exam 3 Flashcards
non probability sampling
a. always produces samples that possess distorted characteristics relative to the population
b. denies the researcher the use of statistical theory to estimate the probability of correct inferences
c. should never be used under any circumstances
d. includes stratified sampling
e. requires the use of sampling frames
b. denies the researcher the use of statistical theory to estimate the probability of correct inferences
You are doing research on hospital personnel—orderlies, technicians, nurses, and doctors. You want to be sure you draw a sample that has cases in each of the personnel categories. You want to use probability sampling. An appropriate strategy would be
a. simple random sampling
b. quota sampling
c. cluster sampling
d. stratified sampling
e. accidental sampling
d. stratified sampling
Stratifying a population prior to drawing a sample
a. generally occurs when the variables used to stratify are known to be associated with the dependent variable
b. eliminates the need for simple random sampling
c. is most useful for studying a homogeneous population
d. eliminates the need for probability sampling
e. is an alternative systematic sampling
a. generally occurs when the variables used to stratify are known to be associated with the dependent variable
The unit about which information is collected and that provides the basis of analysis is called a(n)
a. universe
b. sampling unit
c. unit of analysis
d. sampling frame
e. element
c. unit of analysis
A sampling interval of 5 was used to select a sample from a population of 1000. How many elements are to be in the sample?
a. 5
b. 50
c. 100
d. 200
e. 1000
d. 200
You want to examine the relationship between family size and family cohesion. You use as your sample all the students in your research methods class. What kind of sampling design are you using?
a. simple random sampling
b. quota sampling
c. snowball sampling
d. stratified sampling
e. convenience
e. convenience
Every key element in a list is chosen for inclusion in the sample in
a. simple random sampling
b. systematic sampling
c. disproportionate sampling
d. cluster sampling
e. stratified sampling
b. systematic sampling
The standard error tells us how closely the population parameter is clustered around a single sample statistic. (T/F?)
true
In a sample stratified by gender, the sampling error on this variable is reduced to zero. (T/F?)
true
A confidence interval at the 68% confidence level will be larger than one constructed at the 95% confidence level. (T/F?)
false
Generally, the more heterogeneous the population, the more beneficial it is to use stratified sampling. (T/F?)
true
Sampling error is reduced through an increase in the sample size and an increased homogeneity of the elements being sampled. (T/F?)
true
Kenny performed an experiment on the effects of after‐school educational activities on students’ academic performance. His experimental group involved after‐school educational activities and his control group involved after‐school non‐educational activities. The majority of the children in the control group left the activities. Which source of internal invalidity is reflected?
a. Instrumentation
b. Testing
c. Statistical regression
d. History
e. Attrition
e. Attrition
Feminist performed a study in which he introduced a stimulus (longer coffee breaks) and then measured how often employees left early (the dependent variable). No pretests were done. Which design did he use?
a. Solomon four‐group design
b. Double‐blind design
c. Static group comparison
d. One‐shot case study
e. Classical experimental design
d. One‐shot case study
Lee selected people with only the highest self‐esteem scores for her experiment on the effects of divorce on self‐esteem. She should be particularly alert to which potential source of internal validity?
a. History
b. Maturation
c. Testing
d. Statistical regression
e. Attrition
d. Statistical regression
Dr. Koo did an experiment on children in a classroom. She measured their social anxiety on Monday, randomly assigned half of them to be taught yoga on Wednesday, and measured their social anxiety again on Friday. The half of the children who were not taught yoga are known as the:
a. Independent variable
b. Experimental group
c. Pretest
d. Control group
e. Posttest
d. Control group
Dr. Koo did an experiment on children in a classroom. She measured their social anxiety on Monday, randomly assigned half of them to be taught yoga on Wednesday, and measured their social anxiety again on
Friday. Measuring the children’s social anxiety on Friday is the :
a. Independent variable
b. Experimental group
c. Pretest
d. Control group
e. Posttest
f. Dependent variable
g. Posttest and dependent variable
h. Posttest and independent variable
g. Posttest and dependent variable
- The greatest strength of laboratory experiments lies in the ability to examine numerous variables simultaneously (T/F).
