Resurgence as Choice: Basic and Clinical Studies: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Resurgence as Choice: Basic and Clinical Studies (NCT03423940), a randomized na study. The primary outcome is Number of Participants Showing Resurgence as Predicted by the Quantitative Model -- Highest Response Rate.

2. Clinical Question

The broader research motivation was that relapse can occur after FCT and that Resurgence as Choice theory may help explain this pattern. The study objective was to test the theory's predictions with human participants who exhibit severe destructive behavior. The registry record does not report a clinical treatment-effect question framed as whether FCT is better than a control condition.

3. Trial Design

The record describes a randomized, single-masked, parallel study sponsored by Rutgers, The State University of New Jersey. Its status is completed. The listed study start date is 2018-06-28, and the listed primary completion date is 2023-05-31. The registry lists enrollment as 17; it does not clarify in the information whether that enrollment figure is planned or actual.

The record identifies two arms. Evaluation of Treatment Dosage examined FCT durations in three distinct contexts for each participant, associated with short, moderate, or extended treatment durations. Evaluation of Size of Decrease in Alternative Reinforcement examined whether smaller rather than larger decreases in the availability of alternative reinforcement would reduce the magnitude of resurgence. The arm descriptions are incomplete in the registry record, so detailed procedures, assignment ratios, and the precise meaning of the masking designation are not reported.

Randomized allocation describes how participants or conditions were assigned according to the study design, but the registry record does not report the allocation sequence or the unit randomized. A parallel model identifies the study's overall design classification; it does not establish that every outcome is an independent comparison between arms. The primary outcome explicitly compares responses within participants across resurgence-test conditions.

The primary outcome timeframe is through study completion, described as an average of 4 months. The record lists phase as NA, which does not indicate a conventional drug-development phase. It also lists Problem Behavior, Self-injury, and Communication as conditions. The detailed protocol and the full schedule of assessments 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.

Overall Study

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 Number of Participants Showing Resurgence as Predicted by the Quantitative Model -- Highest Response Rate, measured over Through study completion, an average of 4 months..

The primary endpoint is the number of participants showing resurgence as predicted by the quantitative model under the highest-response-rate prediction. The registry describes the underlying measure as mean destructive responses per minute during resurgence-test conditions, compared within each participant. The outcome asks whether each participant's highest response rate occurred in the test condition that the model had predicted would produce the highest rate.

This endpoint converts a within-participant pattern into a participant count: it records whether an individual's observed response pattern agrees with the model's prediction. Such a count can make a prediction test interpretable at the participant level, but it does not preserve the size of the response-rate differences between conditions. A participant with a small difference and a participant with a large difference may contribute equally to the count if both meet the prediction criterion. The registry record does not report the rule used to classify ambiguous or tied response rates.

The endpoint is used to assess the model's prediction about the relative ordering of response rates, rather than to estimate a treatment effect on behavior in a conventional between-group comparison. The registry reports counts by arm but no inferential analysis, confidence interval, or statistical test for the primary endpoint. The recorded outcome should therefore be read as descriptive evidence about prediction matches, not as a quantified estimate of precision or a treatment benefit.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a participant count, but the observations underlying it are response rates measured across conditions within each participant. That distinction matters: the outcome is not simply an average response rate for each randomized arm. It classifies whether each participant's highest observed destructive-behavior response rate occurred in the resurgence-test condition predicted to produce the highest rate.

What does a within-participant prediction test assess?

In general, a within-participant design compares observations from the same person across conditions. This can account for stable differences between people because each person contributes observations under multiple conditions. The interpretation still depends on how the conditions were ordered, how observations were collected, and how the predicted pattern was defined. Those details are not reported in the registry record. The record says that response rates are compared within-participant and that the outcome concerns whether the highest rate matches the model's prediction, but it does not provide the classification procedure or participant-level measurements.

In general, an outcome based on whether a prediction is met discards information about magnitude. The count treats a qualifying match as an event, but it does not show the distance between response rates, the variability of measurements, or whether a near-tie was considered a match. Those details would be important to assess the robustness of the classification, and they are not reported in the registry record.

