RESEARCH DISCIPLINE
Rehabilitation Research Statistics
Recovery Measured Over Repeated Visits
For rehabilitation and physiotherapy studies that follow recovery with functional tests across repeated visits: change judged against the error of the test, starting level accounted for, and missed assessments handled openly in the analysis.
No Commitment Required
Strictly Confidential
In-House PhD-Level Statisticians
Fixed Project Fee
Same-Day Response
What Does the Analysis of Functional Recovery Involve?
Rehabilitation research statistics concerns studies that follow functional recovery after injury, surgery or illness, using performance tests such as walking distance or timed tasks and clinician-rated scales at repeated visits. It covers how much of a measured change exceeds the error of the test, how to account for starting level and for patients treated by the same therapist or in the same group, and how to handle visits that were missed. OnlineSPSS, a specialist statistical consultancy for academic, clinical and professional research, supports rehabilitation researchers and graduate researchers: an in-house PhD-level statistician carries out functional outcome analysis on your physiotherapy study data and states what the recovery pattern can support. Whether a gain matters to a patient is for the clinical team to say.
WHEN TO ENGAGE
Checkpoints Across a Rehabilitation Study
Planning questions concern which tests and how many visits, and analysis questions concern change, baseline and attrition, so the work differs from one point to the next.
Choosing the Outcome Tests Walking, balance or strength tests are being selected, and their reliability, the visit schedule and the primary outcome are settled so that the study is neither overloaded with measures nor too small.
Visits Completed Baseline and follow-up assessments are in the file, and the structure of the repeated scores, the missed visits and the baseline differences are examined before a model is chosen.
Change to Interpret A score has improved after treatment, and what matters is whether the improvement exceeds the error of the test and what it means for patients, not only whether it is statistically significant.
Therapists or Groups Involved Participants were treated in group sessions or by a small number of therapists, and the clustering, along with the amount of treatment each attended, has to be reflected in the analysis.
WHAT THE SUPPORT COVERS
Building an Analysis Around a Functional Measure
A walking test, a balance score and a clinician rating behave differently over repeated visits, so the analysis is designed around the measure you chose.
Outcome Measures
Functional outcome analysis begins with the type of each measure: distances and times are continuous, whereas clinician-rated scales are ordinal, so summaries and models are chosen with that in mind.
Baseline and Change
Follow-up scores are analyzed with adjustment for the baseline score, which in randomized comparisons is usually more precise than comparing change scores and guards against regression toward the mean.
Repeated Visits
Scores from several visits are modeled with Linear Mixed Models (Multilevel Modeling), or with Repeated Measures ANOVA when every visit is complete.
Clustering and Dose
Patients treated by the same therapist or in the same group are modeled as clustered, and the amount of therapy attended is examined as a factor beside the allocation.
Meaningful Change
A change is compared with the measurement error of the test, and with a minimal clinically important difference where one has been established for the population, not with zero alone.
Reporting Rehabilitation Studies
Findings follow CONSORT for trials or STROBE for observational designs, with intervals, as expected in Health Sciences and, where performance is the focus, Sports & Exercise Science research.
COMMON ISSUES
Weak Points in Repeated Functional Data
Recovery data accumulate over weeks of attendance, and these problems arise from how patients are measured, when they enrolled and what is compared.
One Straight Line for Everyone Recovery is assumed to follow a straight line for all patients, though it often slows over time and patients differ in speed, so a linear model can misdescribe the course for many.
Time Since Onset Ignored Patients enter the study at different intervals after injury or surgery, and natural recovery is often fastest early, so a group recruited sooner after onset can appear to respond better to the same therapy.
Assessor Knew the Group The person measuring knew which treatment a participant received, which can tilt performance tests and ratings alike, so the lack of masking is reported as a limitation.
Dropout Linked to Decline Patients who worsen or tire of traveling stop attending, so those who remain look better, and an analysis of attenders alone overstates recovery.
Several Tests, One Highlighted Several functional tests at several visits are compared and the one that reached significance is featured, though the primary outcome was never named in advance.
Healthy Side as the Control The injured limb is judged against the patient's other limb, though the uninjured side also changes with reduced activity and with training, so the gap between sides can narrow without the injured limb recovering.
