Exposure to Gun Violence in Video Games Increases Interest in Real Guns: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Exposure to Gun Violence in Video Games Increases Interest in Real Guns (NCT03259139), a randomized na study. The primary outcome is Time Spent Holding Gun.

2. Clinical Question

The trial’s title states a directional claim about exposure to gun violence in video games and interest in real guns. The posted primary outcome offers one way to examine that topic: time spent holding a real firearm during a 20 minute play session. The group summaries describe observed mean times, and the comparisons quantify differences relative to the control player group. They do not by themselves establish the result of every question named in the trial objective.

3. Trial Design

The registry record describes a randomized clinical trial with a factorial model and single masking. It lists the phase as not applicable and the status as completed. Enrollment was 250. The listed start date is 2017-07-01, and the primary completion date is 2018-08-24. The sponsor is Ohio State University.

The behavioral intervention was video-game content. The three conditions were a violent game that included guns, a violent game without guns, and a game without violence or weapons. Posted outcome results further distinguish participants as players or watchers in each condition. The primary outcome description says players and observers were collapsed into one category for analysis, although the registry’s arm summaries and pairwise comparisons report player and watcher groups separately. The record does not explain how this difference in presentation relates to any broader model.

The primary outcome was continuous: mean time spent holding a real firearm, measured in seconds during the 20 minute play session. The listed spread is a standard error. The registry record lists three outcome measures, but the other outcome results are not reported in the registry record. It lists no analysis entries. Therefore, the posted group summaries and comparisons do not supply a full account of the trial’s planned factorial analysis or of analyses for other outcomes.

Details about the randomization procedure, the identity of the masked party, allocation concealment, eligibility criteria, and the trial’s original statistical design assumptions are not reported in the registry record. The label single masking does not identify who was masked. The registry record also does not report the protocol details needed to determine how the trial intended to account for the player and watcher roles.

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 both randomly assigned to condition (i.e. non-violent, sword, gun conditions) and to be either a player or watcher.

Overall Study

Analysis population for the primary outcome, as defined in the registry: NOTE: players and observers were collapsed into a single category for analysis

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: Children between the ages of 8 - 12, tested in pairs. Several children were excluded from the analysis, either for not meeting study criteria (n=4), for being identified as outliers (n=4), or for not finding the handguns (n=22). Figures here are after exclusions.

Only the first 6 of 7 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Time Spent Holding Gun, measured over 20 minute play session.

The primary endpoint measures how long a participant held a real firearm during the play session, expressed as a mean number of seconds. Its timeframe is a 20 minute play session. The registry describes the measure as the time spent holding the firearm during that session.

Time is a continuous outcome, so it can capture variation in duration rather than reducing the observation to a simple yes-or-no indicator of whether the firearm was held. Comparing group means provides a compact description of typical holding time in each reported group. The standard error shown with each mean describes the precision of that estimated mean, rather than the spread of individual participants’ times.

This endpoint is relevant to the stated research objective because the trial asks whether video-game exposure is associated with children’s interest in real guns, and firearm-handling duration is one behavioral measure related to that question. It is not a direct measure of all aspects of interest, nor does it answer the other objectives about trigger pulling or pointing the firearm. Results for those other outcomes are not reported in the registry record. The record also does not report further measurement details, such as how the duration was recorded or how observations were handled if a participant did not hold the firearm.

7. Results Posted in the Registry

• Control: No Violence (Watcher): mean 21.14 versus 29.57 on control; difference -8.43 (95% CI -45.24 to 28.38), standardized effect d = -0.11; z-test p = 0.654.
• Experimental: Violence Without Guns (Player): mean 82.86 versus 29.57 on control; difference 53.29 (95% CI -5.46 to 112.04), standardized effect d = 0.41; z-test p = 0.075.
• Experimental: Violence Without Guns (Watcher): mean 48.62 versus 29.57 on control; difference 19.05 (95% CI -37.99 to 76.09), standardized effect d = 0.15; z-test p = 0.513.
• Experimental: Violence With Guns (Player): mean 61.52 versus 29.57 on control; difference 31.95 (95% CI -13.98 to 77.88), standardized effect d = 0.32; z-test p = 0.173.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary measure is a continuous outcome summarized by a mean and standard error. A mean is the arithmetic average of recorded times within a group. A standard error describes the uncertainty in the estimated group mean; it is not the same as the standard deviation, which describes variation among individual observations. The registry record reports the standard errors for each group but does not report individual-level data or standard deviations for the groups.

What does a mean difference represent?

