Trial on the Effect of E-cigarette Advertising on Cigarette Perceptions in Adolescents: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Trial on the Effect of E-cigarette Advertising on Cigarette Perceptions in Adolescents (NCT03786042), a randomized na study. The primary outcome is Implicit Positive Smoking Expectancies, Measured by the Implicit Association Test.

2. Clinical Question

The trial is listed as randomized, with single masking and a factorial model. The registry's overall aim was to investigate whether e-cigarette advertising exposure may lead to combustible smoking initiation in adolescents, while the posted primary endpoint is an immediate measure of implicit expectancies. These are related but distinct questions: an immediate reaction-time measure is not itself an observed smoking behavior or a direct measure of later initiation.

3. Trial Design

The registry identifies this as a randomized trial with a factorial model and single masking. It lists an enrollment of 139. The primary outcome summaries report 86 participants in the e-cigarette ad exposure group and 41 in the non e-cigarette ad exposure group. Those group counts do not account for the full registry enrollment; the reason for the difference is not reported in the registry record. The record does not specify which people were masked or provide further details about the randomization process.

Participants in the e-cigarette ad exposure condition viewed e-cigarette ads on a computer screen while eye movements were tracked. Participants in the comparator condition viewed non e-cigarette ads on a computer screen while eye movements were tracked. The registry lists the intervention types as behavioral for e-cigarette ad exposure and other for non e-cigarette television commercials. It does not give further details about the content, duration, number, or selection of advertisements in these descriptions.

The primary outcome was measured within 5 minutes post intervention. The trial record lists three outcome measures, but no secondary results are posted. It lists zero analyses. The status is completed; the recorded start date is 2019-02-04 and primary completion date is 2021-12-09. These dates describe the registry record and do not supply additional information about the conduct or analysis of the trial.

The record does not report the original sample-size justification, the planned primary analysis model, the allocation ratio intended at design, or the operational details of its factorial structure. Those design details should not be inferred from the posted group counts or from the planning illustrations based on the observed contrast.

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: 139 participants competed the consent process and agreed to participate in this study. However, only 127 completed any study visits because of the suspension of in-person lab visits caused by the COVID-19 pandemic between March 2020 and July 2021.

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 Implicit Positive Smoking Expectancies, Measured by the Implicit Association Test, measured over within 5 minutes post intervention.

The primary endpoint is implicit positive smoking expectancies, measured by the Implicit Association Test and expressed in milliseconds. The task records how long it takes participants to categorize smoking-related words with positive words, such as cool, and with negative words, such as cancer. The registry describes faster categorization of smoking-related words with positive words as evidence of higher positive smoking expectancies.

This endpoint is used to capture an immediate, implicit response to smoking-related concepts rather than relying only on a direct question about attitudes. Reaction time is the measured quantity; its interpretation depends on the task and on the assumption that faster categorization reflects stronger associations. It is not a direct measure of cigarette use, smoking initiation, or future behavior.

The recorded timeframe is within 5 minutes post intervention, so the result concerns a near-term response after the advertising exposure. It cannot by itself establish whether any difference persists or predicts later behavior. The registry does not report the detailed scoring procedure, the exact task structure, handling of unusual reaction times, or criteria for excluding task data. The posted primary endpoint is a continuous measure summarized by a mean and standard deviation in each group.

7. Results Posted in the Registry

• Non E-cigarette ad Exposure: mean 1,030 versus 1,073 on control; difference -43.09 (95% CI -92.34 to 6.16), standardized effect d = -0.28; z-test p = 0.086.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is continuous: it is a reaction-time measure in milliseconds, summarized as a mean and standard deviation for each exposure group. A mean describes the average observed value in a group. A standard deviation describes how dispersed individual values are around that mean. Together, these summaries give a compact description of the group distributions, but they do not show all features of those distributions, such as skewness, extreme values, or subgroup patterns.

How to read the mean difference

The posted contrast is defined as the non e-cigarette ad exposure group relative to the e-cigarette ad exposure group. Its estimate is -43.09 milliseconds. A negative value means the first group's mean is lower than the second group's mean under that contrast convention. Because the endpoint description associates faster categorization with higher positive smoking expectancies, a lower reaction-time mean has a task-specific interpretation. It should not be translated into a claim about smoking behavior or later initiation.

A difference in means retains the original measurement units, which makes it possible to interpret the estimate in milliseconds. The standardized estimate, -0.279, expresses the contrast in relation to the pooled standard deviation, here 154.5. Standardization can help describe the size of a contrast relative to the variability in the measured outcome. It does not make an uncertain estimate precise, and it does not establish whether the measured contrast is consequential outside this task.

