Term
| sociologists WHAT from a population to discover and understand social phenomena. |
|
Definition
|
|
Term
| Two famous examples of sampling illustrate the importance of proper sampling techniques: |
|
Definition
1. 1936 Literary Digest presidential poll
2. 1948 Gallup presidential poll |
|
|
Term
| 1936 Literary Digest presidential poll |
|
Definition
Editors mailed postcards to people in six states, names were selected from telephone directories & automobile registration lists.
Doing so created a sampling frame problem: they only sampled the rich, and therefore did not have an accurate sample to predict the election. |
|
|
Term
| 1948 Gallup presidential poll |
|
Definition
The Gallup samplers used a method called quota sampling, which selects people to match a set of characteristics (e.g. the poor, those living in rural vs. urban environments).
Predictions were based on old census data: the quotas they used were no longer valid indicators of the U.S. population, and Gallup incorrectly predicted Thomas Dewey defeating Harry Truman |
|
|
Term
There are two types of sampling
methods: |
|
Definition
1. Nonprobability sampling
2. Probability sampling |
|
|
Term
|
Definition
| represents techniques where samples are not selected by using probability theory. |
|
|
Term
|
Definition
| are selected according to some sort of random assignment. |
|
|
Term
There are four types of nonprobability
sampling: |
|
Definition
1. Reliance on available subjects
2. Purposive (or judgmental) sampling
3. Snowball sampling
4. Quota sampling |
|
|
Term
| Reliance on available subjects (nonprobability sampling) |
|
Definition
(also called convenience or haphazard sampling) involves using people in close proximity, or people who are gathered in one place at one time.
It is generally considered one of the weaker forms of sampling.
Strengths: easy and efficient
Weaknesses: can be unrepresentative of general populations |
|
|
Term
| Purposive (or judgmental) sampling (nonprobability sampling) |
|
Definition
units to be observed are selected on the basis of a researcher's judgment about which ones will be
most useful or representative.
One may want to interview the entire population of some limited group (e.g. directors of shelters for homeless adults). |
|
|
Term
Three guidelines should be followed when
selecting a purposive sample: |
|
Definition
1. Informants should be knowledgeable about the cultural arena, situation, or experience being studied.
2. Informants should be willing to talk.
3. Informants should be representative of the range of points of view. |
|
|
Term
| In purposive sampling, interviews should be selected until two criteria are met: |
|
Definition
1. Completeness
2. Saturation |
|
|
Term
|
Definition
| Interviews should continue until a researcher is confident subjects have provided an overall sense of the meaning of a concept, theme, or process. |
|
|
Term
|
Definition
| Interviews should continue until one is confident that they are learning little that is new from subsequent interviews. |
|
|
Term
|
Definition
| a nonprobability sampling method whereby each person interviewed is asked to suggest additional people for interviewing. |
|
|
Term
| Snowball samples are appropriate for |
|
Definition
| studying difficult to identify or difficult to locate populations |
|
|
Term
Snowball sampling can be problematic
for the following reasons: |
|
Definition
1. It can result in samples with questionable representativeness.
2. Initial contacts may shape the entire sample and foreclose access to some members of the population of interest. |
|
|
Term
|
Definition
is a type of nonprobability sampling in which units are selected into a
sample on the basis of pre-specified
characteristics, so that the total sample will
have the same distribution of characteristics assumed to exist in the population being studied. |
|
|
Term
| Quota sampling begins with |
|
Definition
| a matrix that describes characteristics of a target population. |
|
|
Term
Quota sampling is similar to probability
sampling, but has two inherent issues: |
|
Definition
1. Quota frame problems
2. Selection of sample elements may be
biased |
|
|
Term
|
Definition
▪ Proportions that different cells in a matrix represent must be accurate.
