PSLE Science Reality Lab Vol No.444 | “Registered Report: Accepted in Principle” — Were the Results Already Confirmed?

Series ID: PSLE-SCI-REALITY-0444

Wait, What? A Journal Can Agree to Publish Before the Main Results Exist

A student opens a research page and sees a serious-looking label: Registered Report — Accepted in Principle. The study is about whether two light schedules affect the growth of young plants. The student says, “If the journal already accepted it, the scientists must already know which light schedule works better.”

But that is exactly what makes the communication object interesting. In a Registered Report, an important part of peer review can happen before the main research outcomes are known. At Stage 1, editors and reviewers examine the research question and planned method. If the plan meets the journal’s requirements, the study can receive in-principle acceptance. The researchers then carry out the study, obtain the observations, analyse them according to the approved plan, and later return for Stage 2 review.

For a Primary 5/6 learner, the evidence-transfer job is precise: read “accepted in principle” as a judgement about an approved research plan and a conditional publication commitment, not as advance confirmation that the later results are true, positive, dramatic or even supportive of the original prediction. That habit fits the current 2026 PSLE Science emphasis on scientific inquiry: making predictions or hypotheses, interpreting and analysing information, evaluating methods and observations, and communicating explanations and reasoning.

Quick Answer

No. “Accepted in Principle” in a Registered Report does not mean the study’s later results were already confirmed. It normally means that a journal has reviewed the Stage 1 research question and planned methodology and has made a conditional commitment to publish the completed work if the agreed protocol and quality requirements are followed and the later report meets the journal’s Stage 2 requirements.

Stage 1 asks whether the question and plan are scientifically worth testing and sufficiently rigorous for the journal’s policy.
In-principle acceptance occurs before the main outcomes are known in the Registered Reports model.
Data collection and the actual results come later.
A result can support the prediction, fail to support it, or reveal a different pattern.
Stage 2 checks matters such as adherence to the approved plan, quality controls and whether interpretation is justified.
Journal policies differ in details, so the specific journal’s Registered Report policy remains the authoritative guide.

The quiet rule is: approval of the plan is not advance approval of the answer.

The Exact Learner Job — and the Boundary Around It

This Reality Lab owns one communication object: how to interpret the label “Registered Report: Accepted in Principle” without confusing Stage 1 review of the planned research with confirmation of the later observations or conclusions.

It does not replace broader owners for prediction, hypothesis, variables, fair testing, peer review, preregistration, method limitations, evidence selection or conclusions. A previous Reality Lab already asks whether a preregistered study must therefore be correct. This page has a narrower job: the reader is looking at a journal publication status that links a reviewed plan to a later Stage 2 article.

Freeze the Timeline Before You Judge the Label

Moment	What exists	What can be judged	What is not yet known
Question formation	A scientific question and possible hypothesis	Whether the question is meaningful and testable	The main study outcome
Stage 1 submission	Introduction, planned method, planned analyses and required quality checks	Whether the plan is capable of testing the stated question under the journal’s policy	The main study outcome
In-principle acceptance	An approved Stage 1 plan plus a conditional publication commitment	The plan has passed the journal’s Stage 1 requirements	Whether the prediction will be supported
Study execution	Actual observations and records begin to accumulate	Whether the approved method is being carried out competently	The final interpretation until analysis is complete
Stage 2 review	Results, analyses, deviations, discussion and quality evidence	Whether the completed work followed the approved route sufficiently and whether conclusions fit the evidence	Nothing should be assumed merely from the old Stage 1 label
Final published Registered Report	The full study record as accepted for publication	The reader can inspect question, method, results, deviations and interpretation together	Universal truth beyond the study’s evidence limits

This timeline is the core protection. If you do not know which stage you are looking at, you can easily attach evidence from the future to a document from the past.

Original Composite Case: The Seedling Light Plan

A fictional research team asks whether two daily light schedules affect the increase in stem length of bean seedlings over 14 days. Before collecting the main study data, the team submits a Registered Report Stage 1 manuscript.

The plan states that the researchers will use 80 seedlings from the same seed batch, randomly assign 40 to Schedule A and 40 to Schedule B, keep water amount, growing medium, container size and starting conditions as comparable as practical, record stem length using the same measurement procedure, define exclusion rules before seeing outcomes, and analyse the difference in mean growth using the stated method. They also specify a quality check: both groups must receive the intended light schedule within a narrow monitored range.

