Men, Mood, and Attention Study: Examination of Alcohol, State Anger, and Emotion Regulation Sexual Aggression: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Men, Mood, and Attention Study: Examination of Alcohol, State Anger, and Emotion Regulation Sexual Aggression (NCT04192448), a randomized na study. The primary outcome is Sexual Aggression Intentions.

2. Clinical Question

The study was designed to examine whether alcohol intoxication, state anger, and emotion regulation relate to men’s intentions in a sexual aggression scenario. The registered design randomized participants across beverage condition and emotion induction, creating a factorial structure in which the planned contrasts could potentially examine each factor and their combination. However, the posted primary outcome data do not fill every arm, and the registry record does not give a complete, interpretable set of factorial effect estimates.

The registry describes a hypothetical scenario and self-reported intentions, rather than observed behavior. The question is therefore about responses to an experimental analog under specified conditions. It is not a direct estimate of the frequency of real-world sexual aggression, nor does the record establish how intentions translate into behavior. The small enrolled population and the unoccupied Sober x Anger arm sharply limit what can be learned about the intended population-level contrasts.

3. Trial Design

The registry lists a randomized, single-masked, parallel design. The brief summary describes a 2x2 arrangement: beverage condition (alcohol or no alcohol) crossed with mood induction (anger or control). The four named combinations were AlcoholxAnger, AlcoholxControl, SoberxAnger, and SoberxControl. The alcohol dose was intended to result in a BAC of .08%; the sober condition involved no alcohol and was described as BAC .00%. The control mood condition used a neutral mood induction.

The record lists Alcohol Administration and Mood Induction as interventions. Participants were intended to complete a sexual aggression analog on the descending limb of alcohol intoxication, reporting how likely they would be to engage in various acts in a hypothetical situation. The primary outcome is described as the mean score of 22 items. The timeframe is 20 minutes. Physical Aggression Intentions and Psychological Aggression Intentions are listed as secondary outcomes.

Enrollment was 3, and the study status is TERMINATED. The primary outcome description states that the analyzed population comprised the three enrolled participants; it also states that no participants were analyzed in SoberxAnger because none were randomized to that arm. The posted arm data show one participant in AlcoholxAnger, one in AlcoholxControl, none in SoberxAnger, and one in SoberxControl. The record lists a start date of 2020-09-01 and a primary completion date of 2021-08-31. The reason for termination and whether the study ended as planned are not reported in the registry record.

Randomization can support causal comparisons when allocation, implementation, and analysis are adequate. In this case, the very limited enrollment and empty arm mean the posted results cannot provide a stable or complete assessment of the planned factorial design. The registry record does not describe allocation concealment, the masking procedures in detail, or the intended statistical model for the factorial effects.

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 enrollment, participants completed baseline measures. Participants were block randomized to condition arm based upon their past sexual assault perpetration status, such that an equal number of participants who had previously perpetrated sexual assault were assigned to each condition arm. No participants were excluded prior to group assignment.

Overall Study

Analysis population for the primary outcome, as defined in the registry: The population analyzed is comprised of the three participants who were enrolled into the study. No participants were analyzed in the Sober x Anger arm because no participants were randomly assigned to that arm.

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.

Population: The baseline population met all the inclusion criteria as outlined in the initial proposal, thus were cisgender men who engaged in moderate to heavy drinking and sexual risk-taking. The number enrolled is substantively different than proposed due to administrative barriers to the study and COVID-related barriers that impacted recruitment (closure of alcohol-serving establishments, community concern regarding participation in in-person studies during the pandemic).

6. Primary Endpoint

The registry lists the primary outcome as Sexual Aggression Intentions, measured over 20 minutes.

The primary endpoint is titled Sexual Aggression Intentions and is measured as a score on a scale. The registry describes it as the mean score of 22 items assessing behavioral intentions, with responses ranging from 1 = Not at Likely to 7 = Very Likely. The items concern whether a participant would engage in sexually aggressive acts against a hypothetical female. The outcome timeframe is 20 minutes.

This kind of endpoint is used to operationalize a response to a standardized scenario. It allows researchers to quantify participants’ stated intentions in relation to the assigned beverage and mood conditions. In general, a multi-item score can summarize responses across related questions, though its meaning depends on how the items are scored and whether the measure performs reliably. Reliability, validation details, and the scoring rules beyond the description of a mean score are not reported in the registry record.

The endpoint measures intentions, not actual conduct. A response to a hypothetical scenario cannot by itself establish what a participant would do in a real-world situation. Nor does the score represent the occurrence of sexual aggression. The registry record describes the intended construct and the scale anchors, but it does not report individual participant values, a measure of variability, or uncertainty estimates for each arm’s mean. The posted arm values therefore need to be read as sparse descriptive summaries, not as precise estimates of the population’s intentions.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does randomization contribute?

