1
Homework Six: Topic: Sampling
Topic: Sampling
Isabel Cabrera
6300.61 Foundations of Research Education
Dr. Alberto Jose Herrera
University of Texas at Brownsville
February 21, 2012
Homework Six: Topic: Sampling
2
Topic 20 Exercise: Biased and Unbiased Sampling
1. In this topic, how is population defined?
The population is the group in which the researchers are ultimately interested.
2. If a researcher studies every member of a population, what type of study is
he or she conducting?
The researcher is conducting a census when he/she studies every member of a
population.
3. How can a researcher draw an unbiased sample?
An unbiased sample can be drawn by giving every member of a population an
equal chance of being included in the sample.
4. Suppose a researcher drew a random sample from a population of college
students but some of those selected refused to take part in the study. Are the
students who participated in the study a “biased or an “unbiased” sample of
the population?
The study would be biased because they were given an equal chance of being
included but have refused to take part in the study; therefore, volunteerism might
bias the sample of students.
5. If a researcher mails questionnaires to all clients of a social worker, and 50%
of them are completed and returned, is the sample “biased” or “unbiased”?
The study would be bias because not all of them participated leaving out
questionnaires or unreported info or data.
6. Suppose a psychologist has his or her clients participate in an experiment
simply because they are readily accessible (not drawn at random from the
population of interest). What type of sample is being used?
The type of sample being used by the psychologist is a sample of convenience.
The sample is readily accessible and they are willing to participate.
7. Briefly describe one way a researcher can draw a simple random sample.
Putting names in a hat and randomly pulling out names for a study is an example
of simple random sample.
Homework Six: Topic: Sampling
3
Topic 21 Exercise Simple Random &
Systematic Sampling
1. Is there a sequence or pattern to the numbers in a table of random
numbers?
Yes, the researcher must give each member of a population a number name,
and each name must contain the same number of digits. For example, giving
members numbers like 00,01, and all the way to number 99 using 2 digit
numbers.
2. This topic explains how to use a table of random numbers to draw what
type of sample?
Random drawing of numbers is an example of simple random sampling.
3. What is the term for errors created by random sampling?
Errors created by random sampling is simply called sampling errors.
4. How can researchers minimize sampling errors?
Researchers can use systematic sampling to minimize errors.
5. Can researchers minimize the effects of a bias in sampling by increasing
the sample size?
Increasing sampling size does not necessarily decrease bias.
6. Suppose a researcher wants to sample from a population of 99 clients and
the random starting point in Table 1 is the first digit in the last row (Row
#26). What are the numbers of the first two clients selected?
According to the Table 1 on p.199 the first two clients would be 38 and 04.
7. Suppose a researcher wants to sample from a population of 500 clients
and the random starting point in Table 1 is the first digit in the fifth row
(Row #5). What are the numbers of the first two clients selected?
According to the Table 1 on p.199 the first two clients would be 332 and 258.
Homework Six: Topic: Sampling
4
8. If a researcher draws every other individual from a list of the population,
he or she is using what type of sampling?
The researcher is doing a systematic sampling when choosing every other one.
9. What is the potential problem with systematic sampling?
The population may have been arranged thus giving the researcher a bias
sample such as more females chosen for the study.
10. How can a researcher get around the problem you named in the answer
to Question 9?
An alphabetical order list could solve the problem because this way there is no
bias in selecting you sample.
Topic 22 Exercise: Stratified Random Sampling
1. What is the technical term for discussing the magnitude of sampling
errors?
The technical term for sampling errors is precision.
2.
Is it possible for a random sample to contain sampling errors?
Yes, because errors can be created by chance by the random sampling process.
3. What is the first step in stratified random sampling?
First, researchers have to divide the population into strata.
4. Does a researcher usually draw the “same number” or the “same
percentage” from each stratum?
The researcher usually draws the same percentage of participants.
5. If the population of freshmen and the population of sophomores on a
college campus are the same in their opinion on a particular issue on
which a researcher will be conducting a survey, will it be to the
researcher’s advantage to stratify by drawing samples separately from
each group?
Yes, to stratify we need to draw the same percentage of participants, just like
in the example of more women than males. It is better to ensure both
freshman and sophomores are represented in the correct proportions by using
stratified random sampling.
Homework Six: Topic: Sampling
6. What does stratification do to precision?
Stratification will improve precision only if it is based on a variable that is
relevant to the issue being studied.
7. Is it possible to stratify on more than one variable?
Yes, for example you may stratify by both gender and age.
8. Is the primary purpose of stratifying to be able to compare subgroups
(such as comparing freshmen and sophomores in Question 5)?
Yes, it is to ensure that the two different subgroups are represented in the
correct proportions.
Topic 23 Exercise: Other Methods of Sampling
1. To conduct a survey on a campus, a researcher drew a random sample of 25
class sections and contacted the instructors, who then administered the
questionnaires in class. This researcher used what type of sampling?
The researcher used cluster sampling.
2. Which type of sampling is based on trust between participants and a
researcher?
