Evidence literacy

Could the aquarium observer see the exposure group?

Follow an allocation key through a visible tank marker in a fictional masked-video study, without claiming that bias actually occurred.

By FishyMag · Published

Original research-appraisal case · Not a clinical review or treatment recommendation.

Quick answer

In the fictional packet, filenames are coded and scoring order is shuffled, but the scorer knows the group meaning of a visible tank marker. Group identity is still available through the frame. That identifies an information path; it does not prove the observer’s scores were biased.

AI-generated illustration of two keepers looking toward an aquarium with a blank hanging tag; no actual research experiment or measured result is shown.
AI-generated editorial illustration. Also used in related FishyMag guides. Not a photograph of FishyMag testing or a treatment demonstration.

Follow the information, not just the word masked

ARRIVE treats awareness of allocation as stage-specific. A person may be unaware during one task and informed during another. This fictional case concerns only video outcome assessment: filenames have been replaced with codes, clips are shuffled, and a coloured tank marker remains inside each frame. The scorer knows that marker’s group meaning. No actual video or investigator is being evaluated.

Complete the identity-path audit

A coded filename closes one route to identity. It does not erase an informative cue in the image. The second branch below changes only the available evidence about the scorer’s knowledge.

Complete the identity-path audit: worked table
Information routeRecord in the main caseBounded finding
Filename to allocationCodes used; scoring list shuffledDirect filename identity hidden as recorded
Allocation key to markerScorer knows the marker keyGroup meaning available
Marker to frameMarker visibleCue reaches the scorer
Frame to group identityKnown key plus visible cueMasking incomplete for this route
Alternative: scorer’s key knowledge absent from packetVisible marker, knowledge unknownActual identity availability unresolved
Bias in the scoresNot measured in this exerciseDo not claim demonstrated observer bias

Ask the question that separates the two branches

When the key-knowledge record is missing, ask whether the scorer could interpret the cue during assessment. Do not replace unknown knowledge with either successful masking or certain unmasking. Retain who held the code, when access occurred, what remained visible and which endpoint was scored. The corrected main-case summary is precise: filenames were masked, while the known marker still disclosed group identity. A later statistical adjustment cannot be assumed to have removed that information path. Whether the path changed scores requires further evidence beyond the existence of the cue.

Use your own details

Could the aquarium observer see the exposure group? worksheet

Download the editable text file, fill it in on your device and keep it with the relevant record. This blank worksheet records your information; it does not verify entries or approve treatment.

  1. Outcome and scoring stage
  2. Filename coding and key access
  3. Visible markers or other group cues
  4. Who knew each cue’s meaning
  5. Frame, metadata and audio review coverage
  6. Scoring order and allocation masking
  7. Information paths still open
  8. Unknown key access and clarification request
Download worksheet (.txt)

Put it into practice

Your checklist

  • Audit identity cues at the actual scoring stage.
  • Separate a visible cue from knowledge of its key.
  • Do not equate incomplete masking with demonstrated scoring bias.

Sources and scope

This article distinguishes manufacturer statements from FishyMag’s own workflows and arithmetic examples. Linked formulation records retain their own evidence dates; publishing this article does not refresh those records.

AI-assisted writing and development. Not hands-on product testing or veterinary review. How we publish →

FishyMag reports what manufacturers say, quoted and linked. It does not diagnose illness or recommend a treatment, and it never infers that a product is safe from the absence of a warning — where no statement exists, it says so.

Formulations change. Always read the label on the bottle in your hand, and contact the manufacturer if your tank contains anything valuable or unusual.

Continue exploring