Shared Flashcard Set

Details

POLI 181 Statistics Exam 1
for the 1st UNC POLI 181 exam: descriptive statistics, bivariate hypothesis tests, theory and causality, experiments
71
Political Studies
Undergraduate 1
10/15/2011

Additional Political Studies Flashcards

 


 

Cards

Term
What are the two big categories of metrics of data?
Definition
Discrete and continuous
Term
What are the sub-categories of discrete?
Definition
Categorical and Ordinal
Term
Categorical variables
Definition
cannot be ranked. e.g., religion, political party affiliation, etc.
Term
Ordinal variables
Definition
relates information about which cases have
similar traits, but allows ranked judgments
Term
Continuous variables
Definition
can be ranked and have equal unit distance across their entire range(e.g. income in dollars- each increase by one dollar means the same thing at every point along the range)
Term
What are two common graphs for categorical variables?
Definition
Pie chart, bar chart
Term
What is the only measure of central tendency for categorical variables?
Definition
The mode- the most frequently occurring category
Term
What are the two major descriptive goals with continuous variables?
Definition
central tendency, dispersion (spread)
Term
What are the two broad classes of descriptive statistics?
Definition
rank statistics, moment-based statistics
Term
What is the measure of central tendency for rank-based statistics?
Definition
The median
Term
What is the measure of dispersion for rank-based statistics?
Definition
The IQR: interquartile range
Term
What is the measure of central tendency for moment-based statistics?
Definition
The mean
Term
Which type of data, rank or moment-based, is more sensitive to outliers?
Definition
Moment-based, because the mean only needs a little bit of an outlier to be thrown off, while the median doesn't break down with outliers
Term
What are the measures of spread for moment-based statistics?
Definition
The variance and the standard deviation
Term
What would happen to the variance if we had no variation in y at all?
Definition
Variance would equal zero- no spread around the mean.
Term
What happens to the variance as data are spread further from the mean?
Definition
The variance increases
Term
What is the standard deviation?
Definition
the average
di erence between values of y and the mean of y
Term
What are the two key dimensions of dependent variables?
Definition
Spatial and temporal
Term
What are two types of research designs (hint: think spatial and temporal)
Definition
Cross-sectional, time series
Term
What is cross-sectional data design?
Definition
Looks at multiple units at one time
Term
What is temporal data design?
Definition
Looks at one unit over time
Term
Why might causality be clearer in a time series?
Definition
we can examine a phenomenon before and after some independent variable changes
Term
What kind of causal theories are common with the physical sciences?
Definition
Deterministic theories: an increase in X by a certain amount will *always* cause an
increase in Y of a certain amount
Term
What kind of causal theories are more common with humans?
Definition
Probabilistic: Increases in X cause increases in Y on average
Term
What is an ecological fallacy?
Definition
inferring individual behavior from
population averages
Term
What is the first hurdle?
Definition
Is there something connecting x and y- does it make sense that they might cause one another in a traceable way?
Term
What is the second hurdle?
Definition
Does y cause x?
Term
What is the third hurdle?
Definition
Do x and y covary?
Term
What is the fourth hurdle?
Definition
Can we eliminate any Z's that might
relate to X and Y and cause Y ?
Term
Finish this phrase: "While correlation does not mean causation,"
Definition
Correlation is necessary for causation to exist.
Term
What is an experiment?
Definition
a research design in which the
researcher both controls and randomly assigns values of the
treatment (key independent variable) to participants
Term
What are the two key components of experiments?
Definition
control, random assignment
Term
What is a control?
Definition
The value(s) of the treatment (key independent variable) X are determined by the researcher and not by the participants or nature.
Term
What two groups comprise the experimental group?
Definition
Treatment group, control group
Term
What does randomization control for?
Definition
for every possible Z regardless of whether we can even list the possible Zs
Term
Experiments have good ____________ validity but bad ____________ validity.
Definition
Internal, external
Term
What is internal validity?
Definition
The extent to which we can accurately state that the observed independent variable produced the observed effect
Term
External validity
Definition
relates to the generalizability of your findings
Term
What can help with the idea that experiments are low on external validity?
Definition
replication
Term
What is an observational study?
Definition
one in which the researcher does not have control over the quantities of the independent variable (or any variable)
Term
What is the population of interest in an experiment?
Definition
the set of units (people, countries, etc. . . ) that the researcher's theory relates to
Term
What are the two ways to get observational data?
Definition
Measure the entire population of interest, measure a sample of the population of interest
Term
What are the two types of data in an experiment?
Definition
Population data, sample data
Term
What is population data in an experiment?
Definition
data about every possible relevant case
Term
What is sample data in an experiment?
Definition
a dataset drawn from a subset of cases of some underlying population
Term
What is a random sample?
Definition
De ned by the characteristic that every
member of the population of interest has an equal probability of being selected for participation in the study.
Term
What is a convenience sample?
Definition
Use participants that are readily at hand
Term
What is the difference between random assignment and random sampling?
Definition
Random assignment refers to the decision about whom to give a potential treatment. We randomly assign people to the treatment group from the larger experimental group.

