Brain Training Games and Mental Exercises
Linas JuozenasShare
Intelligence Unleashed · Cognitive training
Practice with purpose.
Build skills that matter.
Brain games can offer challenge, enjoyment and practice. A useful mental-exercise routine connects those experiences with something you want to learn or do—and checks whether progress carries beyond the exercise itself.
Evidence checked: 4 September 2026 · Practical educational guide · Sources and series links below
Mental activity can have several useful purposes
A crossword, a memory drill, a conversation and learning a new procedure all engage the mind. They do not necessarily produce the same benefits.
Cognitive training usually means repeated exercises intended to improve particular abilities, such as memory, attention or reasoning. A game adds goals, rules and feedback that may make practice engaging. The important questions are what changes and how the change is measured.1
Training a selected ability
Practice a defined task, often with increasing difficulty. Progress might mean remembering longer sequences or responding more accurately under a particular set of conditions.
Broad mental stimulation
Enjoy varied activities such as reading, storytelling, discussion and games. Challenge and social connection can be worthwhile without every activity becoming a scored test.
WHO distinguishes targeted training from this broader stimulation. Its guidance includes everyday activities as well as structured programs, rather than treating a commercial app as the only way to stay mentally engaged.2
A practical goal can call for a different approach
Cognitive rehabilitation focuses on meaningful activities in daily life. A professional may help someone practice a specific routine, use a reminder or adapt the environment. Success is judged against that person's functional goals, rather than requiring a higher score on a generic brain game.3
Enjoyment counts. You can play a puzzle because it is satisfying, share a game because you like the company, or practice a skill because it matters to you. Those reasons do not need an exaggerated promise about intelligence.
The gym analogy: targeted practice and coordinated skills
A useful way to think about brain games is as a practice gym. Carefully chosen exercises provide repeated, adjustable demands that may be difficult to arrange consistently in everyday life. A game can give you many opportunities to retrieve information, maintain attention or revise a strategy, with immediate feedback and gradually increasing challenge.
Practice recall
Briefly view several objects, hide them and recall their locations. Changing the arrangement lets you practice remembering new information rather than memorize one fixed answer.
Practice selective attention
Search for a target among distractors while following a clear rule. Adjusting the display can emphasize careful selection, accuracy or speed within that task.
These exercises emphasize particular processes, but they do not isolate a single “recall muscle” or “concentration muscle.” Demanding mental tasks recruit interacting, partly overlapping brain networks. Shared brain activity also does not, by itself, prove that training one task improves another.4
Learning changes the brain over time
The biological foundation is neuroplasticity: experience can change how neurons connect and how brain networks function. Adaptation can involve connections between existing cells and changes in supporting tissue. Learning unfolds on several timescales, so a quick improvement and a durable skill are not necessarily the same stage of adaptation.5
For example, a small, industry-funded study of adolescent girls learning Tetris found changes in cortical-thickness measures after three months. This illustrates possible structural adaptation during game learning; it did not establish increased neuron numbers or general cognitive improvement.6
Producing new neurons is called neurogenesis. Recent research provides evidence of neurogenic cells in the adult human hippocampus, but it did not test brain games. We cannot conclude that activating a region through an ordinary game makes it grow more neurons. MRI changes are not direct neuron counts, and everyday activity and rest are not an inactive-brain baseline.7, 5
Bring the parts together. A language learner can practice recalling words, following speech and building sentences, then combine those skills in conversation. As with physical skill training, practice the coordinated activity too. The gym analogy is useful for repetition and adaptation; broader fluency needs relevant practice rather than an assumption that isolated gains automatically add together.
Getting better at a task is the beginning of the question
Improvement can come from learning the rules, recognizing familiar patterns, developing a strategy or becoming comfortable with the controls. That is real learning. To understand a broader claim, researchers ask about transfer: whether practice benefits something beyond the trained task.1
| Kind of improvement | Example | What it establishes |
|---|---|---|
| Practiced-task learning | You remember longer sequences in the game you use. | You have improved at that trained activity. |
| Near transfer | You perform better on a closely related task with new items. | Some learning extends to a similar demand. |
| More distant transfer | You improve on a different reasoning or memory measure. | A broader benefit may exist if the comparison is well controlled. |
| Everyday benefit | You follow a work procedure more reliably or remember intended actions. | The change matters in a relevant real-life activity. |
These categories form a continuum. Changing the colors in a task does not make it a completely new ability. A meta-analysis of n-back training found larger transfer to other versions of n-back than to different working-memory or reasoning measures—a useful example of how close a supposedly “untrained” test can remain to practice.8
Match the claim to the outcome. If your goal is remembering what you read, check recall of new reading. If your goal is handling invoices accurately, practice and assess that process. A rising game score answers a narrower question.
