Research methods
1. Survey research:
The city of Warren has over 1,000 full-time police officers and 300 support staff employees. In the past year, there has been an excess number of department employees (police and staff) taking sick leave. The city wants a study done to determine why there has been so much sick time taken. Rumors suggest it has something to do with morale and declining support of the chief. Assume research approval has been given. The city is specifically interested on how you address the following:
1. Will this be an inductive or deductive approach? What is the difference and which will you choose and why?
2. What type of comparative or prior research would you look at in preparing your study ( examples)?
3. What type of research design or questionnaire will you use? In other words, how will you design the study and how will you develop, validate, and administer the questionnaire (consider Likert scaling measures)?
4. Due to costs, you can only do a sampling study. How will you get a random sampling of 500 police subjects and 100 staff for study ( explain your method for doing a random sampling of each)?
5. How will you assure the validity and reliability of the study? Explain the difference between the two and explain the importance of each ( suggest an example of each as well).
6. How will you examine or analyze your findings (how will you present the data), and how will you determine if the results are statistical significant or not?
2. Experimental research:
A researcher wants to determine which method is most effective in reducing weight gain in middle age persons. There are three methods used. One method is to exercise at a gym three evenings a week with a trainer. The second method is a daily drug (pill) administered without exercise and the third is a personal exercise routine without a drug or the use of a trainer. Some argue that trainers are the best, others the pill and others the personal routine, which is referred to as the normal or usual practice of weight loss. Assume you have a pool of 600 volunteers of various weights, age ranges and health issues, etc. to choose for the study. Due to cost you can only select 220 total participants. Develop a study using a classical research design to evaluate the three methods. Include the following in your response.
1. Is this a qualitative or quantitative study ( explain the difference)?
2. How will you address randomization of the participants? In other words, while perfect randomization of subjects may be difficult to achieve, how will you select the participants for the three groups to reduce bias ( e,g, consider health issues, etc.).The participants in the study do not have to be equal numbers in groups.
3. Will you use a classical research design, a non-classical, or pre-experimental design ( explain the difference between them)?
4. Identify the control and experimental groups and pre and post test procedures used in your study (time period of ...
Research methods1. Survey research The city of Warren has o.docx
1. Research methods
1. Survey research:
The city of Warren has over 1,000 full-time police officers and
300 support staff employees. In the past year, there has been an
excess number of department employees (police and staff)
taking sick leave. The city wants a study done to determine
why there has been so much sick time taken. Rumors suggest it
has something to do with morale and declining support of the
chief. Assume research approval has been given. The city is
specifically interested on how you address the following:
1. Will this be an inductive or deductive approach? What is the
difference and which will you choose and why?
2. What type of comparative or prior research would you look at
in preparing your study ( examples)?
3. What type of research design or questionnaire will you use?
In other words, how will you design the study and how will you
develop, validate, and administer the questionnaire (consider
Likert scaling measures)?
4. Due to costs, you can only do a sampling study. How will
you get a random sampling of 500 police subjects and 100
staff for study ( explain your method for doing a random
sampling of each)?
5. How will you assure the validity and reliability of the study?
Explain the difference between the two and explain the
importance of each ( suggest an example of each as well).
6. How will you examine or analyze your findings (how will
you present the data), and how will you determine if the results
are statistical significant or not?
2. 2. Experimental research:
A researcher wants to determine which method is most effective
in reducing weight gain in middle age persons. There are three
methods used. One method is to exercise at a gym three
evenings a week with a trainer. The second method is a daily
drug (pill) administered without exercise and the third is a
personal exercise routine without a drug or the use of a trainer.
Some argue that trainers are the best, others the pill and others
the personal routine, which is referred to as the normal or usual
practice of weight loss. Assume you have a pool of 600
volunteers of various weights, age ranges and health issues, etc.
to choose for the study. Due to cost you can only select 220
total participants. Develop a study using a classical research
design to evaluate the three methods. Include the following in
your response.
1. Is this a qualitative or quantitative study ( explain the
difference)?
2. How will you address randomization of the participants? In
other words, while perfect randomization of subjects may be
difficult to achieve, how will you select the participants for the
three groups to reduce bias ( e,g, consider health issues,
etc.).The participants in the study do not have to be equal
numbers in groups.
3. Will you use a classical research design, a non-classical, or
pre-experimental design ( explain the difference between them)?
4. Identify the control and experimental groups and pre and post
test procedures used in your study (time period of study, etc.)?