▪ It is often difficult to get up-to-date information for this purpose. |
|
|
Term
Selection of sample elements may be
biased. |
|
Definition
| E.g. an interviewer may be instructed to interview 5 people who meet a given set of characteristics, yet still avoid certain people that are representative of the issue at hand. |
|
|
Term
In all research designs, sociologists
distinguish between WHAT and
informants. |
|
Definition
|
|
Term
|
Definition
| allow a researcher to construct a composite picture of the group respondents represent. |
|
|
Term
|
Definition
| s are people well versed in a social phenomenon one wishes to study and who are willing to tell what they know about it. |
|
|
Term
|
Definition
| is a general term for samples selected in accord with probability theory. |
|
|
Term
| Probability samples are often used for |
|
Definition
|
|
Term
Probability sampling can be a very
effective tool in research if... |
|
Definition
|
|
Term
| Nonprobability sampling cannot |
|
Definition
guarantee that a sample is
representative of a population. |
|
|
Term
| Probability sampling is useful |
|
Definition
because it helps ensure that the same
variations that exist in a population
are represented in a sample. |
|
|
Term
|
Definition
| occurs when subjects selected for a study are not typical nor representative of a larger population. |
|
|
Term
| Bias is often WHAT. When one selects a sample based on some background characteristic, they ALWAYS introduce some sort of bias into their sample. |
|
Definition
|
|
Term
| Some degrees of bias are more acceptable than others |
|
Definition
| depending on one’s research design |
|
|
Term
Samples must be WHAT of the
population from which they are selected |
|
Definition
|
|
Term
|
Definition
refers to the quality of a sample of
having the same distribution of characteristics as the population from which it was selected. |
|
|
Term
| A sample is representative of the population from which it is selected |
|
Definition
if all members of the population
have an equal chance of being selected in the sample. |
|
|
Term
There are two advantages to
probability sampling: |
|
Definition
1. Probability samples are typically more representative than other types of samples because biases are avoided.
2. Probability theory permits researchers to estimate the accuracy or representativeness of a sample. |
|
|
Term
In order to understand probability
sampling one needs to differentiate
WHAT and populations: |
|
Definition
|
|
Term
|
Definition
A unit of which a population
is composed and which is selected in a
sample. |
|
|
Term
|
Definition
A specified aggregation of
elements in a study. |
|
|
Term
In probability sampling, one uses
random selection |
|
Definition
| to select a set of elements from a population that accurately portray the total population from which the elements are selected. |
|
|
Term
Random selection ensures that each
element has an |
|
Definition
| equal chance of selection. |
|
|
Term
|
Definition
| provides the basis for estimating parameters of a population. |
|
|
Term
|
Definition
is the summary description of a given variable in a population.
E.g. the mean unemployment rate or age distribution in L.A. are both parameters of the L.A. population. |
|
|
Term
| When researchers generalize from a sample, they use |
|
Definition
sample statistics to estimate population
parameters. |
|
|
Term
An action that involves elements of
probability is called an |
|
Definition
event or trial.
ie
▪ Dealing a card.
▪ Tossing a coin.
▪ Spinning a roulette wheel. |
|
|
Term
| The result of an event or trial is an |
|
Definition
|
|
Term
| An outcome represents the characteristic |
|
Definition
of the event/trial.
For example…
▪ Drawing an ace or a seven or a king, etc.
▪ A coin landing on heads or tails.
▪ A ball landing on double zero (00) in roulette. |
|
|
Term
| robability is used to predict what kind |
|
Definition
| of samples are likely to be obtained from a population. |
|
|
Term
Thus, probability establishes a connection between
samples and populations. Inferential statistics rely on this connection when they use |
|
Definition
| sample data as the basis for making conclusions about populations. |
|
|
Term
|
Definition
number of outcomes classifed as A
--------------------------------
total number of possible outcomes |
|
|
Term
| Two conditions must be met for a sample to be random: |
|
Definition
1. Every individual in a population has an equal chance of being selected.
2. When more than one individual is being selected, the
probabilities must stay constant (i.e. there must be sampling with replacement). |
|
|
Term
| As one selects a bigger sample, |
|
Definition
| the sample mean approaches the population mean. |
|
|
Term
|
Definition
the more accurate its estimation of the
population from which it was drawn. |
|
|
Term
Although samples can have different
means, |
|
Definition
| the sample means should be close to the population mean. |
|
|
Term
The sample means (M1 , M2 , M3 …) should
cluster around |
|
Definition
|
|
Term
Thus, the distribution of sample means
tends to form a normal shape with |
|
Definition
|
|
Term
| Any individual sample mean probably will not be identical to its population mean. Therefore, |
|
Definition
| sampling error exists when sampling. |
|
|
Term
|
Definition
represents the degree of error
to be expected of a given sample design.