Each posted comparison expresses the difference between a reported group mean and the no-violence player group mean. A positive estimate means the comparison group’s reported mean was higher than the reference mean; a negative estimate means it was lower. For example, the estimate for the violent-without-guns player group is positive, while the estimate for the no-violence watcher group is negative. Those directions describe the point estimates, not certainty that the underlying group means differ. Uncertainty is conveyed by the confidence intervals and p-values.

The comparisons also list a pooled standard deviation and a standardized effect estimate, d. A pooled standard deviation combines information about variability across the two groups being compared, under the calculation’s assumptions. A standardized effect expresses a mean difference in standard-deviation units, which can help describe magnitude on a scale that is less tied to seconds. It does not replace the original-unit result: seconds remain the direct interpretation of the primary outcome. The registry record does not provide an analysis specification explaining the role of these standardized values in the trial’s own analysis.

Why does uncertainty matter?

A confidence interval gives a range of values compatible with the observed data under the statistical method used. For a mean difference, an interval that includes zero includes the possibility of no difference. It may also include values on both sides of zero, indicating that the data do not pin down the direction of the difference. The posted intervals are broad, so the point estimates should not be read as exact effects. A p-value measures how unusual data at least as discrepant as those observed would be under a specified null hypothesis and assumptions; it is not the probability that the null hypothesis is true, and it does not measure the size or practical importance of an effect.

The posted p-values are explicitly unadjusted. When multiple comparisons are considered, the chance of at least one small p-value can increase if each is interpreted separately without accounting for the number of tests. Whether and how to adjust depends on the analysis plan and the inferential question. The registry record does not report a multiplicity adjustment or an overall factorial test. The listed comparisons therefore should not be treated as though they constitute a documented, multiplicity-controlled analysis.

What does the factorial label tell us?

In general, a factorial design studies more than one factor or dimension of assignment, and can be used to examine main effects and interactions. An interaction asks whether the effect associated with one factor depends on the level of another factor. The record labels this trial factorial and separately describes game conditions and player or watcher groups. However, it does not report the exact factor structure, the intended model terms, or estimates for main effects or interactions. The separate comparisons against the control player group do not supply those missing model results.

The primary outcome description says players and observers were collapsed into a single category for analysis, but posted summaries and comparisons still separate players and watchers. This creates an important interpretive distinction: group-specific summaries are visible, but the registry record does not explain how they correspond to the stated collapsed analysis population. It is not possible from the record to reconstruct the model or determine whether the reported comparisons are the complete analysis of the primary endpoint.

Why can over-dispersion matter for a duration measure?

As a general statistical point, duration data can vary substantially among individuals. If a small number of observations are much longer than most others, the distribution may be skewed, and a mean-based method can be sensitive to those observations. Over-dispersion refers broadly to variation greater than a model expects; if present, it can make standard errors too small unless the analysis accounts for the extra variation. The registry record does not report a distributional assessment or a model used to address skewness or over-dispersion. This is a general consideration for duration outcomes, not a finding about this trial’s data.

The record’s planning calculations use the observed effect as a basis for illustrations, with stated assumptions described alongside the sample-size material. Those calculations are not the trial’s original design rationale and do not establish the study’s achieved power. The trial’s own sample-size justification, original design assumptions, interim analyses, and detailed statistical plan are not reported in the registry record.

10. Confidence Intervals and Precision

The confidence intervals communicate substantial uncertainty around the posted mean differences. For the no-violence watcher comparison, the interval extends from -45.24 to 28.38 seconds. For the violent-without-guns player comparison it extends from -5.463 to 112.0 seconds, and for the watcher comparison from -37.99 to 76.09 seconds. In the violent-with-guns condition, the player interval ranges from -13.98 to 77.88 seconds and the watcher interval from -6.837 to 84.5 seconds.

Every posted interval includes zero. Each therefore remains compatible with no mean difference, while also allowing differences in either direction across much of its range. The positive point estimates in several comparisons should not be mistaken for precise estimates of a positive difference. The intervals also do not establish that the groups are equivalent: demonstrating equivalence would require a defined margin and an appropriate analysis, neither of which is reported in the registry record.

The uncertainty is consistent with the small group counts reported for these comparisons and the sizeable standard errors accompanying the means, but the record does not report a formal explanation of interval width. The standard errors describe uncertainty in individual group means; the confidence intervals describe uncertainty in the pairwise mean differences. Both are useful, and they answer related but distinct questions. The unadjusted p-values are also consistent with the intervals including zero under the corresponding comparison methods, but should not be read as a measure of precision by themselves.