A persuasive study needs a fair comparison
People often improve when they repeat a test, even without the advertised intervention. They may become more familiar with instructions, feel less uncertain or discover a useful strategy. Comparing a score before and after practice cannot by itself separate those effects.
In a randomized trial of 128 young adults, ten weeks of commercial cognitive training improved performance on the trained games but did not outperform ordinary online games on independent cognitive measures, decision tasks or measured brain activity. A separate retested comparison also improved on the tests. This illustrates why a program's own progress chart is incomplete evidence.9
Look for these features
- An appropriate comparison activity: similar time, contact and engagement where possible.
- Independent outcomes: tests that do more than repeat the trained game.
- A plan made before results are known: clear primary outcomes and transparent reporting.
- Follow-up: evidence about whether any benefit lasts.
- Relevant participants: results from one age group or clinical population may not apply to another.
Expectations deserve attention as well. A study that manipulated expectations independently of working-memory training found that both training condition and expectations influenced cognitive gains. This supports measuring and balancing expectations; it does not show that every training effect is a placebo or that optimism guarantees improvement.10
Brain images are another measurement. A change in activation or another biological marker does not automatically demonstrate better learning, work or independence. The claimed benefit still needs to be measured.
Healthy adults: specific gains are more credible than sweeping promises
The evidence is more informative when it names the task and outcome than when it declares that brain training simply “works” or “does not work.”
A 2024 synthesis
A meta-analysis of healthy young and middle-aged adults found small improvements when working memory was assessed with tasks distinct from training. Improvements were much larger when the assessment closely resembled practice. It found no improvement in fluid intelligence, although some specific executive-function outcomes benefited. Study quality and selective reporting remained concerns.11
Working memory refers to holding and using information over a short period, such as keeping intermediate steps in mind. Fluid intelligence concerns reasoning through unfamiliar problems. Their relationship does not mean that practicing any working-memory game necessarily improves general reasoning.
An earlier large meta-analysis likewise found short-term working-memory improvements but no convincing benefits for intelligence, reading or arithmetic when training was compared with treated control groups. This challenges broad academic promises while leaving room for the narrower skills that participants actually practiced.12
More difficult does not automatically mean more transferable
In an n-back task, a person identifies whether the current item matches one presented a set number of steps earlier. A preregistered 2025 trial found that longer n-back training produced greater gains on the trained level, without detected transfer to its untrained near- or far-transfer outcomes. One protocol cannot settle every training question, but increased training time did not establish broader benefit here.13
The practical implication is to choose difficulty that supports learning. A task can be challenging because it demands the skill you want, or because its interface is confusing and the time limit is uncomfortable. Those are different reasons to struggle.
Keep the useful part. If you enjoy a game and become more skilled at it, that has value. If you want another benefit, give that benefit its own test instead of assuming that it follows.
