Quiz 1 POLI 399 Quantitative Research Methods Flashcards

You may prefer our related Brainscape-certified flashcards:
1
Q

How do you know something is true?

A
  • Expertise
  • Evidence
  • Believe it
  • Perception
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Hallmarks of the Scientific Method

A
  • Empiricism
  • Inter-subjectivity
  • Explanation
  • Determinism
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Empiricism

A
  • Every knowledge claim is verified by systematic observation
  • Assume an objective reality
  • Assumes our senses give us the most accurate information about the world. How we observe and experience reality.
  • Assertion is not enough, need observable evidence
  • Critiques: no objective reality exists, viewpoints are too subjective
  • Empiricism guards against bias, defined as bias: prejudiced against a particular idea/explanation or prejudiced for.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Inter-Subjectivity

A
  • Agreement between the individual and the scientific method (agreement on how the work is done and what counts as evidence and validation)
  • Science as a way of knowing, inter-subjectivity guards against bias (2) through shared standards for determining empirical standard:
  • Transmissibility: steps followed in research that are clear and detailed enough that someone can repeat your research.
  • Replicability: someone who does repeat your work gets the same results. Allows others to test for bias in your work MUST BE REPLICABLE.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Explanation

A
  • the GOAL of the scientific method is explanation
  • Explain political phenomenon using something else which is achieved through variation.
  • Explain why some phenomena are related to others
  • Explanations across time, most useful, must generalize beyond time and place.
  • explanation = general pattern = determinism
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Determinism

A
  • Need to assume that patterns exist which is justified by how much inter-subjective work exists
  • Regularities in politics and political nature, need to assume this pattern and find it.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Scientific Knowledge v.s Common Sense

A
  • Scientific knowledge aims to see patterns and predict
  • Common sense asks questions and wants to know why
  • Difference is that scientific knowledge is conscious and deliberate
  • Common sense: observe accurately, jump to conclusions, overlook contradictory evidence, explain away contradictions
  • Scientific Knowledge: systematically w/criteria and relevance established in advance, avoid over generalizations replication steps before conclusion, vigorously test alternatives, make more observations
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Nature of Scientific Claims

A

1) Never true or proven no matter how many time they have been tested. can make a knowledge claim but always an element of uncertainty.
2) Must be testable or potentially falsifiable.
3) Disconfirming evidence must always be possible

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Traditional Critiques of the Scientific Method

A

Push to become more scientific in the 50’s and 60’s in the States.

1) Human reaction problem
2) Influence of values
3) Complexity of political phenomena
4) Indeterminism (Free Will)
5) Human uniqueness

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Hawthorne Effect

A
  • Tendency to perform or perceive differently when one knows they’re being observed.
  • Problem w/empiricism can’t observe objective reality. It undermines a key assumption of the scientific method.
  • Response to this critique: reactivity is a problem but it is not an insurmountable methodological barrier. You can alter your design.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Influence of Values

A
  • Can never be value free because political scientists are value laden and the individuals studying value-laden behavior
  • Problem is with inter-subjectivity and subsequently bias
  • Response to this critique: recognize and make explicit your own value commitments (Gunnar Myrdal). Inter-subjectivity and testing plausible alternative interpretations and agreement across the political spectrum is powerful.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Complexity of Political Phenomena

A
  • Politics is too complicated and therefore cannot be explained in a generalizable way. Problem with determinism
  • Response to critique: complicated things are studied all the time and this too is not an insurmountable challenge. Polisci has empirical laws that make it less complicated. Incomplete explanations are still useful.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Indeterminism

A
  • Free Will
  • Causation is impossible because humans are free to behave as they wish
  • Problems with determinism
  • Response to critique: humans are confined by the context that they are in and bound by the law. There are a limited number of options available so a pattern will emerge. Completely random behavior is highly unlikely
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Human Uniqueness

A
  • Every person is unique and behaves differently. No two people are alike.
  • Problems with determinism
  • Response to critique: The idea that nothing is shared is not necessarily true.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Statistical Development and Policy

