CJUS 3101 Final Flashcards
Sampling Frame
Probability Sample
Equal chance of selection
Probability Sample
Random Chance or selection
Probability Sample
Generalizability
Probability Sample
Unknown chance of selection
Non-Probability Sample
No sampling frame
Non-Probability Sample
Non-random selection - selected for a reason
Non-Probability Sample
Non-generalizable
Non-Probability Sample
All of the individuals or cases we are interested in studying
Population
A subset or smaller group selected from our population
Sample
The process of applying the results we get from our sample to the population
Generalization
This is what sampling sets out to achieve
Representative Sample
Boundaries of the population being sampled
Parameters
All of the people or elements of the population have an equal chance of being selected for the sample
Probability Sampling
Gold standard of sampling, everyone in the population has an equal and independent chance of being selected for the sample
Simple random sampling
a list of everyone or object in the population
Sampling Frame
A modified form of SRS. Instead of the selection being random there is a patter
Systematic Sampling
- Define target population
- Determine the desired sample size
Sampling Interval
How do you find sampling interval?
Divide your sample size into the number of persons by the number of persons in your sampling frame
- If you had 100 persons in your sampling frame and wanted to sample 20 persons
100
—– = 5
20
Use this type of sampling when the variable is rare and would be hard to obtain through SRS
Stratified Sampling
The population is divided into subgroups and you select from each of the subgroups
Stratified Sampling
This type of sampling is utilized in situations when you cannot construct a sampling frame because it is to difficult.
Cluster Sampling
Use this type of sampling when you have more than one sampling frame of clusters
Multistage cluster sampling
as our sample size increases the sampling distribution will equal the mean of the population distribution
Central Limit Theorem
The standard deviation of the sampling distribution
Standard Error
The range of values that probably include the real value
Confidence Interval
a * denotes or means that our estimate is probably correct and not due to chance or sampling error.
Statistical Significance
This is when we conclude that our statistic represents the population when in fact it does not.
Type 1 Error
The elements do not have an equal chance of being selected
Non-Probability Sampling
This is where you simply use subjects that are easy or convienent for you to use
Convenience Sampling
This is where the researcher selects individuals based on the researchers knowledge or professional judgement
Judgmental Sampling
This is a method typically used with unknown or rare populations that are difficult to locate, situations where it would be impossible to construct a sampling frame.
Snowball Sampling
The way this technique works is that you find one or more subjects that you want to study and ask them to refer you to others.
Snowball Sampling
Is an association between two variables. they occur together
Correlation
A and B tend to be observed at the same time
Correlation
You need to have an association in order to argue for this
Causality
A causes B
Causation
What are the 3 criteria to say that two variables have a casual relationship
- The two variables must be correlated.
- Time ordering
- Non - Spuriousness
A situation where it looks like two variables are related to one another, but in fact they are not
Spuriousness
gives us a way of summarizing the differences so that we can compare the strengths of different tables or variables
Strength Measure
measure the strength of the association between two nominal and ordinal variables
Crosstabulation Table
Something that is required or must be present in order for the effect to occur. A must occur if B is to occur.
Necessary Cause
Guarantees the effect will occur
Sufficient Cause
This type of cause increases the chances that something will happen
Contributory cause
If x does not happen, y will not happen
Necessay Cause
If x happens, Y will happen
Sufficient cause
If x happens, Y may happen or is more or less likely to happen
Contributory Cause
An experiment where you have a group that receives treatment and one that does not and see what happens
Classical Experiment
The independent variable in a classical experiment
The treatment
The group that receives treatment is the
experimental group
The group that does not receive treatment
control group
This is where we collect baseline data
pretest
this is where we measure the dependent variable after treatment
posttest
In order for the design to work it is crucially important that your experimental and control groups are equivalent at the beginning of the experiment. How do you ensure this?
Random Assignment
threats or challenges to our findings and to our potential conclusions that a causal relationship may or may not exist
Internal Validity
These are events that happen during the course of the experiment. These are things that occur outside of the experiment
History
These are changes that take place in the subjects during the course of the experiment. This includes biological, psychological, and physiological changes.
Maturation
Refers to the tendency of subjects to move toward the average over time
statistical regression
This is really just attrition or persons dropping out of the experiment
Experimental mortality
refers to the measures of your variables; can refer to when your measures lack reliability or validity
instrumentation
This is where the subjects learn through their exposure to the pretest learn how to do better on the post test
Testing Effects
This is where the treatment spills over from the experimental group to the control group could be communication
Design Contamination
assessment of wether our results are applicable to other groups
External Validity
situation where the subjects may be exposed to more than one treatment or independent variable
Multiple treatment influence
refers to subjects reacting to the fact that they are in an experiment or research project
reactive effects
this is where you try your best to get a similar control group but they are not randomly assigned
non-equivalent control group design
this is where we observe our dependent variable over time both before the independent variable is introduced
interrupted time series design
surveys that survey respondents at one point in time
Cross - Sectional
surveys conducted over time
longitudinal
these collect information on a variable over time
trend study
this study follows a group that shares the same characteristic over time
cohort study
study that follow the same individuals over time and collect data at intervals
panel study
Levels of measurement
NOIR
- Nominal
- Ordinal
- Interval Ratio