Older adults: encouraging findings need careful boundaries
What long-term training research shows
The ACTIVE trial compared memory, reasoning and speed-of-processing training with a no-contact control in independently living older adults. At ten years, targeted reasoning and speed benefits remained; the memory-training benefit did not. Training groups reported less difficulty with everyday activities, but performance-based measures of everyday functioning did not show a training benefit. Substantial loss to follow-up also limits interpretation.14
The 2026 dementia finding deserves its full context
A twenty-year follow-up using Medicare records found fewer diagnosed dementia cases over time in a specific speed-training group with booster sessions. The overall speed-training arm did not show a statistically significant difference from controls. Booster eligibility depended on completing initial sessions, and diagnoses came from healthcare claims. This is encouraging evidence for further study, not proof that ordinary brain games prevent dementia.15
Current guidance is conditional
WHO's July 2026 guideline says cognitive training may be offered to older adults with normal cognition or mild cognitive impairment, based on low-certainty evidence. It also conditionally encourages cognitive stimulation, with very-low-certainty evidence. These recommendations allow useful activity while acknowledging uncertainty; they do not certify a particular app or guarantee prevention.2
Different clinical situations need different goals
A 2024 review found improvements in several measured memory domains among people with mild cognitive impairment receiving computerized training. Findings in dementia were narrower and less certain, with small studies and substantial variation between them. Better memory-test performance does not establish that progression to dementia has been prevented.16
Stimulation in dementia care
A Cochrane review found small short-term cognitive benefits from cognitive stimulation, often delivered through enjoyable group activities. Communication and social interaction also improved. These are supportive-care findings, not evidence of reversing the disease.17
Rehabilitation around a personal goal
Another Cochrane review found that rehabilitation helped people with mild-to-moderate dementia achieve the specific everyday goals they practiced. Gains in those goals should not be generalized to every activity or underlying cognitive ability.3
A person may benefit from familiar activities, practical cues and assistance that makes participation easier. Repeated failure under time pressure is not a necessary ingredient of good support. Where clinical needs are involved, an appropriate professional can help match the activity and expectations to the person.
Choose the ability you want to use
“Improve my brain” is difficult to act on. “Remember the main points of a chapter” or “select the right method for a calculation” gives practice a direction and creates a way to notice progress.
| Your goal | A useful practice idea | What to check |
|---|---|---|
| Remember useful information | Retrieve it without looking, correct errors and revisit later. | Accurate recall after a delay. |
| Understand a procedure | Explain a worked example, then try a similar problem. | Can you explain why the steps work? |
| Select a strategy | Mix related problem types once the methods are familiar. | Choosing the right method on new examples. |
| Remember an intended action | Connect it to a clear cue and use a suitable reminder. | Whether the action happens in the intended setting. |
| Enjoy a mental challenge | Choose a puzzle, game or creative task you like. | Enjoyment, engagement and skill at that activity. |
Educational research supports principles such as spaced study, retrieval, worked examples and explanatory questions for learning particular material. These strategies can be useful without assuming they expand a general-purpose cognitive capacity.18
Training can be direct. If the goal is learning a language, part of the practice can simply be using that language. There is no requirement to complete an unrelated memory game before beginning the skill you actually want.
Seven mental exercises with a clear purpose
Use material that matters to you. The examples below adapt learning research and ordinary practice ideas; together they are not a tested clinical treatment program.
Recall the main ideas without looking
Read a short passage, close it and write or say its main points. Reopen the source, identify omissions and correct anything inaccurate. Then explain the material again in your own words.
Experiments on studied prose found that retrieval practice improved delayed retention compared with repeated reading, even when reading looked better on an immediate test. Separate research shows how corrective feedback can improve later recall and reduce remembered errors.19, 20
Check: what can you accurately recall later? Familiarity while rereading is a different experience from retrieving the information independently.
Revisit important material after a gap
Choose a few ideas or facts worth remembering. Try recalling them on another day, check the answers and return again later. If recall repeatedly fails, make the material smaller or the gap shorter. If it is reliable, try a longer interval.
Research on factual learning shows that useful spacing depends partly on how long the information needs to be retained. A tomorrow–next week–later pattern can be a starting example; it is not a universal optimum.21
Check: accurate recall across increasingly useful delays, rather than the number of cards you clicked through.
Explain a solution, then attempt a new one
Study a worked example and describe why each step follows. Cover it and solve a similar problem. For a simple arithmetic illustration, increase 100 by 10%, then reduce the result by 10%: 100 becomes 110, then 99. Explain why equal percentages do not cancel when they apply to different starting amounts.
Research with novices learning a technical procedure found benefits from worked examples and example-then-problem practice compared with unsupported problem solving. The aim is to move toward independent use, not continue copying indefinitely.22
Check: can you apply the principle when the numbers or setting change?
Mix related problems after learning the basics
Once you understand several methods, mix examples that require different ones. Before solving each problem, identify what kind it is and why your chosen method fits. For example, combine a few familiar fraction, ratio and percentage problems.
A school mathematics experiment found better later performance when practice problems were interleaved rather than grouped by type. This concerns learning to select a relevant strategy. It does not mean rapidly switching between unrelated tasks or mixing material before you understand any of it.23
Check: choosing the appropriate approach on a new example, alongside getting the answer right.