A
  • Gvt interested in evidence based policy developments but it has its limits. Public does not accept empirical evidence when it does not confirm accepted wisdom. If it is confirmationary of public consciousness it is readily accepted.
  • Stages: 1. Identification of trends requiring a policy response. Use social intelligence * the most important pat of statistical agency. 2. Assessing Causation. Need this to develop an appropriate response. Policy hampered by lack of understanding of causes
    3. Development of appropriate response. Domain of policy analysts
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Intersectional Approach

A
  • Relationship between categories is an open empirical question. Dynamic interaction.
  • Unitary categories bind people into political groups based on a uniform set of experiences. Pre-existing assumption of shared political goals.
  • Critique of group unity. One cannot privilege a single aspect of their identity.
  • Policy solutions addressing multiple identity categories..
  • Multiplicative Categories
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Categories of Difference (Types)

A

Unitary: presumed emphasis on a single category as most explanatory
Multiple: recognizes the role of several categories independent of each other.
Intersectional: challenges relationship between categories from determined to the subject of empirical investigation. interactive, mutually constitutive relationships among categories.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

Fuzzy Set Logic

A
  • Move beyond nominal measures of socially constructed categories.
  • Acknowledges psychological literature on identity development, relies ton mutliple questions to determine category membership and handles causal complexity.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

Feminist Critique of the Scientific Method

A
  • 3 levels of critique, begins because women are invisible in the work (partly due to legal inclusions).
  • Researchers not encountering problems faced by women and people of colour.
  • Critique of methodological norms: Philosophical, Moral Practical.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

3 Feminist Critiques of Methodological Norms

A

Philosophical: the way that value has been represented. Historically, the presence of value and therefore bias was not explicitly states. This critique has been made so forcefully that the ‘science is value free’ statement has been refected

Moral: research ethics. don’t objectify research subjects. subjects should be treated as people first. foundational work has been instituted in response to this critique. Smethod saved: working on this one.

Practical (most forceful now): if you are not including women or people of colour in your generalizations your conclusions will be misrepresentative and distorted. Smethod saved: Some Y some N

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

Alternatives to the Scientific Method

A

Experiential-Inductionist Approach

  • Seeks to avoid objectification
  • Minimizes the power relationship between researcher and object
  • Rejects quantification and structured methods
  • Different and interesting results, theories grounded in observations and the object
  • Can you trust research participants? Are they truly self aware? Harder to discern what produces the experience. Still trying to relate to the research question so researcher imparts meaning.
  • Policy makers aren’t as moved by experiential knowledge.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

Concepts

Universal v.s Particular

A
  • Abstractions that do not actually exist but must have clear definitions. Labels for classification of phenomena. universal descriptive word that refers directly or indirectly to soemthing that is observable
  • 2 Key Functions in scientific method:
    1. foundational units of theories. act as building blocks. variation in concepts lead to theories.
    2. data containers. crucial tool for data collection. directs researchers to what to observe.
  • Universal descriptive word: refers to a class of phenomena. *want this one because want to generalize
  • Particular descriptive word: particular instance
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

Generalization

A
  • Replace particular words with broader ones to describe phenomena based on common characteristics.
  • Conceptualizing each case as a member of a class of events about which meaningful generalizations can be made
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

*Types of Definitions for Concepts

A

Real
Nominal
Operational

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

*Real Concepts

A
  • Truth. essential nature or attributes

- Not used in empirical research

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
26
Q

*Nominal Concepts

A
  • Do not use true or false statements.
  • Start with this type of definition in every research project
  • Names the concept and the properties of the phenomena the concept represents.
  • Precedes operational definitions
  • Basic standard use to judge operational definitions
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
27
Q

*Operational Concepts

A
  • Observations or indicators that will be used to represent the concept empirically.
  • More specific
  • Stems directly from nominal definitions
  • If the nominal and operational definitions match then there is congruence and validity is present. Measuring what researcher said they were going to measure.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
28
Q

Knowledge is gained through empirical research and must ultimately be based on?

A

Observation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
29
Q

*Useful Nominal Definitions

A
  1. Clarity: no implicit aspects in definition. define everything explicitly and state assumptions. will fail on inter-subjectivity if implicitly is present, is not transmissible so can’t guard against bias.
  2. Precision: high accuracy, high precision. categorize observations and indicates what should be excluded.
  3. Non-Circular: cannot use the same ideas to define something. tautology: needless repetition of words that are not adding anything to the functional definition.
  4. Positive: cannot define a concept by the attributes it lacks. no negative statements although negative language can be used for clarification or emphasis.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
30
Q

Concepts provide the basis for..

