Revision 7T · Full Tutorial Edition · Updated 1 September 2026
PWR-205 · RESEARCH full tutorial
Learn to tell whether a neurofeedback study changed its intended signal and produced a benefit beyond placebo feedback
This is a paper-reading tutorial, not neurofeedback training. You will trace what the sensors measured, how non-brain noise was handled, what participants saw, what the comparison group received and which results were named before the data were known. You will then separate three questions: did people change the intended signal, did neurofeedback outperform believable placebo feedback, and did any benefit remain when feedback was absent?
What you will produceThe learner completes every field in a two-study audit, using “not reported” where necessary, and writes separate conclusions for signal learning, the neurofeedback-versus-sham difference, functional benefit and no-feedback transfer.
Method10 numbered Power-specific steps
Practice authorityComplete research method; no operational self-experiment
Full step-by-step individual tutorial · TLU-PWR-205
The learner completes every field in a two-study audit, using “not reported” where necessary, and writes separate conclusions for signal learning, the neurofeedback-versus-sham difference, functional benefit and no-feedback transfer.
Read the two supplied papers and complete the supplied evidence tables.
Record what each paper reports about sensors, noise removal, feedback, comparison group, blinding, outcomes and uncertainty.
Write a research-only conclusion that keeps signal change separate from clinical or functional benefit.
Qualified help is required for
Operating neurofeedback equipment, selecting a protocol or interpreting an individual EEG recording.
Diagnosis, treatment, medication changes or advice based on neurofeedback results.
Research involving participants, clinical data, ethics approval or a formal statistical reanalysis.
Never do this from the page alone
Connect sensors, buy equipment or copy published settings as a home protocol.
Assume that movement of a feedback display proves control of a brain signal.
Call improvement within one group a neurofeedback effect without comparing it with the sham group.
2 · Get ready
Gather what you need and check the starting conditions
What you need
Study A: “Better than sham? A double-blind placebo-controlled neurofeedback study in primary insomnia” — https://pubmed.ncbi.nlm.nih.gov/28335000/
Study B: “Neurofeedback-induced desynchronization of sensorimotor rhythm elicits pre-movement downregulation of intracortical inhibition that shortens simple reaction time in humans: A double-blind, sham-controlled randomized study” — https://pubmed.ncbi.nlm.nih.gov/40800510/
Protocol table with these columns: study; participants; intended EEG feature; sensor location; sampling or filtering; eye, muscle and movement handling; calibration; feedback shown; threshold changes; sessions; trial length; coaching.
Comparison table with these columns: sham feedback; matched contact time; participant blinding; assessor blinding; blinding check; outcome named in advance; signal result; neurofeedback-minus-sham result; uncertainty; adverse events; result without feedback; delayed result.
A conclusion sheet with four headings: “Signal learned?”, “Better than sham?”, “Meaningful function?”, and “Retained without feedback?”
Before you start
Keep this as a paper-only exercise. Do not connect a sensor, choose a protocol or change treatment.
Write these definitions at the top: “artefact” means eye, muscle, movement or equipment activity that can look like EEG; “sham” means believable comparison feedback; “between-group result” means the difference between neurofeedback and sham.
Open each paper's methods, supplement and registration when available. If a required detail cannot be verified, write “not reported” instead of guessing.
Keep the two studies in separate columns; they use different populations, targets and outcomes.
3 · The method
Follow these steps in order
Write the three questions
For each study, ask separately: Did participants change the intended EEG feature? Did neurofeedback outperform sham? Did it improve the named clinical or functional outcome?
Why: A changed EEG feature, a specific treatment effect and a useful outcome are different claims.
Check: The worksheet has three separate questions rather than one general “did it work?” question.
Trace the signal from sensor to score
Copy the sensor type and location, sampling or filtering, intended EEG feature, calibration rule and the rule for turning EEG into a score. Write “not reported” for every missing item.
Why: The reader needs to know what was actually measured before interpreting the feedback.
Check: Another reader can follow the paper from the sensor recording to the reported target-signal value.
Check eye, muscle and movement artefacts
Record how the study detected, rejected or corrected eye movements, facial or scalp muscle activity, body movement and poor sensor contact. Note any sensitivity check.
