RESEARCH DISCIPLINE
COVID-19 Survey Study Analysis
Before and During the Pandemic, Compared Fairly
For surveys fielded during the pandemic: respondents who were reached online, weights built from known population totals, a before-and-during comparison that holds up, and change across survey rounds estimated with its uncertainty stated.
No Commitment Required
Strictly Confidential
In-House PhD-Level Statisticians
Fixed Project Fee
Same-Day Response
What Does a Pandemic Survey Analysis Involve?
COVID-19 survey study analysis estimates how attitudes, behavior, health and circumstances differed during the pandemic from before it, or across its phases, among people who were reached at the time. Pandemic-period survey data are judged first by who responded and how, since many surveys recruited online, and only then weighted, set against a baseline and modeled across repeated rounds. Lockdown impact analysis is reported as a difference with its confidence interval and worded as an association unless the design supports more. In-house PhD-level statisticians at OnlineSPSS, a specialist statistical consultancy for academic, clinical and professional research, carry out the weighting and comparisons and record each choice in the files. The conclusions you draw are yours.
WHEN TO ENGAGE
Survey Situations That Call for Planning
What the analysis can claim is set largely by how the survey was fielded and what it can be compared with, which is best settled before the first model.
A Single Pandemic Survey The survey was fielded once, during the pandemic, and the question is how responses differ from earlier, which needs an outside benchmark or a comparison the design can supply.
A Baseline Round Exists A pre-pandemic survey or cohort round exists and a new round has been collected, so the comparison must allow for changes in who responded and in how items were asked.
Recruited by Invitation Online Recruitment ran through social media, mailing lists or a panel, and the estimates need weights or sensitivity checks that reflect who was reached and who was not.
Restrictions Started at Different Times Measures began on different dates in different places, and the aim is to relate outcomes to that variation instead of drawing a single before-and-after contrast.
SAMPLE AND COMPARISON
Sampling, Weighting and the Pandemic Baseline
Online recruitment, a missing pre-pandemic round and staggered restrictions are handled separately, because each limits a different part of the comparison.
Sample and Response
Who was invited, who answered and who is missing are documented, and the respondents are compared with known population figures such as census totals.
Weighting
Weights are built from known population margins and used with variance estimates that respect them, and unweighted results sit beside them to show what the weights change.
Before-and-During Contrasts
A comparison with a pre-pandemic round or benchmark survey is checked for matching items, mode and population, since a change of mode alone can move answers.
Regional Timing
Frequent measurements around a measure's start date are analyzed with Interrupted Time Series, and with few rounds, regions not yet affected serve as a comparison.
Panels and Cross-Sections
A panel shows change within the same people, whereas repeated cross-sections show only change in the population, and each is analyzed with intervals that fit its design.
Reporting and Wider Support
Reporting follows Public Health conventions, with the sample, weights and response stated, and the wider analysis sits within Quantitative Data Analysis.
COMMON ISSUES
Pitfalls of Comparing Pandemic Responses With the Past
Most problems enter at fieldwork, when the sample, the mode or the timing is fixed, and the analysis must then state what remains comparable.
Online Reach Read as the Population Respondents recruited online often differ from the population in age, education and internet use, and results that ignore this describe the reached group and not the population.
Memory of the Time Before Asking people during the pandemic how things were earlier relies on recall shaped by the present, so a retrospective answer is not the same as a baseline collected then.
Changed Mode, Changed Answers A pre-pandemic interview survey compared with an online one can differ because of the mode alone, especially on sensitive items, whatever the pandemic did.
Weights Without the Variance Weighted percentages are reported with standard errors that ignore the weights and the design, which usually makes the intervals too narrow.
Season and Trend Absorbed A before-and-during gap also contains seasonal pattern and the underlying trend, so with two time points the difference cannot be traced to the pandemic.
Lockdown Credited With Everything Outcomes that moved after restrictions began are described as their effect, though illness, news and economic change moved at the same time and the data cannot separate them.
