RESEARCH DISCIPLINE
Statistical Analysis for Exercise Physiology Research
Consulting for Laboratory & Training Studies
Exercise physiology statistical analysis for crossover experiments, training interventions, graded exercise tests and environmental challenges, with the model fitted to how often each participant was tested and the results prepared to the standard your journal expects.
No Commitment Required
Strictly Confidential
In-House PhD-Level Statisticians
Fixed Project Fee
Same-Day Response
Statistical Analysis for Exercise Physiology Research
Exercise physiology statistical analysis is the statistical design, analysis and reporting of studies that measure how the body responds to exercise, heat, altitude, diet or training, typically in a laboratory with a modest number of participants who each attend several visits. Every participant contributes repeated measurements, the visits differ in order and fatigue, and an early familiarization trial behaves differently from a later one, so a workflow that treats each visit as a separate person reports more precision than the data hold, and one that picks the test from a normality check on a dozen participants rests on a check with almost no power. OnlineSPSS, a specialist statistical consultancy for academic, clinical and professional research, provides exercise physiology data analysis through in-house PhD-level statisticians, who settle the unit of analysis, the treatment of order and the primary outcome with you before any model is fitted.
WHEN TO ENGAGE
Statistical Support at Different Stages of Exercise Physiology Research
Support is most often requested at a few points in a laboratory study, from the sequence of conditions before the first participant attends to a reviewer's question about the final model.
Planning a Crossover or Training Study You are designing a crossover or training study and need the order of conditions, the washout, the primary outcome and the number of participants settled before testing begins.
Analyzing Repeated Laboratory Visits Your visits are complete, and a model is needed for repeated measurements that handles condition, order, time points and the participants who missed a session.
Interpreting a Small Pilot Sample Your sample is small and the estimates look unstable, so you need intervals and effect sizes that show what the data can support before a larger study is planned.
Answering a Methods Critique A reviewer has questioned the handling of order or carryover, normality tests on a small sample or the number of outcomes tested, and the reply needs a corrected analysis.
RESEARCH CONTEXT
Research Designs and Data We Work With in Exercise Physiology
Exercise intervention data analysis starts from a few participants tested many times, and the way visits, conditions and time points are arranged decides which analysis fits.
01
Crossover Experiments
Each participant completes every condition, such as a supplement, a heat or an intensity trial, with the order of conditions balanced across participants and a washout between visits.
02
Training Interventions
Tests before, during and after a training block in one or more groups, with sessions missed and a response that differs from one participant to the next.
03
Graded and Time-Course Protocols
Heart rate, oxygen uptake and lactate recorded across workloads or time within a visit, so that each participant contributes a curve, not a single value.
04
Familiarization and Repeat Visits
Early visits that teach the participant the test and later visits that record it, so that learning and fatigue shift the outcome from one visit to the next.
05
Heat, Altitude and Hypoxia Exposures
Environmental conditions applied to the same participants, with core temperature, perceived exertion and performance measured across each exposure.
06
Laboratory and Field Measures Together
Laboratory values such as maximal oxygen uptake alongside wearable or field data, collected on different scales and sometimes on different days.
METHODOLOGICAL CONSIDERATIONS
Statistical Issues We Consider in Exercise Physiology Research
In physiological outcome statistics, the visit structure and the small sample can change a conclusion more than the choice of test, so these checks come first.
Order, Fatigue and Carryover Sequence and period enter the design and the model, and the washout is justified from the physiology.
The Familiarization Visit A first visit that teaches the test is excluded from the effect estimate or modeled as a visit effect, and the choice is written into the plan, because early performance includes learning.
Visits Within Participants Repeated visits and time points are modeled with participant-level effects or another repeated-measures structure, so that each visit is not counted as a new participant.
Unstable Small-Sample Results With few participants, normality tests and p-values swing and one person can move an estimate. Intervals, standardized effects and a check of each participant's influence show how much rests on one person.
Responders and Noise A claim that some participants responded and others did not is made only where repeated measures or a control condition separate true individual response from day-to-day variation of the test.
Several Physiological Outcomes Oxygen uptake, lactate, heart rate and perceived exertion at several workloads multiply comparisons. One primary outcome is named before testing, and the others are reported as secondary with intervals.
STATISTICAL METHODS
Methods Selected for the Research Question and Study Design
Each method family is chosen for the outcome and the data structure of your study, and the line under each title explains why it is used in exercise physiology.
Longitudinal & Multilevel Analysis
Several visits and workloads per participant are correlated, so mixed models carry a participant-level effect, in generalized form for binary outcomes such as task failure, and keep the data of a person who missed a visit.
Comparative & Hypothesis Testing
Conditions in a crossover are compared within each person, and a training group is compared with a control after adjusting for baseline, each as a difference with its interval.
Multivariate, Scale & Reliability
Day-to-day variation of a test is quantified from repeat visits as a typical error and an intraclass correlation, and ordinal ratings of perceived exertion are analyzed with methods that respect the scale.
Regression & Predictive Modeling
Relations between a physiological marker and performance, such as a threshold and a time trial, are modeled with few predictors for the sample size and reported with intervals.
Bayesian & Computational Methods
When participants are few, Bayesian estimation with stated priors or a bootstrap interval can complement the mixed model, though neither removes the need to defend the assumptions.
Evidence Synthesis & Meta-Research
Trials of heat, altitude and supplements are small and differ in protocol, so a meta-analysis shows heterogeneity, and a synthesis without pooling is used where results are not comparable.
