HOC14:Samples Flashcards
what is sampling
sampling is about selecting the right individuals , objects , or events as representatives for the entire population -> we are interested in the characteristics of the population we sample from
what do we not just use the entire population and instead sample
because it is :
- practically impossible
- time-consuming and costly
- not always known
- not always reliable results
define population
entire group of people, events, things -> ex. study on advertising by computer firms in Brussels: all 100 computer firms in Brussels
define what an element means
single member or subject of the population
define what a sample means
a subset or subgroup of the population -> some members selected from it
-> ex. 30 computer firms drawn from the population
define what sampling unit means
element available for sample selection
-> ex. firm, person, student, organization ( not the same respondent) -> ex. firm = sampling unit , & general manager = respondent
what does accuracy and precision mean in sampling
a sample must be valid by representing the characteristics of the population -> in which accuracy and precision play a key role :
-> Accuracy of the sample means ensuring the lowest degree of sample bias
how can accuracy be achieved in sampling
by having sufficient elements in the sample such that :
- values of sample elements are close to population values
-overestimation or underestimation avoided
- non-response avoided (ex. sending out surveys -> cant make sure that responses are not biased)
what is sample data and population values mean in relation to sampling
a sample must also be valid by representing the characteristics of the population -> this can be done through precision
-> precision of an estimate means avoiding sampling error = ( the result of the sample should be close enough to the population)
what does sampling process mean
Relation of Sample to Population:
Sampling units provide responses.
Responses → Sample statistics → Population parameters.
Sampling Process Overview
what are the factors influencing what sample size is needed?
*Factors:
Greater dispersion in the population requires a larger sample.
Higher desired precision necessitates a larger sample.
More sub-groups of interest in the sample require a larger sample.
If the sample size is > 5% of the population, it may be reduced without sacrificing precision.
*Key Principle:
Selecting a sufficient number of the right elements from the population enables the generalization of properties to the entire population.
what are the major steps of the sampling process?
- define the population
2.determine the sampling frame
3.determine the sampling design - determine the appropriate size
5.execute the sampling process
what does the step of defining the population mean
research objective and scope : key in defining the target population
what does the step of determining the sampling frame mean
- list items from which sample is drawn = population ( or sub-set) list
- names and addresses of people in the population
- ideally , sampling frame is target population , but in practice rarely the case
what does the step of determining the sampling design mean
*Probability (or random) sampling is favored for representativeness and generalizability.
*Non-probability sampling is more practical when time is limited or generalizability is less important.
what does the step of determining the appropriate sample size mean
Difficult decision -> rule of thumb: 30-500.
Factors affecting size decisions include:
research objective,
desired precision,
population dispersion,
cost, and time constraints.
size of the population
*Myth: A larger sample is not always more representative; it depends on the factors mentioned.
what are the two different approaches to probability sampling
unrestricted or simple random sampling
restricted or complex probability sampling
what does unrestricted or simple random sampling mean in probability sampling approaches?
*Definition:
Every element in the population has a known and equal chance of being selected as a subject.
Selection is individual and direct from the population.
*Characteristics:
Population and sample distributions are similar.
Least bias and highest generalizability.
However, it is time-consuming and costly.
Requires a population list (serving as a sampling frame), which is often not available.
what is restricted or complex probability sampling?
Involves a more controlled and specific approach than simple random sampling.
Different techniques are used to ensure representation while addressing practical constraints.
includes :
-stratified random sampling
-proportionate and disproportionate stratified random sampling
-cluster sampling
-particular case: destructive sampling
what is stratified random sampling * in restricted or complex sampling
Population divided into subgroups (strata) based on certain characteristics.
Random samples are then taken from each stratum.
Ensures representation from various subgroups.
what does cluster sampling mean
- used when size of population is very large
- population divided into smaller groups ( clusters) , samples are drawn out at random from these groups
- ex: population at city level → take sample at district level , with same features as population = area sampling
what does a particular case mean in restricted probability sampling
- a technique in which samples are destroyed → population is changed in the random sampling process
- ex. during the test for quality control of light bulbs, car crash tests
what are the different approaches to non-probability sampling
quota sampling
convenience sampling
purposive sampling
snowball sampling
what is quota sampling?
Quotas in the form of percentages are established for different classes of population ( selection criteria: gender , age, nationality…) to make the sample more representative → ex. street-survey of people
what is convenience sampling?
data collected from members of the population who are conveniently available to provide it => easily accessible
what is purposive sampling?
after defining the criteria ( characteristics of the population and objective of the study) → you search for people who meet the criteria
what is snowball sampling
- recruiting participants via the other participants in the network
- when participants are difficult to find or identify
what is a theoretical sampling?
- the process of repeated data collection for generating theory ( see Grounded Theory) → up till the point of saturation
- characteristics of the sample ( and population only become clear at the end of the sampling process