What can the arm counts tell readers?

The registry reports a count and denominator for each arm: 1 of 3 for Evaluation of Treatment Dosage and 5 of 9 for Evaluation of Size of Decrease in Alternative Reinforcement. These entries describe the recorded primary outcome. They are not adjusted estimates and are not accompanied by an inferential comparison. The primary outcome's within-participant structure also means that the arm labels should not automatically be interpreted as two groups in a treatment-effect comparison. The record does not report an analysis that tests a contrast between the arms.

In general, statistical inference would require a prespecified model or test suited to the outcome and design. For a binary participant-level outcome, methods may compare event probabilities or estimate an effect with uncertainty, while respecting the study's assignment and dependence structure. For repeated or within-participant observations, an analysis may need to account for the relationship among observations from the same participant. Which method would be appropriate depends on the full design and analysis plan. The registry lists 0 analyses and does not report the statistical method used, so a specific inferential approach cannot be attributed to this study.

How should precision be understood?

In general, a confidence interval describes the uncertainty around an estimated quantity under specified assumptions. It does not guarantee that the true value lies within the interval, and it does not by itself show that an effect is important. The registry record does not report a confidence interval for the primary outcome. It also does not report a p-value or an adjusted analysis. The posted counts therefore do not support a claim about statistical significance or a precise estimate of the frequency of prediction matches beyond the recorded denominators.

The record includes a sample-size planning illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. It describes the effect to detect as a stated fraction of the observed effect. These are planning assumptions, not a description of the trial's own design rationale or a report of observed power. The registry does not report the trial's sample-size justification. In general, sample-size calculations estimate the number of participants needed under chosen assumptions; they do not show the probability that the observed result is true, and they do not replace a confidence interval or a suitable analysis.

The planning illustration's equal-allocation assumption is an assumption for that calculation, not evidence that study allocation was equal. The record does not report the allocation ratio. Its normal-approximation assumption is likewise a calculation choice, not evidence that the participant counts follow a normal distribution. Planning assumptions should be read separately from the observed outcome counts and from the study's analysis, which is not reported.

Why does the design affect interpretation?

In general, randomization can help balance prognostic factors across assigned groups when it is implemented as intended, but it does not remove uncertainty, guarantee balance in a small study, or address every source of bias. Here the outcome is described as a within-participant comparison across test conditions, and the two arms evaluate different questions. The registry does not report enough detail to determine the randomization unit, the condition order, or how those features entered the analysis. Randomized status alone does not justify treating the two arm counts as a direct causal contrast.

Finally, the registry lists no estimate of response-rate variability, no dispersion measure for the primary endpoint, and no analysis results. In general, variation matters because people can respond differently and repeated observations can fluctuate. When response data vary more than a simple statistical model assumes, over-dispersion can make standard errors too small and confidence intervals too narrow if the model does not account for it. The registry record does not report a model or dispersion assessment, so readers cannot determine whether this issue was assessed. Statistical interpretation should remain descriptive until analysis details and uncertainty estimates are available.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Evaluation of Treatment Dosage", "Evaluation of Size of Decrease in Alternative Rein")
n      <- c(3, 9)
events <- c(1, 5)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Evaluation of Treatment Dosage  3  1  33.3  0.8  90.6  6.1  79.2  6.8  80.6
ARM  Evaluation of Size of Decrease in Alternative Rein  9  5  55.6  21.2  86.3  26.7  81.1  26.2  81.3

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary outcome concerning participants whose highest destructive-behavior response rate occurred in the condition predicted by the model, and one secondary outcome concerning the lowest-response-rate prediction. It reports two outcome measures and no analyses. The record does not report whether the outcomes were analyzed, how they were handled statistically, or whether any additional outcomes were prespecified elsewhere.