DELIVERABLES
Statistical Outputs Prepared for Your Research
The outputs follow the structure of a laboratory study, from the review of the visit schedule to the scripts that regenerate every reported estimate.
Data & Study Review The visit schedule, the order of conditions, the familiarization visits, the coding of each physiological variable and any missed sessions are reviewed before a model is fitted.
Statistical Analysis The agreed mixed or repeated-measures models with period and sequence terms, residual and influence diagnostics and sensitivity analyses such as dropping the familiarization data.
Effect Estimates & Uncertainty Condition differences and training changes as mean differences with confidence intervals, and standardized effects with intervals, set against the smallest worthwhile change where agreed.
Tables & Figures Participant-level plots of every condition, mean and interval plots by time point, a participant flow table and a results table that includes every outcome that was tested.
Statistical Interpretation Each result is explained in physiological terms within what a small crossover or training study supports, and the limits on generalizing to other athletes are stated.
Analysis Files R or SPSS syntax, SAS or Stata programs or Python scripts, with the coding of condition, period and sequence documented so that every table can be regenerated.
REPORTING
Reporting Exercise Physiology Research
The guideline depends on whether your study allocates conditions, and each is applied only where it fits.
A randomized crossover study reports to CONSORT 2025 together with the CONSORT extension for crossover trials, which asks how the sequence was generated, how long the washout was and how period and carryover were handled, and its protocol follows SPIRIT 2025. The exercise, heat or training exposure is described to TIDieR so that intensity, duration and supervision can be repeated by another laboratory. Methods text names the primary outcome, the handling of familiarization data and of participants with missing visits. Participant-level data are shown where numbers are small, each effect carries its confidence interval, and the journal’s requirements on statistics and data sharing are checked at the start.
ILLUSTRATIVE EXAMPLE
How the Analysis May Look in Practice
An illustration of how a small crossover experiment moves from a visit schedule to a result stated with its uncertainty.
ILLUSTRATIVE EXAMPLE · CYCLING HEAT PERFORMANCE RESEARCH · UNITED STATES
A university exercise physiology laboratory wanted to know whether wearing a cooling vest during warm-up changed cycling time-trial performance in the heat, using a randomized crossover study in a small group of trained cyclists.
Participants completed a familiarization trial and then two experimental trials, one with the vest and one without, in a balanced order with a washout between. One participant missed the second experimental visit because of illness. Before outcomes were examined, we fixed mean power in the time trial as the primary outcome, set the familiarization data aside from the estimate and specified a mixed model with condition, period and sequence and a participant-level random effect. Carryover was addressed by the washout and the balanced sequences, not by a preliminary test. The participant with one visit stayed in the model, and a sensitivity analysis removed that participant.
Outcome: The group received the mean difference in time-trial power with its 95% confidence interval, a participant-level plot of both conditions, the crossover flow counts, a methods paragraph and the scripts.
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.
RESEARCHER VOICE
What Researchers Say
Feedback from an exercise physiologist on advice that reflected the realities of the field, on a small sample with unstable estimates and on a result that would stand in review.
“I had worked with analysts before, but this was the first time the statistical advice really reflected the realities of Exercise Physiology. They picked up a relatively small sample and unstable estimates quickly and helped us avoid a result that would have been difficult to defend in review.”
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.
SOFTWARE
Statistical Software for Exercise Physiology Research
Sports lab data analysis is run in the package your laboratory already uses, with syntax or scripts provided where applicable.
Statistical Support Within Your Research OnlineSPSS provides the agreed methodological, analytical, and statistical reporting support. The researcher or research team retains the scientific question, substantive interpretation, conclusions, and overall responsibility for the research.
FREQUENTLY ASKED QUESTIONS
Questions About Statistical Analysis for Exercise Physiology Research
Short answers to what exercise physiologists ask before sharing a study, from crossover analysis and responders to pilot samples, confidentiality and the assessment process.
Which analysis suits a crossover study with few participants?
A mixed model with condition, period and sequence and a participant-level effect uses all the data, including that of a person who missed one condition, and reports the condition difference with its confidence interval. A paired comparison is enough for a simple two-condition design with complete data and negligible period effects. Our Statistical Methods pages describe each family.
Should I test for carryover before analyzing my crossover?
A preliminary carryover test is not a reliable safeguard. In a small study it has little power, and choosing the analysis according to its result distorts the final inference. Carryover is better prevented by the design, with a washout justified from the physiology and balanced sequences, and the model then includes period and sequence.
Can I call participants responders or non-responders?
Only with care. The variation in change scores combines true individual response with the day-to-day variation of the measure, so a spread of changes does not show responders. Repeated baseline measures or a control condition allow the true variation to be estimated, with its uncertainty, and that estimate is what the claim should rest on.
What can a small pilot study show?
A pilot study shows feasibility and gives a rough idea of variability, and it estimates an effect with a wide interval. It cannot confirm that an effect is absent. A pilot effect size is an unreliable basis for the power calculation of a larger study, because it is itself uncertain, so planning usually relies on a smallest worthwhile effect.
How is confidential laboratory and participant data handled?
Visit records, physiological data and analysis files are handled confidentially, and project information is not published. 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 send your research question, the testing protocol and any data you hold. A statistically trained Research Coordinator reviews them and responds the same day, and the analysis, deliverables, timeline and one fixed project fee are confirmed before work begins, so the timeline follows from the agreed scope.
Still have a question? Find more answers in our full FAQ.
NEXT STEP
Discuss Your
Exercise Physiology Research With Us
Share your research question, the testing protocol and the data you have, even if only a design exists. We review how visits and conditions are structured and confirm the approach before work begins.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014