SAFETY 757 Semester Project
Source: Katherine M. Keyes and Sandro Galea, Epidemiology Matters: A New Introduction to Methodological Foundations (2014)
General description
The purpose of the semester project is to apply the seven steps for conducting an epidemiological study, that you are learning in this course, to develop your own research plan. These steps include a plan for: 1) Identifying a population of interest; 2) Measuring exposure and health/safety indicators; 3) Taking a sample; 4) Estimating measures of association between exposures and health/safety indicators; 5) Rigorously evaluating the associations for internal validity; 6) Assessing the evidence for causes working together; and 7) Assessing the external validity of epidemiological study results.
In this exercise, each student will write up a research plan for an epidemiologic study. You will conduct in-depth literature research and will complete a study design on a topic regarding an occupational/environmental health/safety issue of your choice. You will submit it as an MS Word file to D2L-Dropbox-Semester Project. You also need to read several classmates’ study designs and provide constructive feedback and reviews. Grading will be based on the items in the SEMESTER PROJECT GRADING RUBRIC.
Required contents
Title of semester project
The topic should include an occupational/environmental health/safety exposure factor/hazard and a health/safety outcome in a specific environment/population.
Examples:
Age effects in adult lifetime asthma prevalence in California
Deployment, combat, and risk of multiple physical symptoms in the US military
Birth weight, fetal growth, and risk of pediatric rhabdomyosarcoma in California
Introduction
Select a specific topic of your interest. You will choose a topic that you want to understand the causes and distribution of population health/safety so that you may intervene to prevent disease/injury and promote health/safety. You are strongly recommended to find a topic that is related to your work. Have an identified context or place. In other words, consider a particular institution, city, industry, or country when choosing your topic. You should aim to be as specific as possible, but if your topic is too narrow, it will be difficult for you to find relevant information through the literature search.
You can (and should) use the electronic databases on the Andersen Library website (http://library.uww.edu/databases-by-title) to find the scientific journal articles that are relevant to your topic. ScienceDirect (Elsevier Books & Journals Online) is an excellent source for scientific, technical, and medical research. It is one of the largest online collections of published scientific research in the world. You can also search scientific journals by using the Journal Holdings on the Andersen Library website list (http://library.uww.edu/find/journals).
You must read widely on your research topic, and then identify relevant reviews and original research papers from scientific literature (supplemental web resources should be used sparingly). While it is OK to use the occasional web resource (make sure it is from a reputable agency), most of your references should be drawn from updated, peer-reviewed and published scientific papers. All references must be properly cited. Internet citations are okay, but the full web site address must be cited.
Introduction should summarize the purpose and rationale for your study, citing pertinent references. Explain whether you are interested in (a) estimating population parameter(s) or (b) estimating causal effects of exposures on outcomes. Specify your research question of interest by describing specific population parameter(s) or specific causal effects of exposures on outcomes that you would like to investigate.
Recommended length for the introduction section is about 1,000 words (without counting references).
Identifying a population of interest
Define a population of interest for your study. Populations are defined by eligibility criteria, that is, the characteristics of individuals in the population that make them of interest for epidemiological inquiry. Explain the characteristics of the population in which you would like to understand health/safety, using the eligibility criteria of your population including:
geographic area and time period of interest
specific characteristics, events, or exposures for which health/safety-related factors are of interest
specific factors that promote successful study completion
dynamic or stationary population.
Measuring exposure and health/safety indicators
For an epidemiologic study, you need to measure the health/safety status of populations, and define exposures and other factors that may influence health/safety.
Explain the health/safety indicator(s) of your interest, which can be the presence of diseases/injuries, the occurrence of infections, syndromes, symptoms, and biological or subclinical markers associated with disease/injury, quality of life, wellness, and activity, etc. Explain how you will measure the health/safety indicator(s) as variables. Is(are) the health/safety indicator variable(s) binary, ordinal, or continuous?
Explain the exposure variable(s) of your interest that may cause the health/safety indicator(s) or may be associated with variables that cause health/safety indicator(s). Explain how you will measure the exposure variable(s). Is(are) the exposure variable(s) acute, chronic, or time-varying? Is(are) the health/safety indicator variable(s) binary, ordinal, or continuous?
Taking a sample
Once you have accomplished defining a population of interest and conceptualizing and creating measures of exposures and health/safety indicators, you need to take a sample of the population and decide how to study that sample (i.e., follow people forward in time – a cohort study, take a snapshot of the population at a particular time – a cross-sectional study, or sample people based on whether they have the outcome or not – a case-control study).