Why: These signals can change the display without showing control of the intended brain activity.
Check: Every artefact category has a reported method or the words “not reported.”
Reconstruct what feedback participants received
Record the picture, sound or score shown; what signal changed it; how thresholds changed; session count; trial length; rest; and coaching. Do not turn these fields into operating instructions.
Why: The intervention cannot be judged if the feedback rule and exposure are unknown.
Check: The paper audit describes the intervention while remaining explicitly non-operational.
Audit sham and blinding
Write exactly what sham participants saw, whether session time and staff contact matched, who was blinded, and whether participants were asked which group they thought they were in.
Why: Expectation, attention and repeated contact can improve outcomes in both groups.
Check: The table shows whether the sham was believable and whether blinding was tested rather than merely claimed.
Extract outcomes named before the results
Find the registration or protocol. Copy the primary signal and functional outcomes, their time points and responder rules. Label later-added or exploratory outcomes separately.
Why: Choosing an attractive outcome after seeing the data can exaggerate the apparent result.
Check: Every reported conclusion is labelled primary, secondary, exploratory or not verifiable.
Copy the neurofeedback-minus-sham result
For each primary outcome, record the difference between groups with its confidence interval, uncertainty or exact reported statistic. Keep within-group improvement in a separate column.
Why: Improvement in the neurofeedback group alone does not show that neurofeedback caused the improvement.
Check: The conclusion is based on the group comparison, not on one group's before-and-after change.
Check performance without feedback
Look for a trial in which feedback was removed and for a delayed follow-up. Record the result and time point, or write “not tested.”
Why: Controlling a display while it is present is not the same as retaining the skill or using it in daily life.
Check: Device-on performance, no-feedback performance and delayed function occupy separate cells.
Write one verdict per study
Answer the four conclusion headings using one or two sentences each, then list the largest missing detail and any adverse event information.
Why: A structured verdict prevents one positive result from standing in for every unanswered question.
Check: The verdict states what changed, what did not, what remains uncertain and what was not tested.
Compare without mixing the studies
Place the two verdicts side by side. Compare methods only where the fields mean the same thing, and do not import the reaction-time result into an insomnia claim.
Why: Different EEG targets and outcomes do not become interchangeable because both studies use the word neurofeedback.
Check: The final paragraph preserves the population, target and outcome of each study.
4 · Worked example
See the whole method used once
Scenario
A learner audits the supplied insomnia paper first, then uses the sensorimotor-rhythm paper as a transfer check.
Walkthrough
They create separate rows for the intended EEG feature, artefact handling, feedback rule, sham feedback and blinding check.
They record the target-signal result separately from the insomnia outcome and copy the neurofeedback-minus-sham comparison rather than only the neurofeedback group's improvement.
The insomnia row shows that target-signal learning and a sham-specific clinical benefit are not the same result.
Where a no-feedback retention or delayed functional result cannot be verified, they write “not reported” rather than assuming transfer.
For Study B, they record its sensorimotor-rhythm, intracortical-inhibition and simple-reaction-time outcomes in their own column.
They conclude that Study B cannot be used as evidence for insomnia treatment or a general brain-upgrade claim.
Result
The completed audit tells the reader exactly which signal, sham and functional claims each study can support. It supplies no neurofeedback setup or treatment recommendation.
5 · Right and wrong
Compare correct or safer execution with the common wrong version
Right and wrong comparison
Moment
Right / safer
Wrong / riskier
Why it matters
Reading the signal result
Check the sensor-to-score path and the eye, muscle and movement rules before accepting a target-signal change.
Assume that a responsive animation proves control of brain activity.
Non-brain activity and equipment noise can also move the display.
Judging improvement
Use the neurofeedback-minus-sham result and its uncertainty.
Point only to improvement from before to after in the neurofeedback group.
Time, expectation, attention and repeated testing can change both groups.
Interpreting target learning
Report target-signal learning and symptom or functional outcome in separate sentences.
Call a changed EEG feature successful treatment.
Changing the trained feature does not automatically produce a meaningful benefit.
Checking retention
Look for testing without feedback and at a later time, then write “not tested” when absent.
Treat good performance while feedback is visible as a retained everyday skill.
