Benefits of Speed of Processing Training: what the studies show

The clearest benefit of Speed of Processing Training is improvement in the visual skills being practised. Later papers examined how long improvements last and whether they extend to daily activities, driving or dementia outcomes. Read the original ACTIVE trial.

Below, we explain the findings study by study, including the newer 2026 follow-up. These are findings about particular research interventions. SOPT Classic is inspired by this approach; the studies did not test the SOPT app available on this website.

What does the SOPT method practise?

In SOPT Classic, you identify a briefly displayed central target and locate another target around it. The task brings together visual processing speed and attention across different positions. Responding accurately matters as well as handling shorter presentations.

This gives “speed” a specific meaning: extracting the information needed for the task. It is different from tapping as quickly as possible after a single light appears. Our processing-speed practice guide explains that distinction and how to track accuracy alongside timing.

What was the ACTIVE study?

ACTIVE stands for Advanced Cognitive Training for Independent and Vital Elderly. It recruited roughly 2,800 independently living adults aged 65 or older and compared memory, reasoning and processing-speed training with a no-contact control group. The design included cognitive and everyday-function outcomes. Read the trial design: Jobe and colleagues, 2001.

Several papers below analyse the same trial. They examine different outcomes and follow-up periods, so they should not be counted as completely independent replications of one result.

Benefit 1: better performance on trained visual tasks

In the 2002 report, 87% of the speed-training group showed reliable improvement immediately after training. Targeted cognitive gains remained detectable at two years, and booster sessions strengthened speed-training gains. Everyday functioning did not show a training benefit at that early follow-up. Read Ball and colleagues, JAMA, 2002.

For someone choosing an exercise, this is a useful starting point: ask what skill the exercise measures and whether that skill is the one you want to practise. A high score should have a clear meaning before it becomes a goal.

Benefit 2: lasting gains and some daily-function findings

The ten-year ACTIVE report found a maintained benefit on trained processing-speed ability. Training groups also reported less difficulty with everyday activities than controls, although results on performance-based daily-function tests were weak or absent. Self-reported ease and an objectively measured task are different kinds of evidence. Read Rebok and colleagues, 2014.

A useful question when reading “lasting benefits” is therefore: lasting benefits on which measurement? Improvements in the trained ability, a person's reported experience and observed performance deserve separate descriptions.

Benefit 3: encouraging results for older drivers

A study of 908 older drivers used state collision records over approximately six years. The speed-training group had a lower adjusted rate of at-fault collisions per distance driven than controls: rate ratio 0.57, with a 95% confidence interval of 0.34–0.96. Read Ball and colleagues, 2010.

That is an interesting outcome beyond a training score. It does not establish that this app makes someone fit to drive, and an app score cannot replace an appropriate driving or vision assessment.

What about the reported dementia benefit?

The ten-year analysis published in 2017

The speed-training group had a 29% lower estimated dementia hazard than controls: hazard ratio 0.71, 95% confidence interval 0.50–0.998, p = 0.049. Dementia was identified using interview and performance-based criteria. The estimate only narrowly met the conventional statistical-significance threshold. Read Edwards and colleagues, 2017.

The twenty-year follow-up published in 2026

This analysis linked 2,021 participants to Medicare records. Speed-trained participants who completed booster training had a 25% lower estimated hazard of a recorded dementia diagnosis: HR 0.75, 95% CI 0.59–0.95. Those without booster training showed no reduction: HR 1.01. The headline therefore concerns a particular subgroup and diagnosis records. Read Coe and colleagues, 2026.

Why the interpretation still needs care

A subsequent reanalysis reported that the individual findings no longer met statistical significance after a correction for multiple comparisons. The study authors also published a response discussing the analysis. Both belong alongside the original report: read Dickens's reanalysis and the authors' response.

Our reading is that the dementia findings are promising, while their interpretation remains debated. They do not justify advertising this app as a proven way to prevent dementia.

How to read a percentage in a study

A hazard ratio compares the rate at which an event occurs over follow-up. It is not the same as subtracting that percentage from every person's lifetime risk, and it is not an absolute percentage-point reduction. To understand absolute benefit, you also need the underlying event rates and time period.

Confidence intervals describe uncertainty around an estimate. If an interval reaches or crosses 1 for a hazard or rate ratio, it includes the possibility of no difference between groups. A result close to that boundary deserves more care than a headline alone conveys.

Also check who entered the analysis, who completed the intervention and how the outcome was identified. Random assignment to an initial training group does not make every later comparison between completers and non-completers interchangeable.

What this evidence means for the SOPT app

SOPT Classic offers an accessible way to practise a visual task inspired by the speed-of-processing approach. Its central/peripheral target exercise gives you a concrete activity to work on and performance feedback to review. Those are features of this product, rather than medical outcomes established by the papers above.

The distinction is important: the SOPT app itself has not been clinically validated by these studies. Similarity in task design does not establish the same training dose, delivery conditions or long-term outcomes. SOPT Advanced and the other games also need their own evidence; they cannot automatically inherit results from ACTIVE.

If you try the app, choose a personal goal you can observe: understanding the instructions, maintaining accuracy or handling a particular presentation level more consistently. Keep the same mode and device when comparing sessions. A simple record is more useful than interpreting one unusually fast attempt as a broad cognitive change.

You can explore SOPT and its download options, or read our guide to attention and concentration. Persistent or worsening cognitive difficulties deserve a conversation with a qualified health professional rather than interpretation through a game score.

Questions about SOPT benefits

Does faster performance mean better memory or higher intelligence?

That conclusion cannot be drawn from a speed-training score alone. Name the ability being measured and look for a separate assessment of any broader claim.

How much should I train to get the study results?

We cannot give an app schedule that guarantees those outcomes. The papers describe their own interventions; a short daily routine in this product should not be presented as an equivalent clinical protocol.

Are these studies about the UFOV.app website?

No. “Useful field of view” is terminology used in this research area. The research papers did not evaluate our UFOV editorial website or validate it as a clinical test.

Studies and further reading

Each link below goes directly to the paper's PubMed record, which includes the publication details and available full-text links.

1. Jobe et al., 2001: ACTIVE trial design.
2. Ball et al., 2002: initial cognitive-training results.
3. Rebok et al., 2014: cognition and everyday functioning at ten years.
4. Ball et al., 2010: older drivers and collision involvement.
5. Edwards et al., 2017: ten-year dementia analysis.
6. Coe et al., 2026: twenty-year dementia diagnosis analysis.
7. Dickens, 2026: reanalysis of the twenty-year findings.
8. Gross, Rebok and Albert, 2026: response to the reanalysis.