Research Proposal Assignment Help
Obesity is a problem that is far more prevalent within the African American community than other racial and ethnic groups of Americans, and this is particularly true among younger people (Delva, O'Malley, & Johnston, 2007).  The exact explanation for this is elusive and has both genetic and dietary components (Delva, Johnston, & O'Malley, 2007).  The proposed project is in two distinct phases.  The first phase is to collect information retrieval, both through conducting surveys and reviews of the literature, to determine the causes of this common disease.  The second phase of the proposed project is to apply this epidemiological knowledge by creating an educational campaign that the African American community can use to reduce the incidence and severity of obesity.

RESEARCH TITLE

The chosen title of the research is “Attribution of Obesity to Individual and Environmental Factors”.

RESEARCH AIM

The aim of the research is as follows –

1. Develop an integrated etiological model combining all the probable determinants of obesity in American people.
2. Identify the differences in vulnerability between demographic variables (gender, age, ethnicity) when subjected to a set of determinants of depressive disorders

RESEARCH OBJECTIVES

In order to reach a deterministic conclusion, the research aims can be further simplified into following objectives –

1. Determine the factors that may cause obesity
2. Demonstrate the effect of these factors on people varying in demographic attributes (gender, age, ethnicity)
3. Compare the effect of these factors on the research population and establish a cause-effect relationship with probable causes
4. Identify the point of inflection in the career (if any), when these factors start impacting the person adversely and differently

A PICO (T)(Oxford University, 2016) question devised for the aforesaid research would be –

How frequently obese people in California opt for fast food items in a week and to what extent the figure is different for people who are non-obese?

RESEARCH INSTRUMENTS & TECHNIQUES

The research methodology for the proposed research would follow a mixed technique, i.e. both exploratory and confirmatory research will be conducted. Questionnaires, observational techniques, in-depth interviews will be part of the research instruments. The detailed technique would involve the following –

• Research Population: This would include the participants in the survey. This group would comprise of people suffering from obesity as well as non-obese people.
• Sampling Frame: The sampling frame will include the entire California population. As obesity is common from childhood through matured adult age and so on, the sampling frame must contain people aged 10 or above.
• Sampling Method: As the research involves people from varying demography (age, gender, ethnicity), the most suitable sampling technique would be cluster sampling. Cluster sampling divides the entire population into groups which are heterogeneous inside and homogeneous outside, i.e. each cluster can be perceived as a homogeneous and unbiased representation of the entire population. A cluster can be chosen among all the clusters of California population with the simple random sampling technique. The underlying assumption would be the following –
• all the clusters demonstrate similar behavioral traits
• every individual belongs to one and only one cluster
• the number of observations within each cluster Ni is known, and N = N1 + N2 + N3 + ... + Nn-1 + Nn
• Eligibility Criteria: The sampling eligibility criteria would be the following –
• respondents must provide a written consent about participating in the survey
• age of respondents must be more than or equal to 10
• Sample Size: In order to determine the effective sample size, several parameters are needed, such as population size, the margin of error, confidence level is chosen by the researcher and standard deviation. The margin of error determines the extent to which the sample mean may differ from the population mean. Standard deviation is defined by the expected variation in the responses obtained (Smith, 2013).

Sample size (S) = ((Z-score) * StdDev * (1-StdDev)) / (Margin of error)

For e.g. if we consider 95% confidence interval (the Z-score obtained from the standard normal table is 1.96), standard deviation of 0.5 and a margin of error as +/- 5% then

S = (1.96 * 0.5 * (1 – 0.5)) / 0.05 = 385 (approx.)

The standard deviation for the research population can be obtained from syndicate research data during the planning stage.

• Expected Response Rate: We assume that we can achieve an expected response rate of 0.2. This implies out of every 10 people chosen for the survey, only 2 will generate a valid and complete response.