Device-on control does not establish independent or delayed transfer.
Using published settings
Keep electrode and threshold details inside the paper audit.
Copy them into a home neurofeedback procedure.
Equipment operation, clinical selection and treatment require qualified oversight.
6 · Common mistakes
Spot the error and apply the correction
Common mistakes and corrections
Mistake
Fix
Missing methods are filled from another neurofeedback paper.
Write “not reported” and keep each study's protocol separate.
Eye or muscle artefact is not checked.
Add a separate row for eye, facial/scalp muscle, body movement and sensor-contact handling.
Within-group improvement is reported as the treatment effect.
Copy the neurofeedback-minus-sham estimate and its uncertainty first.
A changed EEG target is described as clinical success.
Put signal change and the named functional outcome in different conclusion boxes.
Feedback-on performance is treated as retained learning.
Find a no-feedback and delayed measure or label both as not tested.
7 · Practice
Turn the steps into a usable skill
First session
Write the three separate questions for Study A.
Complete its sensor, preprocessing, artefact, target and calibration fields.
Complete its feedback, dose, sham and blinding fields.
Copy the prespecified outcomes, neurofeedback-minus-sham result, uncertainty and adverse-event information.
Record no-feedback and delayed results or mark them not tested.
Write the four-part Study A verdict.
Repeat all six actions for Study B, then compare the verdicts without mixing their populations or outcomes.
Repeat plan
After at least seven days, start with blank tables and audit Study B again without looking at the first worksheet. Compare the two versions and correct omitted artefact, sham, prespecified-outcome or no-feedback fields. Use the method on a new paper only after every required field is either completed or explicitly marked not reported.
Progress when
Every protocol and result field contains a sourced value or “not reported.”
Signal learning, the neurofeedback-minus-sham result and meaningful function remain separate.
No-feedback and delayed transfer are checked explicitly.
The conclusion contains no equipment setup, diagnosis or treatment instruction.
Do not progress when
The intended signal, sham procedure or primary outcome cannot be identified.
The conclusion relies only on improvement within one group.
A published protocol is being converted into self-treatment or advice for another person.
8 · Check the result
Measure what changed
Completeness and correctness of the two-study signal–sham–function audit
How: Score the sensor path, artefact handling, feedback rule, sham, blinding, prespecified outcomes, between-group result, uncertainty, adverse events, no-feedback test, delayed test and four-part verdict for each paper.
Good result: Every field is sourced or marked not reported, and the verdict separates signal learning, sham-specific effect, meaningful function and retained transfer without offering a protocol.
This does not prove: A correct audit does not prove that neurofeedback works generally, that either result transfers to another condition, or that the learner can operate or prescribe neurofeedback.
Self-check
What exact EEG feature was intended to change?
How were eye, muscle, movement and sensor artefacts handled?
What did the sham group receive, and was blinding checked?
What was the neurofeedback-minus-sham result for the outcome named in advance?
Was performance tested without feedback or after a delay?
Which claim is supported, weakened, unresolved or not tested?
9 · Stop, adapt or get help
Keep the safety boundary practical
Stop and get help
Stop the audit when the intended signal, sham or primary outcome cannot be verified; record the gap rather than inventing it.
Do not convert published sensor locations, thresholds or session schedules into self-treatment.
Seek neuroscience, statistics and clinical expertise before making a decision about a patient, treatment or new participant study.
Accessibility and adaptations
Use one left-to-right table from sensor to artefact handling to feedback to outcome, with every abbreviation expanded on first use.
Provide a text version of figures and allow speech-to-text, a scribe or extra time.
If the signal-processing method cannot be understood reliably, work with a qualified reviewer and preserve the field as unresolved.
10 · Evidence and limits
Why these instructions are here
primary research
In the supplied double-blind primary-insomnia study, learning the intended neurofeedback signal did not establish a clinical benefit beyond sham, showing why signal and treatment outcomes must be separated.
The supplied sham-controlled sensorimotor-rhythm study reports its specified neurophysiological and simple-reaction-time outcomes; those results cannot establish insomnia treatment or broad neurofeedback benefit.
Progress is saved only in this browser on this device.
