CBT Enhanced With Social Cognitive Training: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for CBT Enhanced With Social Cognitive Training (NCT03954392), a randomized na study. The primary outcome is Faux Pas Recognition Test.

2. Clinical Question

3. Trial Design

The registry describes a randomized, single-masked, parallel-group behavioral trial of CBT with social cognitive training versus CBT alone. The listed condition is depression, and the record is marked completed. The sponsor is Vanderbilt University. The recorded study start date is 2019-02-11, and the primary completion date is 2021-08-31. The phase is listed as NA.

The posted primary outcome population is described as 45 youth, ages 12 through 17. Eligibility is described as meeting diagnostic criteria for a major depressive episode or persistent depressive disorder, or having a score of 35 or greater on the Children's Depression Rating Scale. The primary outcome analysis lists 23 participants in the CBT+SCT arm and 22 in the CBT-only arm. These are the arm-specific counts reported with the outcome.

The brief description discusses both an R61 and an R33 study, with participant counts of N=42 and N=82, respectively, and describes different outcome aims for those stages. The outcome results posted in this record are for the Faux Pas Recognition Test at week 12, while the brief description says that the R61 target was social cognitive skills after treatment and the R33 targets included both social cognition and depressive symptoms after treatment and at a 4-month follow-up. How the reported outcome population maps to those described stages is not reported in the registry record. A participant flow, treatment attendance, and details of the masking procedure are also 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: Participants were randomized to condition shortly after eligibility was confirmed.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Participants were 45 youth (64% female), ages 12 through 17 years old (Mean = 14.84, S.D. = 1.54). Youth either met diagnostic criteria for a major depressive episode or persistent depressive disorder or a score of 35 or greater on the Children's Depression Rating Scale.

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 Faux Pas Recognition Test, measured over week 12.

The primary endpoint is the Faux Pas Recognition Test score at week 12. The task includes 10 Faux Pas Stories and 10 Control Stories without a faux pas. It awards a point for each correctly perceived faux pas, with the description giving a lower score of 0 and a higher score of 20. The test is intended to assess the ability to identify and explain situations in which a character unintentionally insults another person because of missing information about that person.

The registry states that higher scores indicate higher theory of mind. That direction makes the endpoint suitable for assessing the social cognitive skill targeted by the training: recognizing another person's perspective and interpreting social situations. The measure therefore focuses on a proposed treatment target, rather than directly measuring depressive symptoms. The Children's Depression Rating Scale - Revised is listed separately as a secondary outcome.

The posted group means are 0.901 for CBT+SCT and 0.854 for CBT-only, while the endpoint description gives a score range from 0 to 20. The registry record does not explain how the reported group values relate to that described score range. This limits interpretation of the values as raw test points, even though their between-arm comparison is reported.

7. Results Posted in the Registry

• CBT-only: mean 0.85 versus 0.90 on control; difference -0.05 (95% CI -0.09 to -0.01), standardized effect d = -0.72; z-test p = 0.016.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Children Depression Rating Scale - Revised

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • CBT-only: mean 34.59 versus 23.74 on control; difference 10.85 (95% CI 5.87 to 15.83), standardized effect d = 1.29; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for CBT-only than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The record posts one secondary endpoint: the Children Depression Rating Scale - Revised at week 12. Its recalculated group summaries show a separation between arms, and the sponsor reports an ANCOVA result with p < 0.05. The registry's brief description identifies depressive symptoms as an R33 target, but the record does not establish that this posted week-12 result represents the R33 follow-up objective.

The primary endpoint concerns social cognitive skill, while the secondary measure concerns depression ratings. A pattern of between-arm differences across these different domains may be relevant to the study's rationale, which proposes enhancing CBT through social cognitive training. However, one posted secondary endpoint is not enough to establish a broad or consistent pattern across depression outcomes. The scale direction, clinically meaningful threshold, adjusted effect, and confidence interval from the sponsor's analysis are not reported in the registry record.

Secondary outcomes generally carry less weight than a prespecified primary outcome because they may be more exploratory and because testing multiple outcomes can increase the chance of a chance finding. The record does not report whether multiplicity was addressed, whether the secondary endpoint was prespecified in the analysis plan, or how the sponsor interpreted it. Its result should therefore be considered supportive or exploratory rather than treated as independent confirmation of the primary outcome.

The brief summary mentions a 4-month follow-up for the R33 study, but the posted secondary result is at week 12. Follow-up results for the secondary outcome are not reported in the registry record. The posted endpoint should not be taken as evidence about durability beyond its recorded time point.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Children Depression Rating Scale - Revised

9. Statistical Interpretation

10. Statistical Methods Explained

The trial is recorded as randomized and parallel-group. Random allocation is intended to make treatment groups comparable, on average, with respect to measured and unmeasured characteristics. In any realized trial, however, the groups may still differ by chance. The record does not describe the randomization sequence, allocation concealment, or whether the analysis followed the assigned groups regardless of treatment received.