5. Differentiate between validity and reliability and how will
each be addressed. You may want to explain each and give an
example if potential validity and validity problems in your
study.
3. 6. What is statistical significance, and how will you determine
which group is the most effective? What type of confidence
levels or Alpha P values would you try to accomplish to show
there is significance to the study?
PAGE
Survey Research
The inductive approach to survey research is the one whereby
the investigator starts by collecting substantial relevant data to
the topic of interest. The investigator then evaluates the
collected data for a pattern in the data so that he/she can
eventually develop a theory to explain the pattern. In inductive
research reasoning, the researcher moves from specific
objectives and observations to broader theories and
generalizations. The approach involves some uncertainty and is
generally a move from specifics to the general. Deductive
approach on the other hand is the opposite of inductive
approach in that it starts from generalization and moves to
specifics. The investigator usually starts with an existing social
theory and then goes on to test the implications of the theory
with relevant data. This may involve reading the existing
theory, studying what others have done and finally testing the
hypotheses that are generated by those theories (Smith, 2009).
This survey will use deductive approach due to the fact there is
a general theory that the police officers and support staff of the
city of warren take sick leave because of the declining support
of the chief and low morale. The investigator will collect the
relevant data from the sample chosen to analyze and conclude
whether lack of morale and low support of police chief causes
the police and support chief to ask for too much sick leave.
In this study the researcher will need to do some prior or
4. comparative research that will guide him/her in formulating the
necessary research questions for collection of the relevant data.
The comparative research will help the investigator to identify
differences and similarities in the social entities in Warren city
police department. The prior research in this study will take
into account the organization of the police department, the
space and time. The researcher will need to compare Warren
police department with one or two other city police departments
and then check on the duration of the sick leaves taken and
within which disciplines in the department. This comparative
study will help the research to formulate and answer some
specific questions regarding this survey.
This survey research study will employ sampling study design
whereby questionnaires will be administered to the selected
sample. The questionnaires will be developed using the five
scale Likert scaling measures to develop questions that will
measure the attitude of police and support staff in Warren city
regarding the support of the police chief and their morale. The
questions will not be āādouble-barreledāā and will include a
declarative statement and an ordered continuum of response
categories. The ordered responses will be possible answers to
close-ended survey questions that will be offered in a
meaningful and specific order. The continuum to be used will
be; strongly disagree, disagree, neutral, agree and strongly
agree that will be assigned numbers 1, 2, 3, 4 and 5
respectively. The respondents will be required to tick once in
each question. Some of the questions to be administered starting
with declarative statement are as follows;
1. There is massive taking of sick leave by police and support
staff of Warren city
2. Most of the police and support staff of Warren city take sick
leave because they are actually sic
5. 3. Lack of morale prompts the police and support staff of
warren city to take sick leave
4. Declining support of chief of police prompts the police and
support staff of Warren city to take sick leave
5. Too much work in the police department of Warren city
prompts the police and support staff to take sick leave.
Validation of the questionnaires and the questions will be done
by administering the questionnaires to a pilot sample of the
police and support stuff so as to identify whether they
understand the instructions and the questions. When proper
adjustment of the questionnaires is done according to the
response of the pilot sample, they will then be administered to a
randomly selected sample of 500 police and 100 support staff
through the police departmentās headquarters.
Out of the 300 support stuff and 1000 police and in Warren
city, a random sample of 500 police and 100 support stuff will
be selected for the study. Simple random sampling technique
will be employed for the police where each individual police
officer will have an equal chance of being selected. The
investigator will assign a serial identifier to all the 1000 police
and then use computer software for random number table to
randomly select 500 police. For the support stuff, the researcher
will use the systematic sampling technique whereby he/she will
select an equal interval of three staff members and select the
third support stuff from the employment list in the department
to finally come up with a random sample of 100 support stuff.
With the help of the department, the investigator will then issue
out the questionnaires to the selected study sample (Kelley et
al, 2003).
6. In survey research, the impact of the results obtained depends
on the concepts of validity and reliability. Validity is the degree
that a questionnaire in survey research intends to measure what
is designed or actually measures it (Etchegaray and Fischer,
2006). Reliability on the other hand is the degree of
repeatability, stability and consistency demonstrated on
repeating the same survey under the same conditions.
