MCRS 6A: sampling survey
Population = set of data that contains every one of the
units the researcher has elected for the study
Sample = segment of population presumed to represent
that population
Ways of sampling
I. Non probability sampling
Based on judgements by the researcher
- Advantage is convenience and providing insight
1. Convenience sampling
b a s e d o n c o n v e n i e n c e t o t h e r e s e a r c h e r
- Constraints on money and time may lead
researchers to use convenience sampling
E.g. Girl may survey her classmates about
their eating behaviors but she has no
intention to generalize her data and her
research does not contribute to science /
scholarly publication
2. Purpose or Judgemental sampling
b a s e d o n i d e a t h a t a s p e c i f i c p e r s o n o r m e d i a
content will meet specific citeria researcher
may have.
- Seeking for a person who exemplify the
problem
E.g. The director of campus dining services
can provide insight on the economic,
nutritional and scheduling decisions that
lead to the menu options students see on day-
to-day basis. It is likely for the researcher
to seek information from the director if he
wants to know something about the diets of
the students.
3. Quota sampling
r e p l i c a t e f e a t u r e s i n a s a m p l e t h a t r e s e a r c h e r
thinks are important in the population
- Not randomly sampled researcher bias!
E.g. If researcher wants to know something
about diets. He can interview 10 selected
participants but they may all have good
kitchen facilities. This is can have a major
influence on their diets compared to
, participants who don’t have good kitchen
facilities researcher bias.
4. Network or Snowball sampling
F o r m o f v o l u n t e e r s a m p l i n g , w h e n y o u r e l y o n
members of a certain network to introduce you to
other members of the network
- Quality and size depends on ability and
willingness of other to identify other
people in their network to you
- Sample may contain individuals with the same
opinion, because friends of a participant are
likely to have the same opinion as the
participant.
- Does NOT capture diversity of opinion within
the population
E.g. Vegan female introducing her vegan
female friends. This leads to vegan males
being underpresented
5. Volunteer sampling
s a m p l e b a s e d o n v o l u n t e e r s w h o a r e v e r y
willing to participate in the research
- BIASED: Does not capture what nonvolunteers
would have said.There is a difference between
individuals who simply agree to participate
and volunteers who are determined to see if
their viewpoint dominates the research
findings.
- Only applies to human participants
E.g. If you’re interested in veganism, you’re
more willing to participate than if you don’t
seem to know much about veganism. This will
lead to the people who don’t know much about
veganism being excluded bias!
Convenience, judgement and quota sampling can be
used
II. Probability sampling
Generated by random sampling
permits us to make statistical generalizations from
our results.
Sampling frames = Master list from which a
probability sample is selected
Sampling units = units selected from sampling
frames (in case of large populations)
, 1. Random sampling
R e m o v e s r e s e a r c h e r a s a g e n t o f s e l e c t i o n a n d
replaces him or her with luck of draw by chance!
- Common misconception: Random sampling will not
produce a sample that is diverse
i n s t e a d r a n d o m s a m p l i n g p r o d u c e s a n e q u a l l y
homogeneous sample
- Disadvantage: may not reflect the population
because randomness does not respect all
categories
2. Stratified Random Sampling
Tries to represent all categories in sample
- Set aside number of places in you sample
relative to the size of the groups in the
population. Then fill those places with random
sampling from those specific subgroups.
E.g. Researcher needs sample of resident and
nonresident students. She knows that 20% of
sample is nonresident. She decides a final
sample size of 100 and she randomly select 20
nonresident students. SIMILAR to quota sampling
DIFFERENCE units in stratified random sampling
are selected randomly and not by the researcher.
3. Systematic sampling
Every nth person of a list
- Use of a sampling interval
- Based on random sampling
- Disadvantage: If a pattern in the original
population matches the sampling interval
possibility to get overweighted or underweighted
sample.
E.g. Use generator to pick on random number to
start with e.g. dorm number 105 and use e.g.
every tenth dorm to analyze starting from 105.
