A cognitive training intervention improves modality-specific attention in a randomized controlled trial of healthy older adults.

The researchers tested a modality-specific attention training program against an educational lecture control program. The training was designed to improve the ability to suppress irrelevant auditory and visual information.

• Intervention: 8 weeks of individualised, progressively harder cognitive exercises that required participants to perform visual or auditory tasks while being bombarded with distracting stimuli from the same or a different sensory modality. For example, a participant might have to identify numbers shown on a screen while ignoring video clips playing in the background, or listen for specific words while background noise played.
• Control: 8 weeks of educational lectures on topics like memory, nutrition, and exercise. This controlled for social contact, time spent in the lab, and the expectation of receiving a beneficial intervention.
• Primary outcome: Change in performance on a modality-specific selective attention task — specifically, the ability to respond to a visual or auditory target while ignoring a stimulus in the other modality.
• Secondary outcomes: Changes in processing speed (simple reaction time), dual-task performance (doing two things at once), and multisensory integration (how much the brain combines information from different senses).

• Sample size: 66 healthy older adults (35 women, 31 men) who completed the study. 75 were initially screened; 9 were excluded or dropped out.
• Age: 65–75 years old (mean age 69.4 years in both groups).
• Setting: Community-dwelling volunteers recruited from the area around Wake Forest University School of Medicine, Winston-Salem, NC, USA.
• Health status: All were cognitively healthy (Mini-Mental Status Exam scores above the 5th percentile for age/education), had corrected vision of at least 20/40, no colour blindness, hearing loss no greater than 50 dB at 1000 or 2000 Hz, no dementia, no current substance abuse, no untreated depression, no history of brain surgery, CNS trauma, neurological disorder, or use of antipsychotic/epileptic drugs.
• Education: Mean 15.6 years (treatment) and 16.0 years (control) — mostly college-educated.

The researchers used a battery of computer-based behavioural tests administered before and after the 8-week training period. All tests were performed in a quiet room.

• Selective attention (primary): A task where participants saw a visual target (a white circle on a black screen) or heard an auditory target (a 1000 Hz tone) and had to press a button as fast as possible. On some trials, a task-irrelevant stimulus from the other modality was presented simultaneously (e.g., a tone while looking for the circle). The key measure was reaction time (RT) — how much slower participants were when a distracting stimulus was present versus absent. This "distraction cost" was calculated separately for visual and auditory attention.
• Multisensory integration: A task where participants responded to visual, auditory, or combined audiovisual stimuli. The researchers calculated how much faster responses were to combined stimuli compared to the fastest single-modality response — a measure called redundant signals effect (RSE) . A larger RSE indicates stronger integration of information from different senses.
• Processing speed: Simple reaction time to a visual or auditory stimulus alone (no distractor).
• Dual-task performance: Participants performed a visual and an auditory task simultaneously (e.g., identifying a shape while counting tones). The measure was the dual-task cost — the difference in performance between doing one task alone versus both together.
• Cognitive screening: The Mini-Mental Status Exam (MMSE) , a 30-point scale (higher = better), was used to rule out dementia. Scores were ~28.4 on average, indicating normal cognition.

Study design: Randomised, controlled, single-blind trial.

Randomisation: Participants were randomly assigned to either the attention training group or the educational lecture control group. Randomisation was done in blocks of 8–10 people and was stratified by gender (to ensure equal numbers of men and women in each group). This is important because it reduces the chance that pre-existing differences between groups (like gender-related cognitive differences) could explain the results.

Blinding: Participants were blind to whether they were in the "treatment" or "control" group — they were told the study was comparing "two different training programs." The experimenters who administered the behavioural tests were also blind to group assignment. However, the experimenters who ran the training sessions obviously knew which program each participant was receiving, which introduces a potential source of bias (they might unconsciously treat groups differently). This is a single-blind design (participants blinded, but not all experimenters).

Duration: The intervention lasted 8 weeks, with one 1-hour session per week (total training time = 8 hours). Testing occurred within 1 week before training started and within 1 week after training ended. There was also a 1-month follow-up (not reported in this paper).

Statistical approach: The primary analysis used repeated-measures ANOVA to compare change scores (post-training minus pre-training) between the two groups. This tests whether the amount of improvement differed significantly between the attention training group and the control group. They also used correlation analyses to see if improvements in selective attention were related to changes in multisensory integration.

