RESEARCH DISCIPLINE
Epidemiology Research Project Statistics
Who Was Studied Decides the Measure
Exposure and outcome studies start with who was observed: the study population defined from its source, the measure of association that the design can support and an estimate reported with its confidence interval, with stratified analysis in epidemiology used where a third factor confounds or modifies the association.
No Commitment Required
Strictly Confidential
In-House PhD-Level Statisticians
Fixed Project Fee
Same-Day Response
What does an epidemiology project analysis at OnlineSPSS cover?
Epidemiology research project statistics concerns how an exposure and an outcome are related in a defined population. Study population data analysis comes first, because the source population, the eligibility rules, the case definition and the start of follow-up decide whom any estimate describes. The measure of association follows from the sampling: risks and rates in a cohort, odds in a case-control study and prevalence in a cross-sectional survey, each with a confidence interval. Where a third factor may confound the association, the strata are combined by a Mantel-Haenszel estimate, the Mantel-Haenszel odds ratio in a case-control study and the matching risk or rate ratio in a cohort, when the stratum estimates agree on that scale. When they do not, effect modification by stratum is reported with each stratum’s estimate and interval. OnlineSPSS, a specialist statistical consultancy for academic, clinical and professional research, provides this support: an in-house PhD-level statistician defines the population with you, selects the measure and carries out the analysis. Whether the association is causal is a question the researcher answers.
WHEN TO ENGAGE
Epidemiology Projects Before the First Estimate
The need appears when an exposure and an outcome are available and it is unclear whom the estimate describes or which measure to report.
Protocol With Loose Eligibility The protocol names an exposure and an outcome but leaves open who is eligible, when follow-up begins and how a case is identified.
Data Collected for Another Purpose A registry, a survey or a clinical database is reused for a new question, and its population must be described before the question is put to it.
More Than One Possible Measure The design allows a risk, a rate, an odds ratio or a prevalence ratio, and the choice depends on what the sampling preserved and what readers need.
A Reviewer Asks Whom It Describes A reviewer asks to whom the findings apply, and the answer depends on how participants were selected from the source population.
WHAT THE ANALYSIS COVERS
From Source Population to Reported Association
Each stage fixes something that the next one relies on, beginning with the people who could have entered the study.
Source Population
The population from which participants arose is described by place, period and eligibility, together with what was excluded, since it fixes the group the estimates can speak for.
Case and Exposure Definitions
Each case and each exposure is defined by explicit criteria applied the same way throughout, with the date from which follow-up counts stated for every participant.
Time Zero and Follow-Up
For cohort data the start of follow-up is aligned with the start of exposure status, so that no period is counted as at risk when the outcome could not yet occur.
Frequencies in the Population
Counts, risks and rates are summarized first, and prevalence or incidence is estimated with the methods of Incidence & Prevalence Analysis.
Measure of Association
The measure is chosen to match what the sampling preserved and is reported with its interval, following Relative Risk & Odds Ratios.
STROBE Reporting
Reporting follows STROBE and the practice of Epidemiology, with the wider family of methods under Clinical, Survival & Epidemiological.
COMMON POPULATION ISSUES
Population Definitions That Distort an Association
A weakness in who was included, or when, cannot be repaired by the model that follows, which is why these faults matter early.
Eligibility Defined After Looking Inclusion criteria are adjusted after the results are seen, which tailors the population to the finding and cannot be undone by later modeling.
Comparison Group From a Different Base Exposed and unexposed participants come from different settings or periods, so a difference may reflect the setting and not the exposure.
Cases Defined Differently Over Time The criteria for a case changed during data collection, so early and late cases are not the same condition and pooled counts are hard to interpret.
Ratios With No Absolute Anchor A relative measure is reported alone, so a large ratio for a rare outcome and a small ratio for a common one cannot be compared by how many people are affected.
Loss to Follow-Up Related to Exposure Participants who leave the study differ in exposure from those who stay, so the people observed to the end no longer represent those who entered.
Short Follow-Up Counted as Full Risks are computed by dividing cases by everyone enrolled although follow-up lengths differ widely, so people observed briefly are counted as if they had been at risk throughout.
DELIVERABLES
Definitions and Counts Under Each Estimate
Each estimate arrives beside the definition of the population it describes and the counts behind it.
Population Definition A description of the source population, the eligibility criteria and the date range, in the form a methods section needs.
Participant Flow Table How many people remain after each eligibility rule is applied to the source file, and why each group dropped out.
Case and Exposure Dictionary The criteria and codes that identify each case and each exposure level, with the dates that anchor follow-up.
Association Estimates Risks, rates, odds or prevalence ratios as the design allows, each with its 95% confidence interval and the counts behind it, by stratum where the analysis is stratified.
Absolute Measures Table Risk differences or rate differences beside the ratios, so that the public health size of an association can be read.
Command Files Commands that recreate the study population from the source file, step by step, and then produce each estimate in the order of the flow table.
ILLUSTRATIVE EXAMPLE
A Case Definition That Changed During the Investigation
A foodborne illness study in which the meaning of a case shifted while the data were collected.