Turn an answer into an explanation
Choose a concept you are learning and explain it to an interested listener, or record a short explanation for yourself. Ask: what causes what, which evidence supports the claim, and what remains unclear? Check the explanation against a reliable source.
In a study of students learning about circulation, prompting self-explanations improved understanding. The finding concerns learning the studied material; it does not establish a general increase in reasoning ability. An explanation can sound fluent and still need correction.24
Check: whether you can answer a fresh “why” or “how” question without simply repeating a memorized sentence.
Make an intention easier to carry out
Choose a specific action and link it to an event you are likely to notice: “When I finish breakfast, I will put the borrowed book in my bag.” Briefly imagine noticing the cue and doing the action. Add a reminder when reliability matters.
A meta-analysis found that these implementation intentions can improve prospective memory—remembering to carry out a planned action. Effects varied, and keeping an intention active can also draw on attention. The method is useful to try, not a guarantee against forgetting.25
Check: whether the action happens. If the cue is easy to miss, change the cue or use an external aid.
Use a puzzle to practice deliberate reasoning
Choose a logic puzzle, a board-game position or a planning problem at a comfortable level. State the rule or constraint, consider two possible approaches and explain why you choose one. Afterward, inspect an error or an alternative solution.
This is a practical way to make your reasoning visible within that activity. You can learn patterns, develop a better strategy and enjoy the challenge without assuming it raises general intelligence.
Check: fewer avoidable errors, clearer explanations or better handling of a new example in the same domain. Start without a timer if speed obscures understanding.
Adapt the format. Speak instead of writing, use larger print, work with a partner or choose material in a familiar language. An exercise should help you engage with the intended skill, rather than mostly test an unrelated barrier.
Build a routine you can comfortably repeat
A short session can be useful when it contains purposeful practice and feedback. The following fifteen-minute example is a scheduling idea, not an evidence-based dose for cognitive enhancement or dementia prevention.
- 2 minutes · ChooseSelect one small goal and suitable material. Recall what you did last time before looking at the answer.
- 7 minutes · PracticeWork on the chosen skill. Keep the challenge manageable enough that you can understand your errors.
- 4 minutes · Check and explainCompare with a reliable answer, identify what changed your understanding and try the difficult part again.
- 2 minutes · Plan a returnNote what to revisit and one place you could use the learning outside this session.
Spread practice across the week
For example, introduce material early in the week, retrieve it later and then try a new application. Keep the same goal long enough to learn something about it. Constantly changing games may feel varied while making progress difficult to interpret.
If an activity is effortless, add a modest challenge: new examples, a longer delay or less prompting. If it becomes confusing, return to a clear example, reduce the amount of material or ask for an explanation. Difficulty is useful when it serves learning, not merely because it feels demanding.18
A missed session is information, not failure. Adjust the time, task or setting if the plan does not fit your life. A routine should support your interests and responsibilities rather than compete with them.
Measure the change you actually care about
Choose a simple starting task before practice begins. After some practice, try a comparable example under similar conditions. If you are learning vocabulary, use delayed recall and fresh sentences. If you are practicing a procedure, observe whether you can perform it accurately with fewer prompts.
Keep a small, useful record
- Task: what did you practice?
- Result: what was accurate, difficult or still unclear?
- Strategy: what helped?
- Application: did anything improve in the relevant everyday activity?
Accuracy and speed should be considered together. Finishing faster while making more errors may not be the improvement you want. Similarly, recalling an answer immediately after seeing it is different from remembering it next week.
Use new examples when possible, and avoid drawing conclusions from one unusually good or bad day. Familiarity with the test, fatigue, interruptions and changes in the task can affect performance. A personal practice log can guide your routine, but it is not a controlled experiment proving a general cognitive gain.
Useful evidence of learning
You explain a new problem more clearly, retain studied information longer or complete a relevant task with fewer mistakes.
A result needing interpretation
The app raises your level, shows a new percentile or assigns a younger “brain age.” Ask how that number was calculated and what it has been validated to measure.
An app score should not be used to diagnose a memory condition. Persistent changes that interfere with ordinary activities deserve an appropriate assessment rather than more attempts to beat a game. There can be different reasons for memory difficulties, and understanding the cause matters.26
Choose games and tools for what they offer
A good tool makes the intended activity clear, provides useful feedback and fits the person using it. It might be an app, a printed puzzle, a deck of cards, a notebook or a conversation partner.
