Social Science Research Foundations

• Social science research aims to make generalizations about reality through social regularities.
• This involves analyzing variables to understand relationships and patterns.
• Researchers use this analysis to make inferences about aggregates rather than individual cases.
• The goal is to develop and test theories that explain empirical generalizations.
• This process aims to provide descriptive and causal inferences

Causal vs Descriptive Inferences

• Causal inferences determine if a change in X directly causes a change in Y. They depend on methods that establish time order and eliminate alternative explanations.
• Descriptive inferences accurately describe characteristics or patterns within data without necessarily establishing a cause/effect relationship.

Foundations of Social Science

• A theory is a statement or series of statements that organize, explain, and predict knowledge.
• It usually implies multiple specific descriptive or causal hypotheses.
• Social regularities are patterns of generalizations about reality, which social science research aims to identify and understand.
• They include descriptive and causal inferences.
• These regularities are typically probabilistic, describing tendencies or likelihoods rather than absolute certainties.
• Social science is generally interested in making inferences about aggregates, not individuals.
• Social science discusses things in terms of variables.
• Variables are made up of attributes.
• For example, party identification (variable), democratic and republican (attributes).

Normative vs Empirical Questions

• Normative questions ask how things should be.
• Empirical questions ask how things are in reality.

What is a Theory?

• It is a statement that organizes, explains, and predicts knowledge.
• Theories explains observations supported by data.
• It is usually widely accepted by the scientific community.

Hypothesis

• An explicit, testable statement indicates how a researcher thinks the phenomena of interest are related.
• Based on theory and other empirical research, it's an educated guess.
• How X influences Y depends on the unit of analysis.
• One value on the independent variable is more likely to have one value on the dependent variable.
• Unit of analysis with a different value on the independent variable.
• A hypothesis is more specific than a thesis.
• A thesis is the main argument of the paper
• A strong thesis must be clear, relevant, arguable, and precise.

Strong vs Weak Hypothesis

• A strong hypothesis:
• Must be an empirical statement, guessing about a realistic relationship.
• Must be a general statement but broad enough to generally explain Y, while not solely based on just one case.
• Must be plausible based on theory and/or sound reason and also be based on other empirical information.
• This is why we do literature reviews.
• Must be general, but also specific which the relationships between X and Y are clearly stated.
• Can be a directional hypothesis (X causes Y), and can be a hypothesis about positive or negative relationships.
• Concepts must clearly defined.
• Should be stated congruently to how the research intends to test it, and be consistent with the data.
• Must be testable by needing distinct concepts for X and Y, and data.
• Must specify the units of analysis (who or what is acting).

Plagiarism, Paraphrasing and Drop-ins

• Plagiarism means using someone else's ideas without citing them.
• Paraphrasing means describing arguments/evidence presented in another's work and citing them.
• Drop-ins are when quotations are dropped into paragraph without properly giving credit to the author.

Important Parts of an Annotated Bibliography Entry

• The research question
• The thesis statement
• Theories underlying the expected answer to the research question
• How the main variables are conceptualized
• Hypotheses being tested
• Methods to test hypotheses
• Findings/results
• Broader research implications
• Any limitations in the research design, methods, or any part of the paper

Qualities of a Good Literature Review

• Qualities include;
• The different schools of thought that have developed in response to your question
• How each school would answer your question
• The strengths and weaknesses of each school
• Which school's argument is best for your purposes and why
• Included in the conclusion of the literature review
• What informs hypothesis
• An independent variable is assumed to have the causal effect.
• A dependent variable is the variable that is being acted on

Mediating vs Moderating Variables

• Mediating/Intervening explains the processes or mechanism where an independent variable (X) influences a dependent variable (Y).
• Moderating influences the relationship between an independent and dependent variable.
• Mediating explains the process or mechanism through which X affects Y.
• Acts as an intermediary step
• Moderating influences the strength or direction of the relationship between X and Y.