false
what is the classical experiment
specific way of structuring research:
3 major components
1. independent and dependent variables
- pretesting and posttesting
- experimental and control groups
2. pretest/posttest
- subjects are initially measured in terms of the DV prior to association with IV (pretested)
- following exposed to IV
- remeasures in terms of DV (posttest)
- differences noted between measurements on the DV are attributed to influence of IV
3. experimental and control groups
- experimental: exposed to whatever treatment, policy, initiative we are testing
- control group: very similar to experimental group, except that they are NOT exposed
- if we see a difference, we want to make sure it is due to the IV, and not to a difference the two groups
how do we use independent and dependent variables
- pretesting and posttesting
- experimental and control groups
traditional use of experimental and control groups
- experimental: exposed to whatever treatment, policy, initiative we are testing
- control group: very similar to experimental group, except that they are NOT exposed
randomization in subject selection
- insures that the experimental and control groups are alike in every way (except for the intervention)
instrumentation (threat to internal)
changes in measurement process
testing (threat to internal)
subject reactions to testing (retesting) itself
maturation (threat to internal)
subject changes/grows during course of study
selection bias (threat to internal)
lack of randomization
- to assignment or subject
-the way the subject are chosen
statistical regression to the mean (threat to internal)
natural return to the average
experimental mortality (threat to internal)
attrition
Participants frequently drop out of experiments whilst they are taking place/before they finish; something that is known as experimental mortality (or experimental attrition).
history of (threat to internal validity)
unpredicatable/ external events may occur during the course of the experiment affecting DV
internal validity
- potential for rival factors
threats: - history
- maturation
- testing
- instrumentation
- statistical regression
- selection bias
- experimental mortality
- causal time order
external validity
- ask the question of whether findings from the current study apply to other populations, times and or other geographic areas
-issue of generalizability - for experiments under heavily controlled conditions rather than natural conditions (reduces internal threats)
construct validity
- asks the question of how well the experiment reflects the causal process In reality
- an issue of generalizability
- can we generalize from this finding in spite of specific measurement concerns
statistical conclusion
- ask the question is the N big enough to support stats (associations)
quasi experimental design factors
- when randomization is not possible
- quasi = “to a certain degree”
uses of experiments in cj
written documents
prior criminal record
- usually see else (besides researcher) collects it for another purpose
- product of direct observation or survey
diret observation
no interaction with subjects
survey
ask questions (includes focus group)
many times indirect measures
direct interaction with subjects
what are the three ways if collecting data
asking questions, direct observation and examining written records.
what does the purpose of the research dictate
decisions abut data collection
what does probability sampling do for the researcher?
helps generalize back to the pop of unobserved cases
probability smapling
each unit in the population has a known chance of selection into the sample
- makes it possible for us to make predictions about the larger unobserved population based on the sample
what does sampling mean?
- selecting units from population to represent that entire pop
- we generalize our finding not only too smaple but entire pop
what is a representative sample ?
sample means that the aggregate characteristics of the sample closely approximate those same aggregate characteristics in the population
sample statistic
summary description of given variable in a sample; we use sample statistics to make estimates or inferences of population parameters
population parameter
summary description of given variable in a population
ex. Median income of student: 10,000
population
specified grouping of study element; whole group
ex. csuf student
sampling frame
list of sampling elements in the population (actual list of units to be selected)
- names of all 400,000 students
sample element
unit about which info is collected and provides for the basis of the analysis; who or what we are studying
ex. student
what are the advantages of probability pf selection method
- minimizes bias of sample
- allows us to estimate the representativeness of the sample to the population
do we want our sample to be homogenous or heterogeneous ?
heterogeneous; we are a society that differs in races
what is a biased sample
sample mens that the sample is not typical of the pop that it is claimed to represent
posttest
an assessment carried out after the application of some intervention, treatment, or other condition to measure any changes that have occurred.
pretest
tested on members of target population/study population, to evaluate the reliability and validity of the survey instruments prior to their final distribution.