Look for helpful design
Clear instructions, adjustable difficulty, readable text, usable controls and explanations of mistakes can make practice more accessible. The ability to pause or remove time pressure may matter more than a large collection of games.
Check the promise
Ask whether research tested the same program being offered, in people relevant to you, using the outcome being advertised. A study of one targeted intervention does not validate every product called “brain training.”
Before paying for a subscription
Decide whether the program offers something useful beyond activities you already have. Look at the actual price, renewal and cancellation terms, and whether core features require extra purchases. Enjoyment and convenience can justify a purchase; a prevention or intelligence claim needs its own evidence.
Consider the data the tool requests. A simple exercise may not need access to contacts, location or a detailed health history. Read what is collected and shared, and choose a format that fits your preferences. Paper-based activities can also be a practical option.
Adapt to the learner
For children, use developmentally suitable material and invite them to explain their thinking. For older adults or people with sensory or motor differences, adjust the presentation and controls. Someone can understand an idea well while finding small targets, unfamiliar interfaces or rapid button presses difficult.
Let the tool serve the goal. If the main achievement becomes protecting a streak or navigating the app, revisit what you wanted the practice to help you do.
Leave room for meaningful learning and a fuller life
Learning a skill can combine memory, attention, movement, decision-making and social exchange in a meaningful setting. A language, musical activity, craft or technical project also produces something you can use or enjoy.
In the Synapse Project, older adults learning demanding photography or quilting skills showed episodic-memory benefits relative to less demanding comparison activities. The program was intensive—roughly three months, with demanding groups averaging about 16.5 hours a week. It does not show that a few daily minutes of any hobby reproduce those results or prevent dementia.27
Cognitive practice sits within a wider pattern
The 2025 US POINTER trial found a modest advantage in global cognitive performance for a structured program combining exercise, dietary guidance, cognitive and social engagement, and health monitoring compared with a less intensive self-guided program. Because several components changed together, it cannot identify a brain-training game as the cause of the benefit or establish dementia prevention.28
A mental-exercise session should fit alongside sleep, movement, meals, relationships and time away from performance goals. Replacing needed rest or valued social contact with more drills may not serve the reason you began.
Continue with learning new skills, memory techniques, critical thinking and problem-solving and healthy lifestyle habits for ways to build these ideas into everyday life.
There is no requirement to turn every hobby into training. Curiosity, pleasure, competence and company are worthwhile outcomes too. Choose activities that give you reasons to return beyond the score.
Questions worth answering clearly
Do brain-training games increase intelligence?
Are crosswords, sudoku and board games still worthwhile?
Yes, if you enjoy them or value the skills and social experiences they involve. They can provide challenge and mastery without needing to be proven treatments for cognitive decline.
How many minutes a day should I train?
There is no single established dose that guarantees broad cognitive improvement. Choose a manageable period for a clear goal, include feedback and revisit material. The example routine above is a planning aid, not a medical prescription.
Can brain games prevent dementia?
Is using notes or reminders cheating?
No. If the goal is reliable everyday functioning, a useful reminder or an adapted routine can be part of the solution. Practicing unaided recall and completing an important task reliably are different goals.
Should practice always feel difficult?
It should offer a useful challenge at your current level. If confusion prevents you from learning from feedback, simplify the task or study an example. If it is automatic, consider a new example or a longer recall delay.
Practice something worthwhile.
Notice what becomes easier.
A good mental-exercise routine joins clear goals, suitable challenge, feedback and opportunities to apply what you learn. Keep the curiosity and enjoyment, and let meaningful progress matter more than a promise printed on an app.
Educational information. These exercises are not diagnostic tests or individualized treatment. Clinical cognitive rehabilitation and assessment of persistent cognitive changes should be matched to the person's circumstances by an appropriate professional.
Evidence & context
Sources and further reading
These references address different outcomes: practiced skills, untrained tests, everyday functioning and clinical conditions. Their populations, comparisons and follow-up periods matter. The exercise examples adapt principles from research rather than reproducing a clinical protocol. External links open in a new tab.
- Simons DJ et al. Do “Brain-Training” Programs Work? Psychological Science in the Public Interest (2016). Evidence and research-design review.
