This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Mental Imagery Therapy for Autism (MITA) - an Early Intervention Computerized Language Training Program for Children With ASD (NCT02708290), a registered phase not stated study. The primary outcome is Improvement of Language as Measured by Autism Therapy Evaluation Checklist (ATEC) and Mental Synthesis Evaluation Checklist (MSEC)..
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT02708290 |
| Phase | Phase not stated |
| Condition | Autistic Disorder, Autism, ASD |
| Allocation | Not Stated |
| Design model | Not Stated |
| Masking | Not Stated |
| Enrollment | 6454 |
| Arms in the primary outcome | n/a |
| Lead sponsor | ImagiRation, LLC |
| Results posted | Yes |
2. Clinical Question
Population
Children with autism spectrum disorder (ASD) who participated in the MITA evaluation. The registry lists an enrollment of 6454. The reported arms each include 1009 participants at risk for the adverse-event summaries, but the record does not explain how these figures relate to the total enrollment or the population used in the outcome analysis.
Intervention
The intervention was MITA Prefrontal Synthesis exercises delivered through an application. The description says each activity adapts to provide a level of difficulty intended to suit a child at a given point in time. The program includes verbal and nonverbal activities.
Comparator
The control arm consisted of participants who did not meet the test-arm criteria. The test arm included participants who completed more than one thousand exercises and made no more than one error per exercise. Test participants were matched to control participants on age, gender, expressive language, receptive language, sociability, cognitive awareness, and health at the 1st evaluation. Allocation was not stated, and the registry cautions that the arms are not a like-for-like control comparison.
Primary question
The stated primary outcome concerns improvement in language measured with the Autism Therapy Evaluation Checklist (ATEC) and Mental Synthesis Evaluation Checklist (MSEC). Parents completed evaluations every three months for up to three years. The registry record does not report the estimated score changes or the between-arm result, so the magnitude and precision of the observed comparison cannot be assessed from the posted results.
Statistically, this is a question about changes in repeated symptom-scale scores over time and whether those changes differ between groups defined by exercise completion and performance. It is not a randomized treatment comparison: the groups were formed using participants’ exercise behavior and outcomes, and the registry does not state that allocation was randomized. Consequently, differences could reflect characteristics associated with completing many exercises, as well as any differences associated with participation in MITA. The record does not report an estimate that would allow those possibilities to be quantified.
3. Trial Design
The registry describes a completed study sponsored by ImagiRation, LLC. The study began in 2015-09, had primary completion on 2020-07-01, and lists enrollment of 6454. Its intervention was MITA Prefrontal Synthesis exercises, described as an adaptive computerized program for children with ASD. The program’s verbal activities progress from vocabulary exercises to more complex language tasks; nonverbal activities use visual puzzles intended to engage mental integration.
The comparison has a test arm and a control arm, but allocation was not stated. The test group included participants who completed more than one thousand exercises and made no more than one error per exercise. The control group included the rest of the participants. Test participants were matched to control participants by age, gender, expressive language, receptive language, sociability, cognitive awareness, and health at the 1st evaluation. The registry explicitly cautions that this is not a randomized comparison and that the arms are not a like-for-like control comparison.
That distinction matters when interpreting group differences. The test-arm definition depends on exercise completion and accuracy, which characterize participant behavior during the program rather than a randomly assigned exposure. Matching on the listed characteristics may address some measured differences at the 1st evaluation, but the registry does not report whether other differences were considered or how matching was performed. It also does not report a protocol for assigning participants, a detailed analysis-population definition, or the number of participants contributing outcome data over time.
The primary outcome was assessed through parent evaluations every three months, for up to three years. The registry says score changes were modeled using a linear model with repeated measures and that least squares mean differences were reported. The outcome results themselves, including the estimated differences and their uncertainty, are not reported in the registry record.
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.