A
  1. Classification: starting point, sorting. defining categories must be exhaustive and mutually exclusive. every observation can only have one category.
    * 2. Comparison: whether we have more/less of something. represent more/less of the concept the researcher is interested in.
  2. Quantification: measuring how much more/less. statistics come in at this point. anything you can count
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
31
Q

Criteria for Evaluating Concepts

A
  1. Empirical importance: must be linkable to observable phenomena. linked either 1. directly observable counterpart (not common), 2. indirectly observable: formed through operational definition but there can be conceptual slippage which results in loss of validity. less congruence between nominal and operational definitions. 3. via relationships. explain variation between concepts
  2. Theory. If concepts cannot be related to other concepts then they cannot be put in to a theory
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
32
Q

In order to have empirical import concepts must be

A

linked to observables

33
Q

When we say concepts provide a basis for comparison what do we mean?

A

concepts help order political phenomena according to whether more or less of something is present

34
Q

What is a theory?

A
  • theories enable the linking of one concept to another by stating the relationship between them. explain why concepts are logically connected
  • if arrived at through deduction, they are called propositions. the relationship is stated before observations
  • if arrived at through induction called empirical generalizations. observations before relationship
  • falsifiable
  • parsimony: simple theories have higher prior probabilities
  • Cannot make theory more restrictive after data collection
35
Q

Theories can be normative or empirical

A

normative: ought
empirical: is in theory. looks for patterns an uses those patterns to explain other patterns

36
Q

*Induction

A
  • Start with an observation and search for patterns.
  • Move from concrete statements about observable phenomena to abstract statements about general relationships.
  • May not use the same data to test the theory that prompted its inception
  • Observation > empirical generalization > hypothesis
37
Q

*Deducation

A
  • Start with logical statements about the relationship between concepts.
  • Theoretical relationships in advance
  • Abstract statements about general relationships to concrete statements about specific behaviors.
  • Axiom > proposition > hypothesis
38
Q

Goals of Scientific Research King, Keohane, Verba

A
  1. Goal is inference, be it descriptive or explanatory on the basis of empirical information collected systematically. infer beyond immediate data to something broader
  2. Procedures are public, explicit and codified so they are replicable. Address limitations.
  3. Conclusions are uncertain
  4. Content it he method and rules not the subject matter..
39
Q

Research Questions King, Keohane, Verba

A
  1. Question should be important in the real world, understanding something that significantly affects many people’s lives
  2. Specific contribution to an identifiable scholarly literature by increasing our collective ability to construct verified scientific explanations of some aspect of the world.
40
Q

Data

A

-Systematically collected elements of information about the world.

41
Q

Which of the following is not a characteristic of a theory?

A

a theory must always include axioms

42
Q

Criteria for Evaluating Competing Theories

A

Tradeoffs between the 5.

  1. Simplicity (parsimony): easier to test and falsify when there are fewer restrictions and a small number of explanatory factors.
  2. Internal Consistency: logically sound. should not contradict itself and should not reach contradictory explanations.
  3. Testability: must be testable. concrete and specific enough to make an empirical observation. if not testable, not falsifiable.
  4. Predictive Accuracy: produce expectations that hold in the real world.
  5. Generalizability: across time and space, applies across time and context.
43
Q

Functions of Theories

A
  1. Explanation: link variation between concepts and explain how and why concepts are linked together.
  2. Organize Knowledge: organize everything that is known and incorporate things that are unknown and how they are logically implied with one another.
  3. Generate new hypothesis: predictive function beyond questions that are motivating the theory initially.
44
Q

Variables

A
  • Is a concept’s empirical counterpart that is not yet an observation.
  • Any property that takes on different values.
  • Cannot be constant, must have variation
  • Must be empirically observable
  • More specific than concepts but every concept represented by lots of variables.
  • No causal relationship between concepts and variables
45
Q

Hypothesis

A
  • conjectural statement of the relationship between two variables. logically implied by a proposition and has clear implications for testing.
  • A statement formed from incomplete information
  • Concepts > multiple variables> logically implied by a proposition that is stated conjecturally in a hypothesis
46
Q

Dependent Variable

A
  • Consequent
  • What we want to explain
  • Values dependent on independent variable
47
Q

Independent Variable

A
  • Antecedent
  • Presumed to explain the dependent variable.
  • If ind. then this is what happens with dep. variable.
48
Q

“The older people are the more they know about politics” Which is the dependent variable?