Large-n Studies

• Characteristics include;
• Examination of a large number of cases, allowing for statistical analysis and generalizations across broader contexts
• Typically quantitative
• Focuses on identifying patterns, testing hypotheses, and making causal inferences by analyzing data from numerous instances of a phenomenon
• The goal is to derive conclusions that are applicable to a wide range of similar cases, enhancing the study's generalizability

Pros and Cons

• Pros include
• Larger generalizability/higher external validity
• Able to identify broad patterns and trends across diverse contexts
• Cons include;
• May lack depth and detailed understanding of specific cases that case studies (small-n) provide
• Sometimes overlook unique or context-specific factors that influence outcomes

Criteria for Nomothetic Causality (or causality in large-n studies)

• Criteria include;
• Establish correlation, by needing to establish that as X changes, so does Y
• Establish time order by proving that X precedes Y
• Nonspurious by eliminating alternatives
• A spurious relationship is correlation, but not causation

Research Designs

• Different types of research designs have strengths, weaknesses, and characteristics, which are Qualitative, Cross-Sectional, Time Series, Time Panel, and Experimental.
• Qualitative designs include:
• Case studies
• Process tracing
• Interviews
• Focus groups
• Ethnographic research

Cross-Sectional Research Designs

• Characteristics include;
• X and Y are measured at the same point in time
• Most survey research interviews each respondent once
• The potential problem lies in the difficulty in establishing time order

Time Series Research Designs

• Characteristics include;
• Multiple measures of X and Y across time.
• This allows researchers to look at the delayed effect of X on Y and controls what effect Yt-1 might have.
• Makes it easier to establish time order

Time Panel Research Designs/Panel studies

• Characteristics include;
• A cross-sectional design with a time element and allows a type of pre and post element to the design.
• Looks at the effect of X₁ on Y₁controlling for Y₂.
• Look at the change in Y, given the change in X.
• Can make stronger claims because can establish time order.

Experimental Research Designs

• Are characterized by randomization control over variables and the ability to establish causal relationships through manipulation of X.
• The strengths lie in strong internal validity.
• The weaknesses lie in limited external validity.
• Findings may not generalize well to broader populations due to factors like selection bias and artificial settings.

Survey Research

• Group Assignment: Researchers assign participants to one of several conditions, or treatments with a treatment condition and control condition.
• Randomization ensures participants are randomly assigned to one of the two (or more) groups
• This eliminates alternative explanations because the treatment and control groups are no different
• This is how to eliminate alternative explanations
• Control over how and who is exposed to the treatment.
• Potential for researchers to not administer treatment until the measures of Y has been taken (before and after).
• Control over other potential effects makes sure that both groups are participating in similar environments, at the same time.
• We want to be sure that X is the only thing that varies across groups.
• Is characterized by the ability to gather data from a large number of respondents.
• Making them effective for achieving external validity and generalizing findings to larger populations.
• Is strong quantifying data and offering flexibility in question topics.
• Challenges internal validity because of difficulties in establishing causality and time order.
• Might struggle with capturing detailed responses and actual behaviors.
• Can be affected by biases such as social desirability and response-set effects.
• Response set effects: The tendency of survey respondents to answer questions in a consistent pattern, regardless of the content.

Experimental Designs

• Might choose answers based on the format or structure of the survey rather than their true opinions or knowledge.

Pre-test Design

• Randomly assigned groups, a pre-test is conducted, treatment, then post-test administered. Researchers examine changes for individuals and entire group after the results are compiled.
• Strengths of being able to examine on both levels giving insight into effect of treatment.
• Weaknesses of the actual pre-tests alerting participants to a subject and skewing results

Post Test Only Design

• Randomly assigned groups, with a dependent variable (Y) measured only after treatment has been administered.
• Differences in the post-test results are due to the treatment with a baseline.
• Randomization ensures compatibility between groups and eliminating the need for a pre-test which could influence participants behavior.
• A strength is avoiding biases from potentially exposing participants to prior interventions
• A weakness with being unable to measure changes from a baseline, making it difficult to asses individual-level changes and potentially limiting insights into the treatment's impact

Internal and External Validity

• Internal shows confidence in the fact that X is directly causing Y and provides with the facts.
• External allows you to generalize broader results.