Reviewers ask the team to improve the randomisation description and explain how they will handle a failed lamp. After revision, the journal gives the Stage 1 manuscript in-principle acceptance.

At that moment, which statement is supported?

Supported: the journal has accepted the research plan in principle under its Registered Report policy.
Supported: reviewers have had an opportunity to challenge important design choices before the main outcomes were known.
Not supported: Schedule A has been proved better.
Not supported: Schedule B has been proved better.
Not supported: the study is guaranteed to produce interpretable data regardless of what happens during execution.
Not supported: every future Stage 2 conclusion will automatically be correct.

The plan has been reviewed. The plants have not yet supplied the answer.

Three Possible Futures After the Same Stage 1 Acceptance

The same Stage 1 acceptance can lead to very different scientific outcomes.

Future A: The prediction is supported

Schedule A seedlings grow more on average than Schedule B seedlings, the quality checks pass, the analysis follows the approved plan, and the Stage 2 interpretation stays within the evidence. The final article may conclude that the study found a difference under the tested conditions.

Future B: The prediction is not supported

The two groups show no clear difference large enough for the planned analysis to support the prediction. That is still a scientific outcome. One purpose of Registered Reports is to reduce the incentive to publish only results that look exciting or positive. A carefully executed study can be informative even when the prediction is not supported.

Future C: A critical quality check fails

One lamp repeatedly switches off, so Schedule B receives much less light than the approved method required. Now the researchers must document what happened. Depending on the journal’s policy and the seriousness of the problem, Stage 2 review may decide that the planned inference is no longer justified or that special treatment is needed. In-principle acceptance is not a magic shield against poor execution.

All three futures began with the same Stage 1 label. Therefore the label cannot tell you which future occurred.

Observed, Claimed and Inferred

Suppose a journal page says:

Registered Report — Stage 1 Accepted in Principle

Observed: the page identifies the work as a Registered Report at Stage 1.
Observed: the journal states that the Stage 1 manuscript has received in-principle acceptance.
Claimed by the status: the research plan has passed the journal’s pre-outcome review requirements and has a conditional publication commitment under that policy.
Not observed: the main study results.
Not established: that the hypothesis will be supported.
Not established: that no protocol deviations or quality problems will occur.
Not established: that the final interpretation will be accepted without Stage 2 review.

The evidence object is telling you something valuable. Scientific discipline means letting it tell you exactly that—and not more.

Representation Check: The Word “Accepted” Can Pull the Reader Too Far

“Accepted” is a strong everyday word. In school, an accepted answer is usually a correct answer. In a Registered Report, accepted in principle belongs to a specific publication workflow. The phrase does not mean “the eventual scientific claim has been accepted as true.” It means the journal has conditionally accepted the planned study for publication according to its Registered Report policy before knowing the main outcomes.

When a badge, icon or headline compresses the phrase to “Accepted,” recover the missing context:

Accepted what?
At which stage?
Before or after the main data were observed?
Under what conditions can the publication commitment be lost?
Is the page showing Stage 1, Stage 2 or the final article?

The shorter the badge, the more important the surrounding policy can become.

Registered Report Is Not Just Another Name for Preregistration

Preregistration and Registered Reports are related but not identical communication objects. A preregistration records aspects of a study plan before specified research steps, according to the registration system used. A Registered Report adds a journal workflow in which the Stage 1 plan is peer reviewed before the main outcomes are known and may receive in-principle acceptance.

Object	Core job	What it does not automatically prove
Preregistration	Creates a time-stamped record of planned research decisions	That the plan was peer reviewed or that the result is correct
Registered Report Stage 1	Presents the question and planned method for pre-outcome journal review	That the later observations support the hypothesis
In-principle acceptance	Signals conditional journal commitment after Stage 1 review	That data collection is complete or that Stage 2 quality is already known
Final Registered Report	Presents the completed study after the Registered Report workflow	That the study is universally correct or applies beyond its evidence

Do not use this table to rank one object as “good” and another as “bad.” They perform different jobs.

Method Check: A Strong Plan Reduces Some Risks, Not Every Risk

Reviewing a plan before outcomes are known can improve the timing of scientific criticism. Reviewers can question sample size, controls, measurements, planned exclusions and analyses before researchers know which choices would make a particular result look stronger. That is valuable.