Snowball sampling is a technique that is based on trust because the study of
participants may be hard to locate or identify.
3. What is a major drawback to cluster sampling?
The major drawback is that each cluster tends to be more homogeneous in a
variety of ways than the population as a whole.
4. Which type of sampling is especially useful when attempting to locate
participants who are hard to find?
Snowball sampling is especially useful when trying to find participants that are
difficult to locate or find.
5. Briefly define purposive sampling.
The researcher purposively selects individual who they believe will be good
sources of information for their research study.
5
Homework Six: Topic: Sampling
6
6. What must researchers do in cluster sampling to obtain an unbiased sample
of cluster?
The researcher must draw the clusters at random to obtain an unbiased sample.
7. Suppose a researcher has identified an individual who has engaged in an
illegal activity to be a participant in a research project, and then the
researcher identifies others who have engaged in the same activity through
the first individual’s contacts. The researcher is using what type of sampling?
Snowball sampling is used based on trust to obtain information for their research
because the participants are hard to locate.
Topic 24 Exercise Sampling and Demographics
1. By collecting demographic information on the participants, a researcher
can provide readers of the research with what?
Demographics provide readers with a picture of the type of individuals who
constituted the sample for the study.
2. Consider the demographic of marital status. Is this demographic likely to
be equally relevant in all studies?
Yes, certain demographics can be more relevant for some studies than others.
3. A researcher can do what two things if he or she compares the
demographics of a sample with its population and finds that a subgroup
is underrepresented?
The researcher can count the responses twice to represent a population for
correct percentage. The researcher can also warn the readers that the sample
differs from the population in terms of its demographics.
4. According to this topic, is it important to compare the demographics of
volunteers with the demographics of non-volunteers?
Yes, such information may be of value to someone who is interested in the
application of the program in their local settings.
5. What provides a unique opportunity for researchers who conduct mailed
surveys to collect demographic information?
The fact that most mail is marked with zip codes provides a unique
opportunity for researchers to collect demographic information.
Homework Six: Topic: Sampling
7
6. If some of the participants drop out of an experiment at midcourse, what
is said to have occurred?
Mortality has occurred, which means participants dropped out of the study,
which now affects the results of the research.
Topic 25 Exercise: Introduction to Sample Size
1.
Is sample size the primary consideration when judging the
adequacy of a sample?
No, the importance of considering bias is the primary consideration when
judging the adequacy of a sample.
2. Does increasing sample size reduce bias?
Yes, if only the sampling is done correctly.
3. Does increasing sample size increase precision?
Increasing sample size is of benefit in research because it increases
precision.
4. Researcher A increased his sample size from 200 to 250, and
Researcher B increased her sample size from 1,000 to 1,050. Which
researcher will get a bigger payoff in increased precision by adding 50
participants?
Researcher A would definitely get the bigger payoff, which gives him/her
a big boost in precision.
5. If a researcher uses a very large sample, is it still possible for the
results to be wrong?
Yes, because the sampling may be bias which leads to errors in the
research. Sometimes the small, unbiased samples may yield more
accurate results than biased samples that are large.
6. Does each additional participant a researcher adds to a sample
make an equal contribution to increasing precision?
Yes, just as long as the participant is unbiased to the sample studied.
7. According to the topic, prestigious national surveys are often
conducted with about how many respondents?
Homework Six: Topic: Sampling
8
They usually use about 1,500 respondents when conduction national
surveys.
Topic 26 Exercise: A Closer Look At Sample Size
1. What are pilot studies?
Pilot studies are those designed to obtain preliminary information on how new
treatments and instruments work.
2. Do researchers usually use “small” or “large” samples in pilot studies?
Pilot studies use small samples of 10 to 100.
3. If a researcher suspects that a trait is rare in a population, should he or she
use a “small” or a “large” sample to identify the incidence of the trait?
Large samples are most useful for observing the incidence of a rare trait.
4. In what type of research might a researcher spend considerable amounts of
time interacting with participants, and what effect might this have on sample
size?
Qualitative research is spent on considerable amounts of time, thus researchers
might have to be content with small samples.
5. Suppose a researcher suspects that there is only a very small difference in the
math abilities of boys and girls at the sixth-grade level. Should the researcher
use a “small” or a “large” sample to measure this difference?
The researcher should use a large sample because the population is very
heterogeneous.
6. Suppose the population consists of church members in the Southern Baptist
Convention. If a researcher believes the members of this population are very
homogeneous in their belief in an afterlife, would it be acceptable to use a
small sample to identify the percentage of who hold such a belief?
A researcher should use a small sample because the population is homogeneous.
7. According to Table 2 on page 201, if there is a population of 1,900 students in
a school, what is the recommended sample size?
The recommended sample size is of 320 participants.
Homework Six: Topic: Sampling
8. According to Table 2, if there are 130 nurses in a hospital, what is the
recommended sample size?
The recommended sample size would be 97 nurses.
9. What is the symbol for population size in Table 2?
The population symbol is (N) and the sample size is represented by (n).
9