What does the reported ANCOVA mean?

ANCOVA, or analysis of covariance, compares group outcomes while adjusting for one or more covariates. In general, adjustment can account for specified baseline or design variables and can make a comparison more precise when those variables are informative. The registry record identifies ANCOVA as the sponsor's method for both the primary and secondary outcomes. It does not identify the covariates, the model specification, the handling of assumptions, or the adjusted effect estimate. The reported p-values alone cannot fill in those missing details.

The posted numerical comparisons are recalculated from the arm-level means, standard deviations, and group counts. These are unadjusted summary-statistic comparisons. For the primary outcome, the difference is expressed as CBT-only minus CBT+SCT, so a negative estimate corresponds to a lower reported mean in CBT-only. The secondary comparison uses the same arm order and gives a positive estimate because the posted CBT-only mean is higher. A subtraction's sign depends on the order of the groups; it does not, by itself, indicate clinical benefit or harm.

What is the role of a p-value?

A p-value describes how unusual data at least as incompatible with a specified null hypothesis would be under the assumptions of the analysis. It does not give the probability that the null hypothesis is true, and it does not measure the size or practical importance of a difference. The recalculated p-values accompany unadjusted comparisons. The sponsor's results report only that the ANCOVA p-values are below 0.05. The registry record does not report the exact sponsor p-values, adjusted effects, or the hypotheses and testing sequence used.

How should standardized differences be read?

The primary comparison includes a pooled standard deviation of 0.065 and d = -0.722. A standardized mean difference expresses a difference relative to within-group variability, making the estimate independent of the scale's measurement units. The sign follows the stated subtraction order. Standardization does not resolve the unexplained relationship between the primary outcome's described 0-to-20 score range and the posted means. It also does not establish a meaningful threshold for individual change, which is not reported in the registry record.

The secondary comparison includes a pooled standard deviation of 8.418 and d = 1.289. The registry lists the Children Depression Rating Scale - Revised as a continuous measure, but does not state its direction or provide a threshold for a meaningful change. Thus, the standardized estimate describes separation of the posted group means relative to the observed variability, but it does not provide a clinical interpretation on its own.

Why does over-dispersion matter in general?

General statistical explanation: for count outcomes, over-dispersion means that variation is greater than a chosen count model expects. If a model ignores that extra variability, standard errors can be too small and confidence intervals too narrow. This trial's posted primary and secondary outcomes are continuous measures reported as means and standard deviations, not event counts. Over-dispersion is therefore not the central issue for these comparisons. The registry record does not give enough model detail to assess the ANCOVA assumptions.

The arm-specific standard deviations describe variation among observed scores within each arm. The standard errors listed for the recalculated comparisons describe uncertainty in the estimated means under the calculation's assumptions; they are not measures of score variability among individuals. These distinctions matter because a group mean can be estimated with some precision even when individual scores vary substantially.

Finally, the reported analysis details are incomplete. The record does not report how missing outcomes were handled, whether any transformations were used, which analysis population was used, or whether the sponsor's analysis adjusted for baseline scores. Those omissions prevent a full audit of the model and make the unadjusted recalculation a limited supplement rather than a replacement for the sponsor's analysis.

11. Confidence Intervals and Precision

The recalculated primary mean difference is -0.047, with a confidence interval from -0.085 to -0.009. The interval describes uncertainty around the unadjusted estimated contrast under the calculation's assumptions. Its endpoints are both negative, so the recalculated interval does not include zero. This is consistent with the recalculated p-value of 0.016, but neither result supplies the sponsor's adjusted estimate.

A confidence interval is a range of values compatible with the observed data and a statistical model, at the confidence level used to construct it. In general, a narrower interval indicates greater precision about the estimated group contrast, while a wider interval indicates more uncertainty. The confidence level for the posted recalculated intervals is not reported in the registry record. The interval is not a range containing most individual participants' scores, and it does not describe the likelihood that a particular treatment effect is true.

The primary outcome's adjusted ANCOVA confidence interval is not reported in the registry record. That absence matters because the adjusted estimate is the sponsor's stated analysis and takes precedence over the recalculation when the record flags a discrepancy. The posted p-value threshold does not show the adjusted effect's size or how precisely it was estimated.

