Determining Learning Ability in People With Aphasia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Determining Learning Ability in People With Aphasia (NCT05119023), a na na study. The primary outcome is SRT Observational Learning Ability.

2. Clinical Question

3. Trial Design

The registry lists this study as completed, with a start date of 2022-06-06 and a primary completion date of 2023-09-01. Enrollment is listed as 18. The model is single group, allocation is NA, and masking is none. The sponsor is MGH Institute of Health Professions. The phase is listed as NA. These design features describe a learning-characterization study, not a randomized comparison of a treatment with a control.

Participants completed computer-based tasks intended to measure observational learning and rule-based learning. The named tasks are SRT Observational Learning, AGL Observational Learning, and AGL Rule-based Learning. Other listed study activities include standardized cognitive-linguistic assessment and brain imaging. The record describes task order as counterbalanced for SRT and AGL observational tasks, and states that the AGL observational task is completed before the rule-based AGL task. The endpoint timeframe refers to study visit 1 or 2.

The registry description frames the work around determining learning phenotype in people with aphasia and examining relationships between learning ability and lesion characteristics, language ability, and cognitive ability. It does not report a treatment allocation or a like-for-like control arm. The posted record lists 8 outcome measures and 12 analyses, but the complete analysis details are not reported in the registry record. The sample-size assumptions describe an illustration based on an observed effect and a two-arm planning scenario; they are not a description of the single-group study design or its observed power.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Characterization of Learning-visits 1&2

Reasons for not completing this period:

Brain Imaging - Session 3

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as SRT Observational Learning Ability, measured over Study visit 1 or 2, AGL Observational task completed before rule-based AGL task. SRT Observational and AGL Observational task order counterbalanced.

The primary endpoint is SRT observational learning ability, expressed as the mean Cohen's d across individual participants. The task records reaction time from target onset until gaze fixation within the target area of interest. Only correct-trial reaction times are examined. A trial is considered incorrect when the fixation does not correspond to the target area.

The registry defines a learning score by comparing reaction times in the final sequenced block, S7, with those in the following pseudorandomized block, PS8. Reaction times more than three standard deviations above the mean reaction time of each block are removed. Cohen's d is calculated for each participant using pooled standard deviations, and the primary measure is the mean of those individual effect sizes. The record specifies that negative values indicate better learning.

As a general statistical concept, a standardized effect size expresses a contrast in relation to variability, which can make the size of a contrast easier to interpret than reaction-time units alone. In this endpoint, the contrast is within a task and is intended to quantify learning-related performance. The registry record does not report the mean Cohen's d, a confidence interval, or other numerical primary results, so the magnitude and precision of the observed endpoint cannot be assessed from the information reported.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is a continuous measure: the registry identifies its parameter as a mean and its dispersion as a standard deviation. The reported unit is Cohen's d. As a general statistical concept, a mean describes the center of a set of values, while a standard deviation describes how spread out those values are around the mean. Here the registry defines the mean as the average of individual participants’ Cohen's d values. The numerical mean and standard deviation are not reported in the registry record.

What does the endpoint contrast measure?

The endpoint is based on reaction times from correct trials in an SRT observational learning task. Reaction time is measured from target onset to gaze fixation within the target area of interest. The task score compares the final sequenced block, S7, with the following pseudorandomized block, PS8. The registry defines negative Cohen's d values as indicating better learning. This direction convention is important when interpreting a reported value: the sign is meaningful only in relation to the stated order of the comparison.

Cohen's d standardizes a difference using a measure of variability. In general, standardization expresses a contrast in standard-deviation units, rather than in the original measurement units. The registry says pooled standard deviations are used for the participant-level calculation. The record does not report the individual effect sizes or enough numerical information to reproduce the mean endpoint. No recalculated estimate is available in the registry record.

Why are reaction-time outliers handled?

The endpoint definition removes reaction times more than three standard deviations above the mean reaction time of each block. As a general statistical concept, unusually long reaction times can have substantial influence on averages and variability. A prespecified or clearly described rule for handling extreme values can limit that influence, but the result can depend on the rule chosen and on how many observations it affects. The registry description gives the threshold and the block-level basis for applying it. It does not report how many reaction times were removed, how many correct trials contributed to each participant’s score, or whether any sensitivity analysis assessed the impact of this rule.

What does over-dispersion mean in general?