ETHICAL CONSIDERATIONS

In the field of medical research, data should not be furnished from an unwilling respondent, i.e. without his consent. The clinical research should be conducted following protocol. The participants must be informed about any medication provided and their outcomes; data solicited from them and the use of that data.

Legally, no other person has the right to touch a patient even without his/her consent. Any such act without permission is termed as “battery” (Richards, 2009), which is a punishable physical assault. In general, battery possesses three legal threats. If the patient is lied about anything, the consent is invalid. The second situation is when the patient is incompetent to consent and receives improper treatment. And finally the third, when the patient is unwilling to consent, and yet the care is forced upon him/her.

Apart from these, the vulnerable groups in the sampled population must be safeguarded. They are the disadvantaged section of the community who require ancillary considerations and utmost care.

PROPOSED ANALYSIS OF DATA & INSTRUMENTS FOR ANALYSIS

Data recorded during the research study, need to be processed and cleaned at first. Thereafter various data mining tools and techniques will be used to predict the outcome of the survey and draw inference on the same.

Based on the above parameters, cross-tabulation can be carried out such as following –

Table no 1. Cross-tabulation of sample data (prepared as a mock analysis)

	Suffering from obesity?		Total
Gender	Yes	No
Male	1000	1250	2250
Female	1750	2000	3750
Total	2750	3250	6000

Correlation analysis interval data and chi-square tests for nominal and ordinal data can be computed between the attributes and variables to figure out whether there is any correlation between them or not.

A hypothesis can be drawn as the following –

Null Hypothesis: A teenager consuming fast food on 4 or more occasions in a week is less likely to be obese compared to a teenager consuming fast food less frequently.

Alternate Hypothesis: A teenager consuming fast food on 4 or more occasions in a week is more likely to be obese compared to a teenager consuming fast food less frequently.

With hypothesis testing (Weisstein, 2016), the null hypothesis can be rejected or may not be rejected. By calculating the p-value, we can decide the same for any confidence interval (for e.g. p-value can be tested for 0.05 for 95% confidence interval).

Apart from these, thematic analysis for categorizing the respondents using words and concepts, decision tree and information gain techniques for the prediction of obesity and causal analysis can be performed to identify any trend in the sample data. The sample statistic can then be estimated for the entire population with suitable estimation techniques.

REFERENCES

1. Delva, J., Johnston, L. D., & O'Malley, P. M. (2007). Availability of more-healthy and less-healthy food choices in American schools: a national study of grade, racial/ethnic, and socioeconomic differences. American Journal of Preventive Medicine, 33(4), S226-S239. Retrieved May 14, 2016, from http://www.ncbi.nlm.nih.gov/pubmed/17884570
2. Delva, J., O'Malley, P. M., & Johnston, L. D. (2007, October). The epidemiology of overweight and related lifestyle behaviors: racial/ethnic and socioeconomic status differences among American youth. American Journal of Preventive Medicine, 33(4), S178-S186. Retrieved May 14, 2016, from http://www.ncbi.nlm.nih.gov/pubmed/17884566
3. Oxford University. (2016). P I C O: Formulate an Answerable Question. Retrieved May 11, 2016, from learntech.physiol.ox.ac.uk: http://learntech.physiol.ox.ac.uk/cochrane_tutorial/cochlibd0e84.php
4. Richards, E. (2009, April 19). Battery—No Consent. (T. L. Site, Editor) Retrieved May 11, 2016, from biotech.law.lsu.edu: http://biotech.law.lsu.edu/map/BatteryNoConsent.html
5. Smith, S. (2013, April 8). Determining Sample Size: How to Ensure You Get the Correct Sample Size. Retrieved May 11, 2016, from qualtrics.com: https://www.qualtrics.com/blog/determining-sample-size/
6. Weisstein, E. (2016). Hypothesis Testing. (MathWorld--A Wolfram Web Resource) Retrieved May 11, 2016, from mathworld.wolfram.com: http://mathworld.wolfram.com/HypothesisTesting.html