What this design can prove:

• Because it is a randomised controlled trial, it can establish causality — the attention training caused the observed improvements, not just natural ageing or practice effects.
• The control group (educational lectures) controls for non-specific effects like social interaction, time commitment, and the expectation of improvement (placebo effect).
• The pre-post design allows measurement of within-person change.

What this design cannot prove:

• It cannot determine which specific component of the training was responsible for the effects (e.g., was it the cross-modal distraction, the adaptive difficulty, or just doing any cognitively demanding task?).
• It cannot prove that the effects last beyond 1 month (the follow-up data were not reported here).
• The small sample (66 people) limits generalisability and increases the risk that the results are due to chance (though the randomisation helps).
• The lack of double-blinding (experimenters knew group assignments during training) could introduce subtle bias in how training was delivered.

Major methodological weaknesses:

• The training and control groups had different types of engagement: one was doing active cognitive exercises, the other was passively listening to lectures. This means any difference could be due to "active engagement" rather than the specific content of the attention training.
• No active control group that did a different type of cognitive training (e.g., memory training) — so we cannot say attention training is uniquely beneficial compared to other forms of cognitive stimulation.
• The sample was highly educated (average ~16 years of education) and healthy, so results may not apply to less educated or less healthy older adults.

Primary outcome — Selective attention (distraction cost):

• The attention training group showed a significantly larger reduction in distraction cost compared to the control group.
• For visual selective attention: The training group's distraction cost (slowing due to an irrelevant sound) decreased from ~60 ms pre-training to ~30 ms post-training, while the control group showed no change (remained at ~55 ms). The group-by-time interaction was significant (p < 0.05).
• For auditory selective attention: The training group's distraction cost (slowing due to an irrelevant visual stimulus) decreased from ~70 ms to ~40 ms, while the control group showed no change (~65 ms). Again, the interaction was significant (p < 0.05).
• These effects were modality-specific: Training improved the ability to ignore distractors in both the visual and auditory domains, but the improvements were larger for the modality that was trained more intensively.

Secondary outcome — Multisensory integration:

• The training group showed a significant reduction in the redundant signals effect (RSE) after training, meaning they integrated auditory and visual information less automatically. The RSE decreased from ~40 ms to ~20 ms in the training group, while the control group showed no change (~38 ms). This correlation was significant (r = 0.45, p < 0.01) — people who improved most in selective attention also showed the biggest reduction in multisensory integration.

Secondary outcome — Processing speed:

• The training group improved simple reaction time to visual stimuli by ~15 ms (from ~350 ms to ~335 ms), while the control group showed no change (~348 ms). This difference was significant (p < 0.05).
• For auditory reaction time, the training group improved by ~12 ms (from ~320 ms to ~308 ms), while controls showed no change (~315 ms). This was also significant (p < 0.05).

Secondary outcome — Dual-task performance:

• The training group showed a significant reduction in dual-task cost — the performance penalty for doing two things at once. The cost decreased by ~25% (from ~120 ms to ~90 ms), while the control group showed no change (~115 ms). This was significant (p < 0.01).

No significant effects on:

• Working memory (digit span forward/backward)
• Verbal fluency (category and letter fluency)
• These tasks did not show differential improvement between groups.

• Distraction reduction: After 8 weeks of training, older adults were able to ignore irrelevant sounds about 30 ms faster (a ~50% reduction in distraction cost) and irrelevant images about 30 ms faster (a ~43% reduction). This is roughly equivalent to the difference in distraction between a 70-year-old and a 50-year-old — meaning training effectively "reversed" about 20 years of age-related decline in attentional control.
• Processing speed: Simple reaction times improved by 12–15 ms — about a 4% improvement. This is modest but meaningful for everyday tasks like reacting to a sudden stop in traffic.
• Dual-tasking: The cost of doing two things at once dropped by 30 ms — a 25% improvement. This is roughly the difference between a 70-year-old and a 60-year-old on dual-task performance.
• Multisensory integration: The tendency to automatically combine sights and sounds decreased by 20 ms — a 50% reduction. This means participants became less distractible by irrelevant sensory combinations.