But no research format can make all later problems impossible. During execution:

equipment can fail;
samples can be lost or contaminated;
participants or specimens can differ unexpectedly;
a manipulation can fail to produce the intended condition;
measurement quality can be weaker than expected;
protocol deviations can occur;
new exploratory questions can emerge;
the data can reveal that the original model was incomplete.

The Registered Report format does not ask scientists to pretend surprises never happen. It creates a clearer record of which decisions were planned before the outcomes and which analyses or explanations were added afterward.

Why Outcome-Neutral Quality Checks Matter

Suppose a study asks whether fertiliser type changes seedling growth. The researchers predict that Fertiliser A will increase growth. A bad quality check would be “the prediction must come true.” That confuses scientific quality with getting the hoped-for result.

A better quality check is independent of whether A wins: for example, the planned watering amounts were actually delivered, the growth measurement method worked within its useful range, the seedlings were alive at the starting point, and the treatment labels were applied correctly. These checks ask whether the experiment was capable of testing the question.

This is one reason Registered Report guidance often discusses outcome-neutral checks. The learner lesson is broader: judge the quality of an investigation by whether it can answer the question fairly and clearly, not by whether it produces the expected answer.

Worked Case 1: The Prediction Fails but the Study Succeeds

A team predicts that seedlings exposed to a patterned light cycle will grow taller than seedlings under a constant light cycle. Their Stage 1 plan receives in-principle acceptance. They conduct the experiment carefully, the planned quality checks pass, and the two groups show nearly the same average increase in height.

A student says, “The experiment failed because the prediction was wrong.”

That confuses prediction outcome with investigation quality. If the study was capable of testing the prediction and was executed adequately, a result that does not support the prediction can still be scientifically useful. The Stage 1 acceptance was not a promise that the hypothesis would win.

Worked Case 2: The Prediction “Wins” but a Quality Check Fails

A fictional study predicts that Treatment P increases the rate of a reaction. The final numbers appear to support P strongly. However, the temperature-control log shows that the P samples were consistently 8°C warmer than the comparison samples because one incubator malfunctioned.

The exciting result does not repair the failed control condition. The alternative explanation—temperature rather than Treatment P—must be considered. A Registered Report label does not allow the reader to skip the method evidence.

Transfer: a desirable outcome cannot compensate for a broken comparison.

Worked Case 3: A Necessary Protocol Deviation

During a plant experiment, a sensor model is recalled for a safety fault. The researchers replace it with another instrument. They document the change, test whether readings are comparable and explain the deviation in Stage 2.

The existence of a deviation does not automatically make the study invalid. The scientific questions are: Why did the deviation occur? Was it documented? Could it change the measured outcome? What evidence was gathered to understand its effect? How does the journal’s policy treat the change?

Rigidly chanting “follow the plan” is not scientific reasoning. Transparent deviation plus evidence is more informative than pretending nothing changed.

Worked Case 4: Exploratory Finding After the Planned Analysis

The approved plan compares average stem growth between two light schedules. After completing that analysis, the researchers notice that leaf angle may also differ between the groups. They explore the pattern and report it clearly as an exploratory analysis rather than pretending it was the original prediction.

This is not automatically bad science. Exploration is how new questions can begin. The evidence problem appears when a later-discovered pattern is presented as though it had been predicted and pre-specified all along. A good scientific record keeps planned confirmation and later exploration distinguishable.

Worked Case 5: A Stage 1 Screenshot Circulates After Stage 2 Exists

A social post shares an old Stage 1 page saying “Accepted in Principle” and claims, “The journal has endorsed this result.” Meanwhile, the full Stage 2 article has already been published with a result that did not support the original hypothesis.

The error is partly one of status freshness. A maintained publication record can change as the study moves through its stages. Always check the current journal page and linked final article before interpreting an old screenshot.

Provenance Check: Find the Stage, Policy and Date

When you encounter a Registered Report label, reconstruct its provenance.

Which journal issued the label?
Is the object Stage 1, Stage 2 or the final article?
What date was that status issued?
Does the journal link its Registered Report policy?
Is there a DOI or persistent identifier connecting stages?
Has the Stage 2 article appeared?
Were deviations or quality-check failures reported?
Which analyses were planned and which were exploratory?

This is not paperwork for its own sake. The stage determines what evidence could possibly have existed when the label was issued.

