Midterm Flashcards
(78 cards)
1
Q
Normative measurments
A
- names not numbers
- have values assigned to them that represent different categories but cannot be ranked in any order
2
Q
ordinal measurments
A
- in additionto values that signify different categories ordinal measuremnts have values that can be ranked
- but distance is unknown or not meaninful between categories
3
Q
interval measuremtns
A
- distances between categories have real meaning
- we can rank and we can say by how much
- e.g events in a year or degrees celcius
4
Q
Is political science a science?
A
- yes because you can use empirical data to explain variables
- political phenomonea can be treated that same as a sciecne using research methods, hypotheseis
- scientific method is used
5
Q
what are non scientific ways of knowing
A
- hunch feeling
- intution
- faith
- personal experience
6
Q
what are some elements of a good research question
A
- specific
- clear direction of research
- examines a relationship between varaibles
- deals with empirical political phenomena
- ‘why’ questions are usuallly good ones
- ‘so what’ test
7
Q
how do we bring different variable together to specify an explanation
A
- definiing the variables and establishing the relationship between them whether they it is a postiive relationship or a negative one
- independent variable
- dependent variable
- antecentdent variable
- intervening variable
8
Q
what is the goal of a hypothesis
A
-seeks to make an educated prediticton of a empirical political phenomon
9
Q
what are characterisitcs of a good hypothesis
A
- seeks to make an educated prediticton of a empirical political phenomon
- expalisn general rather than particular
- can be inductive from particular case
- or deductive from another hypothesis or existing literature
- specific in which it states the direction of the hypothesis
- clear units of analysis (usually 1)
10
Q
what are characteristics of a good theory
A
- large explanatory power
- satisfying our curiosities
- applicable
- uses a little variables as possible
- falsible
- clearly framed with defined concepts
11
Q
what is a strong test of theory
A
- case studies
- statistical analysis
- experiments
12
Q
what issues need to be considered when defining concepts?
A
- clear
- accurate
- precice
- informative
- not abstract or open to interpreation
13
Q
what are units of analysis?
A
-most hypothesis use only one unit of analysis such as individuals, countires, groups, organizations or states
14
Q
frequency distrubution
A
-table that shows number of obersvation for each value or category
15
Q
what are descriptitive statisitsics
A
- gives concise summaries of variables
- describes large amounts of data with one number
16
Q
central tendencie measurments
A
- what is average or typical about the statistics
- mean,median, mode
17
Q
ways of variabliltiy and dispersion
A
-range (intercortile range that is resistant to outlyers
18
Q
what is normal distribution
A
- perfectly symettrical
- mean, median, mode have some numbers
- half the data lies on one side of the median and half on the other
19
Q
what does a squared, skewed right, and skewed right shaped distribution mean?
A
- right: mean higher than median
- left: mean lower than median
- squared: each value has same number of observations
20
Q
what must we consider when making statistical inferences?
A
- even if the sampel was recorder using proper technique and processs there can still be statisitacl probabliilty that the sample is different then the population
- forming judgments of a population and the realibility of the relationships
21
Q
null hypothesis
A
- a hypothesis that the research tries to disprove which is contract to a hypothesis
- a statment that
- that there is nothing interesting going on
22
Q
what are the steps to statisitcal hypothesis testing
A
- null hypothesis
- alternative hypothesis
23
Q
alternative hypothesis
A
- that somthing is interesting going on
- can be direectional (one sided)
24
Q
types of corelation/relationships
A
- postive/negative monotonic
- positive/negative linear
- strong/weak positive relationships
25
what is statistical inference
-taking a sample and making inferences on the base of sample measurements
26
population
-well defined set of objects such as states organization,s countires
27
parameter
-desribes a secific feature of a population
28
sampel statistic
-numerical summary of a sample that corresponds to a population parameter of interest
29
why do we do hypothesis testing
-we do this to be sure that a sample is result of an undlyign phenomoen and not sampling error
30
face validity
-matter of judgement and the measure of concept appears like it was intenended
31
content validity
- full meaning of a concept and making sure the concept and its entiry is iincluded in the measuremnt
- e.g deomcracy is more that just about free elections
32
construct validity
-a measure compared to another that is though to be related
33
ratio measurment
- same as interval except zero actually means zero (absense of a variable)
- e.g 10 is twice as much as 5
- interval and ratio give us the most information about the data
34
What is a causal relatinoshiop and its 3 checks
- one phenomenon causes another in complete regularity
1. covariation
2. time order
3. elimination of altenrative explainations
35
spurious relationship
-a relationship when two variable co-vary because of the involvement of a third
36
probability sample
-each element has a known probability of being included in the sample
37
nonprobbility sample
- each element has a unknown probability of being selected int he sample
38
simple random sample (probability)
-guarentted equal chance of selection (probability)
39
systematic selection (probability)
- generated by a list by a predetermines interval
- start point should be chosen at random
- list should be randomized instead of ranked by a pattern
40
stratified (probability)
- drawn from a population that has been divided into subpopulations
- elements are then selected in accordance to representation in the entire population
41
disproportionate strateified (probability)
- elemetns are drawn disportionaly from the sub populations