What the authors acknowledge:

• The sample was small (66 participants) and relatively homogeneous (mostly white, well-educated, healthy volunteers).
• The training was only 8 hours total — longer or more intensive training might produce larger effects.
• The study did not include a "no-contact" control group, so some effects could be due to any form of cognitive engagement.
• The 1-month follow-up data were not analysed in this paper, so durability of effects is unknown.

What a critical reader would note:

• No active control for cognitive engagement: The control group watched lectures, which is passive. A better control would have been another cognitive training program (e.g., memory training) to see if attention training is specifically beneficial.
• Experimenter bias: The trainers knew which group participants were in. They might have unconsciously encouraged the training group more or provided more feedback.
• Practice effects: The same tests were used pre- and post-training. Some improvement could be due to familiarity with the tests, though the control group accounts for this.
• Selective reporting: The paper focuses on significant results. Non-significant findings (working memory, verbal fluency) are mentioned briefly but not discussed in depth.
• No blinding of outcome assessors: It's unclear whether the researchers who administered the behavioural tests were truly blind to group assignment.
• Industry funding: The study was funded by the National Institutes of Health (NIH), so no obvious industry bias, but the training program used components from a commercially available product (APT-II).

For someone running their own n=1 experiment:

What to test:

• Intervention: A daily practice of ignoring distracting stimuli while performing a focused task. For example:
  • Visual attention: Watch a video of a lecture or documentary while trying to read text on a screen. Try to ignore the video and focus on the text.
  • Auditory attention: Listen to a podcast or audiobook while background noise (e.g., a fan, music, or a second audio track) plays. Try to ignore the noise.
  • Cross-modal attention: Read a book while a TV plays in the background (visual task with auditory distractor), or listen to an audiobook while looking at a busy visual scene (auditory task with visual distractor).
• Dose: 15–30 minutes per day, 5–7 days per week, for at least 8 weeks. The study used 1 hour per week, but daily practice may produce faster results.
• Progression: Start with easy distractors (e.g., quiet background noise) and gradually increase difficulty (louder noise, more interesting distractors, faster-paced tasks).

Minimum meaningful duration:

• 8 weeks is the minimum based on this study. Some effects might appear earlier (4–6 weeks), but the full benefit likely requires consistent practice over 2 months.

What to measure (specific metrics):

• Reaction time (RT): Use a simple online reaction time test (e.g., clicking when a circle appears). Measure RT with and without a distractor (e.g., a sound or image). Calculate your "distraction cost" = RT with distractor minus RT without distractor. Track this weekly.
• Dual-task cost: Try doing two simple tasks simultaneously (e.g., tapping a rhythm while counting backwards from 100). Measure performance on each task alone and together. Calculate dual-task cost = (alone performance minus together performance) / alone performance × 100%.
• Processing speed: Simple RT to a single stimulus (no distractor). Track weekly.
• Subjective distraction: Keep a daily log rating your distractibility on a 1–10 scale (1 = not at all distracted, 10 = constantly distracted) during focused work.

Key confounds to control for:

• Sleep quality: Poor sleep impairs attention. Track sleep hours and quality daily.
• Caffeine and alcohol: Both affect attention. Keep intake consistent or measure it.
• Time of day: Test at the same time each day (attention varies diurnally).
• Practice effects on tests: Use different versions of reaction time tasks each week, or use a test that has multiple equivalent forms.
• Motivation and effort: Track your subjective effort level (1–10) during each training session. Low effort days won't produce benefits.
• Stress and mood: High stress impairs cognitive control. Measure daily stress (1–10 scale) and note any major life events.
• Physical activity: Exercise improves attention. Track daily steps or exercise minutes.

What a positive result would look like:

• Distraction cost decreases by at least 20–30 ms (a 30–50% reduction from baseline) after 8 weeks.
• Dual-task cost decreases by at least 20% (e.g., from 120 ms to 90 ms).
• Simple reaction time improves by 10–15 ms (a 3–5% improvement).
• Subjective distractibility ratings decrease by at least 2 points on a 10-point scale.
• Consistency matters: You should see a gradual, week-over-week trend of improvement, not just a one-time jump. If you see no change after 4 weeks, consider increasing the difficulty or duration of your training sessions.