Multiplicity arises when several outcomes, contrasts, or analyses are examined. In general, conducting multiple statistical tests can increase the chance of a false-positive finding if each is interpreted without considering the set of tests. Researchers may address this through a prespecified hierarchy, an adjustment, or a clear distinction between confirmatory and exploratory analyses. The registry record does not report a multiplicity strategy, so it is not possible to determine whether or how multiplicity was addressed.

Missing-data handling and analysis populations are also not reported in the registry record. Although the enrollment entry is 17 and the primary outcome entries have arm denominators of 3 and 9, the record does not explain how those denominators relate to enrollment, assignment, or follow-up. No participant flow, exclusion criteria after enrollment, missingness reasons, or imputation method is reported. It is therefore not possible to determine whether the posted counts include all eligible participants, a defined analysis population, or a subset with available outcome data.

In general, missing-data methods depend on why observations are missing and on assumptions about the relationship between missingness and outcomes. A complete-case count can be difficult to interpret if participants with unavailable outcomes differ from those observed. The registry record does not provide the information needed to assess that possibility. Nor does it identify an intention-to-treat or per-protocol population. These omissions limit assessment of both the primary outcome and the relationship between the posted counts and the study's planned analysis.

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.

The registry's adverse-event entries list zero serious adverse events for Evaluation of Treatment Dosage, with 4 participants at risk, and zero for Evaluation of Size of Decrease in Alternative Reinforcement, with 11 at risk. The entries for other adverse events are also zero, with 4 and 11 participants at risk, respectively. Deaths are listed as zero in both groups. These are the recorded counts in the registry.

These figures should be read as registry-reported event counts, not as proof that no adverse events occurred outside the recorded definitions or reporting period. The record does not report serious-event terms, and the top serious-event terms list is empty. It does not provide event definitions, ascertainment procedures, duration of safety follow-up, or details about how the at-risk counts were determined. The adverse-event denominators also differ from the primary-outcome denominators, and the registry does not explain why.

In general, an event count of zero does not establish that an intervention is risk-free. Safety interpretation depends on how events were sought and defined, the number of people observed, and the length and completeness of follow-up. The record describes the primary outcome timeframe as through study completion, an average of 4 months, but does not specify that this is the adverse-event follow-up period. The posted figures support only the narrow statement that the registry records zero serious adverse events, zero other adverse events, and zero deaths in the listed groups.

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

The primary outcome is a test of whether observed response-rate ordering agrees with a model prediction. The registry records 1 of 3 participants in Evaluation of Treatment Dosage and 5 of 9 in Evaluation of Size of Decrease in Alternative Reinforcement as showing resurgence as predicted for the highest-response-rate outcome. This is descriptive information about prediction matches in the registered outcome record.

The counts do not establish that the theory is correct or incorrect. The registry does not report uncertainty estimates or a statistical analysis, and it does not give participant-level response rates. Without those details, the reader cannot assess the strength or precision of the evidence, the closeness of the observed responses to the predicted ordering, or the role of any missing observations. The counts should not be converted into an unreported conclusion about statistical support.

The study's two arms investigate different questions: treatment duration and the size of decreases in alternative reinforcement. The registry does not report a direct statistical comparison between the arms. Accordingly, the counts do not show that one strategy is superior to the other, nor do they establish the effect of FCT compared with no treatment or an alternative intervention. The record describes a behavioral intervention and model-prediction tests, not a conventional treatment-versus-control efficacy estimate.

The results also do not establish effects on outcomes beyond the stated endpoint. They do not quantify changes in communication, destructive behavior outside the resurgence-test conditions, or long-term outcomes. The registry identifies communication and self-injury among the conditions, but the primary endpoint is specifically about destructive-behavior response rates and model predictions. Claims about other clinical outcomes are not reported in the registry record.

In general, interpretation of a trial result requires attention to the outcome definition, design, analysis, and uncertainty, as well as the population studied. The eligibility criteria describe a specific group of children and require stable educational placement and medication circumstances. The record does not establish that the findings apply to people outside those criteria. The appropriate conclusion is limited to the posted participant counts and the outcome definition; broader clinical or theoretical conclusions are not established by the registry entries alone.

