Film creditsThe same people. Different questions.
A repeated-measure analysis plan for separating changes within a specimen or participant from differences between them.
Proposed evidence-engineering direction | External sources and defined methods; scope, rights and delivery capacity to be agreed.
Decide which association matters
Decide whether the question concerns changes within one specimen or differences between specimens before interpreting a pooled association.
A material comparison may ask whether a change in condition alters the response within the same specimen, or whether specimens with one property tend to have another. Those questions can produce different associations in the same dataset. The proposed analysis begins by naming the estimand rather than calculating a pooled correlation immediately. A formulator studying several conditions on each batch needs to retain batch identity. An application team studying several liquids with each participant needs to retain participant identity. Without that structure, the apparent relationship can reflect differences between units rather than the change the proposed application seeks to exploit.
Examine the development decision
The same people. Different questions.
Open the sequence to examine what a development decision needs.
Identify the exact preparation, its source, and the properties a second laboratory must be able to reproduce.
Explanatory geometry · no simulated or measured performance
The spreading data have a repeated design
Follow participant and liquid identifiers through the spreading records so that repeated trials retain their actual experimental relationships.
Fehlberg et al. (2025) deposit measurements for the same 11 participants across 15 liquids. The reviewed analysis retains each published trial median and the coded participant. Selecting two liquids therefore does not create two independent cohorts. The proposed comparison would make paired observations visible and keep the participant dimension available when asking a within-person question. It would also distinguish that question from the physical explanation of friction across speed, load and rheological conditions. The dataset offers a useful analysis case, but a selected view cannot establish broad consumer preferences or an independent validation population.
Separate the unit mean from its local deviation
Decompose a predictor into unit means and local deviations to expose the different scientific questions hidden inside one trend.
A useful decomposition expresses a predictor as its unit-level mean plus the observation’s deviation from that mean. It makes the within-unit and between-unit questions visible before choosing a model. Bland and Altman (1994) show why indiscriminately mixing repeated measurements can mislead correlation analysis. The proposed record would retain both components and state whether the intended interpretation concerns changes within a unit or differences across units. The decomposition itself does not remove confounding or validate a causal claim. It is a way of making the statistical question precise enough for the scientific team to challenge.
Separate unit-level mean and within-unit deviation without changing the original predictor.
- Predictor in observation j from unit iOriginal predictor unit
- Mean predictor within unit iOriginal predictor unit
- Repeated-unit identifier, such as participant or preparationIndex
- Observation identifier within unit iIndex
Model conditions. Repeated units identified correctly Arithmetic decomposition only; no causal interpretation or model fit asserted
Keep the conditions attached to the result.
Make scientific records useful through source-linked comparisons, explicit units and a reproducible decision trail.
Paired differences need the same endpoint
Check endpoint and protocol compatibility before using paired trial differences to describe a meaningful change for the same experimental unit.
A participant-level difference between two liquid trials can be meaningful only if the measured endpoint and relevant protocol are compatible. Speed and force may differ even within a participant, so a simple friction difference may mix material and behavioural variation. The proposed analysis would record those conditions and distinguish an observed paired difference from an adjusted model estimate. The buyer could then decide whether the physical question requires controlled speed/load experiments or whether a descriptive natural-spreading comparison is sufficient. A paired label is not a guarantee that every other part of the experiment is matched.
Change the conditions · examine the reference
More readings—or more evidence?
Change the independent preparations and repeat readings. See which source of variation the next measurement can actually reduce.
The data, equations and controls remain available below if 3D cannot load.
Defined model · illustrative conditions
Points are discrete design calculations, joined for comparison. No experimental observations or confidence intervals are shown.
Fund the preparation or repeat that addresses the uncertainty relevant to the development decision.
The defined comparison
Balanced random-intercept scenario with between-preparation SD 10 response AU and within-preparation SD 6 response AU. Preparations are independent; residual readings are independent with common variance. The naive estimate treats all BR readings as independent and ignores clustering. The 3D groups depict a design count, not measured sample values. No confidence interval, power or precision guarantee is calculated. Drift, pairing, operators, serial correlation and unequal group sizes need additional structure.
Inspect the calculated values
Calculation grid, not measured observations. Inputs and assumptions are stated above.
| Independent preparations / count | Hierarchical design | All readings treated as independent |
|---|
A sweep is not automatically a repeated-unit population
Treat a rheology sweep as observations across a condition axis, preserving its distinction from a population of independently prepared materials.
Woodbridge et al. (2025) provide many points along shear or oscillatory axes for source specimens. Their within-sweep structure is important, but the files do not identify a population of independent batches. A model of response across an experimental axis may be useful for a specimen without supporting inference about a broader preparation population. The proposed record would separate axis points, sweeps and independent sample identifiers. This prevents a general repeated-measures method being applied mechanically to any long CSV, and helps a partner decide which additional specimens are needed before asking about between-lot variation.
Validate the model at the unit level
Keep the intended new unit outside model fitting when asking whether a relationship transfers to another batch or participant.