- World Health Organization. Risk reduction of cognitive decline and dementia: WHO guidelines, second edition. Recommendations (2026). Conditional guidance on training and stimulation.
- Kudlicka A et al. Cognitive rehabilitation for people with mild to moderate dementia. Cochrane (2023). Benefits for personally selected everyday goals.
- Fedorenko E, Duncan J, Kanwisher N. Broad domain generality in focal regions of frontal and parietal cortex. PNAS (2013). Overlapping recruitment across demanding tasks.
- Zatorre RJ, Fields RD, Johansen-Berg H. Plasticity in gray and white: neuroimaging changes in brain structure during learning. Nature Neuroscience (2012). Mechanisms and limits of MRI interpretation.
- Haier RJ, Karama S, Leyba L, Jung RE. MRI assessment of cortical thickness and functional activity changes in adolescent girls following three months of practice on a visual-spatial task. BMC Research Notes (2009). Small, industry-funded Tetris study.
- Disouky A et al. Human hippocampal neurogenesis in adulthood, ageing and Alzheimer’s disease. Nature (2026). Human tissue study; not a gaming intervention.
- Soveri A et al. Working memory training revisited: A multi-level meta-analysis of n-back training studies. Psychonomic Bulletin & Review (2017).
- Kable JW et al. No Effect of Commercial Cognitive Training on Brain Activity, Choice Behavior, or Cognitive Performance. Journal of Neuroscience (2017). Randomized comparison with ordinary games.
- Parong J, Seitz AR, Jaeggi SM, Green CS. Expectation effects in working memory training. PNAS (2022).
- Rodas JA, Asimakopoulou AA, Greene CM. Can we enhance working memory? Bias and effectiveness in cognitive training studies. Psychonomic Bulletin & Review (2024).
- Melby-Lervåg M, Redick TS, Hulme C. Working-memory training and measures of intelligence and far transfer: meta-analytic review. Perspectives on Psychological Science (2016).
- Zhong LA et al. Near and far transfer effects of working memory training: A preregistered, double-blind, randomized-controlled study. Acta Psychologica (2025).
- Rebok GW et al. Ten-Year Effects of the ACTIVE Cognitive Training Trial on Cognition and Everyday Functioning in Older Adults. Journal of the American Geriatrics Society (2014).
- Coe NB et al. Impact of cognitive training on claims-based diagnosed dementia over 20 years: evidence from the ACTIVE study. Alzheimer’s & Dementia: Translational Research & Clinical Interventions (2026).
- Chan ATC et al. Computerized cognitive training for memory functions in mild cognitive impairment or dementia: a systematic review and meta-analysis. npj Digital Medicine (2024).
- Woods B et al. Cognitive stimulation to improve cognitive functioning in people with dementia. Cochrane (2023).
- Pashler H et al. Organizing Instruction and Study to Improve Student Learning. Institute of Education Sciences practice guide (2007).
- Roediger HL, Karpicke JD. Test-Enhanced Learning: Taking Memory Tests Improves Long-Term Retention. Psychological Science (2006).
- Butler AC, Roediger HL. Feedback enhances the positive effects and reduces the negative effects of multiple-choice testing. Memory & Cognition (2008).
- Cepeda NJ et al. Spacing effects in learning: a temporal ridgeline of optimal retention. Psychological Science (2008).
- van Gog T, Kester L, Paas F. Effects of worked examples, example-problem, and problem-example pairs on novices’ learning. Contemporary Educational Psychology (2011).
- Rohrer D, Dedrick RF, Stershic S. Interleaved Practice Improves Mathematics Learning. Journal of Educational Psychology (2015). Official ERIC record.
- Chi MTH et al. Eliciting Self-Explanations Improves Understanding. Cognitive Science (1994).
- Chen XJ et al. The effect of implementation intention on prospective memory: a systematic and meta-analytic review. Psychiatry Research (2015).
- National Institute on Aging. Memory Problems, Forgetfulness, and Aging. Official information on assessment and possible causes (2023).
- Park DC et al. The Impact of Sustained Engagement on Cognitive Function in Older Adults: The Synapse Project. Psychological Science (2014). Intensive new-skill learning.
- Baker LD et al. Structured vs Self-Guided Multidomain Lifestyle Interventions for Global Cognitive Function: The US POINTER Randomized Clinical Trial. JAMA (2025). Combined lifestyle intervention.