Random sampling refers to drawing at random of a sample to study. Usually done (or attempted) in survey research.
Term
Without experimental control and random assignment, crossing hurdles ___ and ___ of causal evaluation is difficult.
Definition
2, 4
Term
What kind of test do you perform if both variables are categorical?
Definition
Tabular analysis
Term
What kind of test do you perform if both variables are continuous?
Definition
Correlation coefficient (Pearson's r)
Term
What kind of test do you perform if the dependent variable is continuous and the independent variable is categorical?
Definition
Difference of means
Term
What kind of test do you perform if the dependent variable is categorical and the independent variable is continuous?
Definition
N/A
Term
What do p values range between?
Definition
0 and 1
Term
What does the p value show?
Definition
the probability of randomly finding a
relationship in the sample that does not exist in the population
Term
As the p value approaches ____ we get more confidence that there is a real relationship between the two variables in the population
Definition
Term
The more data we have, the _____ our p values will be.
Definition
Lower
Term
If a p value is less than ____, the relationship is said to be _______ ________.
Definition
0.05, statistically significant
Term
What are the three steps to trying to interpret cross tabs?
Definition
1. Figure out what defines the rows and columns
2. Figure out what each cell tells you
3. Look for general patterns
Term
When is Pearson's chi squared (x^2) statistic used? Write it down.
Definition
Used to find the relationship between two categorical variables.
Term
How do you calculate the degrees of freedom for the chi-squared statistic?
Definition
df = (r - 1)(c - 1) where r is the number of rows in your table and c is the number of columns
Term
When do you use a t-test? Write down the formula.
Definition
When the dependent variable is continuous and the independent variable is discrete.
Term
What does the numerator of the t-test formula tell you?
Definition
the greater the di erence between the means, the higher the value of t will be
Term
The denominator of the t test requires the _____ ______ of the difference of the two means. Write down how this is calculated
Definition
Standard error
Term
How do you calculate degrees of freedom for the t-test?
Definition
Subtract one from the smaller of the two n's
Term
When do you use Pearson's r (correlation coefficient?)
Definition
When both the independent and dependent variables are continuous
Term
What does covariance mean?
Definition
That the variables change together
Term
How do you calculate the covariance between two variables? Write it down.
Definition
N/A
Term
If ___ is systematically higher than its mean for the same observations in which ___ is higher than its mean, we'll get a ________ contribution to the covariance.
Definition
x, y, positive
Term
How is Pearson's r calculated? Write it down.
Definition
N/A
Term
What t-statistic is used to determine whether two continuous variables have a higher correlation than we would expect at random? Write it down. How are degrees of freedom calculated?
Definition
df = n-1
Supporting users have an ad free experience!