Text Ch.4 Flashcards
endogenous
exogenous
variations in a concept are based on, or depend, on variations in other elements inside the model
variations in the independent concepts are hypothesized to depend upon factors outside the model
T or F
In most cases, statistical techniques can assess the strength of association, not the direction of causation
T
categorical vs. continuous concept/variable
.is one in which the concepts characterisitcs are separate and distinct
-a level of measurement in which variables are given numerical values(1,2,3,etc) that represent a difference in kind rather than in degree(ex: 2 is not higher than 1 but simply different)
.is one in which the categories are joined or connected in a sequential manner
-a variable whose responses can be arrayed along the entire range of the measure, with no gaps or breaks in the distribution
multidimensional concept
is one in which more than one factor or dimension, exists with a concept
Partisanship
is defined as a psychological attachment to a political party , which can vary along the dimensions of strength, stability, and consistency of attachment
Canadian Election Studies(CES)
a series of academic surveys conducted on a sample of the Canadian electorate at the time of federal elections from 1965 to the present-day provide a rich source of data on political beliefs and voting behaviour
Inductive Reasoning
using empirical evidence to help for the definition of a concept
extrapolation
.borrow from other fields of study
deductive reasoning
.a type of reasoning that shows or attempts to show that a conclusion necessarily flows from a set of premises; in political science, often used in rational choice analysis, where individuals are assumed to be utility maximizers and their actions are assessed in relation to what they would be if this premise was true
-begins by the identification of one or more postulates, followed by the derivation of expectations and conclusions based on the claims
Explain:
Internal Validity:
External Validity:
Measurement Validity:
- internal validity:the extent to which the researcher has produced results reflective of reality, as measured within the confines of the study. Concerned with causality.
- external validity:the extent to which the findings drawn from the cases under examination may be used to make generalizations about phenomena outside the original study. Concerned with representativeness.
- measurement (construct) validity:the extent to which the measurement of a particular concept matches its operational definition. Prerequisite for internal and external validity.
Explain:
Internal Validity:
External Validity:
Measurement(construct) Validity:
- internal validity:the extent to which the researcher has produced results reflective of reality, as measured within the confines of the study. Concerned with causality.
- external validity:the extent to which the findings drawn from the cases under examination may be used to make generalizations about phenomena outside the original study. Concerned with representativeness.
- measurement (construct) validity:the extent to which the measurement of a particular concept matches its operational definition. Prerequisite for internal and external validity.
Define Theory
3 things theories try and state
Theory: system of ideas; condenses and organizes knowledge
- That events are related
- How events are related
- Why events are related
What is Theory oriented researcher’s goal?
Goal: identify causal relationships, build/support/advance theory
define concept
define hypothesis
- Concept = abstract representation of phenomena
* Hypothesis = statement of anticipated relationship between concepts
How is a hypothesis important to the scientific method?
- Concept = abstract representation of phenomena
* Hypothesis = statement of anticipated relationship between concepts
How is a hypothesis important to the scientific method?
“the key characteristic of a scientific approach to understanding the world is to use observations to test whether a prediction, or an interrelated set of predictions about how the world works, actually holds up when subjected to empirical scrutiny.”
3 Characterisitcs of a good hypothesis
- Clear statement that relationship exists – Correlation
- Statement of how concepts are related – What is the effect? – Which comes first? (temporal order)
- Falsifiable – i.e., can be proven wrong
It is important to figure out whether there is a ________ between the two concepts observed in the data that is either due to a relationship or random luck
correlation
4 errors in causal reasoning!!!
- Assuming a relationship means causality
- Need to think through plausibility
- does one really cause the other? - Mistaken Temporal Order
- Failing to Identify Spurious Relationship
- Inability to Reproduce Results, not consistent
How are concepts related? What is there relationship?
1.Positive relationship: both concepts move in same direction
E.g., increased political knowledge and increased political interest; decreased political knowledge and decreased political interest
2.Negative relationship: concepts move in opposite direction
E.g., increased political knowledge and decreased political interest; decreased political knowledge and increased political interest
3.Patterns
E.g., Group A has greater political knowledge than Group B
- What causesparents to choose to vaccinate or not vaccinate their children against preventable diseases?
- Why do parents choose to vaccinate/ not vaccinate their children against preventable diseases?
Which of these 2 questions would be harder to complete?
- Explanatory
- looking to explain
2.Descriptive
Question 1 because it entails a personal aspect, qualitative
What are the 5 criteria for causal arguments?
- Correlation
- Plausible
- Temporal Order
- Not Spurious
- Consistent(reproductibility)
What is the post-hoc fallacy?
Errors in Causal Reasoning:
Assuming Temporal Order Means Causality
Explain difference between a Null and Alternative Hypothesis?
Null:
• Any pattern between the two concepts observed in data is due to chance
• “Nothing going on”
• Default assumption
Alternative:
• Pattern between the two concepts observed in data is not due to chance
• “Something going on”
• Burden of proof is on Ha: must gather evidence
• Reject/fail to reject H0 – Metaphor: guilty/not guilty. Not guilty ≠ innocent.
Give a Null and Alternative hypothesis for:
Political Interest–>Political Knowledge
- Null hypothesis: there is no relationship between political interest and political knowledge.
- Alternative hypothesis: as political interest increases, political knowledge increases
____ Hypothesis: There is no relationship between false information and vaccination.
____ Hypo: False information increases anti vaccination rates
What is wrong with the second one?
Null
Alternative
- needs to give type of relationship
- as false information increases, vaccination decreases(negative)
3 characteristics of useful hypothesis
Characterisitcs of a useful hypothesis:
- correlation-the things discussed have to have a relationship in the first place
- either no relationship or relationship “Nothing going on” or “Something Going On” - how are they related, which comes first in temporal order?
- effect: what is the direction of movement
- how are concepts are related in terms of temporal order: for A to cause B it must come first - Falsifiable?
-in order for a hypothesis to be useful there has to be a scenario where we can prove it wrong!
-by empirically testing theories we move from the abstract to concrete
Ex: Student effort-Acedemic performance vs. Hours Studying-GPA
Possible explanantions for correlation:
A caused B
B caused A
Or
Something else caused it: C causing A and B’s relationship
-this is called a ______ variable
Ex: ice cream sales and rioting go up at the same time, why? Heat wave, 3rd variable
confounding
confounding variable
extraneous variable that affects both of the correlated variables and makes it seem like there is a relationship between them.
What type of validity does the causal criteria work for?
Internal Validity
Observational Vs. Experimental Designs
Observational:
• Data collection does not directly interfere with (manipulate) how the data arise.
• Can only be used to establish association (correlation)
–its hard to determine temporal order and whether it is spurious
Experimental
• Subjects are randomly assigned to groups (e.g., treatment, control)
• Data collection directly interferes with (manipulates) how the data arise.
• Used to establish causality
– Temporal order
– Control confounding variables