Comparison Check: AIP Versus Ordinary Final Acceptance

Two articles can both contain the word “accepted” while referring to different decisions.

Label	Timing	Main object being judged
Registered Report: Accepted in Principle	Before the main outcomes are known	Question, design, planned analyses and specified quality conditions
Final article accepted	After the completed manuscript is reviewed	The manuscript as completed, including actual results and interpretation

Do not assume every journal uses exactly the same labels or workflow. Read the specific journal policy. The table protects the central timing distinction, not a universal vocabulary rule.

Alternative Explanations: Why Might a Final Result Differ From the Prediction?

If a well-planned study does not support its prediction, many scientific possibilities remain.

The hypothesised effect may be absent under the tested conditions.
The effect may be smaller than expected.
The measurement may not be sensitive enough to detect the relevant difference.
The tested population or specimens may respond differently from those in earlier work.
The causal mechanism may require another condition not included in the original model.
Ordinary variation may make the observed pattern uncertain.
A quality problem may have weakened the study’s ability to test the question.

“The hypothesis was wrong” is only one possible interpretation. The evidence decides how narrow or broad the conclusion can be.

What Would Strengthen Confidence in the Final Registered Report?

The Stage 1 plan is accessible or clearly described.
The main analyses match the approved plan or deviations are transparently explained.
Outcome-neutral quality checks pass or failures are disclosed and handled appropriately.
The measurements actually represent the claimed variables.
Missing data and exclusions are explained.
Exploratory analyses are distinguished from planned analyses.
The Stage 2 conclusions match the observed evidence and uncertainty.
Independent later studies test whether the result transfers.

What Would Weaken a Strong Public Claim Based Only on AIP?

The speaker has seen only a Stage 1 badge but talks as if final results exist.
The main data had not yet been collected when the quoted status was issued.
The public claim says the journal “proved” a hypothesis before the experiment.
A Stage 1 plan is presented as though it were a completed result.
The final Stage 2 article is ignored even though it is now available.
Protocol deviations or failed quality checks are omitted.
An exploratory finding is described as the original registered prediction.

How Far Can the Label Travel?

Across stages: Stage 1 approval does not automatically describe Stage 2 evidence.
Across results: AIP does not tell you whether the prediction was supported.
Across methods: an approved method can still encounter execution problems.
Across journals: details of Registered Report policies vary, so read the relevant policy.
Across scientific claims: a well-run Registered Report is still one contribution to a wider evidence base.
Across populations and conditions: final results remain bounded by what was studied.

The label has real meaning. Its meaning is just narrower than “science has already confirmed the answer.”

Tempting but Invalid Reasoning

“Accepted in principle means the hypothesis was accepted.” The Stage 1 plan was accepted under the journal’s policy, not the unknown outcome.
“Peer reviewed means the result must already exist.” Registered Reports deliberately move important peer review earlier.
“If the final result is negative, the Registered Report failed.” Publication is designed not to depend on whether the result is positive or supports the prediction, provided required quality and reporting conditions are met.
“The method was reviewed, so execution cannot go wrong.” A good plan and a good execution are separate evidence layers.
“Any change from the protocol destroys the study.” Deviations need transparent evaluation; their importance depends on what changed and how it affects the claim.
“Exploratory analyses are forbidden.” They can be valuable when clearly distinguished from the pre-specified analyses.
“A final Registered Report is automatically true.” It still has measurements, assumptions, uncertainty and scope limits like other scientific work.

Model and Measurement Limits Still Apply

A Registered Report can improve when decisions are scrutinised, but it cannot make a weak measuring instrument magically stronger. If stem length is measured with a method too coarse to detect the expected difference, Stage 1 review should ideally catch that problem. If it does not, the limitation remains. If a model assumes that one measured signal perfectly represents a biological process when it does not, the Registered Report label does not erase that model limitation.

Keep asking the ordinary PSLE Science questions:

What was changed?
What was measured?
What was kept comparable?
What alternative explanation remains?
What does the evidence actually show?
How far can that conclusion travel?

The publication format helps organise evidence. It does not replace scientific inquiry.

PSLE-Style Transfer Case

A fictional journal gives Stage 1 in-principle acceptance to this plan:

Compare the cooling time of identical cups wrapped with Material P or Material Q. Use equal starting water volumes and temperatures, the same room, the same cup type and the same endpoint temperature. Predict that Material P will produce a longer cooling time.

