The Science Behind BrainSharp 50+
What the BrainSharp Score actually measures, how we built it, and — just as important — what we don't claim.
What BrainSharp Trains
Cognitive aging is not a single trait — it's a constellation of distinct abilities that change at different rates. BrainSharp targets six of the most heavily studied domains in aging research. Each lesson is designed to load one or more of these domains under time pressure, the way real life does.
Memory Recall
Holding, encoding, and retrieving sequences, names, lists, and source information.
Processing Speed
How quickly you take in new information and produce a correct response.
Attention & Focus
Sustaining concentration, filtering distractors, switching between tasks under load.
Reasoning & Logic
Following multi-step chains of reasoning, spotting patterns, evaluating arguments.
Word Retrieval
Producing the right word at the right time — fluency, vocabulary, lexical access speed.
Spatial Processing
Mentally rotating, navigating, and reading space — maps, rooms, routes, distances.
These six are not arbitrary. They are the cognitive domains most consistently linked to functional independence in adults 50+: managing money, recognizing scams, driving safely, following medication instructions, and holding a conversation that crosses topics.
How the BrainSharp Score Is Computed
Every answer in every lesson produces three signals: accuracy (did you get it right), response time (how fast), and region weights (which cognitive domains were actually loaded). We blend them into a region-level score on a 0–100 scale.
Per-question score
For a single question the blended score follows the same shape across every lesson:
where rtScore = max(0, 100 − max(0, rt − floor) / k)
Accuracy weight (wa) is typically 0.75–0.80 and response-time weight (wrt) is 0.20–0.25, with per-lesson tuning. The response-time penalty has a generous floor (3–5 seconds) so a careful reader is not punished, but a 30-second answer scores lower than a 5-second one.
Per-region score
A single lesson may train two regions at once (e.g. Analogical Reasoning loads both reasoning and word retrieval). We project the blended question score into each region using that question's region weights, then average across questions:
Trend over time
Your dashboard shows a rolling region score using your most recent sessions, weighted equally across lessons (no single lesson can dominate). Lesson history and session history are both tracked, so the score recovers smoothly if you skip a day. A bad question doesn't tank your score; a long streak of weak performance moves it.
What BrainSharp Does — and Does Not — Claim
| What we claim | What we don't claim |
|---|---|
| YES. The exercises are real cognitive tasks drawn from the same families used in cognitive-aging research. | NO. BrainSharp does not prevent, treat, or cure Alzheimer's disease, dementia, mild cognitive impairment, or any other medical condition. |
| YES. Practicing a domain improves performance on that domain. This is well-established (the "practice effect"). | NO. The BrainSharp Score is not a clinical assessment and is not equivalent to neuropsychological testing. |
| YES. Six months of daily training will move your region scores, often substantially. Your dashboard shows it. | NO. We do not claim transfer to untrained tasks beyond the practice effect. Far-transfer research remains contested. |
| YES. Several of our lessons (scam detection, financial reasoning, health literacy, driving hazard perception) target skills with strong real-world stakes. | NO. Brain Age and percentile estimates are training metrics, not biological age. |