Randomization assigns participants to study conditions by chance, with the aim of making comparison groups similar on average with respect to factors that could affect the outcome. In a factorial design, randomized factors can be studied separately and in combination, provided the design is implemented and analyzed as planned and there are adequate observations in each combination. Randomization does not guarantee balance in a very small study, and it cannot create information for an arm with no participants.

The registry describes randomization across beverage and emotion-induction conditions, but the posted primary outcome data include no participant in SoberxAnger. As a result, the data do not populate all combinations of the intended design. The registry record does not report details about the random sequence, allocation concealment, or the planned handling of an empty cell. Without those details, the strength and intended scope of the randomized comparisons cannot be fully assessed.

What does a group mean describe?

A mean is an average of the observed scores in a group. It is a concise summary of the center of those scores, but it does not show their spread or the range of individual responses. In this study, each nonempty primary outcome arm has one participant, and no standard deviation or other within-arm spread is posted. Consequently, the arm means do not reveal how varied responses might be among participants in the intended population. With one observation in an arm, the value is the score for that participant rather than a stable estimate of a group average.

Means also need a clearly specified comparison to answer a research question. A difference between two arm means has an interpretation only when the contrast is defined and the measured outcome and analysis population are clear. The registry’s two posted analyses both list the same three groups, but the record does not clearly map each result to a named outcome or explain the precise contrast. That makes it unsafe to interpret either reported difference as a particular alcohol, anger, or combined effect.

How should the t-test results be read?

A t-test evaluates a specified comparison of means under assumptions about the data and the sampling process. A one-sided test evaluates evidence in a prespecified direction, rather than treating effects in either direction symmetrically. The registry lists the method as “t-test, 1 sided” for both reported analyses. The directional hypothesis, the outcome attached to each analysis, and the exact contrast are not reported in a way that permits a full independent interpretation.

The listed p-values are .28 and .39. A p-value measures how unusual data at least as incompatible with a specified null hypothesis would be under that null and the test assumptions. It is not the probability that the null hypothesis is true, and it does not measure the size or importance of an effect. Interpretation also depends on the test’s direction, the comparison definition, and the quality of the data. The registry does not provide enough detail to reconstruct those elements for the two analyses.

What information does a confidence interval add?

A confidence interval expresses uncertainty around an estimate under a statistical procedure and its assumptions. A narrow interval indicates greater precision than a wide interval, all else being equal. The reported 95% intervals are -29.82 to 26.19 and -33.01 to 35.01. Their broad ranges, together with the very small number of participants, show that the estimated differences are highly uncertain. The record does not report arm-level variability, and the reported analyses are not clearly linked to the primary outcome means, so the intervals cannot repair the lack of a clear comparison.

Confidence intervals should not be read as ranges of individual participant scores, nor as a guarantee about the true effect. Their interpretation depends on the analysis model and sampling assumptions. A confidence interval can be wide because data are limited or variable; in this record, the enrolled population is exceptionally small, while the relevant variability information is not posted. The intervals therefore do not support a precise conclusion about the experimental conditions.

Why does over-dispersion matter in count outcomes?

General statistical explanation: for count outcomes, a simple model may assume that the variability follows a particular relationship to the mean. Over-dispersion occurs when the observed variability is greater than that assumption allows. If ignored, standard errors can be too small and confidence intervals too narrow, making results appear more precise than the data warrant. Analysts may use methods that account for extra variability when the outcome and study design justify them.

The primary endpoint in this record is a mean score on a scale, not a count endpoint. The registry also lists physical and psychological aggression intentions as secondary outcomes, but it does not describe their specific measurement distributions or the statistical methods used for them. No claim about over-dispersion or a count model can be made from the registry record. General discussion of count-data methods should not be mistaken for a description of the analysis that was actually performed.

What can sample-size planning illustrations establish?

General statistical explanation: sample-size planning asks how many observations might be needed to detect a specified effect under stated assumptions about significance, allocation, and the statistical model. Such calculations are prospective illustrations. They are not observed results and cannot compensate for a study that enrolled very few participants. A planning calculation based on an observed effect is especially dependent on the unstable effect estimate used as its starting point.

The registry description of planning assumptions specifies a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and effects defined as fractions of an observed effect. These are described as planning illustrations, not as the trial’s own design. The record does not report a sample-size justification for the trial or an analysis plan connecting those assumptions to the randomized factorial design. A planning value should therefore not be described as observed power or as evidence that the study was adequately sized.

What can be concluded from these methods?