4. Multistage Cluster sampling
Sampling larger units and then sampling from the
larger units to smaller units and repeat
- P r o v i n c e s t o w n s c i t y b l o c k s
- Advantage: relative ease of identifying
people
- Disadvantage: Potential bias in final sample
increases because no two “states” are
identical, so any sample will have attributes
over of underpresented
Population = set of data that contains every one of the
units the researcher has elected for the study
Sample = segment of population presumed to represent
that population
Ways of sampling
I. Non probability sampling
Based on judgements by the researcher
- Advantage is convenience and providing insight
1. Convenience sampling
b a s e d o n c o n v e n i e n c e t o t h e r e s e a r c h e r
- Constraints on money and time may lead
researchers to use convenience sampling
E.g. Girl may survey her classmates about
their eating behaviors but she has no
intention to generalize her data and her
research does not contribute to science /
scholarly publication
2. Purpose or Judgemental sampling
b a s e d o n i d e a t h a t a s p e c i f i c p e r s o n o r m e d i a
content will meet specific citeria researcher
may have.
- Seeking for a person who exemplify the
problem
E.g. The director of campus dining services
can provide insight on the economic,
nutritional and scheduling decisions that
lead to the menu options students see on day-
to-day basis. It is likely for the researcher
to seek information from the director if he
wants to know something about the diets of
the students.
3. Quota sampling
r e p l i c a t e f e a t u r e s i n a s a m p l e t h a t r e s e a r c h e r
thinks are important in the population
- Not randomly sampled researcher bias!
E.g. If researcher wants to know something
about diets. He can interview 10 selected
participants but they may all have good
kitchen facilities. This is can have a major
influence on their diets compared to
, participants who don’t have good kitchen
facilities researcher bias.
4. Network or Snowball sampling
F o r m o f v o l u n t e e r s a m p l i n g , w h e n y o u r e l y o n
members of a certain network to introduce you to
other members of the network
- Quality and size depends on ability and
willingness of other to identify other
people in their network to you
- Sample may contain individuals with the same
opinion, because friends of a participant are
likely to have the same opinion as the
participant.
- Does NOT capture diversity of opinion within
the population
E.g. Vegan female introducing her vegan
female friends. This leads to vegan males
being underpresented
5. Volunteer sampling
s a m p l e b a s e d o n v o l u n t e e r s w h o a r e v e r y
willing to participate in the research
- BIASED: Does not capture what nonvolunteers
would have said.There is a difference between
individuals who simply agree to participate
and volunteers who are determined to see if
their viewpoint dominates the research
findings.
- Only applies to human participants
E.g. If you’re interested in veganism, you’re
more willing to participate than if you don’t
seem to know much about veganism. This will
lead to the people who don’t know much about
veganism being excluded bias!
Convenience, judgement and quota sampling can be
used
II. Probability sampling
Generated by random sampling
permits us to make statistical generalizations from
our results.
Sampling frames = Master list from which a
probability sample is selected
Sampling units = units selected from sampling
frames (in case of large populations)
, 1. Random sampling
R e m o v e s r e s e a r c h e r a s a g e n t o f s e l e c t i o n a n d
replaces him or her with luck of draw by chance!
- Common misconception: Random sampling will not
produce a sample that is diverse
i n s t e a d r a n d o m s a m p l i n g p r o d u c e s a n e q u a l l y
homogeneous sample
- Disadvantage: may not reflect the population
because randomness does not respect all
categories
2. Stratified Random Sampling
Tries to represent all categories in sample
- Set aside number of places in you sample
relative to the size of the groups in the
population. Then fill those places with random
sampling from those specific subgroups.
E.g. Researcher needs sample of resident and
nonresident students. She knows that 20% of
sample is nonresident. She decides a final
sample size of 100 and she randomly select 20
nonresident students. SIMILAR to quota sampling
DIFFERENCE units in stratified random sampling
are selected randomly and not by the researcher.
3. Systematic sampling
Every nth person of a list
- Use of a sampling interval
- Based on random sampling
- Disadvantage: If a pattern in the original
population matches the sampling interval
possibility to get overweighted or underweighted
sample.
E.g. Use generator to pick on random number to
start with e.g. dorm number 105 and use e.g.
every tenth dorm to analyze starting from 105.
4. Multistage Cluster sampling
Sampling larger units and then sampling from the
larger units to smaller units and repeat
- P r o v i n c e s t o w n s c i t y b l o c k s
- Advantage: relative ease of identifying
people
- Disadvantage: Potential bias in final sample
increases because no two “states” are
identical, so any sample will have attributes
over of underpresented