Two factors that have major implications for the study design are whether we are interested in (a) estimating population parameters or (b) estimating causal effects of exposures on outcomes. The samples needed to answer these two questions are different. One of the most fundamental roles of epidemiology is to document population parameters. Population parameters include estimates of proportions, means and standard deviations, and other indicators of the health/safety status of populations. You need a representative sample if you are interested in documenting the burden of disease/injury in populations. On the other hand, if you want to know whether there is a causal effect of a certain exposure or exposures on a particular health/safety outcome or health/safety outcomes, you need to take purposive samples to achieve comparability between groups of exposed and unexposed persons so that we can identify causes.
Explain what type of study design (cross-sectional, cohort, or case-control) you will use and what type of sample (representative or purpose) you will take. Describe how to take the sample in detail.
If you will take a representative sample, explain how you will get each member of the population as having an equal and independent probability of being selected into the sample.
If you will take a purposive sample, explain the sample eligibility criteria on the basis of exposure status or outcome status that you will use.
Describe who will be the exposed group and unexposed group in your study.
Estimating measures of association between exposures and health/safety indicators
We examine whether there is an association between exposures and health/safety indicators of interest. To test associations, we first need to assess the amount of the health/safety indicator that occurs in both the exposed and unexposed group.
Describe what standard measures (presence/absence, value, mean/variance, median, mode, counts, proportion, rate, prevalence, risk/incidence, incidence rate, etc.) you will use to provide informative summaries of disease/injury occurrence and frequency in your samples.
Describe what standard measures (presence/absence, value, mean/variance, median, mode, counts, proportion, rate, etc.) you will use to provide informative summaries of exposure status in your samples.
We can now move to comparing the occurrence of health/safety indicators between two groups, thus estimating measures of association. This can be done by assessing measures of disease/injury occurrence and frequency conditionally on exposures. Once we have those two measures, we need tools to compare them to decide whether the overall disease/injury measure is higher amongst one group than another.
Describe what measure of association (risk ratio, rate ratio, odds ratio, risk difference, rate difference, etc.) you will use to quantify the magnitude of difference between two measures of disease/injury occurrence.
Rigorously evaluating the associations for internal validity
When we conduct epidemiologic studies and derive associations between exposures and health/safety outcomes, a new question emerges: Does the association that we measure in our data reflect the amount of excess disease/injury that occurred due to the effects of the exposure (i.e., would the risk of disease/injury have been different if those who were exposed had not been exposed), or could there be alternative explanations for the study findings other than a causal explanation?
Also, non-comparability between exposed and unexposed on other causes of health/safety indicators is at the root of many noncausal associations in epidemiologic studies. There are four ways in which non-comparability between exposed and unexposed groups in an epidemiologic study can arise: (a) because of chance in the sampling process, (b) because of associated causes, (c) because of improper selection of study subjects or loss to follow-up, and (d) because of misclassification of exposure and disease/injury status.
From the literature review you have done in the Introduction section, or from your own experiences, did you find that other associated causes, in addition to the exposure variable(s) of your interest, may be associated with the amount of excess disease/injury in your data? Explain what the associated causes are and how they affect your study findings.
Do you expect any of the other ways which non-comparability between exposed and unexposed groups in your study can arise? Explain it.
Assessing the evidence for causes working together
Causes of disease/injury rarely act in isolation. For nearly all health/safety indicators that are of interest to public health/safety, our exposures of interest are embedded in a host of other component causes that work together to activate the exposure’s effect. When multiple component causes work together to produce a particular health/safety indicator, we term this process interaction, as in the two component causes interact to cause a particular outcome. This frequently manifests in our data as an effect of an exposure on a health/safety indictor (e.g., risk or rate ratio, risk or rate difference) varying across levels of another component cause.
Do you expect any of interaction factors in your epidemiologic study? How will you address if there is interaction or not in your data? Explain it.
Assessing the external validity of epidemiological study results
We can now assess the extent to which the results of our study are applicable in populations outside of the underlying population base of our particular study. We now understand how to sample from a defined population, count cases, estimate associations, rigorously test those associations for validity by mitigating non-comparability, and assess the evidence for causes working together. And yet our work is still not finished. In any study, we need to think through the characteristics of our population of interest and determine how robust the study findings might be across populations with similar or different characteristics. The process of evaluating our findings beyond the study sample is often termed generalizability, or synonymously, external validity.