The range of plausible differences matters for interpretation. In some comparisons, the interval permits values below zero as well as sizeable positive values. Thus, the posted results do not sharply determine the direction or magnitude of the group differences. The precision is limited for conclusions about this primary endpoint, and no precision estimate for the other stated behavioral outcomes is reported in the registry record.

11. 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 sample-size figures are planning illustrations based on the observed effect, not the trial’s original design assumptions. The registry record’s stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. For this continuous outcome, the listed figures are participants per arm, not a total across arms and not total events. The time-to-event exception mentioned in the planning notes does not apply to this endpoint.

The 80% power column gives the planned number of participants per arm for a calculation with an 80% chance of detecting the assumed effect under those assumptions, if that effect were the true one. The 90% power column gives the corresponding planning number for a 90% chance. These are conventional design targets, not guarantees that a study of that size will identify an effect. They also do not describe the probability that the observed result is true.

The rows for 100%, 75%, and 50% of the observed effect show how the illustration changes when the assumed effect is reduced. Smaller effects need more participants to detect under the same planning assumptions. For example, the violent-without-guns player comparison lists 92 participants per arm for 80% power at 100% of the observed effect, and 368 participants per arm at 50% of that observed effect. The values differ across comparisons because each illustration is based on its own observed effect.

These calculations are not observed power. They are prospective planning illustrations using effects estimated from the trial results, and the registry record does not report that these were the assumptions used when the trial was designed. The accompanying curve should be read as a visual representation of how the planning requirement varies with the assumed effect, rather than as evidence about the trial’s achieved power.

12. 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 = -8.430; pooled standard deviation σ = 78.566.
2. n = 2 × (1.96 + 0.842)² × 78.566² / -8.430² = 1,363.5, rounded up to 1,364 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 <- -8.43                      # observed difference in means
sigma <- 78.5664304903818                     # 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(3259139)
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 = 1,364 (80% power), 1,826 (90% power)
 75% of observed effect: n per arm = 2,425 (80% power), 3,246 (90% power)
 50% of observed effect: n per arm = 5,455 (80% power), 7,302 (90% power)
power.t.test: 1364.5 
Simulated power at n = 1,364 per arm: 0.804 (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.

13. Multiplicity, Missing Data and Analysis Populations

The record lists three outcome measures, but results for only the primary endpoint are described in the posted summaries. The outcomes involving trigger pulling and pointing the firearm at oneself or another person are mentioned in the objective, but their results are not reported in the registry record. The registry record lists no analysis entries. It does not provide an analysis plan showing which outcomes were designated primary or secondary for inferential testing beyond the posted primary outcome label.

Multiple comparisons can complicate interpretation when several groups or outcomes are tested. A multiplicity plan specifies how the analysis will account for the number of comparisons, or which comparisons are intended to support confirmatory conclusions. The posted p-values are unadjusted, and an adjustment strategy is not reported in the registry record. The five listed contrasts use the no-violence player group as their reference; no overall comparison across the three game conditions is reported. It is therefore important not to treat the individual p-values as though they were adjusted tests of a single prespecified family.

Missing-data methods are not reported in the registry record. The summaries include group counts, but the record does not explain whether any randomized participants lacked an endpoint measurement, how such observations were handled, or whether imputation or sensitivity analyses were used. The analysis population is also not fully described. In particular, the outcome description says players and observers were collapsed into one category for analysis, while the posted summaries and contrasts separate these roles. The record does not explain the relation between these descriptions or identify an intention-to-treat population.

The record does not report interim analyses, stopping rules, protocol deviations, or sensitivity analyses. These details are relevant to judging the full analysis process, but their absence from the registry record does not establish that such procedures did or did not occur. Without them, the posted statistics can be interpreted as reported comparisons, but not as a complete account of the prespecified statistical strategy.

14. 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 adverse-event summaries report zero serious adverse events and zero other adverse events in each listed player or watcher group, with zero deaths reported in each group. The at-risk counts are listed separately for each group: 38 for each violent-with-guns group, 37 for each violent-without-guns group, and 35 for each no-violence group. These figures describe what the registry record reports; they should not be added together to create a participant total.

A zero event count means that no event in that category was reported for the listed group in this registry record. It does not establish that the intervention is free of risk or that an event could not occur in another setting. The record does not provide additional detail on adverse-event definitions, collection methods, observation duration, or whether events were assessed as related to the intervention. The reported figures are therefore a limited safety summary and should be interpreted within those reporting constraints.

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

The posted results describe mean time spent holding a real firearm during a 20 minute play session, for groups assigned to different video-game conditions and identified as players or watchers. Several group means are higher than the no-violence player mean, and several posted mean-difference estimates are positive. However, the confidence intervals for all listed comparisons include zero and extend across a range of possible differences. The results therefore do not precisely determine the size or direction of the underlying differences.