Step-by-step learner mode
Each activity includes its success check, a nearby accessible alternative and an “I’m stuck” correction path. Alternatives preserve the target where possible; when they change the task, Titan labels them as related rather than equivalent.
01
Write the three questions
For each study, ask separately: Did participants change the intended EEG feature? Did neurofeedback outperform sham? Did it improve the named clinical or functional outcome?
Why this step exists
A changed EEG feature, a specific treatment effect and a useful outcome are different claims.
Success check
The worksheet has three separate questions rather than one general “did it work?” question.
I’m stuck on this step
Reset: Re-read this authored instruction — “For each study, ask separately: Did participants change the intended EEG feature? Did neurofeedback outperform sham? Did it improve the named clinical or functional outcome?” — and its success check, then attempt only this step.
Possible snag: A changed EEG target is described as clinical success.
Correction: Put signal change and the named functional outcome in different conclusion boxes.
Stop / get help: Stop the audit when the intended signal, sham or primary outcome cannot be verified; record the gap rather than inventing it.
02
Trace the signal from sensor to score
Copy the sensor type and location, sampling or filtering, intended EEG feature, calibration rule and the rule for turning EEG into a score. Write “not reported” for every missing item.
Why this step exists
The reader needs to know what was actually measured before interpreting the feedback.
Success check
Another reader can follow the paper from the sensor recording to the reported target-signal value.
I’m stuck on this step
Reset: Re-read this authored instruction — “Copy the sensor type and location, sampling or filtering, intended EEG feature, calibration rule and the rule for turning EEG into a score. Write “not reported” for every missing item.” — and its success check, then attempt only this step.
Possible snag: The result from “Copy the sensor type and location, sampling or filtering, intended EEG feature, calibration rule and the rule for turning EEG into a score. Write “not reported” for every missing item.” does not yet meet this declared check: Another reader can follow the paper from the sensor recording to the reported target-signal value.
Correction: Return to the start of “Trace the signal from sensor to score”, reduce complexity or pace, and repeat only the part needed to satisfy: “Another reader can follow the paper from the sensor recording to the reported target-signal value.”
Stop / get help: Stop the audit when the intended signal, sham or primary outcome cannot be verified; record the gap rather than inventing it.
03
Check eye, muscle and movement artefacts
Record how the study detected, rejected or corrected eye movements, facial or scalp muscle activity, body movement and poor sensor contact. Note any sensitivity check.
Why this step exists
These signals can change the display without showing control of the intended brain activity.
Success check
Every artefact category has a reported method or the words “not reported.”
I’m stuck on this step
Reset: Re-read this authored instruction — “Record how the study detected, rejected or corrected eye movements, facial or scalp muscle activity, body movement and poor sensor contact. Note any sensitivity check.” — and its success check, then attempt only this step.
Possible snag: Eye or muscle artefact is not checked.
Correction: Add a separate row for eye, facial/scalp muscle, body movement and sensor-contact handling.
Stop / get help: Stop the audit when the intended signal, sham or primary outcome cannot be verified; record the gap rather than inventing it.
04
Reconstruct what feedback participants received
Record the picture, sound or score shown; what signal changed it; how thresholds changed; session count; trial length; rest; and coaching. Do not turn these fields into operating instructions.
Why this step exists
The intervention cannot be judged if the feedback rule and exposure are unknown.
Success check
The paper audit describes the intervention while remaining explicitly non-operational.
I’m stuck on this step
Reset: Re-read this authored instruction — “Record the picture, sound or score shown; what signal changed it; how thresholds changed; session count; trial length; rest; and coaching. Do not turn these fields into operating instructions.” — and its success check, then attempt only this step.
Possible snag: The result from “Record the picture, sound or score shown; what signal changed it; how thresholds changed; session count; trial length; rest; and coaching. Do not turn these fields into operating instructions.” does not yet meet this declared check: The paper audit describes the intervention while remaining explicitly non-operational.
Correction: Return to the start of “Reconstruct what feedback participants received”, reduce complexity or pace, and repeat only the part needed to satisfy: “The paper audit describes the intervention while remaining explicitly non-operational.”
Stop / get help: Stop the audit when the intended signal, sham or primary outcome cannot be verified; record the gap rather than inventing it.