Before assignment: MITA was made available gratis at all app stores from 2015 to 2020. From this pool of potential study participants, we selected participants for the observation clinical study based on the following criteria: 1. Consistency: at least three ATEC evaluations completed. 2. The interval between the first and the last evaluation was six months or longer. 3. Diagnosis: ASD 4. Age: 2 to 12 years This resulted in 6454 participants, from whom we have selected the test and control groups
Overall Study
| Stage | Test Arm: Completed More Than One Thousand Exercises | Control Arm: the Rest of Participants |
|---|---|---|
| STARTED | 1009 | 1009 |
| COMPLETED | 1009 | 1009 |
| NOT COMPLETED | 0 | 0 |
Analysis population for the primary outcome, as defined in the registry: The framework for evaluation of score changes over time was explained in Mahapatra et al. Autism Dev Disord, 2018. In short, changes in the score were modeled by applying the Linear Model with repeated measures. Least squares means differences generated by the model are reported below. Participants were matched using propensity score analysis.
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.
| Characteristic | Unit | Test Arm | Control Arm | Total |
|---|---|---|---|---|
| Participants analysed | 1009 | 1009 | 2018 | |
| Age, Customized | Participants | 1009 | 1009 | 2018 |
| Sex: Female, Male: Female | Participants | 262 | 252 | 514 |
| Sex: Female, Male: Male | Participants | 747 | 757 | 1504 |
| Race and Ethnicity Not Collected | Participants | — | — | 0 |
6. Primary Endpoint
The registry lists the primary outcome as Improvement of Language as Measured by Autism Therapy Evaluation Checklist (ATEC) and Mental Synthesis Evaluation Checklist (MSEC)., measured over up to three years, assessed at 3 months intervals.
| Element | Registry entry |
|---|---|
| Result type | Least Squares Mean |
| Measure of spread | Standard Error |
| Unit | units on a scale |
| Endpoint classification | Continuous Endpoint |
The primary endpoint concerns score changes on the Autism Therapy Evaluation Checklist (ATEC) and Mental Synthesis Evaluation Checklist (MSEC). Parents completed evaluations every three months, with assessment continuing for up to three years. The record describes five subscales: receptive language, expressive language, sociability, cognitive awareness, and health. It gives score ranges of 0 to 40 points for receptive language, 0 to 28 points for expressive language, 0 to 40 points for sociability, 0 to 36 points for cognitive awareness, and 0 to 75 points for health.
For all five subscales, a lower score indicates lower severity of ASD symptoms and a higher score indicates more severe symptoms. Receptive language is based on the MSEC evaluation; the other listed subscales are based on ATEC. These scales allow repeated assessments to track reported symptoms across multiple domains rather than relying on a single language task. Repeated measurements can also describe the course of scores over time.
The stated analysis uses least squares mean differences from a linear model with repeated measures. Such model-based means can account for the repeated-measure structure when estimating group patterns over time, subject to the model and data assumptions. However, the registry record does not report the actual score estimates, differences, standard errors, confidence intervals, or statistical tests. It is therefore not possible to describe the direction, size, or precision of the endpoint result from the posted results.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The record identifies a clinically relevant measurement question: whether scores on language-related and other ASD symptom subscales change over repeated parent assessments, and how the modeled changes compare between the test and control groups. But the essential numerical outcome results are not reported in the registry record. No least squares mean difference, score change, standard error, confidence interval, or p-value is listed. Without those values, the observed direction and magnitude of the comparison and its statistical precision cannot be described.
The design also limits what a group comparison can establish. Test-arm membership was defined by completing more than one thousand exercises while making no more than one error per exercise. Controls were the rest of the participants, and the groups were matched on specified characteristics at the 1st evaluation. Because allocation was not stated and the registry identifies the comparison as nonrandomized, a difference in score trajectories would not by itself isolate an effect of MITA. Engagement, performance, or other group differences could also be related to subsequent scores. The record does not report an adjusted estimate or information that would allow the influence of these factors to be judged.