14. Why This Trial Matters Statistically

This study is statistically distinctive because the primary endpoint translates a model prediction into a participant-level count. The research question is not simply whether average behavior differs between treatment groups. It asks whether, within a participant, the test condition predicted to produce the highest destructive-behavior response rate is in fact the condition with the highest observed rate. That structure makes the outcome a test of predictive ordering rather than a conventional between-group treatment effect.

That distinction is useful for clinical researchers and biostatisticians. It highlights the importance of matching the analysis to the scientific question. If the purpose is to assess whether a quantitative theory predicts a pattern, the analysis needs to preserve and evaluate the pattern the theory specifies. A count of prediction matches offers a compact summary, while the underlying response-rate data and the rule for classifying matches are important for assessing how informative that summary is. The registry record does not provide those participant-level details or the analysis.

The two arms also address different elements of FCT and alternative reinforcement. Treating their outcome counts as a direct treatment comparison would conflate different scientific questions. The record's randomized classification does not remove the need to understand what was randomized, which conditions were compared, and what estimand the analysis targeted. Those design and analysis details are not reported in the registry record.

Finally, the posted record illustrates why outcome counts should be read alongside denominators, uncertainty, missing-data information, and the analysis plan. The primary outcome counts are available, but the record does not report confidence intervals, p-values, analysis populations, or a participant flow. This limits the strength of statistical conclusions that can be drawn, while leaving the underlying question—whether a behavioral model predicts resurgence patterns—of methodological interest.

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 registry record leaves several trial-specific details unclear. Although it labels allocation randomized, it does not report the randomization unit, sequence generation, concealment, or allocation ratio. It labels masking as single but does not identify who was masked or how masking was maintained. These details affect assessment of potential bias and how the design should be represented statistically.

The arm descriptions are incomplete. The dosage arm description identifies short, moderate, and extended FCT durations across distinct contexts, while the other arm refers to differently sized decreases in alternative reinforcement and counterbalancing of their order. The full procedures and condition definitions are not reported. This makes it difficult to assess implementation, reproducibility, or whether the conditions were comparable for the intended analyses.

The enrollment entry is 17, but the primary-outcome denominators are 3 and 9 and the adverse-event at-risk counts are 4 and 11. The registry record does not reconcile these figures or explain participant flow, exclusions, or missing outcomes. It also does not report participant-level observations, response-rate variability, the classification rule for prediction matches, or a statistical analysis. Confidence intervals and p-values are not reported.

The record does not report the sample-size justification, an interim-analysis plan, a statistical analysis plan, an analysis population, or a multiplicity strategy. The planning illustration assumptions are not a substitute for those details. The primary completion timeframe is described as through study completion, an average of 4 months, but the duration and completeness of outcome and safety follow-up are not further specified. These gaps constrain evaluation of precision, bias, and reproducibility.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Resurgence as Choice: Basic and Clinical Studies — NCT03423940.
• Registry record last updated 2024-09-03; status last verified 2024-04.
• Results first posted 2024-09-03.
• 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

NCT03423940, titled Resurgence as Choice: Basic and Clinical Studies, is listed as a completed randomized, single-masked, parallel study sponsored by Rutgers, The State University of New Jersey. The registry lists enrollment as 17 and identifies functional communication training as the behavioral intervention. Eligibility concerned children with destructive behavior maintained by social consequences and specified intellectual, adaptive-behavior, medication, and educational criteria.

The primary outcome counts participants whose highest destructive-behavior response rate in resurgence-test conditions matched the quantitative model's prediction. The registry reports counts of 1 among 3 for Evaluation of Treatment Dosage and 5 among 9 for Evaluation of Size of Decrease in Alternative Reinforcement. It lists one secondary outcome concerning the lowest-response-rate prediction and no analyses.

The registry records zero serious adverse events, zero other adverse events, and zero deaths in both listed safety groups. Statistical comparisons, uncertainty estimates, participant flow, missing-data methods, and detailed analysis plans are not reported in the registry record. The outcome counts are therefore descriptive rather than inferential.