ILLUSTRATIVE EXAMPLE · FOODBORNE ILLNESS INVESTIGATION RESEARCH · UNITED STATES
A doctoral researcher in epidemiology analyzed a gastrointestinal illness reported by guests at a catered event, to find which foods were associated with illness among the people who attended.
The source population was the list of attendees, the attendees who were interviewed formed the denominators, and the risk of illness was compared between those who ate each food and those who did not, using risk ratios with confidence intervals. The complication was the case definition: early interviews counted any stomach complaint, while later ones required specific symptoms within a set window after the event, so the cases were not the same condition. The definition was fixed before the food-specific results were examined, probable and confirmed cases were separated, and the main analysis used the stricter definition, with the broader one reported as a check. Risk differences were given beside the ratios, and because not all attendees were interviewed, the report described who responded.
Outcome: The researcher received the population definition, the participant flow table, the case dictionary, risk ratios with 95% confidence intervals and risk differences for each food, and the commands to rerun them.
This example does not describe a real client or project. We never share details of actual client work, in line with the confidentiality we promise every researcher. This illustrative example is included to show the type, scope, and standard of work involved in this kind of analysis.
SOFTWARE
Packages Suited to Risks, Rates and Ratios
Each of these packages estimates the common epidemiological measures, so the data source and the reviewers' expectations usually decide.
HOW WE ASSESS YOUR PROJECT
We Review Your Actual Research
Your assessment is based on a review of your actual research materials and analytical requirements, not a generic project category.
You Submit Your Project Details
Share whatever is available at the current stage, such as the research aim, protocol or proposal, dataset, variable list, manuscript or statistical requirements.
We Review Your Research and Define the Scope
Your statistically trained Research Coordinator reviews the materials to understand the study design, research questions, data structure, analytical requirements and expected deliverables. If clarification is needed, additional documents or information can be requested before the assessment is finalized.
Your Analysis Is Completed and Delivered
Once the scope and fixed project fee are agreed, the in-house statistician completes the work and the Research Coordinator remains the communication point through delivery, with the analysis, deliverables and timeline as agreed.
RESEARCH RESPONSIBILITY
The Scientific Question Is Still Yours
OnlineSPSS carries out the agreed definition of the population and the estimation, while the study question and its causal interpretation rest with the researcher.
Agreed Statistical Support OnlineSPSS provides the agreed methodological, analytical and statistical reporting support, defined in a scope that is confirmed before any work begins.
No Intellectual Contribution OnlineSPSS does not claim any intellectual contribution to your project. Authorship of the dissertation, protocol or article remains the researcher's.
Your Research Decisions The research question, the choice of exposure and outcome, the public health meaning of the findings and the conclusions remain yours, with overall responsibility.
Confidential Handling Research data and project details are handled confidentially, project information is not published, and an NDA is available where required.
FREQUENTLY ASKED QUESTIONS
Questions on Study Populations and Associations
The source population against the study sample, participants who already had the outcome, and the methods wording for the population.
What do epidemiology research project statistics cover?
OnlineSPSS defines the study population, chooses the measure of association that the design supports and estimates it with its confidence interval, stratified by a confounder or modifier where the question calls for it. Each design, from outbreak investigations to cohorts, is analyzed on its own terms.
What is the difference between the source population and the study sample?
The source population is the group from which participants could have come, defined by place, time and characteristics, and the study sample is the subset who actually entered and were analyzed. An association measured in the sample carries over to the source population when selection into the study did not depend jointly on the exposure and the outcome. When it did, as when exposed people with the outcome were more willing to take part, the estimate can be biased, and the limitation is stated.
Can participants who already had the outcome join a cohort?
Usually not, for a question about new disease. People who already have the outcome at the start cannot develop it, so including them mixes existing and new cases and blurs what is being estimated. They are excluded or analyzed separately, and the same logic applies to exposure: a person exposed for years differs from a new user. Where existing cases are the focus, the design is a prevalence study and the measure changes accordingly.
What details of the study population does a reviewer expect?
At minimum: the setting and dates, the source population, the eligibility criteria, how participants were selected, the definitions of cases and exposures, how follow-up was counted and how many people were excluded at each step. STROBE lists these items for observational studies, and a participant flow table presents the counts compactly.
How do you protect my research data and confidentiality?
Your data and project details are handled confidentially, and project information is not published. Researcher names and institutions are not used publicly without permission. An NDA is available where required and can be arranged before sensitive materials are shared. Read more about confidentiality and data security.
What happens when I request an assessment, and how long does it take?
You share your research question, design and relevant materials, such as the protocol, the case definition and the source data, and a statistically trained Research Coordinator reviews the request and responds the same day. The analysis, deliverables, timeline and one fixed project fee are then confirmed before any work begins, so the timeline follows from the agreed scope.
Still have a question? Find more answers in our full FAQ.
NEXT STEP
Population, Exposure and
Outcome in Your Words
An odds ratio, a risk ratio and a prevalence ratio describe different quantities, and which of them a project can report was settled when its participants were selected.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014