Repeated assessments are potentially informative because they can show how scores develop over time rather than only comparing a single measurement. A repeated-measures model can use a participant’s series of observations to estimate an average pattern, with least squares means representing model-based means. As a general statistical point, the usefulness of such a model depends on its specification and on how missing observations and within-participant correlation are handled. Those details are not reported in the registry record.
The registry describes standard errors as the measure of dispersion for the continuous outcome, but it does not give numerical standard errors. A standard error expresses the sampling uncertainty of an estimate; a confidence interval would show a range of values compatible with the data and model under its assumptions. Neither type of numerical uncertainty can be interpreted without the actual results.
The adverse-event summaries list 1009 participants at risk in each arm and zero affected for serious adverse events, other adverse events, and deaths. These are the posted counts, not evidence that no events occurred outside the populations or reporting process represented in those summaries. The registry does not provide further event detail.
Any sample-size illustrations based on an observed effect are planning scenarios, not observed power. Since the outcome effect estimate is not reported, the numerical basis for such illustrations cannot be evaluated from the registry record. Overall, the record documents the study framework and measurement approach, but it does not provide the primary outcome results needed to judge how scores differed or how precisely any difference was estimated.
9. Statistical Methods Explained
The registry describes a linear model with repeated measures. In general, repeated-measures methods are used when the same participants contribute measurements at multiple times. The measurements from an individual are usually related to one another, so treating every assessment as independent can misrepresent uncertainty. A model can represent the time pattern and account for the correlation structure specified by the analyst. The registry does not report the model formula, covariance structure, or other details needed to reproduce the analysis.
What does a least squares mean represent?
A least squares mean is a model-based average. Unlike a simple average of recorded scores, it is calculated from the fitted statistical model and can reflect the model’s treatment of time and other terms. A difference between least squares means is an estimate of the contrast defined by that model. In this study, the registry says least squares mean differences were generated by the repeated-measures model. It does not report the numerical estimates or define precisely which time points or contrasts those estimates compare.
For continuous outcomes, an estimate is more informative when accompanied by a measure of uncertainty. The registry identifies standard error as the dispersion measure. In general, a standard error describes how much an estimate would vary across repeated samples under the statistical model. A confidence interval uses the estimate and its uncertainty to express a range of values compatible with the data and assumptions. The record does not provide numerical standard errors or confidence intervals, so the precision of the primary result cannot be assessed.
Why does the nonrandomized comparison matter?
Random allocation, when used, is intended to make groups comparable on average with respect to measured and unmeasured characteristics. The registry does not state that allocation was randomized. Instead, the test group was defined by completing more than one thousand exercises and making no more than one error per exercise, while the control group consisted of the rest of participants. Matching was reported for age, gender, expressive language, receptive language, sociability, cognitive awareness, and health at the 1st evaluation.
As a general statistical principle, matching can make groups more similar on the characteristics used in the matching process, but it does not guarantee balance on unmeasured factors. In this design, group definition is also connected to exercise engagement and performance. Those features may be associated with participant characteristics or score trajectories. Therefore, a modeled group difference would be an association in this nonrandomized comparison and should not automatically be interpreted as an effect attributable only to the intervention. The registry record does not report details of the matching procedure or a full account of adjustment for possible confounding.
What is over-dispersion, and is it relevant to this outcome?
Over-dispersion is a general issue for outcomes whose variability is greater than a particular statistical model assumes. It often arises with count data, where an overly narrow variance assumption can make uncertainty appear smaller than it is. The primary outcome is described as continuous, with scores on assessment subscales, and the named approach is a linear model with repeated measures. The registry does not report a dispersion assessment or a model adjustment for over-dispersion. This concept should not be treated as a reported finding for this trial.
What can be said about planning calculations?
The registry describes sample-size illustrations based on an observed effect, with a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. It states that the effect to detect is a specified fraction of the observed effect, and that non-time-to-event calculations are expressed as participants per arm. Such calculations are planning illustrations: they describe sample sizes under stated assumptions, not the probability that a completed study has detected a true effect. They are not observed power. The primary effect estimate is not reported in the registry record, so the effect used for those illustrations cannot be independently evaluated from the posted results.