DELIVERABLES
Study Outputs for a Rehabilitation Report
The outputs show how recovery was measured and modeled, and they are delivered so that a clinical coauthor can check them against the patient records.
Outcome Definition Sheet Each test with its units, direction of improvement, timing and source of reliability, and the primary outcome named before the models are shown.
Recovery Curves Plots of mean or model-predicted recovery across visits with confidence bands, drawn on the actual times of assessment.
Visit Completion Summary A table of who was assessed at each visit, with the reasons given for missed visits and a comparison of completers with those lost.
Adjusted Difference Tables Between-group differences at each follow-up with 95% confidence intervals, adjusted for baseline and for therapist or group where relevant.
Meaningful Change Appendix The change in each outcome set against the thresholds you specify, with the source of each named.
Rerunnable Model Code The code behind every model and figure, with a short methods note recording the repeated-measures approach and the software version.
ILLUSTRATIVE EXAMPLE
Walking Capacity After Cardiac Events
An illustration of how a walking test measured at three visits is analyzed when patients are treated in groups.
ILLUSTRATIVE EXAMPLE · CARDIAC REHABILITATION PROGRAM RESEARCH · UNITED STATES
A physiotherapy researcher wanted to compare gains in six-minute walking distance between patients attending a supervised group program after a cardiac event and patients following a home-based program, with assessments at baseline, at discharge and six months later.
The plan named walking distance as the primary outcome and fixed the baseline-adjusted model before any result was inspected. The complication was structural: supervised patients were treated in cohorts run by a few therapists, while home-based patients exercised alone, so clustering existed in one arm only. The model included a random effect for cohort in the supervised arm, the timing of each walk test was taken from the records instead of the planned schedule, and a sensitivity analysis that ignored the cohorts showed how much the structure mattered. The gain was set against the minimal clinically important difference reported for this population, and the nonrandomized comparison was described as an adjusted difference.
Outcome: The researcher received the adjusted difference in walking distance at each visit with its 95% confidence interval, the recovery curves, the visit completion summary, the sensitivity analysis and the model code.
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
Software for Repeated Functional Measures
Repeated assessments sit in wide or long files, and the model is set up in the package your clinical team can open and rerun after delivery.
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
Clinical Meaning Is Decided by the Research Team
OnlineSPSS supports the statistical analysis of the assessments, and what a change means for patients and for practice is decided by the research team.
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 trial report, thesis or article remains yours.
Your Research Decisions The research question, the clinical interpretation, the choice of meaningful thresholds, the discussion 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 About Analyzing Functional Recovery Data
Not every patient can be tested, not every device is worn and not every gain is real, and the answers here deal with each of those situations.
What do statistics for a rehabilitation study cover?
Rehabilitation research statistics covers the analysis of functional outcomes measured at repeated visits: outcome definitions, baseline adjustment, models for repeated and clustered scores, missed visits and the interpretation of change. It suits trials, cohort studies and theses in rehabilitation and physiotherapy.
How is a functional test scored when a patient cannot perform it?
A patient too impaired to walk or stand for a test has no score, but the gap reflects poor function, not chance, so dropping the patient makes the group look better than it is. Options include ranking such patients below everyone who completed the test, analyzing a composite of completion and performance, or using a scale that covers lower levels of function, and the number unable to perform the test is reported at each visit.
How are wearable activity data turned into a study outcome?
Daily step counts and activity minutes depend on how long the device was worn, so the minimum hours of wear that make a valid day, and the minimum number of valid days per patient, are set before the data are examined. Values are then summarized over valid days, models suited to repeated or skewed daily data are considered, and the wear rules are reported so that another team could apply them.
What if the improvement is smaller than the error of the test?
Measurement error describes how far repeat scores vary in a stable patient, so it is the yardstick for whether one person’s change is larger than noise, while a minimal clinically important difference describes a change patients would notice. A group difference can be statistically significant yet small against both, so it is reported with its interval, the two thresholds are kept distinct, and the wording stays modest.
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 a protocol, the outcome tests and assessment schedule or a dataset, 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
Separate Recovery
From Test Noise
Real change, test error and missed visits are tangled together in recovery scores until the analysis pulls them apart.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014