Over-dispersion is a general concern when observed variability is greater than a statistical model assumes. If a model understates variability, uncertainty estimates can be too narrow and tests can appear more decisive than the data warrant. The endpoint uses continuous reaction times and participant-level standardized effects; the registry record does not report a model-based over-dispersion assessment. This general concept should not be mistaken for a reported finding about this study.

What do the planning assumptions mean?

The registry’s sample-size assumptions describe a planning illustration computed from an observed effect, using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The stated effect to detect is a fraction of the observed effect. The record specifies that non-time-to-event sample-size entries are participants per arm, while time-to-event entries would be total events. No numerical planning values are included in the registry record for interpretation.

These assumptions are a planning illustration, not the study’s own design justification. In particular, the study is recorded as single group, while the illustration assumes equal allocation between two arms. Such illustrations are not observed power. As a general concept, power calculations depend on assumptions about the effect, variability, allocation, and analysis; they do not measure the probability that a completed study’s result is true. A sample-size justification, interim-analysis plan, and detailed statistical analysis plan are not reported in the registry record.

How should multiple analyses be considered?

The record lists 8 outcome measures and 12 analyses. When several outcomes or analyses are considered, the chance of obtaining at least one apparently unusual result can increase if each is assessed separately without adjustment. As a general statistical concept, multiplicity procedures can help control this issue, but they may also affect precision or the thresholds used for inference. The registry record does not report whether any adjustment was made, which results were designated primary for each analysis, or how the 12 analyses correspond to the listed outcome measures. No multiplicity strategy should be assumed.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Characterization of Learning")
n   <- c(14)
m   <- c(-0.03)
sp  <- c(0.34)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Characterization of Learning  14  -0.030  0.340  -0.226  0.166

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 8 outcome measures and 12 analyses. The primary endpoint is identified as SRT Observational Learning Ability, and the listed secondary outcomes include language severity, cognitive composites for attention, working memory, and executive function, and percent spared tissue per region of interest. The registry record does not provide complete numerical results for these outcomes or all analysis details.

When multiple outcomes and analyses are examined, multiplicity is a general statistical concern: separate tests can collectively increase the chance of a nominally notable result arising by chance. Whether that concern was addressed depends on the analysis plan and the role assigned to each outcome. A correction method, hierarchy of tests, or other multiplicity strategy is not reported in the registry record. The significance level in the planning assumptions does not establish which tests were conducted or how multiple tests were handled.

Missing-data rules are not reported in the registry record. The number of missing observations, any exclusions from the endpoint calculation, and the method used to handle missing values are also not reported. The endpoint description states that outlier reaction times are removed, but this is not a missing-data procedure and the record does not quantify those removals.

The analysis population is not reported in the registry record. It is therefore not possible to determine from the recorded details whether all enrolled participants contributed to each outcome or whether different analyses used different subsets. The registry’s enrollment figure and group safety figures do not establish the analysis denominator for the primary endpoint.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

For the Characterization of Learning group, the registry lists 18 at risk for serious adverse events and 0 serious adverse events. It also lists 18 at risk for other adverse events and 0 other adverse events, with 0 deaths. These are the safety figures recorded for the study group.

As a general statistical concept, a zero count means that no events of that type were recorded in the stated safety data; it does not prove that the underlying risk is zero. Interpretation depends on the amount and duration of observation, how events were defined and collected, and whether participants were followed completely. The follow-up period and detailed safety-collection methods are not reported in the registry record.

These figures should be read separately from the learning endpoint. They do not indicate whether participants learned in the tasks, and they do not establish a treatment benefit or harm comparison because no separate control group is listed. The record’s safety counts are descriptive, not an estimate of a comparative safety effect.

13. What the Result Does, and Does Not, Mean

The registry record describes an investigation of learning ability in people with aphasia. Its primary endpoint is designed to quantify SRT observational learning using a within-task contrast in reaction times, standardized as Cohen's d for each participant and averaged across participants. The endpoint definition identifies the direction of better learning as negative. The primary numerical result, however, is not reported in the registry record.

Without the estimate and a measure of uncertainty, the observed magnitude and precision cannot be judged. The record also does not report numerical results for the stated relationships between learning ability, lesion characteristics, language ability, and cognitive ability. As a result, the registry entries do not establish whether those relationships were present, how large they were, or how uncertain they were.

The study is single group and has no separate comparator. The task contrast used to construct the primary endpoint does not turn the study into a randomized comparison. The record therefore does not establish that any intervention caused a change in learning, nor does it compare the effectiveness of treatments. The listed learning tasks and assessments are study activities for characterization, not evidence of a treatment effect.

