Sampling Flashcards
What is a sample?
A smaller set of cases a researcher selects from a larger pool and generalizes to the population
What is a sampling element?
The name for a case or single unit to be selected (eg. person, group, organization)
What is a population?
Name for the large general group of many cases from which a researcher draws a sample and which is usually stated in theoretical terms (EVERYONE)
What is a target population?
Large general group of many cases from which a sample is drawn and which is specified in very concrete terms (eg. women in undergraduate residences)
What is a sampling frame?
- A specific list of cases in a population, or the best approximation of it
- Operational definition of an abstract concept (changing population)
What is a parameter?
A characteristic of the entire population that is estimated from a sample (eg. % of people who smoke)
What is a statistic?
A numerical estimate of a population parameter computed from a sample
What is a sampling ratio?
The number of cases in a sample divided by number of cases in the population
What is nonprobability sampling?
Sampling elements are selected using something other than a mathematically random process
What are the types of nonprobability sampling? If possible, give examples.
- Haphazard/Convenient (eg. TV interviewers on the street)
- Quota
- Purposive (eg. seeking out dropouts who are from stable two-parent, rich families)
- Snowball (eg. studying members of an organized crime family)
- Sequential or Theoretical
What is the principle, pros, and cons of haphazard/convenient sampling?
- Researcher selects anyone they happen to come across
- Can produce ineffective and highly unrepresentative samples
- Cheap and quick
- Not recommended
What is the principle, pros, and cons of quota sampling?
- Get a preset number of cases in each of the several predetermined categories that will reflect the diversity of population, using haphazard methods
- Still possible to misrepresent a population; you pay attention to number in sample or face value of participants (ie. “do I have a minority here, check!”) but less attention to DEPTH of information (saturation)
What is the principle, pros, and cons of purposive sampling? Discuss deviant case sampling.
- Get all possible cases that fit particular criteria (often specific and difficult-to-reach population), using various methods
- Until time, resources, or energy is exhausted
- Expert uses judgement in selecting cases; never knows whether cases selected represent population
- Generalizability is not really the goal
- Often used in exploratory and field research
- May result in stereotyping and false sense of security in representation [SAME AS QUOTA]
- Deviant case sampling: researcher selects unusual or nonconforming cases (outliers) purposely as a way to provide greater insight into social processes or a setting
What is the principle, pros, and cons of snowball sampling? What is a sociogram?
- Get cases using referrals from one or a few cases, and then referrals from those cases, and so forth (NETWORK; multi-stage technique!)
- Selection/volunteer bias may apply, can be exclusionary
- Sociogram: a diagram that shows the network of social relationships, influence patterns, or communication paths among a group of people/units
What is the principle, pros, and cons of sequential/theoretical sampling? Discuss theoretical saturation.
- Get cases until there is no additional information or new characteristics (often used with other sampling methods)
- Sample size is determined when data reach theoretical saturation (point at which no new themes emerge from data and sampling is complete)
- Requires that the researcher continuously evaluate all collected cases
- Expensive, time-consuming, hard
What is the experience sampling method (daily diary study method)? Discuss a limitation.
- Intensive longitudinal approach involving DAILY reports on thoughts, feelings, experiences, behaviours, context, environment
- Can address response fallacies (social desirability bias, forgetting)
- More of a data collection technique; but can be considered sampling as recruits must be willing
- Difficult to analyze/code!
What is probability/random sampling?
Researcher uses mathematical random process so that each sampling element in the population will have an equal probability of being selected
What is sampling error?
- Relevant to random sampling
- How much a sample deviates from being representative of the population
What is margin of error?
- Relevant to random sampling
- An estimate about the amount of sampling error that exists in a survey’s results
What must you consider to get a representative sample if your sample is small?
- It is possible to get a representative but small sample using the right sampling methods/frame
- The smaller the population, the bigger the sampling ratio (more in the sample) has to be more an accurate sample
What are the types of probability sampling?
- Simple Random
- Systematic
- Stratified
- Cluster
- Random Digit
What is the principle, pros, and cons of simple random sampling?
Define:
- Sampling distribution
- Sampling distribution of sample means
- Central limit theorem
- Researcher creates a sampling frame and uses a pure random process (ie. random number table) to select cases
- Sampling distribution: a distribution created by drawing many random samples from the same population; to get true idea of population
- Sampling distribution of sample means: a distribution of sample means created by drawing many random samples from the same population
- Central limit theorem: a law-like mathematical relationship stating that when many random samples are drawn and plotted, a normal distribution forms whose centre is equal to its population parameter
- Allows for generalization as researcher can calculate probability of statistic being off from parameter
What is the principle, pros, and cons of systematic sampling?
Define sampling interval and discuss how it is calculated.
- Researcher selects every kth case in the sampling frame using a sampling interval
- Sampling interval: tells researcher how to select elements from a simpling frame by skipping elements in the frame before selecting one for the sample; calculated as the inverse of the sampling ratio
- Likely yields equivalent results are simple random sample
- Cannot be used if elements in a sample are organized in a pattern
What is the principle, pros, and cons of stratified sampling?
- Divides population into strata and then draws random sample from each subpopulation using simple random or systematic sampling
- Fixes proportion of different strata within a sample to guarantee representiveness (if stratrum information is correct)
What is the principle, pros, and cons of cluster sampling? Define cluster.
- Done in multiple stages and is often used to cover wide geographic areas in which clusters are randomly selected; samples are then drawn from the sampled clusters
- Cluster: unit that contains final sampling elements but can be treated temporarily as a sampling element itself
- Eg. if no list of residents in a city, sample city blocks, households, and then residents
- Less expensive and can be used when there is no sampling frame for dispersed population
BUT each stage introduces sampling errors
What is the principle, pros, and cons of random digit sampling?
- General public is interviewed by telephone
- Population is telephone numbers, not people with telephones
- Numbers are randomly dialled (not directory)
- Cost effective way to reach many
What are the 2 methods of cluster sampling?
- Proportionate/unweighted cluster sampling
- Size of each cluster is the same - Probability proportionate to size (PPS)
- An adjustment made in cluster sampling when each sample does not have the same number of sampling elements
- Eg. giving each university an equal chance of being selected when each university has different numbers of students
Discuss statistical power.
- The desired sample size depends on how high we need statistical power to be to analyze data
1. Degree of accuracy required
2. Degree of diversity in a population (homogenous = less error)
3. # of variables being analyzed
What is inferential statistics?
Researcher makes precise statements about level of confidence they have in results of a random sample being equal to population parameter