Reliability is important because it helps to control inaccuracies
in scoring, motivation of respondents, order in which questions
are asked, uniformity of interpretation by respondents and
clarity of questions. Validity is also important survey research
due to the fact that it adds strength to the generated data after
the data collection process creating confidence in survey results
interpretation. To ensure validity in this study the researcher
will have to design questions that will actually lead to the
acceptance or rejection of the theory that the increased
application of sick leave by police and support staff is due to
lack of morale and low support from the police chief. The study
design and sampling process of the respondents will ensure that
if another researcher conducts the same research in Warren city
police department, the same results will be obtained thus
ensuring reliability.
The data collected in this survey research will be analyzed and
presented using histograms. Results from each question in the
questionnaire will be used to draw a histogram and the measures
of central tendency will be obtained from the histograms. The
mean, median and mode will be obtained and then the range,
variance and standard deviation will be calculated to obtain
conclusive results. To determine whether the survey results are
statistically significant or not, the investigator will compute
confidence intervals and do a statistical test of means. A 95%
confidence level will ensure low standard error precision in
determining the real cause of increased taking of sick leave by
police and support stuff in Warren city.
7. Experimental Research
This will be a quantitative study that seeks to get statistical
result of the best method of losing weight in middle aged
individuals. Quantitative study differs with a qualitative study
in a number of ways. In the objective, qualitative study seeks to
understand and interpret relationships, while quantitative study
is involved in data quantification and generalization of the
results from sample to the entire population. In this case,
determination of the best method of weight loss will be
determined by the average weight lost within each experimental
group compared to the control sample. The sample in qualitative
study is carefully selected and the type of data collected is
through unstructured techniques such as interviews and requires
no statistical analysis. Large sample size that has been
randomly selected is used in quantitative study and structured
techniques are used to collect data that is analyzed to provide
conclusion basis.
The subjects to be included in each of the groups will be
randomly selected to eliminate any bias in the study. Since there
exist variations in a number of aspects in the 600 individuals
eligible for inclusion in the study, a proper method of
randomization is necessary. To select the 220 individuals to be
distributed in the three groups, various exclusions have to be
made. Individuals with any health condition that may be
exacerbated by exercising or taking of pills and individuals who
have conditions that may result in weight loss over the study
time will be excluded from the study. Randomization will
therefore be to distribute individuals without discriminating the
age, height and weight, however, gender will be predetermined,
where each group will have almost equal number of both sexes
and a BMI value that warrants weight loss. Females in the group
will be assigned number such as 1f, 2f up to their exact number
8. and random selection generated using randomization software to
select 110 of them. Males among the 600 will be randomly
selected using the same method. The randomly selected 220
individuals will be assigned numbers and their inclusion in the
three experimental groups and a control group determined with
software that will randomly select 55 for exercise with a trainer,
55 for normal exercise, 55 for pills and 55 for the control group.
Classical research design will be used in this case, where a
control group and experimental groups are compared after the
treatment. Classical research design involves the use of a
control and an experimental group, where the assignment of
cases is by random selection. A pretest and posttest are taken
prior and after the introduction of the independent variables
respectively. A non-classical research design involves the
investigator describing a group or investigates the relationship
between pre-existing groups unlike classical where experimental
groups are created. The individual in non-classical design are
not randomly assigned and the independent variable is not
manipulated by the investigator unlike classical where the
researcher administers the variable to the study group. Pre-
experimental design is more of a classical design but lack a
control group and/or a pretest. Individuals to be assigned to the
experimental groups are not randomly selected and thus control
of extraneous variables is not taken into consideration unlike in
classical design. Pre-experimental design differs from the non-
classical in that the investigator manipulates the independent
variable (treatment administration) and the experimental group
is selected, but not in non-classical, where pre-existing groups
are studied for their relationship with un-manipulated variables.
The control group will be composed of 55 individuals randomly
selected and they will not be involved in any form of exercise
or treatment with pills for the period of study. Experimental
groups will be three and the treatment will be exercising in a
9. gym with a trainer thrice per week, normal exercise at home and
without a trainer and taking pills to reduce weight without a
trainer or any form of exercise. The pretest will involve weight
measurement from all members of each experimental group and
the control and recording it accordingly. Posttest will involve
retaking of their weight after a period of approximately three
months after which the results will be compared with those from
pretest and an conclusion drawn.
Validity is where the instrument used in the study measures
what is supposed to measure and this is determined testing its
applications in previous studies or testing it on standard
parameters. Moreover, validity may also involve the integrity of
the entire study where the results are quality and can be
generalized to the population from which the sample was gotten
(external validity). Internal validity of the entire study involves
the capability of the study to address the research question.
