Factorial Optimization Trial to Test Effects of Coping Intervention Components: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Factorial Optimization Trial to Test Effects of Coping Intervention Components (NCT05822687), a randomized na study. The primary outcome is Mean Child-Reported Interparental Emotional Security Score on the SIS-SF at 3-Month Follow-Up.

2. Clinical Question

3. Trial Design

This is a randomized, single-masked factorial behavioral trial, registered as completed. The registry lists the phase as not applicable. Its enrollment is 148. The study title describes an optimization trial: rather than evaluating only one fixed package against education, the design varies the presence of three digital coping components—reappraisal, distraction, and relaxation—to assess their main and interactive effects.

Every group receives an educational digital module. The education-only group receives no skills-based module; the remaining groups receive one or more skills modules. The combinations cover each component individually, pairs of components, and all three together. This structure can support questions about the effects associated with including a particular component across combinations, as well as whether effects may differ when components are combined. The registry record does not report the detailed allocation procedure beyond randomized allocation, nor does it identify who was masked.

The primary outcome is the mean child-reported SIS-SF score at approximately 3 months after posttest. The registry lists seven outcome measures and 49 analyses. The primary outcome’s descriptive results are based on 78 participants with observed scores at follow-up, with numbers varying by arm because of missing follow-up data. The posted Bayesian intent-to-treat analysis included 147 of 148 randomized participants; one participant was excluded for missing baseline coping. The record does not report a sample-size justification, a planned interim-analysis scheme, or a complete statistical analysis plan.

Recruitment began on 2023-02-18, primary completion is listed as 2026-03-30, and the record status is completed. These registry entries describe the study’s design and status; they do not by themselves establish how fully each planned analysis was reported.

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: After eligibility and baseline assessments, all 148 enrolled children were randomized to one of eight conditions in the 2 × 2 × 2 factorial design. Parents/legal guardians were not enrolled as participants; they provided permission and completed parent-report measures about their child as informants. Parents were not randomized or assigned to an intervention.

Only the first 6 of 8 groups are shown.

Overall Study

Analysis population for the primary outcome, as defined in the registry: The descriptive outcome table includes 78 participants with an observed score at the 3-month follow-up; numbers vary by arm because of missing follow-up data. The Bayesian intent-to-treat analysis included 147 of 148 randomized participants; one participant was excluded for a missing baseline coping-style covariate. Missing outcome and baseline-outcome data were jointly modeled.

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.

Only the first 6 of 9 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Mean Child-Reported Interparental Emotional Security Score on the SIS-SF at 3-Month Follow-Up, measured over approximately 3 months after posttest.

The primary endpoint is the mean child-reported score on the Security in the Interparental Subsystem Scale–Short Form (SIS-SF) at approximately 3 months after posttest. It is a continuous scale outcome: the score is the mean of 17 items, each rated from 0 (not at all true) to 3 (very true for me). Higher scores indicate greater interparental emotional security, which the registry describes as the better outcome.

This endpoint is relevant to the study’s stated focus on children exposed to interparental conflict after separation or divorce. It measures children’s reported sense of security in relation to the interparental subsystem, rather than a clinical diagnosis or an event count. The registry cites the measure as SIS-SF and identifies Holt et al., 2020, but further information about its measurement properties is not reported in the registry record.

The posted arm-level results are observed mean scores and standard deviations. These descriptive summaries should be distinguished from the factorial analyses, which adjust for baseline SIS-SF and prespecified covariates. The descriptive outcome table includes 78 participants with an observed score at follow-up, and the number varies by arm because of missing follow-up data. The registry record does not report a confidence interval for each arm’s mean. As a result, the arm summaries alone do not express the precision of each mean or provide a complete adjusted comparison among factorial components.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a factorial design estimate?

In general, a factorial design varies multiple intervention components within one experiment. Here, the components are digital modules for reappraisal, distraction, and relaxation, on a shared educational-module background. This arrangement can estimate component effects across the combinations in which each component appears, while also allowing investigation of interaction effects. An interaction means that the association for one component may depend on whether another component is present. The registry’s stated primary aim includes both main and interactive effects, but the model details and the specific estimands are not reported in the registry record.

The existence of multiple combinations does not make every simple arm-to-arm comparison the central factorial question. A contrast between two individual combinations can differ from a model-based component effect averaged across other components. To interpret a factorial estimate, readers need to know how components and interactions were represented, what comparison was defined, and which outcome direction was coded. Those details are not reported for the listed standardized contrasts.