Question 1: What does the in-principle acceptance tell us?

Answer: It tells us that the planned research has passed the journal’s Stage 1 review requirements for a Registered Report. It does not tell us whether Material P actually produces a longer cooling time.

Question 2: The final experiment finds similar cooling times for P and Q. Does that contradict the fact that Stage 1 was accepted?

Answer: No. Stage 1 acceptance concerned the plan before the outcome was known. A result that does not support the prediction is an allowed scientific outcome if the study is executed and interpreted appropriately.

Question 3: During the experiment, cups with Material P start at 85°C while cups with Material Q start at 70°C. Why does this matter?

Answer: Starting temperature is no longer comparable, so a difference in cooling time could be explained partly by the different starting condition. The quality of execution must be evaluated even though the plan was reviewed earlier.

Question 4: After seeing the results, the researchers notice that cup colour may matter and run an extra analysis. Can they report it?

Answer: They can report exploratory work if the journal permits it and it is clearly distinguished from the analysis planned before seeing the outcomes. The reader must know which was planned and which was discovered later.

Explained Practice: Identify the Stage Before the Claim

Read each fictional statement and decide what is safe to conclude.

“Stage 1 Registered Report accepted in principle.” The plan has passed pre-outcome review under the journal’s policy; the main study result is not implied.
“Data collection complete; Stage 2 submitted.” Results now exist, but final review may still be in progress.
“Final Registered Report published.” Read the completed results and discussion; do not infer the outcome from the earlier prediction.
“Exploratory analysis found a new pattern.” Treat it as a later-discovered result unless the record shows it was part of the approved Stage 1 plan.
“Quality check failed.” Investigate whether the study could still answer the planned question before relying on the final claim.

Delayed Independent Return

Tomorrow, draw six boxes on a blank page: Question → Stage 1 Plan → Stage 1 Review → In-Principle Acceptance → Data and Analysis → Stage 2 Review. Without looking back, write one piece of information that becomes available at each box and one thing that still remains unknown.

Then invent three endings to the same approved plan: the prediction is supported, it is not supported, or a critical quality check fails. Explain why all three endings can begin from the same in-principle acceptance.

If you can do that, you have stopped treating publication labels as magic quality words and started reading them as evidence about a scientific process.

Useful eduKateSengkang Routes

Parent and Tutor Teaching Guide

Teach this page with a timeline rather than a definition list. Write a prediction on the left side of a sheet and leave the right side blank. In the middle, place a card saying Stage 1 Accepted in Principle. Ask the learner to write everything that is known at that point. Stop them if they try to invent the result.

Next, prepare three sealed ending cards. Card A says the prediction was supported. Card B says it was not supported. Card C says a critical control failed. Ask the learner whether any of those endings would make the earlier Stage 1 label impossible. The answer is no: the label was issued before the ending was known.

Then introduce a protocol deviation. For example, one temperature probe breaks and is replaced. Ask two questions instead of one: Was there a deviation? and Does the deviation damage the claim? The first is factual; the second requires evidence.

Finally, give the learner a planned analysis and a surprising extra pattern noticed after the data are seen. Ask them to label one planned and the other exploratory. Emphasise that exploratory does not mean useless. It means the timing of the idea is different and should be communicated honestly.

Authoritative Sources and Further Reading

The Center for Open Science describes core Registered Report features as peer review before the research outcomes are observed and in-principle acceptance that is not revoked merely because of the outcome, while still depending on quality assurance, adherence to the registered protocol and acceptable reporting. Nature Human Behaviour similarly explains that in-principle acceptance occurs before the research outcomes are known. Specific journal rules should always be checked because implementation details vary.

The Quiet Return

Science becomes easier to misunderstand when one short label is asked to carry an entire investigation. “Accepted in Principle” carries a real and useful message: a planned study has passed an important review stage before the main outcomes were known, and the journal has made a conditional publication commitment under its Registered Report rules.

It does not tell you which result the plants, sensors, samples or people will produce. It does not excuse a failed control. It does not transform an exploratory pattern into a prediction. It does not replace Stage 2 review.

So when you see the badge, locate it on the timeline. Ask what had happened by then, what had not happened yet, and what later evidence must still be checked.

That small act—putting evidence back into time—is one of the most reliable ways to read scientific communication without being pulled beyond what the evidence can actually say.

Small Group Tutorials