The trial title makes a broad statement about gun violence in video games and interest in real guns, but the primary endpoint measures time holding a firearm in the trial setting. It is one behavioral measure, not a complete measure of interest, future behavior, or real-world consequences. The registry record’s objective also mentions trigger pulling and pointing behavior; results for these outcomes are not reported in the registry record. The posted endpoint cannot stand in for those unreported outcomes.

The comparisons are against one control group and are reported separately for players and watchers. The registry identifies a factorial model but does not report an overall factorial analysis or interaction estimates. As a result, the group-specific comparisons do not establish whether the effect of violence depends on gun content, whether player and watcher role modifies the response, or whether a single overall contrast would differ from the individual comparisons.

The findings should be understood as estimates with uncertainty, rather than as proof that a particular game condition increases or does not increase interest in guns. The registry record does not provide enough detail about the analysis population, missing data, multiplicity, or original statistical plan to resolve those methodological questions. These observations concern the trial’s measured behavior and do not demonstrate what would happen in other populations or circumstances.

16. Why This Trial Matters Statistically

This trial illustrates how a behavioral question can be translated into a measurable continuous endpoint. Time holding a firearm retains more information about duration than a binary outcome of whether it was held, but the mean can be difficult to interpret without uncertainty and information about the distribution. The registry reports standard errors and comparison intervals, which help distinguish the point estimates from the range of values compatible with the data.

It also shows why the description of a design and the reported analysis need to be read together. The record labels the study factorial and says players and observers were collapsed for the primary analysis, while the posted results separate player and watcher groups and compare them with a single control group. Without the model specification or an overall factorial result, those contrasts do not answer every question implied by the design label. Reporting the analysis population and model terms would help readers understand which estimands the results represent.

Multiplicity and precision are also central. The comparisons address several groups, and the reported p-values are unadjusted. Their confidence intervals are broad and include zero, so the estimates are compatible with a range of interpretations. This is a reminder that statistical conclusions depend not only on the observed means but also on uncertainty, the planned family of comparisons, and the model used.

Finally, the registry objective names several behaviors, but only the holding-time outcome is represented in the posted results. Clear reporting of all intended outcomes and analysis decisions is important for readers assessing whether the evidence speaks to the full research question. The trial is statistically informative as an example of how outcome definition, comparison structure, and reporting completeness shape what can be concluded from a randomized study.

17. 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 does not report the age range, eligibility criteria, recruitment setting, or participant characteristics for this outcome. These omissions limit assessment of how broadly the group results might apply. It also does not provide detail about the firearm-handling procedure, how holding time was measured, or the distribution of individual durations. For a time-based outcome, such information would help readers judge the influence of unusual values and the appropriateness of mean-based comparisons.

The outcome description says players and observers were collapsed into one category for analysis, but the posted summaries and comparisons distinguish those roles. The record does not reconcile these descriptions or explain the model behind the contrasts. Although the design is labeled factorial, no overall effects or interaction estimates are reported. This limits interpretation of how game violence, gun content, and player or watcher role relate to one another.

The posted comparisons have broad confidence intervals that include zero, and their p-values are unadjusted. The registry record does not report a multiplicity strategy, missing-data methods, analysis population, interim analyses, or sensitivity analyses. It also does not report the trial’s original sample-size justification or detailed statistical plan. These gaps limit assessment of whether the reported comparisons align with prespecified analyses.

Only the primary holding-time results are available in the registry record, while other behaviors named in the objective are not reported. The adverse-event summaries give counts but do not describe collection methods or assessment of relatedness. These are limitations of what the record reports; they do not establish that the missing procedures or analyses were absent from the conduct of the study.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Exposure to Gun Violence in Video Games Increases Interest in Real Guns — NCT03259139.
• Registry record last updated 2020-05-11; status last verified 2020-04.
• Results first posted 2020-05-11.
• 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.

21. Record Summary

The completed randomized clinical trial enrolled 250 and examined video-game content in relation to children’s interest in real guns. Its primary outcome was mean time spent holding a real firearm, in seconds, during a 20 minute play session. The registry describes a factorial model and single masking, but does not report the identity of the masked party or the full analysis specification.

Posted group means and comparisons distinguish players and watchers across the violent-with-guns, violent-without-guns, and no-violence conditions. The comparisons use the no-violence player group as the reference. All listed confidence intervals include zero, and the p-values are unadjusted. Results for the other behaviors named in the objective are not reported in the registry record. The registry record also does not report multiplicity handling, missing-data methods, or the original design assumptions.