What do the means and standard deviations describe?

For a continuous measure, a mean is the average observed scale score among the participants represented in that summary. A standard deviation describes how spread out the observed scores are around the mean; it is not a confidence interval and does not show the uncertainty in the estimated mean. The posted arm summaries are descriptive and use observed follow-up values. The registry notes that the numbers vary by arm because follow-up data are missing.

In general, an adjusted analysis can differ from an unadjusted summary when it accounts for baseline outcome values and other prespecified covariates. The registry says its factorial analyses adjust for baseline SIS-SF and prespecified covariates, but does not identify those covariates or give model specifications. Thus the observed means should be read as a straightforward description of the available follow-up scores, not as the adjusted estimates from the factorial model.

How should standardized contrasts be read?

Cohen’s d expresses a contrast on a standardized scale, so the difference is represented relative to variability rather than in the original SIS-SF units. In general, standardization can help compare effects measured on different scales, but it does not remove the need to know which groups or component effect are being contrasted. Interpretation also depends on the direction convention: a positive or negative value is meaningful only when the reference and outcome coding are clear.

The registry lists seven standardized contrast values from Bayesian factorial regression, but supplies no contrast labels, interval estimates, or p-values. It is therefore not possible to determine from the record which component effect each value represents, or to judge how precisely each was estimated. A point estimate alone does not show the range of values compatible with the analysis. For Bayesian analyses, posterior intervals are a common way to convey uncertainty; none are reported for these entries.

Why does missing follow-up matter?

Missing outcome data can affect both the number of observations contributing to descriptive summaries and the assumptions required for model-based analyses. In general, an analysis can address missingness in different ways, and conclusions may depend on whether the method and assumptions are appropriate. The record identifies 78 participants with observed primary follow-up scores and notes variation by arm. It also reports that the intent-to-treat Bayesian analysis included 147 of 148 randomized participants, with one exclusion for missing baseline coping. It does not describe how missing follow-up SIS-SF values were handled in that analysis, or whether additional missing-data sensitivity analyses were performed.

Intent-to-treat is commonly used to describe an analysis that seeks to retain participants according to their randomized assignment, but the posted description is not a full account of the handling of every missing measure. The specific reason one participant was excluded is reported as missing baseline coping. Further details on any imputation, model-based handling of missing outcomes, or sensitivity analysis are not reported in the registry record.

What can be recalculated from these summaries?

A simple unadjusted two-arm comparison requires the relevant group summaries and a clearly specified pair of groups. The registry’s comparison data are not available for a two-arm calculation of this continuous endpoint, as indicated by the record’s note that values are missing or not estimable. Such a recalculation cannot reproduce the adjusted Bayesian factorial regression in any case: it would answer a different, unadjusted question and would not account for the factorial structure or the stated covariate adjustment.

In general, a confidence interval communicates uncertainty around an estimate, while a p-value addresses compatibility with a specified null model under frequentist assumptions. Neither is a substitute for effect size or clinical context. Because no p-values or uncertainty intervals accompany the posted standardized contrasts, these summaries do not support a threshold-based significance interpretation. The registry record also does not report the full statistical plan, multiplicity procedure, or posterior decision criteria. Any interpretation should therefore distinguish the descriptive data from the incompletely documented model-based results.

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("Education", "Education + Reappraisal", "Education + Distraction", "Education + Relaxation", "Education + Reappraisal + Distraction", "Education + Reappraisal + Relaxation")
n   <- c(8, 9, 7, 11, 11, 11)
m   <- c(1.01, 0.88, 1.29, 0.81, 1.16, 1.28)
sp  <- c(0.56, 0.5, 0.69, 0.8, 0.67, 0.42)
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  Education  8  1.010  0.560  0.542  1.478
CARM  Education + Reappraisal  9  0.880  0.500  0.496  1.264
CARM  Education + Distraction  7  1.290  0.690  0.652  1.928
CARM  Education + Relaxation  11  0.810  0.800  0.273  1.347
CARM  Education + Reappraisal + Distraction  11  1.160  0.670  0.710  1.610
CARM  Education + Reappraisal + Relaxation  11  1.280  0.420  0.998  1.562

11. Analyses Reported by the Sponsor

The registry reports the sponsor’s Bayesian factorial regression results as standardized contrasts, with values of -0.21, -0.34, -0.10, 0.16, -0.26, -0.03, and 0.25. These entries list the full set of intervention groups, but do not identify the specific contrast associated with each value. The registry record does not report confidence intervals, confidence levels, or p-values for them. Accordingly, the listed estimates cannot be linked to particular component effects with confidence, and their precision cannot be assessed from the posted information.