In other words, there is “error” between M and μ.
Some sample means will be relatively close to μ and others will be relatively far away. |
|
|
Term
| The standard error provides a measure |
|
Definition
| of the standard distance between M and μ. |
|
|
Term
| The law of large numbers: |
|
Definition
| The larger the sample size (n), the closer the sample means should be to the population mean (μ). |
|
|
Term
The formula for standard error reflects
the intuitive relationship between |
|
Definition
standard deviation, sample size, and
“error:” |
|
|
Term
|
Definition
| standard error decreases. |
|
|
Term
|
Definition
| standard error increases. |
|
|
Term
|
Definition
| is the estimated probability that a population parameter lies within a given confidence interval. |
|
|
Term
|
Definition
| is the range of values within which a population parameter is estimated to lie. |
|
|
Term
|
Definition
| is a list of units that compose a population from which a sample is selected. |
|
|
Term
If a sample is to be representative of a
population, it is essential |
|
Definition
| that the sampling frame include all members of the population. |
|
|
Term
There are different ways to sample
using probability theory, including… |
|
Definition
1. Simple random sampling
2. Systematic sampling
3. Stratified sampling |
|
|
Term
| Simple random sampling is a type |
|
Definition
of probability sampling in which the units composing a population are assigned
numbers. |
|
|
Term
| Simple random sampling is beneficial in many ways, but it is |
|
Definition
not necessarily the most accurate
sampling method. |
|
|
Term
| Simple random sampling has |
|
Definition
| A set of random numbers is generated and the units having those numbers are included in the sample. |
|
|
Term
|
Definition
| is a type of probability sampling in which every kth unit in a list is selected for inclusion in a sample. |
|
|
Term
| Generally, systematic sampling is |
|
Definition
| more accurate than simple random sampling. |
|
|
Term
Two terms are frequently used in
connection with systematic sampling: |
|
Definition
1. Sampling interval
2. Sampling ratio |
|
|
Term
|
Definition
| is the standard distance between elements selected from a population in the sample. |
|
|
Term
|
Definition
| is the proportion of elements in a population that are selected to be in a sample. |
|
|
Term
| Stratified sampling is a modification of |
|
Definition
either simple random or systematic
sampling. |
|
|
Term
|
Definition
refers to the grouping of units
composing a population into homogenous
groups (strata) before sampling. |
|
|
Term
| Stratified sampling is slightly more accurate |
|
Definition
| than simple random sampling. |
|
|
Term
|
Definition
| is a multistage sampling technique in which natural groups are sampled initially with the members of each selected group being sub-sampled afterward. |
|
|
Term
| Cluster sampling is used when it is |
|
Definition
| not practical or possible to create a list of all elements that compose a target population. |
|
|
Term
| Cluster sampling is efficient, but |
|
Definition
|
|
Term
| In multistage cluster sampling, |
|
Definition
stratification techniques can refine and
improve a sample being selected. |
|
|
Term
By stratifying a sample in multistage
cluster sampling, |
|
Definition
| sampling error can be reduced. |
|
|
Term
| Probability proportionate to size (PPS) sampling |
|
Definition
| is a type of multistage cluster sample in which clusters are selected not with equal probabilities, but with probabilities proportionate to their sizes—as measured by the number of units to be sub-sampled. |
|
|
Term
| PPS is a more sophisticated form of |
|
Definition
|
|
Term
Depending on the nature of one’s
population, one may wish to sample a |
|
Definition
disproportionate amount of one or
more elements, or use weighting to
sample. |
|
|
Term
|
Definition
| assigning different weights to cases that were selected into a sample with different probabilities of selection. |
|
|
