Shared Flashcard Set

Details

Elem survey sampling
GWU - Stat 187
35
Mathematics
Not Applicable
01/28/2005

Additional Mathematics Flashcards

 


 

Cards

Term
variance
Definition
measure of variability or measuring the deviation from the mean
Term
Central Limit Theorem
Definition
As the number in the same increases the Sum of all x's across n tends toward normal distribution
Term
Tchebysheff's theorem
Definition
for any k ge 1, at (1-1/k squared) of the measurements in any set must lie with in k SD of their mean
Term
sampling distribution
Definition
closely approximates a normal distribution if n is reasonably large but no more the 20 percent of the total population
Term
2 measures of variable dependence
Definition
1. covariance
2. simple coefficient of linear correlation
Term
coefficient of correlation
Definition
between -1 and 1
Term
purpose of sample surveys
Definition
to estimate numerical characteristics called parameters like mean, total, or variance
Term
Element
Definition
object on which a measurement is take
Ex. a registered voter in community
Term
Population
Definition
a collection of elements about which we wish to make an inference
Ex. collection of voters in a community
Term
sampling units
Definition
nonoverlapping collection of elements from the population that covers the entire population
Ex. households
Term
frame
Definition
list of sampling units
Ex. list of all registered voter's households
Term
sample
Definition
a collection of sampling units drawn from a frame or frames
Term
probability samples
Definition
sample designs based on planned randomness
Term
simple random sampling
Definition
selecting a group of n sampling units in a way that each sample of size n has the same chance of being selected
Term
stratifeid random sample
Definition
dividing the population elements into 2 or more groups (strata) and select a simple random sample from each group
Term
cluster sampling
Definition
divide the population into groups by simple random sample then sample items within the cluster
Term
systematic sample
Definition
selecting every nth person
Term
quota samples
Definition
selecting randomly until reach a limit that matches the population (in all determining factors)
Term
Gallup telephone samples
Definition
unclustered directory-assisted, random-digit telephone samples - proportuate stratified sampling design
Term
"listing" bias
Definition
differences in population between unlisted and listed telephone customers systematicly effects survey results of only listed telephone numbers
Term
Errors of observation
Definition
-when recorded data deviates from the truth
-by sampling coverage or nonresponse due to interviewer, respondent, instrument, or method of data collection
Term
sampling error
Definition
the deviation between an estimate from an ideal sample and the population environment
Term
errors of coverage
Definition
the sampling frame does not matchup perfectly with the target population
Term
nonresponse
Definition
3 ways:
1. inability to contact element
2. inability for respondent to answer
3. refusal to answer
Term
What are the benefits and problems of personal interviews?
Definition
benefits: usually answer questions and able to note reaction and eliminate misunderstandings
problems: cost and trained interviewers
Term
What are the benefits of telephone interviews?
Definition
interviews can be supervised
less cost
random digit dialing (only 20% of numbers are residential)
Term
Benefits and problems of Self-Administered Questionaires
Definition
lowest rate of response
low costs
Term
Direct observations
Definition
usually not studies with people or using medical records
Term
What are some methods to reduce errors in surveys?
Definition
1. Callbacks - use different times
2. rewards to participate - offered only after being randomly selected
3. train interviewers
4. check data
5. carefully planned questionaires
Term
Sample survey design
Definition
the procedure for selecting the sample
Term
simple random sampling
Definition
a type of sampling procedure in which a sample of size n is drawn from a population of size N such that every possible sample of size n has the same chance of being selected
Term
stratified random sample
Definition
seperating the population elements into nonoverlapping groups and then selecting a simple random sample
Term
What are three reasons to stratify?
Definition
1. produce a smaller bound on the error of estimation
2. cost
3. to find information about each strata (domain)
Term
What three factors are used to allocate samples when stratifying?
Definition
1. total number of elements in each stratum
2. the variability of observations within each stratum
3. the cost of obtaining an observations from each stratum
Term
What are the three methods of stratifying?
Definition
1. proprotional
2. Neyman
3. Optimal
Supporting users have an ad free experience!