The available data do not support a recalculated two-arm comparison for this continuous endpoint: the registry indicates that the required comparison values are missing or not estimable. Any simple comparison of observed arm means would also be unadjusted and would not reproduce the sponsor’s factorial model, which the record says adjusts for baseline SIS-SF and prespecified covariates. The two approaches would address different questions.

The registry record does not report enough model detail to independently reconstruct the sponsor’s analysis or verify how each contrast was formed. The descriptive arm means and standard deviations remain useful for understanding the observed scores, but they should not be treated as a replacement for the model-based factorial analysis. The posted account is therefore best read as reporting the sponsor’s analysis results while leaving their contrast definitions and uncertainty summaries unavailable.

12. Multiplicity, Missing Data and Analysis Populations

The study evaluates multiple outcomes and analyses: the registry lists seven outcome measures and 49 analyses. When many outcomes or contrasts are assessed, the chance of finding apparently notable results can increase unless the analysis plan accounts for multiplicity. In general, multiplicity procedures may control error rates or organize which outcomes are primary and which are exploratory. The registry record does not report a multiplicity adjustment, a hierarchy for interpreting the analyses, or a detailed prespecified analysis plan. It is therefore not possible to determine how the reported analyses were prioritized or how multiplicity was handled.

The primary SIS-SF outcome has two kinds of information: descriptive follow-up summaries for 78 participants with observed scores and Bayesian intent-to-treat analyses including 147 of 148 randomized participants. The registry notes that observed follow-up numbers vary by arm because of missing data, and that one participant was excluded from the Bayesian analysis for missing baseline coping. It does not describe how missing follow-up outcomes were handled, whether imputation was used, or whether sensitivity analyses assessed the impact of missingness.

These populations should not be conflated. Descriptive summaries based on observed scores characterize participants with available measurements; an intent-to-treat analysis is intended to retain the randomized comparison, but its actual handling of absent measurements depends on the model and assumptions. Those details are not reported in the registry record. The record also does not specify the precise randomization procedure or the identity of the masked party. These omissions limit assessment of analysis reproducibility and of the assumptions underlying the reported estimates.

Finally, the seven reported standardized contrasts lack labels and uncertainty intervals, while p-values are also not reported. Without contrast definitions, multiplicity information, and uncertainty summaries, readers cannot determine which findings were primary among the model-based estimates or how stable they are. This is a reporting limitation, not evidence that a particular result is or is not robust.

13. 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.

The serious-adverse-event fields report zero serious events in each listed group. The registry also reports zero other adverse events and zero deaths for each group. These are the figures entered in the safety record; they should not be interpreted as proof that no adverse experiences occurred outside the events captured and reported there.

The numbers at risk differ by group in the adverse-event record. Those safety denominators are not the same presentation as the observed primary-outcome counts, and they should not be substituted for them. The registry does not list serious adverse-event terms, and the top serious-event terms field is empty. The adverse-event record therefore gives event counts but little descriptive detail about monitoring, definitions, or collection procedures.

In general, a zero count in a trial record is a statement about reported events in the recorded observation and reporting process. It does not establish that an intervention has no risks, particularly when information about event ascertainment and follow-up is limited. The posted safety figures should be read narrowly as registry-reported counts, not as a complete characterization of safety.

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

The record describes a randomized factorial study of digital coping components for children exposed to high interparental conflict after parental separation or divorce. Its structure is intended to assess reappraisal, distraction, and relaxation modules individually and in combination, with an education-only group as a comparator. The primary measure is child-reported emotional security at approximately 3 months after posttest. This makes the reported outcome relevant to the trial’s stated focus, but it remains a scale score rather than a diagnosis or a direct measure of every mental-health outcome named in the study aims.

The posted descriptive scores show variation among intervention combinations, while the Bayesian model results are given as standardized contrast values without labels or interval estimates. The record defines higher SIS-SF scores as greater emotional security, but does not provide enough information to map each contrast’s sign to a specific component effect. The available information therefore does not establish which component or combination has a more favorable adjusted result, nor how precisely the reported estimates are known.