DELIVERABLES
Documents Behind a Survey Comparison
The files record who answered, how they were weighted and what was compared, so that each estimate can be traced back to the questionnaire and the fieldwork.
Sample and Response Report A description of recruitment, response and nonresponse patterns set against population figures where available, naming the groups most under-represented.
Weighting Documentation The weighting scheme, the margins used, the range of weights and their effect on the main estimates, with weighted and unweighted results side by side.
Period Comparison Tables Differences between periods or regions with 95% confidence intervals, the respondents in each comparison and the items checked for consistent wording.
Round-by-Round Trend Figure Estimates for each survey round plotted with intervals, the dates of key events marked and the sample size shown beneath every point, so that a reader sees the movement and its precision together.
Limitations Note A note on selection, recall, mode and timing, stating plainly what each limitation does to the comparison and to the claims it permits.
Weighted Analysis Scripts Scripts that declare the weights and design settings and rebuild every table from the survey file, so that a supervisor or reviewer can rerun the full analysis.
ILLUSTRATIVE EXAMPLE
Social Contact Before and During Restrictions
An illustration of how two survey rounds collected in different modes are compared without crediting the restrictions with the whole difference.
ILLUSTRATIVE EXAMPLE · REGIONAL SOCIAL CONTACT SURVEY RESEARCH · UNITED STATES
A public health researcher at a research institute wanted to know whether weekly social contact reported by adults in a regional survey differed between a pre-pandemic round and a round fielded during restrictions.
The two rounds were fresh cross-sections, and the later one had been collected online while the earlier one used home interviews. Because mode could shift answers by itself, the plan compared the rounds on items asked identically, weighted both to the same population margins and reported weighted and unweighted results together. The complication was the lower response of the oldest adults in the online round, which the weights repaired only in part, so a sensitivity analysis showed how the difference moved under alternative weighting choices. The difference was worded as a change between rounds, not as the effect of the restrictions.
Outcome: The researcher received the weighted difference between rounds with its 95% confidence interval, the sensitivity analysis, the response report, the tables, the scripts and a note on how mode and selection limit the comparison.
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
Declaring Weights and Design in the Software
Survey weights and design variables must be declared to the program, and the choice follows the data format and the tools your team already uses.
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
Pandemic Findings Stay Your Interpretation
What a pandemic-era difference means for public health, and what the study concludes from it, are decisions for the research team to make.
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 research. Authorship of the report, thesis or article remains yours.
Your Research Decisions The survey design, the choice of items, the interpretation 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
Weighting, Recall and Round Questions
Weighting has limits, memory is a weak baseline and two rounds are not a trend.
What does COVID-19 survey study analysis cover?
COVID-19 survey study analysis covers the description of who responded, weighting to known population totals, comparisons with a pre-pandemic baseline or across regions, and change across survey rounds, reported with confidence intervals. Cross-sectional and panel surveys are both covered, and so are surveys fielded once during the pandemic with no earlier round.
Can weighting correct for who is missing from a pandemic survey?
Only in part. Weights can reduce differences in measured characteristics such as age and education, but they cannot correct for differences in unmeasured traits that also relate to the outcome. The usual response is to report weighted and unweighted results, name the under-represented groups and word the findings as describing respondents unless the evidence supports more.
Is recalling the time before the pandemic a usable baseline?
It can supplement a baseline, but it is weaker than one collected at the time, since memory is shaped by present circumstances. Where a pre-pandemic survey exists, comparison with it is preferred, with items and mode checked for consistency. Recalled change is otherwise reported as perceived change and not as measured change.
Can two survey rounds show the impact of a lockdown?
Two rounds give a before-and-after contrast that cannot separate the measure from trend, season and other events. More rounds around the start date, or a comparison group not yet affected, strengthen the design, though the finding is still worded as an association. A long series suits interrupted time series methods.
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 questionnaire, the sampling method and the survey file, 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
A Fair Baseline
for Your Pandemic Survey
How respondents were reached often limits a pandemic survey more than any modeling choice, which is why recruitment and fieldwork dates come first in the statistical plan.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014