05
Audit sham and blinding
Write exactly what sham participants saw, whether session time and staff contact matched, who was blinded, and whether participants were asked which group they thought they were in.
Why this step exists
Expectation, attention and repeated contact can improve outcomes in both groups.
Success check
The table shows whether the sham was believable and whether blinding was tested rather than merely claimed.
I’m stuck on this step
Reset: Re-read this authored instruction — “Write exactly what sham participants saw, whether session time and staff contact matched, who was blinded, and whether participants were asked which group they thought they were in.” — and its success check, then attempt only this step.
Possible snag: The result from “Write exactly what sham participants saw, whether session time and staff contact matched, who was blinded, and whether participants were asked which group they thought they were in.” does not yet meet this declared check: The table shows whether the sham was believable and whether blinding was tested rather than merely claimed.
Correction: Return to the start of “Audit sham and blinding”, reduce complexity or pace, and repeat only the part needed to satisfy: “The table shows whether the sham was believable and whether blinding was tested rather than merely claimed.”
Stop / get help: Stop the audit when the intended signal, sham or primary outcome cannot be verified; record the gap rather than inventing it.
06
Extract outcomes named before the results
Find the registration or protocol. Copy the primary signal and functional outcomes, their time points and responder rules. Label later-added or exploratory outcomes separately.
Why this step exists
Choosing an attractive outcome after seeing the data can exaggerate the apparent result.
Success check
Every reported conclusion is labelled primary, secondary, exploratory or not verifiable.
I’m stuck on this step
Reset: Re-read this authored instruction — “Find the registration or protocol. Copy the primary signal and functional outcomes, their time points and responder rules. Label later-added or exploratory outcomes separately.” — and its success check, then attempt only this step.
Possible snag: The result from “Find the registration or protocol. Copy the primary signal and functional outcomes, their time points and responder rules. Label later-added or exploratory outcomes separately.” does not yet meet this declared check: Every reported conclusion is labelled primary, secondary, exploratory or not verifiable.
Correction: Return to the start of “Extract outcomes named before the results”, reduce complexity or pace, and repeat only the part needed to satisfy: “Every reported conclusion is labelled primary, secondary, exploratory or not verifiable.”
Stop / get help: Stop the audit when the intended signal, sham or primary outcome cannot be verified; record the gap rather than inventing it.
07
Copy the neurofeedback-minus-sham result
For each primary outcome, record the difference between groups with its confidence interval, uncertainty or exact reported statistic. Keep within-group improvement in a separate column.
Why this step exists
Improvement in the neurofeedback group alone does not show that neurofeedback caused the improvement.
Success check
The conclusion is based on the group comparison, not on one group's before-and-after change.
I’m stuck on this step
Reset: Re-read this authored instruction — “For each primary outcome, record the difference between groups with its confidence interval, uncertainty or exact reported statistic. Keep within-group improvement in a separate column.” — and its success check, then attempt only this step.
Possible snag: Missing methods are filled from another neurofeedback paper.
Correction: Write “not reported” and keep each study's protocol separate.
Possible snag: Within-group improvement is reported as the treatment effect.
Correction: Copy the neurofeedback-minus-sham estimate and its uncertainty first.
Stop / get help: Stop the audit when the intended signal, sham or primary outcome cannot be verified; record the gap rather than inventing it.
08
Check performance without feedback
Look for a trial in which feedback was removed and for a delayed follow-up. Record the result and time point, or write “not tested.”
Why this step exists
Controlling a display while it is present is not the same as retaining the skill or using it in daily life.
Success check
Device-on performance, no-feedback performance and delayed function occupy separate cells.
I’m stuck on this step
Reset: Re-read this authored instruction — “Look for a trial in which feedback was removed and for a delayed follow-up. Record the result and time point, or write “not tested.”” — and its success check, then attempt only this step.
Possible snag: Feedback-on performance is treated as retained learning.
Correction: Find a no-feedback and delayed measure or label both as not tested.
Stop / get help: Stop the audit when the intended signal, sham or primary outcome cannot be verified; record the gap rather than inventing it.