When the proposed application concerns a new batch or participant, validation should keep that unit out of the fitting set rather than scattering its observations across both sets. Otherwise the model may benefit from information it would not have for a genuinely new unit. NIST’s residual guidance provides one part of model assessment, but unit-level prediction also needs a defined holdout task (NIST/SEMATECH, n.d.). The proposed plan would state the prediction target and show where unit-level information enters. It would not treat an attractive in-sample relationship as evidence that the model predicts unseen people, lots or applications.
What a buyer could commission
Commission a dependence-aware comparison that identifies the association of interest, available design and observations needed to challenge it.
The proposed deliverable would include a repeated-unit map, the intended association, a descriptive display and an analysis specification appropriate to the available design. It could identify whether a question is answerable from deposited observations or requires a new controlled comparison. A formulation or application-science partner could use that record to avoid commissioning the wrong statistical analysis. The scope would name endpoints, units, potential adjustments and exclusions, with uncertainty about unreported conditions retained. Additional model fitting or experimental work would require a separate delivery agreement; the existence of paired observations alone does not create a validated service outcome.
The next proposed comparison.
A within/between estimand specification and descriptive comparison
The next check should distinguish explanations
Test whether the apparent pooled relationship survives a clear separation of within-unit and between-unit explanations before extending its use.
If a pooled trend looks strong, ask whether it survives when the scientific question is stated within and between units. If it disappears within units, the application may require another explanation or a different experimental design. If the source lacks the necessary identifiers, recover them through an authorised source route or narrow the question. The proposed next check should therefore reveal the structure carrying the association before a sponsor relies on it. That is useful commercial analysis: it tells a team whether to investigate a material mechanism, collect independent units or stop treating a correlation as a development proposition.
Evidence engineering · Proposed development discussion
A development path
for this research.
The association a repeated-measure dataset can support
The people this could serve.Formulation, application-science and research-data teams.
The decision behind the project.
Your intended application, present alternative and the constraint behind “Within and between specimen effects”. A useful starting point is the change that would justify another development step.
A proposed scope centred on a within/between estimand specification and descriptive comparison.
Agree the comparison, acceptance basis and responsible expertise before discussing a paid package.
Specialist methods and a defined contribution.
Bring relevant analytical or experimental expertise, a comparator suggestion, and the conditions under which your method could resolve this question.
A jointly scoped proposal could connect your contribution with a within/between estimand specification and descriptive comparison.
Define contributions, required access, independent preparations and permitted outputs before work or protected information is shared.
The commercial purpose of the next milestone.
Bring the application or asset thesis, the buyer you expect to value it, and the scientific or commercial uncertainty that must be resolved before another commitment.
The proposed decision record would connect the buyer question with a within/between estimand specification and descriptive comparison. It would identify what a subsequent comparison needs to demonstrate.
Demand, development cost, resources, rights and a decision-changing milestone would need examination. No financing terms or investment outcome are represented here.
What a useful comparison
would record.
- Record predictor in observation j from unit i (Original predictor unit) with the preparation, method and conditions needed to interpret it.
- Record mean predictor within unit i (Original predictor unit) with the preparation, method and conditions needed to interpret it.
- Retain the source identifier, preparation identity, conditions and permitted reuse.
What the proposed
brief could leave behind.
- A within/between estimand specification and descriptive comparison
- A preparation, comparator and endpoint record tied to this question.
- A source-linked account of the evidence gaps, study assumptions and next review point.
The proposed discussion brief
A brief you
can take away.
Use non-confidential notes to connect this topic with your own decision. Entries stay in this browser tab; saving creates a text file on your device. Nothing is submitted.
A scoped evidence/data decision brief or qualified comparison programme; recurring or API access only after delivery capacity and repeat buyer value are established.
Notes are optional. The brief includes this page’s proposed question and public source links.
Proposed scope; resources, methods, rights and delivery terms would be agreed before work. Begin with non-confidential information.
Sources and reading notes.
External work is identified below. Access notes describe the material inspected for this page; company research questions and proposed work are not presented as findings from these sources.
Fehlberg et al, 2025Fehlberg et al. (2025) A Touch of Stribeck—Finger-Pad Friction in Viscous Liquid Spreading: deposited dataset
Trial medians, coded participant and liquid identities, rheological parameters and source metadata.
Scope and limitations: 165 repeated trial medians from 11 people; 77-row fit is descriptive, not predictive validation.
Open original sourceBland and Altman, 1994Bland and Altman (1994) Correlation, regression, and repeated data
Separates within- and between-subject questions; repeated measurements are dependent.
Scope and limitations: Statistical note, not evidence of a Bio effect or licence to republish its tables.
Open original sourceWoodbridge et al, 2025Woodbridge et al. (2025) Rheological Characterisation of a Carbopol Microgel and a Commercial Body Lotion: Steady, Oscillatory, and Parallel Superposition Measurements
Acquired six source CSVs: 392 flow, amplitude and frequency rows for two source specimens; source units and README conditions.
Scope and limitations: CC BY 4.0 version 1; two studied specimens, not independent lots or a current commercial product specification. No parallel-superposition or slip data analysed.
Open original sourceNIST/SEMATECH, n.d.NIST/SEMATECH (n.d.) e-Handbook: How can I tell if a model fits my data?
Residual definition and why R² alone cannot validate a model.
Scope and limitations: General modelling guidance, not a Bio model-validation result.
Open original source