Describe for what other populations you might use your study results and evidence to intervene. How robust might your study findings be across the other populations? Explain how you would address the generalizability, or external validity of your study findings.
References
SEMESTER PROJECT GADING RUBRIC
Accomplished 100% Competent 80% Developing 60% Unaccept.
1. Context, Justification, Objectives Explicitly and fully addresses the assignment. Clearly delineates context, justification, and objectives. Mostly addresses assignment. Mostly identifies context, justification, and objectives Partially addresses assignment. Lacked identification of context, justification, and objectives
Sub-score: __/20 20 18 16 7 12 10 <10
2. Analysis, Methods, Results, Interpretation Demonstrates a plan for complete, accurate, and creative application of methods and presentation of results. Demonstrates a plan for critical and logical understanding; insightful understanding and explanation. Mostly demonstrates a plan for complete, accurate, and creative application of methods and presentation of results. Mostly demonstrates a plan for critical and logical understanding; insightful understanding and explanation. Issues with accuracy and application of methods and presentation of results. Lacked critical or logical understanding, limited explanation or justification. Recommended actions without support.
Sub-score: __/40 40 36 32 14 24 20 <20
3. Coherence and Organization
Focused and logical development throughout. Cohesive writing, with smooth transitions between ideas and concepts. Fully developed introduction and conclusions Mostly focused and logical development. Some issue with writing, unsmooth transitions between ideas/concepts. Lacked a fully developed introduction and conclusion Lacked focus and logical development.
Writing needs improvement, and transitions were abrupt. Very limited development in introduction and conclusions.
Sub-score: ___/10 10 9 8 7 6 5 <5
4. Language Use and Conventions;
Appropriate choice of perspective (3rd person). Uses formal (no slang), objective, accurate and concise language. Closely adheres to rules of Standard Written English. Occasionally shifted from 3rd person perspective. Mostly uses formal (no slang), objective, accurate and concise language. Occasionally inaccurate, repetitive or imprecise word choice. Mostly adheres to rules of Standard Written English but may include some minor errors Primarily uses an inappropriate perspective. Uses informal, inaccurate or, in general, improper language. Major issues with grammar, punctuation, and spelling.
Sub-score: ___/10 10 9 8 7 6 5 <5
5. Evidence
and Documentation
Sources are properly cited using APA format or another professional citation style. Ideas are consistently supported by well-chosen evidence. References are credible, accurate and presented with an appropriate level of detail for the assignment. Sources are mostly cited using APA format or another professional citation style. Ideas are occasionally supported by well-chosen evidence. References are mostly credible, accurate and presented with an appropriate level of detail for the assignment. Sources are listed but using incorrect format. Some ideas are supported by evidence. References lack credibility, accuracy or are presented with insufficient level of detail.
Sub-score: ___/5 5 4.5 4 3.5 3 2.5 <2.5
6. Constructive Feedback and Reviews
All feedback and reviews are relevant and reflect understanding of assigned classmates’ semester projects. All feedback and reviews are helpful in improving own research design and classmates’ semester projects. Some feedback and reviews are relevant and reflect understanding of assigned classmates’ semester projects. Some feedback and reviews are helpful in improving own research design and classmates’ semester projects. Feedback and reviews are provided but some are irrelevant and do not reflect understanding of assigned classmates’ semester projects.
Sub-score: __/15 15 13.5 12 10.5 9 7.5 <7.5
Total score: _______/100
Last Completed Projects
topic title | academic level | Writer | delivered |
---|
jQuery(document).ready(function($) { var currentPage = 1; // Initialize current page
function reloadLatestPosts() { // Perform AJAX request $.ajax({ url: lpr_ajax.ajax_url, type: 'post', data: { action: 'lpr_get_latest_posts', paged: currentPage // Send current page number to server }, success: function(response) { // Clear existing content of the container $('#lpr-posts-container').empty();
// Append new posts and fade in $('#lpr-posts-container').append(response).hide().fadeIn('slow');
// Increment current page for next pagination currentPage++; }, error: function(xhr, status, error) { console.error('AJAX request error:', error); } }); }
// Initially load latest posts reloadLatestPosts();
// Example of subsequent reloads setInterval(function() { reloadLatestPosts(); }, 7000); // Reload every 7 seconds });