For the secondary outcome, the recalculated interval runs from 5.866 to 15.83. It also excludes zero, but the registry record does not report the scale direction, so interval direction cannot be translated confidently into a symptom interpretation. Neither interval addresses uncertainty about generalizing beyond the enrolled youth, missing outcomes, or the analysis choices that are not documented.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm, not total events. They are illustrations based on the observed effect, not the trial's original design assumptions. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; each planning scenario uses the stated fraction of the observed effect. The original trial's sample-size assumptions and justification are not reported in the registry record.

For planning at 100% of the observed primary effect, the figures are 31 participants per arm for 80% power and 41 participants per arm for 90% power. Planning for 75% of the observed effect gives 54 participants per arm for 80% power and 72 participants per arm for 90% power. Planning for 50% of the observed effect gives 121 participants per arm for 80% power and 162 participants per arm for 90% power.

In general, power is the probability that a planned study will detect an effect of a specified size under the assumptions used. The 80% and 90% columns therefore describe different target probabilities for detecting the stated planning effect, not certainty that a future study will find a statistically significant result. Smaller effects need more participants to detect with the same target power, which is why the participant requirements rise as the assumed effect fraction decreases. These figures are planning illustrations, not observed power.

The calculations use an effect estimated from the posted trial results, so the planning figures inherit uncertainty in that estimate. They should not be treated as a validated design recommendation or as a reconstruction of the trial's own design. The registry record does not report the rationale behind its recorded enrollment or whether the trial used a different planning method.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -0.047; pooled standard deviation σ = 0.065.
2. n = 2 × (1.96 + 0.842)² × 0.065² / -0.047² = 30.2, rounded up to 31 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -0.047                      # observed difference in means
sigma <- 0.0651372433030158                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(3954392)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 31 (80% power), 41 (90% power)
 75% of observed effect: n per arm = 54 (80% power), 72 (90% power)
 50% of observed effect: n per arm = 121 (80% power), 162 (90% power)
power.t.test: 31.1 
Simulated power at n = 31 per arm: 0.797 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The registry reports ANCOVA as the sponsor's analysis for the primary and secondary outcomes. For the primary outcome, the sponsor's reported result is p < 0.05; an adjusted mean difference and confidence interval are not reported. The secondary outcome is likewise listed as an ANCOVA with p < 0.05, without an adjusted estimate or interval.

The recalculated comparisons use only the posted arm-level summary statistics and are unadjusted. They therefore answer a more limited question than an ANCOVA that adjusts for covariates. The record flags a discrepancy, so the sponsor's own analysis takes precedence over the recalculation. This means that the recalculated estimates and p-values should be treated as supplementary descriptions of the posted summaries, not as substitutes for the sponsor's analysis.

The registry record does not identify the covariates in the ANCOVA, give enough detail to reproduce the adjusted analysis, or explain the flagged discrepancy. As a result, readers cannot determine from the record alone whether differences arise from adjustment, analysis populations, missing data, or another analytic choice. The sponsor's p-value threshold indicates a reported result under its analysis, but without an adjusted effect estimate its magnitude and precision remain unclear.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when a study evaluates multiple outcomes, time points, contrasts, or analyses. With repeated statistical tests, the chance of at least one apparently notable result can increase unless the analysis plan accounts for the testing structure. This is a general statistical concern, not evidence that any particular result in this trial is spurious.

The registry lists a primary endpoint and one posted secondary endpoint, and it describes different aims for R61 and R33, including follow-up assessment in the brief summary. It does not report a multiplicity adjustment, a testing hierarchy, or which outcomes were included in any prespecified family of tests. It is therefore not possible to determine whether the reported p-values were adjusted for multiple testing or how the sponsor intended the primary and secondary findings to be interpreted together.

Missing-data handling is also not reported in the registry record. The posted outcome summaries give arm-specific counts, but the record does not explain whether those counts include everyone randomized, whether any participants lacked outcome data, or how missing values were handled in the ANCOVA. There is no report of imputation, complete-case analysis, or sensitivity analysis. Without that information, the potential influence of missing outcomes cannot be assessed.

The analysis population is not identified. The record does not say whether the sponsor analyzed participants according to randomized assignment, according to treatment received, or using another defined population. Nor does it report protocol deviations, adherence, or the handling of participants who did not complete treatment. These details can affect interpretation of randomized comparisons and reproducibility.

The record also omits the ANCOVA covariates, model diagnostics, and adjusted estimates. These gaps do not negate the posted results, but they limit independent evaluation. The sponsor's analysis is the relevant reported analysis, while the unadjusted recalculations are only a secondary way to describe the arm summaries.

16. 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 posted adverse-event figures list 0 serious adverse events in the CBT+SCT group, with 23 at risk, and 0 in the CBT-only group, with 22 at risk. The record also lists 0 other adverse events in each group, with the same respective numbers at risk, and 0 deaths in each group. No serious adverse-event terms are listed.

