Sampling And Representativeness Flashcards
What is a population?
Entire group of people of interest
What is a sample?
Subgroup of the population
What is a sampling frame?
The group from which the sample is chosen.
Not always the same as the population of interest
What occurs if the sampling from differs from population of interest?
Under-representation
Describe the sampling process?
Category of interest Population Sampling frame Sample Individual participants
What dose generalisation from a sample to a population depend on?
Representativeness
What two types of bias affects representativeness of a sample ?
Selection bias
Response bias
What is selection bias?
Over representation of segment of population
-occurs from the sampling method based on convenience.
What is response bias?
- problems for questionnaire studies or surveys.
- underrepresentative sample
What is volunteer bias and who was it introduced by?
Rosenthal & Rosnow
- More highly educated and have a higher need for approval
- more social and less authoritarian
- More ‘arousal seeking’
- Females more likely than males
How do you reduce volunteer bias?
- Making appeal for particpants as interesting and non-threatening as possible
- Theoretical and practical importance
- avoiding ‘stressful’ research.
- Request made by an person of status as high as possible and preferably by a woman’.
What is self sampling?
A problem for under representation.
- as people who only see the advert
- interested in doing research or interested in the effects of stress
- free during the day
——will take part in the study
What sampling methods do we have?
Probability sampling:
-random sampling
- Stratified sampling
- Cluster sampling
Non-probability sampling:
-Opportunity/convenience sampling
- purposive sampling
- quota sampling
What is probability sampling?
- uses some form of random selection
- each member of the population has a specific probability of being chosen as a participant
- researcher can Estimate likelihood that sample findings will differ or be similar from population
- High degree of representativeness
What is random probability sampling?
Simple random sampling:
-sample chosen by chance.
What is systematic random sampling?
Sample members from a population are selected at a fixed sampling interval
-sampling interval is calculated:
Population divided by sample size.
-Everyone in the population has an equal chance of selection
Why is a true random sample very difficult to achieve?
-difficult if population is large.
What is stratified sampling?
Population is divided into subgroups/strata
Random sampling within each strata
Divided into subgroups according to variables likely to affect the dependent variable.
What is cluster probability sampling?
Divide population into groups or clusters, then select random sample of clusters
Every member of selected cluster is part of the sample.
-typically used when you cant obtain list of population
What is non probability sampling?
Probability of any member of the population being chosen is unknown and cant be calculated
- convenient and easy
- probably not representative
- Generalisation problems
What three kinds of non-probability sampling do we have?
Opportunity/convenience sampling
Purposive sampling
Quota sampling
What is opportunity/convenience sampling?
Participants for sample are chosen if they are conveniently available at the time.
-this can cause bias in the sample
What is purposive sampling?
Particpants are selected on the basis of belonging to a specific group or having certain experiences.
-selective is not random
What is quota sampling?
Population is divided into subgroups
Participants selected by researcher until quota is reached
-non-random sample as selection is done by researcher
- High potential for bias:
- Friendly looking people
- convenient location
Why is correct sampling important?
-High ecological validity( can generalise result to real life behaviours)
- Reliability:
- a representative sample will produce the same/similar results each time
Why is sample size important?
More participants more likely to have normally distributed data= more powerful statistical tests.
The greater the power of statistical tests=increased likelihood of finding statistical differences between groups.
However too many participants make very small differences become significant
What is a exclusion criteria?
-Getting rid of variables which may have a large effect on the DV.
What is stiles sampling?
Choosing various time intervals for observations.
Time intervals can be chosen randomly or systematically
What is event sampling?
Selecting pre determined specific events/ behaviours
- used in natural disasters as well
- event defines when the observation takes place
What is situation sampling?
Observing specific behaviours in different locations
Increases external validity
What is WEIRD sampling?
Participants that are:
- Western
- Educated
- Industrialised
- Rich countries
- Democratic courtiers
-96% of particpants which only house 12% of the worlds population
What is standard error?
- measure of dispersion
- standard deviation of multiple sample means
- it measures the accuracy with which a sample represents a population.
- How much a sample deviates from the actual mean of a population
What is the central limit theorem?
States that the sampling distribution of the sample means is likely to be normally distributed as long as the Sam size is large enough (>30)
What is the problem collecting multiple samples?
Expenses, funding, time and resources
Why do we conduct studies on multiple samples?
More samples better understanding of the population.
What are confidence intervals?
- we use standard error to find 95% confidence intervals.
- they give us boundaries around are sample mean which indicates what the population mean will be which determines representativeness of study
SE x 1.96 then add or subtract this value from the mean
1.96 comes from z scores
What are Z scores?
They are measures of standard deviation
They indicate how many standard deviations your score is from the mean.
- If the z score is 0 your score is identical to the mean score.
- Positive and negative scores show the number of standard deviation that the score is either above or below the mean.
What is standard deviation?
The average amount of variation or dispersion from the mean of a set of data values.
What do z scores allow us to do?
Allow us to compare different data sets:they use s