A

Political knowledge

49
Q

“Female legislator tend to get less media coverage than male legislation” Which is the independent variable?

A

Gender

50
Q

Hypothesis Formulation

A
  • Can be inductive or deductive
  • State a relationship between two variables, specifies how they are related and has clear implications for testing.
  • Null hypothesis means there is no relationship
  • If IV+DV = compare or quantitative (more/less terminology must be present)
51
Q

Forms of Hypothesis Formulation

A
  • If IV + DV = categorical: state which category of the DV is most likely to occur with which category of the IV.
  • If IV= categorical + DV = comparative/quantitative: state which category of the IV will result in more of the DV
  • If IV + comparative/quantitiative + DV = categorical: state which category of he DV most likely to occur when IV increases.
52
Q

7 Most Frequent Mistakes

A
  1. 1 Variable, need a relationship between at least 2.
  2. Not stating the relationship between them, not stating how they are related so it cannot be questioned.
  3. *Incomplete specification: most common. haven’t compared to anything, must make reference categories explicitly.
  4. *Misspecification or Improper specification, most common: comparison using dependent variables. Must introduce/ define relationship using independent variable. lacking comparison
  5. Using “should”. Cannot be normative. Needs to be empirical.
  6. Not generalizable. cannot be applied broadly and generally.
  7. Totalogy. IV and DV not distinct. Talking about the same thing.
53
Q

Why are Hypotheses Important?

A
  • Bridge theory and observation.
  • Theories help us predict and hypotheses tell us what to look for. Type of relationship being tested.
  • Direct investigation
  • Guard against dta mining
  • A priori rationale for the relationship. (justification before observation.
  • Useful when disconfirmed, opens up new avenues for research, cannot turn disconfirming evidence in to confirming evidence.
54
Q

Operationalization

A

-Move from concepts to indicators
-Select observable things to represent abstract concepts.
-Outcome = set of indicators.
Degrees of definition: Concept > proposition > variable> hypothesis> indicator> working hypothesis.

55
Q

Levels of Measurement

A
  • Hierarchical and Cumulative. Stevens Classification (1946): Nominal, Ordinal, Interval, Ratio
  • Measurement: process of assigning numerals to observations according to rules. Can be placeholders or have quantitative meaning if it is given.
  • Must be exhaustive, able to categorize each observation and mutually exclusive, place each in one category.
  • Values of a variable refer to numerals.
  • As the level of measurement increases, the statistical measurement available
  • Depends on choice of data collection procedures.
  • Basis for classifying, comparing and quantifying.
56
Q

Nominal

A
  • Lowest level, do not do arithmetic with them.
  • Naming, puts observations in to two or more categories.
  • Numerals are just labels, have no quantitative meaning.
  • Categories are interchangeable.
  • Do not assign the same numeral to different categories or different numerals to the same category.
57
Q

Ordinal

A
  • Most common for political science
  • Likert Scale
  • All nominal rules apply, ordered hierarchies
  • Natural ordering, cannot say how much or or less
  • Numerals do not hae meaning but are hierarchicical.
  • More/less but not HOW MUCH more/less
  • Not equidistant
58
Q

Interval

A
  • Ordered categories with the same intervals
  • Equidistant
  • Can say how much more or less
  • Arbitrary zero means that you cannot multiply or divide.
59
Q

Ratio

A
  • Mutually exhaustive, ordered, equidistant and meaningful zero-point
  • Anything that is countable
  • Can add, subtract, multiply and divide.
  • Can say twice as much/less
60
Q

Ordinal level measurement allows us to?

A

order observations

61
Q

Highest level political cynicism can be measured?