The record lists one outcome measure and one analysis, while the described outcome spans five subscales and repeated assessments. The registry does not explain how the reported analysis count relates to those subscales or to any possible contrasts across time. It also does not report a model diagnostic assessment, the treatment of baseline scores in the model, or whether estimates were adjusted for additional covariates. These absences prevent a detailed appraisal of the model’s implementation.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists one outcome measure and one analysis, while the primary outcome description includes five subscales assessed every three months for up to three years. Those multiple domains and repeated time points can create multiple opportunities for statistical comparisons, depending on which contrasts are tested. The registry record does not report a multiplicity adjustment, a hierarchy among the subscales, or a prespecified selection rule for the primary contrast. It also does not explain how the single listed analysis relates to the several domains and time points described.
Missing-data handling is not reported in the registry record. The record does not state how many participants contributed assessments at each time, how incomplete follow-up was treated, whether participants with no later assessments were included, or what assumptions were made about missing observations. As a general statistical point, repeated-measures models can use incomplete series under particular assumptions, but the appropriateness of that approach depends on why data are missing and how the model is specified. The registry does not provide enough detail to assess those assumptions.
The analysis population is also not defined in sufficient detail. The registry describes the test and control groups and states that matching was based on characteristics at the 1st evaluation, but it does not report a full rule for inclusion in the outcome analysis or whether all enrolled participants were analyzed. This matters because the arm definitions depend on exercise completion and accuracy, and because the adverse-event summaries list 1009 at risk in each arm while total enrollment is 6454. The record does not explain the relationship between these counts. No inference about exclusions, follow-up completeness, or analysis-population membership can be made from the counts alone.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Test Arm | 0 | 1009 | 0.0% |
| Control Arm | 0 | 1009 | 0.0% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Test Arm | 0 of 1009 |
| Control Arm | 0 of 1009 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Test Arm | 0 of 1009 |
| Control Arm | 0 of 1009 |
The registry lists 1009 participants at risk in the Test Arm and 1009 in the Control Arm for serious adverse events. It reports zero participants affected in each arm and zero deaths in each arm. The same at-risk and affected counts are listed for other adverse events: 1009 at risk and zero affected in each arm. No top serious-event terms are listed.
These figures should be read as the adverse-event counts recorded in the registry, not as proof that no adverse events occurred in every participant or setting. The record does not give further detail about event collection, reporting procedures, follow-up for safety, or how the at-risk counts relate to the total enrollment of 6454. It also does not describe any event terms beyond the empty listing of top serious terms. The reported zero counts therefore describe the posted summaries and their stated denominators; they do not supply a detailed account of safety monitoring or event ascertainment.
The test and control arms are nonrandomized groups, and the registry cautions that they are not a like-for-like control comparison. The adverse-event counts alone do not establish that group membership caused or prevented events. They are descriptive counts, with no further comparative safety analysis reported in the registry record.
12. What the Result Does, and Does Not, Mean
The registry record establishes that MITA was evaluated in a completed study involving children with ASD, using repeated parent assessments of five symptom subscales. It also describes a model intended to analyze score changes over time and a comparison between participants who met the test-arm exercise criteria and the rest of the participants. These details define the question and the general analysis framework.
The record does not report the numerical primary outcome results. There is no listed score change, least squares mean difference, standard error, confidence interval, or p-value. As a result, the posted results do not establish the direction of the comparison, its magnitude, or its precision. It is not possible to determine from the record whether the groups’ scores changed differently, or whether any estimated difference was compatible with a wide or narrow range of values.
Even if an outcome estimate were available, the design would still require careful interpretation. The groups were not stated to be randomized, and the test group was defined using exercise completion and accuracy. Matching on characteristics at the 1st evaluation addresses only the listed matching dimensions and does not by itself eliminate possible differences on other characteristics. A group contrast in this setting would not automatically isolate the effect of the program.