The methods and results support descriptive statements about the three posted arm values and the two analyses as recorded. They do not support a robust estimate of the planned factorial effects. The record does not explain the intended model, the exact analysis contrasts, or how the unoccupied arm was handled. The distinctions matter: a statistical test only answers the question encoded in its model and contrast. Without that information, reported numerical results should be described faithfully but not assigned a causal meaning that the record does not establish.

10. Analyses Reported by the Sponsor

The registry reports two sponsor analyses as one-sided t-tests. Both are labeled Mean Difference (Final Values) and list the three groups AlcoholxAnger, AlcoholxControl, and SoberxControl. The first reported value is 1.82, with a 95% confidence interval from -29.82 to 26.19 and p = .28. The second is -1.0, with a 95% confidence interval from -33.01 to 35.01 and p = .39.

The record does not clearly identify the outcome and exact contrast for each analysis. It also does not provide enough detail to connect those estimates to a particular primary endpoint comparison. The posted primary means are 2.18, 4.73, and 1.64 for the three nonempty arms, but the sponsor analyses cannot be interpreted as direct contrasts among those means without further specification.

No recalculated two-arm comparison is reported for the primary endpoint. The registry record indicates that values were missing or not estimable for a two-arm comparison for the count-other endpoint type, and the outcome is not reduced to one comparison. Because the discrepancy indicator is false, there is no recorded disagreement for which a recalculation would displace the sponsor analysis. The sponsor’s reported results remain the relevant inferential results in the record, while their unclear mapping and the small enrollment limit interpretation.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists three outcome measures in total, including Sexual Aggression Intentions as the primary outcome and Physical Aggression Intentions and Psychological Aggression Intentions as secondary outcomes. It also lists six analyses. Multiple outcomes and analyses create a multiplicity issue: the more questions tested, the greater the chance that some results may appear notable by chance if no adjustment or prespecified hierarchy is used. Adjustment methods and a prespecified testing hierarchy are not reported in the registry record. The relationship between the six analyses and the named outcomes is also not clearly described.

Missing-data procedures are not reported in the registry record. The primary outcome description states that three enrolled participants were analyzed, and that there were no participants in SoberxAnger because no one was randomized to that arm. This is an unoccupied randomized condition rather than an individual outcome value identified as missing. The record does not say whether any other outcome data were missing, how incomplete questionnaires would be handled, or whether any imputation method was planned or used.

The analysis population is described for the primary outcome as the three participants enrolled. The registry does not report whether the analyses followed an intention-to-treat principle, whether participants were excluded from any analysis, or how adherence to assigned conditions was assessed. The record also does not report a protocol-based analysis population definition. These omissions make it difficult to determine whether the posted analyses address all randomized participants under a prespecified strategy or only a subset defined after enrollment.

The combination of multiple outcomes, six analyses, unclear outcome-to-analysis mapping, and missing statistical-plan details means that the reported p-values should be interpreted cautiously. The record does not establish whether any multiplicity adjustment was applied, and it does not provide enough information to reconstruct a comprehensive analysis plan. These limitations are separate from the sparse sample itself: even a clearly specified multiplicity strategy would not resolve the lack of observations in one planned arm or the uncertainty caused by the small enrolled population.

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 serious adverse event figures report zero serious adverse events in AlcoholxAnger, AlcoholxControl, and SoberxControl. Each of those groups lists one participant at risk. SoberxAnger lists zero participants at risk and zero serious adverse events. Deaths are listed as zero in every group. The record also lists zero other adverse events in the groups with participants at risk.

These entries describe what was recorded among the very small number of participants in the posted safety groups. A zero count means no event was reported in the listed group; it does not establish that an event could not occur or that a condition is free of risk. With only one participant at risk in each nonempty group, the figures provide little information about how often adverse events might occur in a larger population. The empty SoberxAnger group provides no participant experience from which to assess adverse events.

The registry record does not list serious adverse event terms, and the serious-event term list is empty. These safety results should be read as sparse event reporting, not as a comparative safety analysis or evidence that the conditions have equivalent safety profiles.

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

The registry record documents a randomized study intended to examine alcohol, anger induction, and men’s intentions in a hypothetical sexual aggression scenario. Its primary endpoint is a self-reported scale score. The posted means are observations from one participant in each of three arms, with no participant in SoberxAnger. The two reported statistical analyses have broad confidence intervals and are not clearly mapped to particular outcomes and contrasts.

These data do not establish a dependable effect of alcohol, anger, or their combination on the primary endpoint. The incomplete set of arm observations means the planned factorial structure cannot be fully assessed from the posted primary results. The very small number of participants also makes arm summaries and inferential estimates highly uncertain. The reported p-values do not demonstrate that there is no effect, just as the mean differences do not by themselves establish an effect.