A

ordinal variable

62
Q

Questions to ask when assessing level of measurement

A
  • Exhaustive and mutually exclusive?
  • Variable dichotomy?
  • Categories ordered?
  • Equidistant?
  • Meaningful zero?
63
Q

Scale Construction

A
  • Treat ordinal like interval.
  • Add scores from category scheme.
  • Approximation of what would you would get if distances were known.
64
Q

Discrete, Continuous, Discrete-Continuous

A

Discrete: specific values distance form other possibilities. bar
Continuous: ordered, no gap in its distribution. bell curve
Discrete-Continuous: well ordered and categories are broken down but gaps in distribution. histograph

65
Q

Social class is an

A

ordinal variable

66
Q

Validity

A

-does our indicator represent our target concept?

67
Q

Reliability

A

-measurement process assign values consistently.

68
Q

Measurement Error

A
  • Difference in values assigned to observations attributable to flaws in the measurement process.
  • Systematic or Random
69
Q

Systematic Measurement Error

A

Indicator is picking up on other property in addition to the one that it is supposed to measure.

  • Systematic biases
  • Problem with validity, not measuring what you are saying you are measuring.
  • Can be reliable but not valid.
70
Q

Random Measurement Error

A
  • Chance fluctuations in measurement do not reflect true differences in property being measured.
  • Affect each observation differently.
  • Necessary but insufficient condition for measurement validity. Affected the same every time in the same way.
71
Q

Content Validity

A

Content: substance. content of the measure match what we think we are measuring.
-Complete and appropriate.
-Face validity and Sampling validity. Have to do with criterion appropriateness.
Face: appropriate indicator of target concept. can a knowledgeable person be persuaded. problem: about persuasion, subjective judgement. no replicable rules for evaluating measure.
Sampling: does the measure represent the full range of meaning of the target concept. is it complete? problem: subjective, no replicable rules, harder to represent the content completely.

72
Q

Criterion Related Validity

A
  • Assess whether the scores produced by an indicator are empirically associated with scores for other variables called criterion variables which are considered direct measures of the phenomenon of concern.
  • Valid indicator if empirical correspondence between results obtained using the indicator and results obtained using another indicator of the same concept that is already known to be valid.
  • Does it match what we already know?
  • Concurrent or Predictive.
  • Correlation between indicator and criterion.
  • Convergent discriminant: indicators of at least two different concepts each measured using at least two different methods. when these indicators are correlated:
  • SEE TABLE
  • Problems: why not use existing indicators. how do we know the standard is valid? few criterion variables available as real measures.
73
Q

Construct Validity

A

-Does our indicator produce the relationships with indicators of other concepts that our theoretical understanding of the target property would lead us to predict.
-Indicators of other concepts must not be among those to test the theory.
AHEM: Assume Hypothesis Evaluate Measure
-Problems: indicator not valid or theoretical framework is flawed.

74
Q

Assessing Reliability

A
  • Test Retest Method
  • Alternative Forms/ Split-Half Method
  • Coefficient Alpha
  • Subsample Method
75
Q

Test Re-test

A
  • Assessing reliability
  • Take measure and use it multiple times to see if yo get similar results.
  • Advantages: non-reactive methods of data collection and content analysis.
  • Disadvantages: not always feasible (in cases of human contact), reactivity, real change in the cases
76
Q

Alternative/Parallel Forms Method/Split-Half

A
  • Parallel forms of the same method.
  • Advantages: no reactivity, no time elapse, no confounding effect from possible changes in the cases, feasibility.
  • Disadvantages: difficulty ensuring that the two forms are truly parallel. difficulty of coming up with two measuring instruments.
  • Split-half: assessed by randomly dividing the times in half and comparing the results.
77
Q

Internal Consistency Alpha Coefficient

A
  • Average correlation for every possible combination of items in to half
  • Low correlation items are deleted
  • 0 is no correlation, 1 is perfectly correlated. 0.4 is the Polisci minimum.
  • Advantages: no reactivity, no time elapse, no confounding effect from possible changes in case feasibility
78
Q

Subsample

A

-Same items are given to each subsample, multiple
-Assessed by similarity of responses across subsample.
-Advantages: no reactivity, no time elapse, no need to come with double the items
Disadvantages: large sample required