The observed adverse-event summaries report zero affected participants and zero deaths in both arms among 1009 at risk per arm. Those counts describe the registry entries but do not resolve questions about the scope of event collection or explain the relationship between the adverse-event populations and total enrollment. Similarly, planning illustrations based on an observed effect are not a substitute for reporting the observed estimate and its uncertainty, and they are not observed power.
In short, the record supports a description of the study’s population, comparison, outcome instruments, and stated modeling approach. It does not support a quantitative conclusion about the primary outcome because the primary estimates and their uncertainty are not reported.
13. Why This Trial Matters Statistically
This trial is statistically notable because its comparison groups are defined by participants’ exercise behavior and performance rather than by stated random allocation. The test arm required completion of more than one thousand exercises and no more than one error per exercise; the control arm comprised the rest of participants. That structure makes selection and comparability central to interpretation. Matching on age, gender, expressive language, receptive language, sociability, cognitive awareness, and health at the 1st evaluation is relevant, but the registry does not report enough detail to determine how the matching was conducted or what differences remained.
The outcome is also longitudinal. Parent evaluations were scheduled every three months for up to three years, across receptive language, expressive language, sociability, cognitive awareness, and health. Repeated-measures models can represent patterns across time and account for the fact that observations from the same participant are related. But meaningful interpretation requires details about model specification, missing observations, and the contrasts of interest. Those details are not reported in the registry record.
Most importantly, the record describes the planned type of estimate without reporting the primary numerical results. Without an estimate and its uncertainty, statistical readers cannot assess how large a score difference was observed or how precisely it was estimated. The case therefore illustrates that a detailed endpoint description and a named model do not replace transparent reporting of effect estimates and uncertainty. The nonrandomized design, repeated assessments, and absent numerical outcome results all shape what can be concluded from the registry entry.
14. 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 primary limitation for interpretation is that numerical results for the stated endpoint are not reported in the registry record. The outcome description names least squares mean differences and standard errors, but gives no estimates, standard errors, confidence intervals, or p-values. Without these, the direction, size, and precision of the primary comparison cannot be assessed.
The comparison is nonrandomized, with allocation not stated. Test-group membership depended on completing more than one thousand exercises and making no more than one error per exercise. Such criteria may identify participants with different engagement or performance patterns from the rest of participants. Matching was reported on several characteristics at the 1st evaluation, but the matching method and any additional adjustment are not described. The registry itself cautions that the arms are not a like-for-like control comparison.
The outcome includes five subscales and repeated assessments every three months for up to three years, but the record does not clarify the primary contrast across these domains and time points. It does not report whether multiplicity was addressed. The handling of missing assessments, participant withdrawal, analysis-population membership, and model assumptions is also not reported. Those omissions make it difficult to assess how complete follow-up and analytic choices may affect the estimates.
The registry lists enrollment of 6454, while adverse-event summaries report 1009 at risk in each arm. It does not explain how these numbers relate or how many participants contributed to the outcome analysis. The record also does not provide further information about safety-event ascertainment. These limitations are specific to the reported design and results and constrain conclusions drawn from the registry record.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Mental Imagery Therapy for Autism (MITA) - an Early Intervention Computerized Language Training Program for Children With ASD — NCT02708290.
- Registry record last updated 2020-11-27; status last verified 2020-11.
- Results first posted 2019-12-20.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
NCT02708290 is a completed study of MITA, a computerized intervention for children with ASD, sponsored by ImagiRation, LLC. The registry lists enrollment of 6454, a start of 2015-09, and primary completion on 2020-07-01.
The test group was defined by exercise completion and accuracy; controls were the rest of participants. Matching was reported on characteristics at the 1st evaluation, but allocation was not stated and the registry cautions that the comparison is not like-for-like. The primary outcome used parent-reported ATEC and MSEC subscales assessed every three months for up to three years, with score changes modeled using a linear model with repeated measures.
Numerical primary outcome estimates and their uncertainty are not reported in the registry record. Adverse-event summaries list 1009 at risk per arm, zero affected for serious and other adverse events, and zero deaths in each arm.