examples of conscious sampling bias
Presidential election voters. If you poll 1000 middle-class, blue collar voters, the sample will be heavily biased because it won’t be diverse enough to paint the whole picture.
what is the basic principle of probability sampling
all members of the population have an equal chance of being selected for the sample
what is a sample distribution
range of sample statistics obtained from many samples.
what is a bell shaped curve
- bell curve is a symmetric curve centered around the mean or average of all the data points being measured
- set of chosen values across a specified group that tend to have a central, normal values, as peak with low and high extremes tapering off relatively symmetrically on either side
what is a normal curve
- symmetrical bell shaped curve repressing the probability density function of a normal distribution
what is the roll of standard error
used to estimate the efficiency, accuracy and consistency of a sample
standard error
a measure of sampling error; we can estimate the degree to be expected
-indicates how closely the sample estimate reflects the population parameter
how do you decrease the presence of standard error
- having a large sample size, decreases sample error
what are confidence intervals and how are they important
confidence lvs: express the accuracy of our sample statistics in terms of lv of confidence that stats fall in
confidence intervals: what the statistics fall within the specified interval from the parameter
99% confidence
+ or - 3 standard error
95% confidence
+ or - 2 standard error
68% confidence interval
+ or - 1 standard error
snowball sample
- (non prob)
- contact of an initial subject to informant who will then refer other subjects pr informants
- done often with criminals, subcultures, hard to reach and hidden populations
ex. study of persistent thieves. have contact of several thieves and use them to gain access to several more
convenience sample
- non prob
- researcher selects sample based on convenience
ex. csuf stop and ask to answer questions
purposive
- non prob
- researcher or another “expert” selects the sample elements done based on some type of criteria or characteristics based on purpose of study
ex. study of gang behavior in prison - select 3 of 7 possible gangs be they perform certain characteristics
non probability designs
- purpose or judgement sample
- convenience
- Snowball
disproportionate stratifed
- probability
- purpose of sampling technique is to ensure that sample has higher rep of sample elements based on arable of interest
-when a characteristic of pop is not that common - sample is purposefully created to include more of the unusual characteristics
ex . large supermarket may account for 20% of all grocery stores but account for 80% of grocery sales - need a disportioncate sample to represent the large supermarket to reflect the 80% rather than number of stores
stratified sampling
- probability
- can lower standard error
- you select certain # from homogenous subsets from pop sample is stratified on certain variable and represent of variable in sample will be improved
- ex race, gender, age
ex freshman, sophomore, junior , senior, grad
+ 40,000 students - choosing someone from each category to get the right representative
- grouping from each one lowers the probability of standard error
systematic sample
- probability
- no attached to # only the “list”
- ex of 40,000 students
- just uses the order of list
- every 2nd element
Cody
Alex +
jody
Adrain + and so on - however may be a problem with certain characteristic such as ethnicity or alphabet order
ex: will always get black
Hispanic
white
black +
hispanic
white
black +
simple random
- there is first sampling Frame on this a # is assigned of list
ex. 40,00 students listing each name of student - then assign numbers and randomly select the sample from the list
what are the different types of probability sampling designs
- simple random
- systematic random
- stratified
- disproportioate
what is the difference between sampling designs used by the National Crime Vic Survey and British Crime Survey
the NCVS relies on a sample rather than a census of the entire U.S. population, weights
difference between probability and non probability sampling techniques
- probability each unit in the pop has a known chance of selection into the sample; involving random selection
- non random selection baed on convince or other criteria; main purpose is to collect data easy
ambiguous causal time order
the dependent variable actually caused the change in the stimulus
experimental mortality (threat to internal)
attrition
Participants frequently drop out of experiments whilst they are taking place/before they finish;
sample statistic
summary description of given variable in a sample; we use sample statistics to make estimates or inferences of population parameters
sample statistic
summary description of given variable in a sample; we use sample statistics to make estimates or inferences of population parameters
posttest
tested on members of target population/study population, to evaluate the reliability and validity of the survey instruments prior to their final distribution.
