selecting a sample. to define sampling in both: qualitative research & quantitative research
TRANSCRIPT
CHAPTER 5EDUCATIONAL RESEARCH
SELECTING A SAMPLE
LEARNING OUTCOMES:
To Define sampling in both:
QUALITATIVE RESEARCH &QUANTITATIVE RESEARCH
KEY TERMS: SAMPLE, POPULATION, QUANTITATIVE, QUALITATIVE,
SAMPLEA sample is a selected group (that when properly selected) provides information the same as the population.
The representation of the information from the sample group is intended to be the same as the population.
POPULATIONThe entire group of interest which the
researcher would like to get their study results
from. A population may be of any size, and usually
begins with the word “ALL”
Every member of the population has an equal and independent chance of
selection for the sample.
The researcher has no control over the selection.
PROBABILITY IS EQUALIZED
ERROR & BIAS ARE MINIMALIZED
Generally, it is not possible to conduct an experiment on all the units of a population. An entire population is usually not available.
To save the time and money of the researcher, a portion of the population is used.
The results collected from a study on a sample are generalizable to the entire population.
Why do a sample, and not a whole population?
RANDOM SAMPLING STRATEGIES
SIMPLE STRATIFIED
CLUSTERSYSTEMATIC
SIMPLE RANDOM SAMPLINGEveryone in the population has an equal chance
of selection for the sample.
The researcher has no control over the selection.
The selection of an individual does not effect the selection of any other individual (independent)
• You should have at least 30 samples.• The sample size should be less than 10% of the
entire sample.
Random Numbers
The sample size formula for the infinite population is given as :
The number of observation in a given sample population is known as Sample size. The sample size plays an important part in any study which helps us to find the difference between the population from the given sample. Sample size can be smaller and larger, but the larger sample size gives us the more accurate results and in the lower case it is denoted by 'n' and the sample size in upper case is denoted by 'N' .
Sample Size Problems
Back to Top Below are few problems based on Sample size:
Solved Examples
Question 1: Find the Sample size for finite and infinite population, when percentage of 4300 population is 5, confidence level 95 and confidence interval is 0.04? Solution: From the given data: Z = 3.8416 ( from the z table, we the value of confidence level, that is 1.96) by applying given data in the formula
SS = Z2p(1−p)C2
SS = (1.96)20.5(1−0.5)0.042 = 600.25
SS=600 (after rounding to nearest whole numbers) Now lets calculate the sample size for the finite population.
New SS = SS1+(SS−1Pop)
New SS = 6001+(600−14300) New SS = 527 Question 2: Find the Sample size for finite and infinite population using the given data below, when percentage of 7800 population is 5, confidence level 90 and confidence interval is 0.04? Solution:
From the given data: Z= 2.7060( from the z table, we the value of confidence level, that is 1.645) by applying given data in the formula SS = Z2p(1−p)C2
SS = (1.645)20.5(1−0.5)0.042 = 422.812 SS = 423 (after rounding to nearest whole numbers) Now lets calculate the sample size for the finite population. New SS = SS1+(SS−1Pop)
New SS = 4231+(423−17800) = 401.28 New SS = 401 (after rounding to nearest whole numbers)
THE SAMPLE SIZE FORMULA FOR THE FINITE POPULATION IS GIVEN AS :
Here,SS = Sample size.Z = Given z value
p = Percentage of populationC = Confidence level
Pop = Population
STRATIFIED RANDOM SAMPLING
Stratified RS is a way to guarantee representation of relevant subgroups within
a sample.Population are subdivided into subgroups
(strata) on a certain variable. From each group proportional or equal
numbers of subjects are selected randomly to form a sample
STEPS IN STRATIFIED RS
1. Identify and define the population.2. Determine the desired sample size.3. Identify the variables and subgroups
(strata).4. Classify all members of the population
into subgroups.5. Randomly select an equal or
proportional number of individuals from each subgroup (using table of random
numbers).
CLUSTER SAMPLING intact groups (clusters), not GROUPS are
randomly selected.
All the individuals of the selected clusters are included.
May be the only feasible method of selecting a sample when the researcher is unable to obtain a list of all members of an intended
population.