These figures describe the events recorded in the registry, not a guarantee that no event occurred outside the recorded reporting process. The registry record does not provide definitions, collection methods, observation duration, or details about whether all participants were assessed for adverse events. The counts should therefore be read as the posted safety record for these groups, not as evidence that the interventions are risk-free or as a comparison that can establish relative safety.

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

The primary result describes a week-12 difference in a social cognition measure between randomized groups receiving CBT with social cognitive training and CBT alone. The posted means and recalculated unadjusted contrast indicate that the CBT-only group had a lower mean on the Faux Pas Recognition Test. The test description says that higher scores indicate higher theory of mind. The sponsor reports an ANCOVA result with p < 0.05, but the adjusted contrast and its precision are not reported.

This result does not establish the size of an adjusted effect, the importance of the difference to youth or families, or whether the same pattern would be seen in other settings. The discrepancy flag means the sponsor's own analysis takes precedence over the unadjusted recalculation. The unclear relationship between the stated 0-to-20 test range and the reported means also limits direct interpretation of the primary scores in test-point units.

The record includes a secondary depression rating measure at week 12, but its scale direction is not stated. The numerical contrast therefore cannot be translated from this record alone into a clear statement about the direction of symptom change. The record's brief description discusses a 4-month follow-up for R33, but no follow-up results are posted with the outcomes.

Nor does the record establish which part of the intervention, if any, accounts for the observed group differences. It compares two treatment approaches as assigned, and does not provide enough information about treatment exposure, adherence, or component-specific effects to isolate mechanisms. The results are evidence about the posted outcomes and analyses in this trial; they are not medical advice, a prediction for an individual, or proof of benefit or lack of benefit.

18. Why This Trial Matters Statistically

This trial raises a statistical question about how to evaluate an intervention that adds a skills-focused component to an established behavioral treatment. Its primary measure targets social cognition, the domain the added training is intended to address, while a depression rating measure is also posted. Keeping those outcome domains distinct helps readers avoid treating a target-engagement measure as if it were itself a direct measure of depressive symptoms.

The record also illustrates why effect estimates, uncertainty, and analysis methods need to be considered together. The unadjusted recalculation gives a numerical contrast and confidence interval, while the sponsor reports ANCOVA significance without an adjusted effect estimate or interval. Because the record flags a discrepancy, the sponsor analysis takes precedence; at the same time, the missing adjusted estimate leaves the magnitude and precision of that analysis difficult to assess independently.

More broadly, transparent reporting of outcome scales, analysis populations, covariate choices, missing-data methods, and multiplicity plans is essential for clinical interpretation and reproducibility. In this record, the primary scale description and posted means do not readily align, and several analysis details are absent. These issues show why a reported p-value alone is not a complete account of a clinical trial result.

19. 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 endpoint description gives a score range from 0 to 20, but the posted means are 0.901 and 0.854. The registry record does not explain whether the posted values are transformed, normalized, or otherwise expressed on a different scale. This makes it difficult to interpret the primary contrast in the test's described units.

The sponsor's ANCOVA estimates, confidence intervals, covariates, and model details are not reported. Since the record flags a discrepancy, the sponsor's analysis takes precedence over the unadjusted recalculation, but the information needed to independently assess that analysis is missing. The record also does not report the analysis population, missing-data approach, treatment adherence, or participant flow.

The brief description refers to R61 and R33 cohorts and a 4-month follow-up, while the posted outcome results are reported at week 12 for a population of 45 youth. How the posted results relate to the described study stages is not reported in the registry record. Follow-up outcome results are not included in the posted data analyzed.

Only one secondary endpoint is posted, and its direction is not specified. The record does not state whether multiplicity was addressed. The study is recorded as single-masked, but does not say who was masked. Finally, the results pertain to youth meeting the listed age and depression eligibility description; the record does not establish how well they generalize to other populations or settings.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: CBT Enhanced With Social Cognitive Training — NCT03954392.
• Registry record last updated 2023-03-28; status last verified 2023-02.
• Results first posted 2023-03-28.
• 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.

23. Record Summary

NCT03954392 is a completed, randomized, single-masked, parallel-group behavioral study sponsored by Vanderbilt University. It compares CBT with social cognitive training against CBT alone in youth with depression. The primary endpoint is Faux Pas Recognition Test score at week 12; the posted secondary endpoint is the Children Depression Rating Scale - Revised at week 12.

The sponsor reports ANCOVA results with p < 0.05 for both outcomes, without adjusted effect estimates or confidence intervals. The registry flags a discrepancy, so the sponsor's analysis takes precedence over the unadjusted recalculations. The record does not report key details including the analysis population, missing-data handling, multiplicity procedures, or the relationship between the primary scale description and its posted means.