What the p-value means

The unadjusted p-value is 0.086. In general, a p-value describes how surprising a test statistic at least as extreme as the observed one would be if a specified null model were true. It is not the probability that the null hypothesis is true, nor is it the probability that the finding occurred by chance. The word unadjusted indicates that the posted value is not presented as corrected for multiple comparisons. The registry record does not report the exact test procedure or model underlying the p-value, so additional analytic details should not be assumed.

In general, a p-value is sensitive to sample size, variability, and the magnitude of the contrast. It should be considered together with the estimate and its confidence interval. A single threshold-based reading would discard information about direction and plausible magnitudes. Here, the interval includes zero and the p-value is reported as unadjusted; both reinforce that the data do not pin down the contrast precisely.

Why variability and allocation matter

The standard deviations are 171.9 milliseconds in the e-cigarette ad exposure group and 108.6 milliseconds in the non e-cigarette ad exposure group. The group sizes are 86 and 41, respectively. Unequal group sizes can affect the precision of a comparison, particularly when the amount of variability differs between groups. The posted pooled standard deviation is 154.5, which is used for the reported standardized contrast. The registry does not describe whether the primary analysis used this pooled estimate or another variance model.

In general, an analysis of means can be affected by skewed data, influential extreme reaction times, and differences in variance. Reaction-time tasks can produce distributions that are not symmetric, so the analysis may depend on how the task data were processed. The registry record does not report transformations, outlier rules, missing-task handling, or diagnostic checks. These omissions do not prove that such procedures were absent; they mean the registry record does not document them.

In general, over-dispersion means that variation is greater than a model assumes. For continuous outcomes, a poor variance assumption can make standard errors and confidence intervals unreliable. The posted standard deviations describe observed spread, but no diagnostic information is reported that would establish whether the analysis assumptions were checked. It is therefore appropriate to interpret the posted contrast and interval as given, while recognizing that the statistical plan and diagnostics are not documented in the registry record.

Randomization and the factorial label

Random assignment can, in general, support a causal comparison by balancing measured and unmeasured characteristics on average, provided the trial is implemented and analyzed appropriately. The registry labels the trial randomized and factorial. It does not give the randomization sequence, concealment procedures, specific factorial factors or cells, or the analysis used to account for the factorial design. The two posted exposure groups identify the contrast summarized, but do not justify reconstructing the full design.

Single masking is recorded, but the record does not specify who was masked or how masking was maintained. Nor does it state whether participants could identify the advertising condition. Such details matter when outcomes involve a task that requires participant responses, but they cannot be inferred from the masking label alone.

What is not documented

The registry lists zero analyses and does not report the original statistical analysis plan, covariate adjustment, interim analyses, missing-data methods, or analysis populations. It also does not provide a sample-size justification for the trial. These items are not reported in the registry record. The absence of a posted analysis description limits assessment of whether the primary endpoint, contrast, and statistical procedure followed a prespecified plan.

General statistical interpretation separates estimation from decision-making. The mean difference estimates the contrast, the confidence interval describes uncertainty under the analysis, and the p-value summarizes evidence relative to a null model. None of these statistics independently establishes a mechanism, persistence over time, or a behavioral consequence. The endpoint is an immediate implicit-association measure, and conclusions should remain within that scope.

10. Confidence Intervals and Precision

The estimated mean difference is -43.09 milliseconds for non e-cigarette ad exposure relative to e-cigarette ad exposure. Its confidence interval extends from -92.34 to 6.155 milliseconds. A confidence interval is a range produced by a statistical procedure that, under its assumptions, reflects uncertainty about the underlying contrast. It is not a range containing a known proportion of individual participants' reaction times, and it does not assign a probability to the true effect after the data have been observed.

The interval includes zero and covers contrasts in either direction. Its lower end represents a more negative mean difference, while its upper end allows a small positive difference. Thus, the point estimate alone does not capture the uncertainty about whether the group means differ or how large the contrast might be. The estimate should be interpreted with the reported interval rather than as a precise standalone value.

The group summaries have different standard deviations, 171.9 and 108.6 milliseconds, and different participant counts, 86 and 41. Both observed spread and group size contribute to the precision of a comparison. The registry record does not report the detailed method used to construct the interval, so its model assumptions cannot be evaluated from the listed results alone.