CLUSTER: STEPS1.Identify and define the population.2.Determine the desired sample size.3.Identify and define a logical cluster.
4.List all clusters.5.Estimate the average number of population
members per cluster.6.Divide the sample size by the estimated size of cluster to determine the number of
clusters.7.Randomly select the needed number of
clusters.8.Include all population members in each
selected cluster.
SYSTEMATIC SAMPLINGSelecting every Kth individual from the list
of the population.K = Number of Individuals on the
list/Number of individuals desired for the sample
All members don’t have an independent chance of selection.
It is considered random sampling if the list of the population is randomly ordered.
Process may cause certain subgroups of the population to be excluded from the
sample* NOT USED VERY OFTEN
STEPS: SYSTEMATIC SAMPLING
1. Identify and define the population.2.Determine the desired sample size.
3.Obtain a list of the population.4.Determine K by dividing the size of the population by the desired sample size.5.Start at some random place in the
population list.6.Take ever K th individual on the list.
7.If the end of the list is reached before the desired sample is reached, go back to the top
of the list.
NONRANDOM SAMPLING STRATEGIES
CONVENIENCE SAMPLINGPURPOSIVE SAMPLING
QUOTA SAMPLING
AKA: non-probability sampling.
Independent or biased free selection of the individuals will not happen.
Useful when the population can’t be described.
CONVENIENCE SAMPLING
AKA accidental Sampling or haphazard sampling.
Sample includes available individuals; “whoever is available”
Volunteers Pre-existing groups
Difficult to describe the population from which the sample was drawn and to whom results can be generalized
PURPOSIVE SAMPLINGAKA judgment sampling.
Selection based on the researcher’s experience and knowledge of the individuals being sampled.
Researcher select the criteria to select the individuals.
Main weakness is the imperfections in the researcher’s criteria.
QUOTA SAMPLING
Process based on required, exact numbers, or quotas of individuals or groups with varying characteristics.
Mostly used in wide-scale survey research when listing all members of the population is not possible.
Data obtained from easily accessible individuals.
People who are less accessible are underrepresented due to their own unavailability.
QUALITATIVE SAMPLINGINTENSITY SAMPLING
HOMOGENOUS SAMPLINGCRITERION SAMPLINGSNOWBALL SAMPLINGRANDOM PURPOSIVE
SAMPLING
QUALITATIVE SAMPLING
In qualitative research the sampling is mainly purposive. Selecting process designed to select a small number of individuals that will be good key informants. “QUALITY instead of Quantity”
The researcher first identifies the potential participants of the research.
Participants are selected on some criteria according to their knowledge, experience, characteristics and willingness
Intensity Sampling- good and poor, experienced and inexperiencedHomogenous Sampling- similar
subjects in experience, perspective & outlook
Criterion Sampling- according to some specific criterion
Snowball Sampling- first select small number, then get additional people
from them. Random Purposive Sampling- Select
more participants than needed.
Intact groups are randomly selected:
a. Simple Samplingb. stratified Samplingc. cluster samplingd. systematic
Intact groups are randomly selected:
a. Simple Samplingb. stratified Samplingc. cluster sampling *d. systematic
Everyone has an equal chance of being selected:
a. Simple Samplingb. stratified Samplingc. cluster samplingd. systematic
Everyone has an equal chance of being selected:
a. Simple Sampling *b. stratified Samplingc. cluster samplingd. systematic
Similar subjects in experience, perspective, & outlook:
a. Intensityb. Homogenousc. Criterion Sampling d. Snowballe. Random Purpose
Similar subjects in experience, perspective, & outlook:
a. Intensity b. Homogenous *
c. Criterion Sampling
d. Snowball e. Random Purpose
Select more participants than needed:
a. Intensity b. Homogenous c. Criterion
Sampling d. Snowball *
e. Random Purpose
Good & Poor, Experienced & Inexperienced:
a. Intensity b. Homogenous
c. Criterion Sampling d. Snowball
e. Random Purpose
Good & Poor, Experienced & Inexperienced:
a. Intensity * b. Homogenous c. Criterion
Sampling d. Snowball
e. Random Purpose