- used to overrepresent a group
- cannot generalize about a population
42
cluster (probability)
-identifiying groups in populaiton to sample such as cities, provinceies, blocks
43
convenience (non-probability)
- gerneralizations of alarger populatino of uknown accuracyu
| - drawn from convenience
44
quota (non-probability)
- first identify relevent categories of people
- then decide how many to get in each category
- usually in proportion to representation in the population
45
purposive (non-probability)
- for specialied samples that are especially informative
| - cannot be used for statistical gernalization of a larger population
46
snwoball (non-probability)
- relies on elements in a target population to reach other elements
- useful for hard to locate populations
47
bivariate cross tabs-
- analyzing the relationship of two variables that have in organized in table
- can be used for any measurement but more appropriate for nominal or ordinal variables
- by converting data into different measurement, losing accuracy
48
Measures of Association
- summarizes strenth of a relationship between 2 variables in a single number
- when 1 or more variables is nominal (Lamba)
- when both variables are ordinal: gamma
49
chi square test
-goodness of fitness test, to see if the sample data matches a population
50
difference of means test
- test to see if there is a differnece of means between two categories of the independent variable
- appropriate for inteval or ratio DV
- appropratie for nominal or ordinal IV
51
test of significene
-t test (compare to critical value)
52
Analysis of Variacnce
- ANOVA, analysis for testing difference of means across more than 2 categories of IVs
- appropriate for interval or ratio DV
- nominal or ordiannl IV
- f test of signficance
53
bivariate regression
- tool to describe the realtionship between two interval or ratio variables
- f or t test
54
Pearsons coefficition
-R explains the strength and direction of the correlation between IV and DV
55
R square
- how much variation of the DV we can explain with the IV
| - or how much we can explain using the IV as an indicatior rather than just the DV mean value
56
regression coefficient
-the B value, tell us how much the DV will chance per 1 unit change in the IV
57
Mulivariate Cross Tabs
- technique for analayzing the relationship between the DV and multiple interval or ratio levle IV including control variable organized into a table
- usually used for ordinal or nominal data
58
multivariate regresssion
- tool for describing the realtionship between dv and many IVs
- appropriate for interval or ratio, dicotromy, or ordinal or nomibal recoded into dummy variable of 0, 1
59
f test
- tests the model as a whole
| - tests the null that all coefficient in the regreesion coefficient equaiton equal zero
60
t test
- tests each individual IV
| - tests the null that each partial coefficient in the population equals zero
61
partial regression coefficient
-how much the DV changes with the IV changes by one unit provided all other IVs are held constant
62
standardized regression coefficient
-some researchers use to to tell the relative influnce that the IV has on the DV compared to the other IVs
63
variance inflation factor
-VIF: indicates a problem of mulicollinearity for IV when VIF > 5
64
What kind of guidlines have been established to address research ethics?
-Tri Council Policy Statement: Regarding ethical conduct for for research involving humans
65
What does the Tri-Council Policy Statement include?
- respect for human dignity
- respect for justice and inclusiveness
- balancing harms and benefits (minimizing harm and maximizing benefits)
- respect for free and informed consent
- respect for vulnerable persons
- respect of privacy and confidentially
66
What was the Milgram experiment and what ethical questions did it raise?
- Milgrams obedience to authority experiment: originally to test the how the hollocaust occured by testing human soical pressures to obey authority
- tested people to see how much they would electricute another person
67
What was the Tuskegee experiment and what ethical questions did it raise?
-In Tuskegee Alabama, uneducated African Americans were studied for to research advanced forms of syphillis
=studies continued and patients were not told about having syphillis nor given treatment for the disease
-even after penicilion was found out to be an important cure, they were not given it.
68
What ethical questions arose from the Tuskegee experiment?
- no respect for justice and inclusiveness: discriminated against a specific population
- no respect for vulnerable persons (uneducated)
- high levels of harm
- no level of consent
69
what ethical questions arose from the Milgram experiment?
- high levels of mental harm
- no informed consent
- harms do not outweight he benefits (experiement had already been conducted before )
- no respect for human dignity
70
What does Neuman argue about ethic research
- ethical research rests with the individual researcher
- do no exploit subect for personal gain
- do no coerce or minipulate subjects
- use desception only if needed
71
Percent vs Valid percent
- percnetage of a sample that gave a particular response
| - valid is a percentage of the frequencies on only those population with data included
72
frequency
-number of respondants that gave a particular response
73
empirical research
-resarch based on objective observation in order to achieve scientifc knowledge about a phenomena
74
what is the scienfitifc method?
- objective systematic observation verified through public inspection and methods and results
- goal is to use varificable results to construct a causal relationship that explain a phenomena
75
what consists of the scientific method ?
- explanatory
- prediction
- empirical generalization
- probalistic explaination
76
What is the process of the scientific method?
- research question
- hypothesis formulation
- emirical research
- interpretation and decision
- modify, expand, rebrand
77
type 1 error
- when null is true but still rejected
| - level of significance could be too high
78
type 2 error
- when null is false but failed to be rejected
| - false negative