These results do not, on their own, show that a coping module works or does not work. The registry’s summaries have missing follow-up data, and the analysis details needed to understand the model, contrast definitions, missing-data handling, and multiplicity are not reported. Nor should the absence of reported adverse events be interpreted as proof of no risk. The record supports a description of the trial’s design, measured outcome, observed summaries, and listed model estimates; broader conclusions require a more complete account of the analyses and their uncertainty.

General statistical context matters: estimates are more informative when paired with clearly defined contrasts and uncertainty intervals. Those elements help readers distinguish the direction of a comparison from the range of values that remain compatible with the data. Without them, the posted results are best treated as incomplete evidence about the stated factorial questions rather than as a definitive ranking of intervention packages.

15. Why This Trial Matters Statistically

Factorial trials address a practical statistical challenge: intervention packages often contain several components, and studying only a complete package can leave unclear which elements contribute to observed outcomes. This trial varies three coping strategies across combinations while retaining an educational module in all groups. In principle, that structure can help separate component effects from the effect of receiving a particular package, and can investigate whether combinations behave differently from individual components.

The design also illustrates why factorial interpretation depends on more than comparing the observed means of each group. A component effect may be estimated across multiple combinations, and interactions can make an effect conditional on what else is included. Model definitions, contrast labels, uncertainty estimates, and multiplicity handling are therefore central to understanding the results. The registry’s stated aims include main and interactive effects, but the posted analysis details do not fully identify the estimates.

The study also highlights a common reporting distinction between observed outcome summaries and model-based analyses. The SIS-SF means describe available follow-up scores, whereas the factorial regression is described as adjusting for baseline SIS-SF and prespecified covariates. The two forms of information answer related but not identical questions. Reporting both, with complete definitions and uncertainty, would help readers assess what the data say about the component-optimization question.

As a statistical case, the record is informative about the design’s structure and the limits of interpreting incomplete results. It shows why a clearly randomized, multi-component design still needs transparent reporting of estimands, missing-data methods, and uncertainty for its conclusions to be independently evaluated.

16. 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 interpreting the efficacy results is incomplete reporting of the Bayesian factorial analysis. Seven standardized contrast values are listed, but the record does not identify which component effect or comparison each represents. It also does not report confidence intervals, posterior intervals, confidence levels, or p-values. Without contrast definitions and uncertainty summaries, neither direction nor precision can be established for particular components.

The descriptive primary-outcome results are based on observed follow-up scores, with 78 participants represented and numbers varying by arm because of missing follow-up data. The registry reports that the Bayesian intent-to-treat analysis included 147 of 148 randomized participants, with one excluded for missing baseline coping, but does not explain how missing follow-up SIS-SF scores were handled. The missing-data assumptions and any sensitivity analyses are not reported.

The registry lists seven outcomes and 49 analyses, but does not report how multiplicity was addressed or how analyses were prioritized. It also does not provide the full statistical analysis plan, detailed factorial model specification, or the identities of the prespecified covariates. These gaps limit reproducibility and make it difficult to assess whether the reported estimates correspond to the intended main and interaction effects.

Other trial details are also limited in the registry record: the precise randomization procedure and the identity of the masked party are not stated. The serious-adverse-event section reports zero events in each group but does not give event terms or detailed ascertainment information. The record does not provide a sample-size justification. These are limitations in what can be evaluated from the registry record, not proof that the study did not use procedures that are simply unreported.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Factorial Optimization Trial to Test Effects of Coping Intervention Components — NCT05822687.
• Registry record last updated 2026-09-30; status last verified 2026-09.
• Results first posted 2026-09-30.
• 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.

20. Record Summary

NCT05822687 is a completed randomized, single-masked factorial behavioral trial sponsored by Arizona State University. It enrolled 148 participants and tested combinations of digital reappraisal, distraction, and relaxation modules alongside an educational module.

The primary outcome is child-reported mean SIS-SF score at approximately 3 months after posttest; higher scores indicate greater interparental emotional security. Descriptive scores are reported for 78 participants with observed follow-up data, with counts varying by group. The registry lists seven Bayesian factorial standardized contrast values, but does not report their individual contrast definitions or uncertainty intervals. The adverse-event record reports zero serious events, other events, and deaths in each group. Important details about missing follow-up data, multiplicity, model specification, and analysis uncertainty are not reported in the registry record.