Reliability is the consistency, repeatability and stability of the
study results. This is the ability of another investigator to
obtain similar results if the study is conducted using a different
sample in the same population at a different occasion. The
example of validity in this study is the ability of the
participants to remain in the study for the entire period and
without any indulgence in behaviors or habit that will lead to
wait loss or excessive weight gain. An outbreak of a disease
may compromise both the validity and reliability as the measure
will not determine weight loss due to treatment but due to
disease. The results will not be repeatable in another occasion
as the health conditions of the sample that will be selected will
be different from the one used to generate the current results.
Statistical significance means that the statistic obtained is
reliable; the relationship or the difference noted between the
groups is not by chance but due to a specific variable or the
treatment administered. Weight loss in the study group and the
control group will be compared and higher weight loss in the
10. experimental group compared to control will indicate that both
exercising and pills are effective methods as suggested at the
beginning of the study. A significance level of 95% or Ī± P value
of 0.05 will be used to determine the presence of significant
difference between control and treatment group using ANOVA.
The group with the highest weight loss will indicate the best
method, which may be recommended to the population.
PA 550 Research Methods answer question 2:
When it comes to medical research then it is very important to
be very caution and professional too. In this research we will
make sure to use the latest tools and consider the latest studies
made in this field. It will be very dangerous to try in such
situation without strong facts.
1- The first step is to match and select the sample before
starting and that is going to be using group clustering, this
method will help us to have patients with similar level on the
stage of the cancer to be in one group. The stages will be
considered to be four groups as per the following; 1) Early stage
where the patients just discovered the cancer and is not yet
expanded to the full lung (up to 20%). 2) Middle stage and this
group will have patients with lung cancer but the cancer already
expanded to big part of the lung (up to 50%). 3) Advance stage
where cancer is already reached more than 80% of the lung. 4)
Very advance group where cancer is almost 100% of the lung is
affected by cancer. We will have three experiment groups which
will be groups 1, 2, and group 3. Group number 4 will be the
control group since the cancer already covered 100% of the
lung.
2- The time frame for the study will be 12 months divided to
four courses each course will be three month. During these
courses we will measure the effect of the new drug on the
patients based on 24 measurement stage for the all period of the
experiment and thatās mean every patient will have to be tested
once every two weeks. It is very important to keep detailed
record of the improvement or the negative affect of the new
11. drug very closely. What will be measured is the development of
the cancer and monitors any other side effects to any part of the
body.
3- We will assure study validity by using the most advance
technology in the field of medical and will also keep very
detailed record for each patient containing details of the drug
doses and the amount and also the side effects happened during
the use of this new drug. All these details will help us and other
research centers to consider and will help us to ensure
reliability of the experiment if we decided to do it again on
other patient.
4- We will consult the most famous and latest reports from
other research centers that are approved from the FDA and all
other studies made for similar drug use. We will study and
analyze these reports and consider the results of them to help us
start with high level of knowledge and confident of what to
expect to get from this drug.
5- In regard to those patients who may drop or die during the
study will make sure to exclude the record and results of those
patients to make sure we have the maximum accurate result and
outcome from this study.
6- The probability level we will use is going to be Pā¤ .01
It is very important to have this probability level used in our
study because it is life threatening case and the life of people is
very important and this is the accepted level by the FDA to
accept the official use of our new drug.
RESEARCH METHOD
I will follow the experimental research to handle this study.
Initially, I would try to randomize the group that I will choose
to eliminate any bias in the study. Since we have a big number
of populations I would like to choose about 250 representatives
to run my research on and they will represent of whole
population. To make a proper randomization and study as well,
many individuals will be excluded as the people they fear of
12. firing them from job to apply the study on them because thy
will make indulgence of the behavior due to to their postion. In
addition, the representatives who will be withdrawn during the
study due to their desires or their administrative report must be
excluded. The investigators need to pay attention to prevent any
kind of discrimination as ethnicity, age, gender, and color. For
example, female can attribute by 1f, 2f, 3fā¦ till their exact
number and they will be distributed by software which it
distributes them into equally to the groups.
In this research I will divide the experimental sample into three
experimental groups. The representatives who attended more
than six workshops are consider a first group. Second, the
representatives who attended between 3-6 workshops will be a
group. Third, the representatives who attended two workshops
and less are last group. The control group is the rest sample that
I choose and will not involve in any kind of experiment
(practicing their work naturally). So the first group will be 55
representatives, the second one is 55, the third one is 55 and the
control group will be 55.
The research question is to measure the quantity of attending
the workshops on the representativeās performance and how the
workshops decline the number of complaints from customers.