Field Experiments

• Lacks random assignment of subjects, but still involves the manipulation of the experimental variable
• Efforts are made to control the selection of subjects and the manipulation of the independent variable, despite the challenges posed by the real-world context
• Strengths lie in the higher external validity because they're conducted in real-world settings, making them applicable to broader contexts
• Weaknesses include; generalization issues, lack of control over variables, and ethical concerns regarding environmental manipulation

Confounding Factors

• Are variables that can affect both the independent and dependent variables in a study, potentially leading to a false assumption of causation.

Survey Research

• Focuses on whether questions are open or closed which have their own advantages and disadvantages

Open vs Closed Questions

• Open allow respondents to answer however they want.
• Advantages of allowing respondents to answer questions while also being to provide more detailed questions
• Disadvantages are time to conduct and provide more room for interpretation
• Closed focuses on multiple choice answers
• Advantages; takes little time and provides a wider, more detailed explanation for sensitive answers.
• Disadvantages; may force respondents to chose a more accurate answer which doesn't accurately represent their attitude.

Wording Influences

• Wording of response categories can sway the results

Potential Problems:

• Double-barreled questions include more than one attitude or stimuli.
• Ambiguous questions ask for evaluation for questions not clearly defined.
• Unbalanced statements have options for only support or opposition.
• Overly favoring respondents to agree is more harmful than good.
• People tend to agree on all issues with the lack of a medium section to identify actual issues.
• Failure to Provide Middle Positions
• Response Acquiescence forces to respondents from actually providing legitimate evidence for issues.
• Incompetent and unwilling Respondents is when they competent to work alone and provide legitimate answers.

Desirability

• People may lie about some things (race,voting) which may be harder to identify.

Question Ordering

• May also influence the validity or reliability of measures being surveyed.

Framing Effects and Pretesting

• Previous interviews are not useful for later questions due to differences in the same case leading to inconsistent responses.
• Researchers that don't conduct pre-testing will ask themselves about the thought process and measuring questions properly.
• All groups are sampled and distributed in proportion to their sample size within the population

Samples

• Each group is sampled in proportion to their size in the population.
• Used to make inferences about separate groups without regard to their size in the population.
• Simple Random Sample: Each element has an equal probability of being selected.
• Systematic Sample: Elements are selected at predetermined intervals from a list.
• Stratified Sample: Elements are grouped by characteristics, and samples are taken from each group proportionally.

Multistage Cluster Sample

• Involves sampling clusters, such as cities or blocks, and then sampling within those clusters.
• Includes purposive, convenience, quota, and snowball samples, where the probability of selection is unknown.

Sample vs Populations

• Sample refers to who the data collection is on vs a population is the type of make of an inference

Quantitative vs Qualitative

• How to select cases with qualitative research:
• Emphasizes non-numerical data vs quantitative
• Focuses numerical and analytical data, aiming to test wider populations
• Qualitative focuses on understanding, whereas quantitative focuses on numerical data with larger samples.

Different Tests

• Smoking tests (X is sufficient to explain Y) vs Straw test (not sufficient to explain Y).

Qualitative and Qualitative Reseach Basics

• Basics Qualitative Research
• Understanding methods, unstructured groups, and active engagement
• Emphasizes idopgraphic thinking and specific content
• Considers the importance of epistological frameworks

Similar Research Strategies

• Similar strategies are selecting different variables within the group in order to solve for different groups.

Pros and Cons

• Better control, provides for a more detailed explanation
• Risk in overlooking potentially biased conclusions

Differing Research

• Selecting two different groups, with the outcome of finding the results that lead to a same outcome.
• Allows researchers to identify similar factors

Process Tracing

• Focus on how and why variables are acting within certain pathways.
• Involves detailed data analysis to solve these specific problems.

Four categories of tracing/analyzing

• Patterns
• Sequencing
• Traces Accounts

How to Select Cases for the Process

• Focus on four types with quadrant 1 being the most important.
• Single, multi, and interviews are all different types of research methods to finding this conclusion.

Single Multi Reviews

• Single is deeper, but lacks the breadth, mulit focuses on a more broad topic being solved.

Interviews

• Is the ability to find more specialized data that needs to be looked into
• Includes Face to face for richer data, while email can be more in-depth and convenient