The study’s stated motivation is to consider learning ability when thinking about how aphasia interventions might be aligned with individual learning characteristics. That rationale is not itself a result. The record does not show that a learning phenotype predicts response to therapy, that matching a therapy to a learning profile improves outcomes, or that any particular clinical approach is appropriate. Such claims would require evidence beyond the study purpose and endpoint description.

The safety record lists no recorded serious adverse events, other adverse events, or deaths for the group, but the observation details are not reported. These descriptive counts do not resolve questions about safety beyond the recorded data. Overall, the registry record outlines what the study set out to measure, but the missing numerical efficacy and association results limit interpretation of what the study found.

14. Why This Trial Matters Statistically

This study is statistically notable for separating learning characterization from a conventional treatment comparison. The registry describes a single-group design and defines a participant-level learning measure through a contrast between sequenced and pseudorandomized task blocks. That distinction matters: the task contrast can quantify performance according to the endpoint definition, but it is not a randomized control-group comparison.

The endpoint also illustrates how a standardized effect size can summarize a contrast whose raw measure is reaction time. Using pooled standard deviations places the individual contrast on a standardized scale, while averaging individual Cohen's d values gives the registry’s stated group-level parameter. Interpretation still depends on direction, variability, and uncertainty. The record specifies that negative values indicate better learning, but does not report the numerical mean or its precision.

The study’s stated objectives connect learning measures with lesion characteristics, language ability, and cognitive ability. That makes the analysis conceptually relevant to research on individual differences. Statistically, however, such relationships require reported estimates and uncertainty, as well as clear information about analysis populations and the handling of multiple outcomes. Those details are not reported in the registry record.

The record lists several outcomes and analyses, making transparency about outcome selection, missing data, and multiplicity important for interpretation. It also includes a planning illustration whose two-arm assumptions do not match the listed single-group design. Recognizing that distinction helps prevent planning calculations from being mistaken for a description of the completed study or for observed power.

The statistical importance of the record therefore lies in its measurement framework and research question, not in an effect estimate that is absent from the registry record. The available information describes a way to characterize learning, while leaving the size, precision, and practical implications of the findings unresolved.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The principal limitation for interpreting results is that numerical primary endpoint results are not reported in the registry record. The record does not give the mean Cohen's d, its standard deviation, a confidence interval, or a p-value. It also does not report numerical results for the stated relationships between learning ability and lesion, language, or cognitive characteristics. The size and precision of any findings therefore cannot be evaluated.

The study is listed as a single group, with allocation recorded as NA and no separate comparator. Findings from the task contrast should not be interpreted as randomized treatment effects. The registry does not report a detailed statistical analysis plan, a sample-size justification, or an interim-analysis plan. The sample-size assumptions describe a two-arm planning illustration rather than the listed study design.

The endpoint definition specifies removal of reaction times more than three standard deviations above each block’s mean, but the number of removed observations and the number contributing to each participant’s score are not reported. Missing-data rules, the analysis population, and the handling of incomplete task data are not reported in the registry record. These omissions limit assessment of how robust the endpoint is to exclusions or incomplete observations.

The registry lists 8 outcome measures and 12 analyses, but does not report a multiplicity strategy or full details linking analyses to outcomes. Safety counts are recorded, but the duration and completeness of safety observation are not reported. Finally, the record’s stated focus and endpoint definition do not establish whether learning measures predict response to aphasia therapy or whether treatment alignment improves clinical outcomes.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Determining Learning Ability in People With Aphasia — NCT05119023.
• Registry record last updated 2025-03-10; status last verified 2025-02.
• Results first posted 2025-03-10.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

NCT05119023 is a completed, single-group study titled Determining Learning Ability in People With Aphasia, sponsored by MGH Institute of Health Professions. The registry lists enrollment of 18, allocation as NA, and masking as none. The study start date is 2022-06-06, and the primary completion date is 2023-09-01.

Study activities include SRT and AGL observational learning tasks, rule-based AGL learning, standardized cognitive-linguistic assessment, and brain imaging. The primary endpoint is mean Cohen's d for SRT observational learning, with negative values defined as indicating better learning. The numerical primary result and uncertainty estimates are not reported in the registry record.

The record lists 8 outcome measures and 12 analyses. For the Characterization of Learning group, it records 0 serious adverse events, 0 other adverse events, and 0 deaths. The registry record does not report detailed analysis-population, missing-data, or multiplicity methods.