So the method that I will use the classical method which is
about taking the number of complaints for these groups who are
various number of attending the workshops prior the study and
post of the experiment and comparing these result with control
group before and after as well. The frame time for this study
can be identified as six months and will be divided to three
courses (course per two months).
The type of data analysis that I will use is quantitative data
analysis because it relates to measure the relationship on data
basis or numbers (digits). The quantitative has structured
method rather that qualitative method which is unstructured like
using the interview. In addition, the advantage of using the
quantitative method is the randomization samples which means
the result of the study we can be generalize on the population
13. the sample is belonging to unlike the qualitative is selected
sample.
Research Methods
Question 2
In this nonprofit health clinic, they provide a number of
services. The board of directors wants to know what services
are most popular and how satisfied they are with the services.
This research is inductive type of research because the board
does not know what is popular or not. Since there is no
hypothesis, as the researcher you have to find out what services
are popular and stratification levels.
Deductive is where there is a hypothesis and you are just
proving it. Also this research would involve both quantitative
and qualitative (mix of both). The mixture of both is most
popular with social researchers to not only find the answers
they need but also help with the why. In order to do that I
would have to use survey research to address the questions
below.
How would I collect the data?
I would use likert style of questionnaires to gather the
information. Likert is a systemic type of questionnaires that
involves questions and the person answering responds with a
agree, strongly agree, strong disagree, or disagree. These
answers are assigned a value and at the end the values are added
up to give a result. This would help narrow down what services
are popular and how satisfied they are. For example after people
answering that most people strong agree that screenings are the
most popular, it would be safe to assume that it might be the
most popular exam.
Now to get satisfied levels is a bit trickier. You could still
use likert questionnaires to gather information. You can ask are
you satisfied with blank services, and the test taker could say
14. strongly agree, disagree. But you could also use personal
interviews to back up the questionnaires. The drawback to
interviews is that they tend to be time consuming and
expensive. If this survey has a limited budget then settling for
questionnaires would be enough. If interviews or random focus
groups are feasible for the research, it would be effective in the
post testing to get some validity.
For the pretest, I would need to gather a small group, like
coworkers or a group of 5 or 10 people to go over the survey. In
the pretest I would have them check for loaded, misleading,
biased questions that could mess up my data gathering.
What is sampling design?
The sampling design would have to a randomization
process. Under some situations like if it was employees being
studied you can take a list of people, assign them a value and
then randomly select them. You would have to make sure all
participants can be equally picked by using a computer program
or some numbering system that is located in the back of the
assigned Research Methods book.
For this case though since all services or anonymous bases
then it would be have to done during or after people come in for
services. It would have to be offered to people after they are
done and ask if they want to fill out a quick survey. This style
is not perfect because it may be tough to get a response from
various individuals at a public health clinic who might not want
to be identified. So the researcher would have to give incentives
to the participants such as promising not to keep personal
information, chance to win gift certificate or free food.
What ethical considerations are there?
IRB for researchers are needed in most research. This
approval is given usually by a company, school or some other
research group to give approval for the research. The reason
why is to protect the researcher, protect the participants, to
make sure the research is valuable and safe. If the research
involves over personal information or some form of physical,
mental or mental harm then it needed to be denied.
15. Once approval is given there is usually someone
overseeing the research to make sure that ethical choices are
made. Also informed consent is needed and the information
provided needs to be secured and when the research is done
usually destroyed to protect the participants. Informed consent
gives the participant the rules, tells them at any time they can
stop, that the information is private. An example given by
Professor Hoang at CLU is some research may appear safe but if
it becomes dangerous to the participants (example if the
information of participants being unethically released ends in
being fired, physically harmed or killed) or possible harm to the
researcher (threats, harmed, fired ect. ). In this case there are
medical regulations to consider such as keeping peopleās
medical information confidential so I would have to make sure
that my research does not violate patients privacy or any federal
laws.
In conclusion after all the data is gathered, I would end the
posttest with a debriefing, follow-up or some sort of interview
(if possible depending on resources). In order to further verify
to the board I would need to show validity and reliability. To
show reliability the easiest way is to re test with future samples.
Reliability is about being able to repeating getting the same
response. I could retest with future patients to see if I keep
getting the same answers. Validity is to see if the results are
likely to happen and not a rare occurrence. There are several
tests such as interview various individuals for face validity. I
can look up similar studies, previous studies if any, popular
tends, interview doctors to get their feedback, review records to
see what services are requested the most.