09
Write one verdict per study
Answer the four conclusion headings using one or two sentences each, then list the largest missing detail and any adverse event information.
Why this step exists
A structured verdict prevents one positive result from standing in for every unanswered question.
Success check
The verdict states what changed, what did not, what remains uncertain and what was not tested.
I’m stuck on this step
Reset: Re-read this authored instruction — “Answer the four conclusion headings using one or two sentences each, then list the largest missing detail and any adverse event information.” — and its success check, then attempt only this step.
Possible snag: The result from “Answer the four conclusion headings using one or two sentences each, then list the largest missing detail and any adverse event information.” does not yet meet this declared check: The verdict states what changed, what did not, what remains uncertain and what was not tested.
Correction: Return to the start of “Write one verdict per study”, reduce complexity or pace, and repeat only the part needed to satisfy: “The verdict states what changed, what did not, what remains uncertain and what was not tested.”
Stop / get help: Stop the audit when the intended signal, sham or primary outcome cannot be verified; record the gap rather than inventing it.
10
Compare without mixing the studies
Place the two verdicts side by side. Compare methods only where the fields mean the same thing, and do not import the reaction-time result into an insomnia claim.
Why this step exists
Different EEG targets and outcomes do not become interchangeable because both studies use the word neurofeedback.
Success check
The final paragraph preserves the population, target and outcome of each study.
I’m stuck on this step
Reset: Re-read this authored instruction — “Place the two verdicts side by side. Compare methods only where the fields mean the same thing, and do not import the reaction-time result into an insomnia claim.” — and its success check, then attempt only this step.
Possible snag: The result from “Place the two verdicts side by side. Compare methods only where the fields mean the same thing, and do not import the reaction-time result into an insomnia claim.” does not yet meet this declared check: The final paragraph preserves the population, target and outcome of each study.
Correction: Return to the start of “Compare without mixing the studies”, reduce complexity or pace, and repeat only the part needed to satisfy: “The final paragraph preserves the population, target and outcome of each study.”
Stop / get help: Stop the audit when the intended signal, sham or primary outcome cannot be verified; record the gap rather than inventing it.
Correct versus incorrect execution
These accessible process diagrams are built from the tutorial’s own right/wrong teaching. They are not anatomical illustrations and do not add technique beyond the canonical tutorial.
Reading the signal result — Non-brain activity and equipment noise can also move the display.
Correct / safer
Check the sensor-to-score path and the eye, muscle and movement rules before accepting a target-signal change.
Wrong / riskier
Assume that a responsive animation proves control of brain activity.
Judging improvement — Time, expectation, attention and repeated testing can change both groups.
Correct / safer
Use the neurofeedback-minus-sham result and its uncertainty.
Wrong / riskier
Point only to improvement from before to after in the neurofeedback group.
Interpreting target learning — Changing the trained feature does not automatically produce a meaningful benefit.
Correct / safer
Report target-signal learning and symptom or functional outcome in separate sentences.
Wrong / riskier
Call a changed EEG feature successful treatment.
Checking retention — Device-on control does not establish independent or delayed transfer.
Correct / safer
Look for testing without feedback and at a later time, then write “not tested” when absent.
Wrong / riskier
Treat good performance while feedback is visible as a retained everyday skill.
Using published settings — Equipment operation, clinical selection and treatment require qualified oversight.
Correct / safer
Keep electrode and threshold details inside the paper audit.
Wrong / riskier
Copy them into a home neurofeedback procedure.
Method-structure checklist
10 of 10 structural checks present
✓ Ordered, Power-specific instructions — present
✓ Every activity has a success check — present
✓ Materials or supplied records are declared — present
✓ Measurement or assessment rule is present — present
✓ Tutorial-specific troubleshooting is present — present
✓ Stopping or escalation boundary is present — present
✓ Every activity has an adjacent alternative — present
✓ Correct-versus-incorrect comparison is present — present
✓ Evidence context is bound to the Power record — present
✓ Planning metadata is present — present
The method-readiness band and presence checklist assess tutorial presentation and are separate from evidence quality for the underlying Power. They are automated editorial aids, not human approval.
Manual editorial sign-off: Pending. This tutorial must not display a human-approved state until an identified editor signs the exact content hash.