Precision is distinct from importance. Even a precisely estimated difference would describe this reaction-time endpoint and timeframe, not automatically a meaningful change in later smoking behavior. Conversely, the interval's inclusion of zero does not prove that the groups are identical. It indicates that the posted data leave uncertainty about the direction and magnitude of the contrast.

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 planning figures are participants per arm, not total events. They illustrate sample sizes for detecting specified fractions of the observed effect under a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set at the stated fraction of the observed effect. These are planning assumptions for an illustration, not the original trial design assumptions; those original assumptions are not reported in the registry record.

For planning at 100% of the observed effect, the illustration gives 202 participants per arm for 80% power and 271 participants per arm for 90% power. At 75% of the observed effect, it gives 359 participants per arm for 80% power and 481 participants per arm for 90% power. At 50% of the observed effect, it gives 808 participants per arm for 80% power and 1,081 participants per arm for 90% power.

In general, power is the chance that a study using a specified design and analysis will detect an effect of a specified size, if that effect is present and the model assumptions hold. The 80% and 90% columns therefore represent different planning targets: the higher target calls for more participants per arm under these assumptions. The figures also illustrate that smaller effects need more participants to detect than larger effects, because a smaller signal is harder to distinguish from variability.

These calculations use the observed effect as their reference point and should not be read as a recommendation or as evidence about the trial's original planning. This is not observed power. The registry record does not report whether the trial was originally designed using these assumptions or whether its planned sample size was justified using another approach.

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 = -43.090; pooled standard deviation σ = 154.500.
2. n = 2 × (1.96 + 0.842)² × 154.500² / -43.090² = 201.8, rounded up to 202 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 <- -43.0900000000001                      # observed difference in means
sigma <- 154.4996437148                     # 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(3786042)
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 = 202 (80% power), 271 (90% power)
 75% of observed effect: n per arm = 359 (80% power), 481 (90% power)
 50% of observed effect: n per arm = 808 (80% power), 1,081 (90% power)
power.t.test: 202.8 
Simulated power at n = 202 per arm: 0.799 (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 registry record lists three outcome measures, but zero secondary results are posted and zero analyses are listed. The primary endpoint is the Implicit Association Test measure of implicit positive smoking expectancies. Other concepts listed in the record include attentional bias, smoking cues, social norms, and smoking susceptibility, but results for these concepts are not posted as secondary results in the record described.

Multiplicity arises when several outcomes, comparisons, time points, or analyses are examined. In general, conducting multiple tests can increase the chance of at least one apparently unusual result if each is interpreted independently. Methods such as prespecifying a primary endpoint or adjusting for multiple comparisons can address aspects of this issue, but the registry record does not report a multiplicity plan or an adjustment procedure. The posted p-value is unadjusted, so it should be understood as applying to the stated comparison without a documented correction for other tests.

The record does not report whether there were interim analyses or whether other analyses were conducted but not posted. No assumptions should be made about either possibility. The absence of posted secondary findings means the primary comparison cannot be placed in a complete results context across the full set of outcome measures.

Missing-data handling and analysis populations are also not reported in the registry record. The enrollment is listed as 139, while the primary outcome summaries list 86 participants in one group and 41 in the other. The record does not explain the difference between the enrollment figure and the numbers contributing to these group summaries. It also does not state whether the results use an intention-to-treat population, a per-protocol population, or another definition, nor does it describe exclusions or missing outcome values.

These documentation gaps limit assessment of the analysis pathway and the completeness of the posted comparison. They do not establish that a particular procedure was or was not used. Interpretation should be confined to the results and analytic details documented in the registry record.

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

The posted result describes a comparison of average reaction times on an implicit-association task shortly after two advertising exposures. The mean was 1073.0 milliseconds after e-cigarette ad exposure and 1030.0 milliseconds after non e-cigarette ad exposure. The estimated difference, expressed as non e-cigarette minus e-cigarette exposure, was -43.09 milliseconds. The interval from -92.34 to 6.155 milliseconds indicates uncertainty that includes zero, and the unadjusted p-value is 0.086.

The endpoint description links faster categorization of smoking-related words with positive words to higher implicit positive smoking expectancies. That task-specific description helps explain the direction of the measured response, but it does not turn a reaction-time difference into evidence of smoking initiation. The measure was collected within 5 minutes post intervention. The result therefore addresses an immediate response, not whether participants later smoked or whether any response persisted.