The outcome is a stated intention in response to a hypothetical situation. It is not a measure of actual behavior, and the record does not report how well the scale predicts behavior outside the study setting. The results therefore should not be generalized to real-world conduct or to populations beyond the intended study population. The registry also does not give enough information about the analysis plan, missing-data procedures, or the reason for termination to resolve these uncertainties.

In general, a small randomized study can be useful for testing procedures or generating hypotheses, but the registry record does not state that this trial was designed as a feasibility study or report such objectives. The posted results should be understood narrowly as limited observations from the enrolled participants. They cannot answer the broader public-health question on their own, and they do not support a treatment recommendation or a conclusion about individual behavior.

14. Why This Trial Matters Statistically

This trial is statistically notable because its intended design and observed data are very different in scale. The registry describes a randomized 2x2 design crossing alcohol condition with anger or control induction, but the primary outcome population includes only three participants and one planned combination has no participant. That contrast between design complexity and sparse data is important when interpreting any calculated estimate: a factorial design requires enough observations in its component conditions to support the comparisons it is meant to make.

The record also illustrates why an outcome label, a group summary, and an inferential analysis need to be connected explicitly. Three arm means are posted, while two analyses report mean differences for the same three groups without a clear outcome or contrast mapping. Statistical results are meaningful only when readers can identify what was compared, which endpoint was analyzed, and which population contributed data. Without that information, even correctly quoted values can invite interpretations the registry does not support.

A further lesson concerns precision. Confidence intervals accompany the reported analyses, but their breadth and the limited enrollment show why a p-value alone cannot summarize the strength of evidence. An estimate may have a direction, yet remain too uncertain to distinguish among materially different possibilities. Planning illustrations based on observed effects are also not a substitute for a prespecified design justification or for adequate enrollment.

Finally, the record distinguishes intentions from behavior and reports no serious adverse events among the very small number at risk. Neither type of result should be overread. The trial’s statistical importance lies in making clear how design completeness, analysis transparency, and sample size shape what can be learned—not in providing a definitive answer about the study’s broader behavioral question.

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 most immediate limitation is the enrollment of 3 participants across a design intended to include four combinations of beverage and mood induction. The primary results include one participant each in AlcoholxAnger, AlcoholxControl, and SoberxControl, and none in SoberxAnger. As a result, the record has no observed primary outcome for one planned combination and very little information for the other arms. The posted means cannot describe likely variation among participants within any arm.

The registry reports that the study was TERMINATED but does not state why. It does not report a planned recruitment target or the rationale for the sample size, so it is not possible to judge how far actual enrollment fell short of the intended target. A reason for termination, recruitment history, and any changes to the study plan are not reported in the registry record.

The two posted analyses are difficult to interpret because the record does not clearly connect each mean difference to an outcome and exact comparison. The registry does not report the full statistical plan, the directional hypotheses for the one-sided tests, adjustment for multiple analyses, or procedures for missing outcome data. It also does not provide arm-level variability for the primary score. Those omissions constrain independent evaluation of the estimates in addition to the limited enrollment.

The primary endpoint is a self-reported intention score based on a hypothetical scenario. The registry record does not report psychometric validation details, reliability estimates, or evidence connecting the score to behavior outside the study. The description of the scale anchors and item count alone cannot establish those properties. Finally, the eligibility criteria target a specific population, and the results cannot be assumed to apply to people outside that intended group. These limitations are specific to the design, enrollment, outcome, and reporting in this record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Men, Mood, and Attention Study: Examination of Alcohol, State Anger, and Emotion Regulation Sexual Aggression — NCT04192448.
• Registry record last updated 2024-05-29; status last verified 2024-05.
• Results first posted 2024-05-29.
• 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

The registry record describes a randomized, single-masked, parallel study of alcohol administration and mood induction. Its planned 2x2 design crossed alcohol versus no alcohol with anger versus control induction. The intended population was non-monogamous men who identify as male, whose biological sex is male, who have sex with women, and who are ages 21-35. The primary endpoint was a 22-item score of sexual aggression intentions in a hypothetical scenario.

The study enrolled 3 participants and is listed as TERMINATED. Primary outcome data are posted for AlcoholxAnger, AlcoholxControl, and SoberxControl, with one participant in each; SoberxAnger had no participant. The corresponding means are 2.18, 4.73, and 1.64. The registry reports two one-sided t-test analyses, but does not clearly map them to outcomes and contrasts. Their estimates are 1.82 and -1.0, with broad 95% confidence intervals and p-values of .28 and .39.

The record does not support a precise or complete assessment of the planned factorial question. The small enrollment, unoccupied arm, unclear analysis mapping, and limited statistical-plan details require cautious interpretation.