinion

Do NOT talk about your theory and hypotheses; it biases the information you will get

Use the interview for other contacts

Stay safe and leave if the situation starts becoming weird

Question 35

What is mixed methods research ? Author?

Answer 35

Using different research methods to study a phenomena

Mixed or multi-methods research is based on the practice of triangulation, i.e. “the combination of methodologies in the study of the same phenomenon.” (Jick)

Question 36

What are the types of quantitative data (6)?

Answer 36

Surveys

Indices

Cross-national political and economic characteristics

Text

Output of political processes that produce repeated outcomes

Network data

Question 37

What are the types of qualitative data (5)?

Answer 37

Interviews

Focus groups

Text

Cross-national political and economic characteristics (sometimes found in archives)

Participant observation (ethnography)

Question 38

What is each method (quantitative, qualitative) good for?

Answer 38

Quantitative methods are good for generalization and hypothesis testing; qualitative methods are good for nuance and hypothesis generatio

Question 39

What is confirmation v validity

Answer 39

(external validity) Using mixed methods to confirm your findings (quant+qual)

Using both to arrive at conclusion

Question 40

What are the drawbacks of mixed methods (4)?

Answer 40

Commensurability problem: Doing an MM study is the same as doing two studies on one subject. Not directly comparable because different methods are based on different conceptualization/measurement of the main concepts used.

Contributes to conceptual stretching

Labor intensive

Will lead to less specialization in the discipline