The trial's broader stated aim concerns a possible pathway from e-cigarette advertising exposure to combustible smoking initiation in adolescents. The posted primary outcome is narrower than that aim: it measures implicit expectancies rather than smoking behavior. No secondary results are posted to provide an outcome on initiation or to show results for the other listed concepts. Accordingly, the registry result does not answer whether exposure changed later smoking behavior, social norms, or susceptibility.

The reported group contrast is an estimate, not a definitive classification of the exposure as beneficial or harmful. The interval's inclusion of zero does not prove no difference; the point estimate does not establish a meaningful difference. It leaves uncertainty about the exact size and direction of the population contrast under the analysis used. The record also does not report enough statistical-plan detail to assess all assumptions or alternative analyses.

The findings should be understood as descriptive statistical evidence from the posted comparison, not as medical advice or a claim that an exposure does or does not cause smoking. The result cannot independently establish a causal pathway or a lasting effect. Any interpretation beyond the immediate task measure would require evidence from outcomes and follow-up that are not reported in the registry record.

15. Why This Trial Matters Statistically

This trial is statistically interesting because it connects a randomized exposure contrast with a behavioral science measure that is indirect and immediate. The primary endpoint is not a self-reported intention or observed smoking event; it is reaction time during an implicit association task. That distinction matters when aligning the statistical question with the broader public-health motivation described in the registry.

The mean difference is expressed in milliseconds, while the standardized estimate relates the contrast to observed variability. Reading both helps distinguish the raw scale from a relative scale. The confidence interval extends across zero, so the reported point estimate should not be mistaken for a settled direction or a precise magnitude. The unadjusted p-value adds information about the test, but it cannot substitute for estimating the size and uncertainty of the contrast.

The trial also highlights how registry completeness affects statistical interpretation. Randomized allocation and a stated primary endpoint identify important design features, but information about the factorial structure, original sample-size rationale, analysis model, missing data, and multiplicity is not reported in the registry record. With no secondary results posted, the primary contrast cannot be assessed alongside results for the other outcomes in the registry listing.

Finally, the planning illustrations show how a study's required size depends on the effect it aims to detect and its chosen power target. Those figures are separate from the original trial's design rationale and do not describe the precision actually achieved as a planning decision. Keeping the distinction between trial results, planning scenarios, and the broader research question clear is essential when evaluating this kind of evidence.

16. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The posted outcome summaries cover the immediate implicit-association endpoint and do not report later smoking behavior. The timeframe is within 5 minutes post intervention, so persistence of any measured response is not established. The broader research aim mentions combustible smoking initiation, but no initiation result is posted in the primary comparison described.

The registry lists enrollment of 139, while the arm summaries contain 86 participants in the e-cigarette ad exposure group and 41 in the non e-cigarette ad exposure group. The registry record does not explain the difference, identify the analysis population, or describe missing outcome data. The reason should not be inferred. The group counts are also unequal, and the observed standard deviations differ, which are relevant features when considering precision.

The record labels the design randomized, single masked, and factorial, but does not explain allocation procedures, the identity of the masked parties, or the factorial components and analysis. It does not report the detailed reaction-time scoring and data-cleaning rules, eligibility criteria, participant demographics, or advertisement content and exposure duration. These details would help evaluate generalizability and the construction of the endpoint.

There are zero analyses listed and zero secondary results posted, despite the record listing three outcome measures. The statistical model, covariate adjustments, interim analyses, missing-data approach, multiplicity handling, and original sample-size justification are not reported in the registry record. The posted result is therefore informative as an estimate with an interval, but the available documentation does not support a fuller assessment of analysis choices, outcome completeness, or the full set of trial findings.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Trial on the Effect of E-cigarette Advertising on Cigarette Perceptions in Adolescents — NCT03786042.
• Registry record last updated 2025-09-26; status last verified 2025-09.
• Results first posted 2025-09-26.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

20. Record Summary

The registry record describes a completed randomized, single-masked trial with a factorial model and an enrollment of 139. Participants viewed either e-cigarette ads or non e-cigarette ads on a computer screen while eye movements were tracked. The primary endpoint was implicit positive smoking expectancies measured by the Implicit Association Test in milliseconds within 5 minutes post intervention.

The posted group means were 1073.0 milliseconds for e-cigarette ad exposure and 1030.0 milliseconds for non e-cigarette ad exposure. The mean difference comparing non e-cigarette with e-cigarette ad exposure was -43.09 milliseconds, with an interval from -92.34 to 6.155 milliseconds and an unadjusted p-value of 0.086. The record lists three outcome measures but posts no secondary results and lists zero analyses. Several design and analysis details, including the original sample-size justification and missing-data methods, are not reported in the registry record.