A Website to Teach Children Safety With Dogs: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Website to Teach Children Safety With Dogs (NCT02299427), a randomized phase 1 study. The primary outcome is Children's Behavior With Dogs on Standardized Objective Scale.

2. Clinical Question

This question focuses on behavior observed in a structured encounter, rather than on dog bites, injuries, or longer-term behavior in everyday settings. The registry description says the website aimed to teach children safe interactions and also to support cognitive and behavioral change in children and parents. However, the primary outcome is specifically the children's behavior during the live-dog assessment. The registry record does not report a primary outcome measuring bite incidence or establish whether observed behavior predicts later injuries. The comparison is therefore best read as an assessment of a particular behavioral scale after two different educational website experiences, in the specified analysis population.

3. Trial Design

The registry identifies this as a randomized, parallel-group Phase 1 study with no masking. The study status is completed. The registry enrollment is 69. The recorded study start was 2015-03, and primary completion was 2016-01. The sponsor was the University of Alabama at Birmingham.

Participants were assigned to one of two behavioral interventions. The Dog Safety group used a website developed for this research for 2 weeks; the Transportation Safety group used a publicly available child transportation safety website for 2 weeks. Because both groups received website-based education, the design compares two types of educational content and activity. It does not isolate the effect of using any website against receiving no website intervention.

The primary outcome was assessed post-intervention, about 2 weeks after the pre-intervention assessment. The registry describes coding behavior during a semi-structured interaction with a live therapy dog. Although the study's broad enrollment is recorded as 69, the outcome-specific analysis populations differ: the primary measure includes only children with known noncompliance in the dog safety group, and the arm summaries list 12 for Dog Safety and 34 for Transportation Safety. The registry record does not explain the relationship between these outcome populations and overall enrollment, nor does it give a complete participant flow for this outcome.

The registry's brief description presents the study as developing and evaluating an interactive website for child dog-bite prevention, with an intended target age range of 4-6. The primary result concerns a short-term observed behavior measure, rather than the full set of aims described in that summary. The protocol's original sample-size rationale, detailed allocation procedures, randomization method, and planned analysis strategy are not reported in the registry record. No masking was used, so the registry does not indicate that participants or assessors were blinded.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: In the dog safety group, only children with known noncompliance were analyzed for this measure

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 Children's Behavior With Dogs on Standardized Objective Scale, measured over post-intervention (about 2 weeks after pre-intervention assessment).

The primary endpoint is the mean score on the Children's Behavior With Dogs on Standardized Objective Scale at post-intervention, about 2 weeks after the pre-intervention assessment. It is based on coded behavior during a semi-structured interaction with a live therapy dog. The assessment considered patterns across 15 tasks, activities, or decisions, including touching or approaching the dog, handling the dog's toys, and interrupting the dog during its rest time.

Seven behaviors were selected after factor analysis, standardized, and averaged to form the scale. A linear transformation made all values positive. The theoretical range is 0-infinity, although the registry says most children scored between 0-4 in practice. Higher scores indicate higher risk-taking. The scale's individual items had an average intercorrelation of .50 and Cronbach's alpha of .65. These values describe how the items related to one another and their reported internal consistency; they do not by themselves establish that the score predicts dog bites or captures all aspects of safe interaction.

Behavioral observation is relevant to the study's educational aim because it assesses what children did in a dog interaction, rather than only what they knew or reported. At the same time, the result is limited to the assessment setting and the constructed scale. The registry record does not report a validated threshold for an unsafe score, nor does it report a link between this score and later injuries. Interpretation also needs to account for the stated analysis population: only children with known noncompliance in the Dog Safety group were analyzed for this measure.

7. Results Posted in the Registry

• Transportation Safety: mean 1.60 versus 1.40 on control; difference 0.20 (95% CI -0.19 to 0.59), standardized effect d = 0.27; z-test p = 0.315.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a continuous measure summarized by a mean and standard deviation. A mean gives the average score in an analyzed group, while the standard deviation describes how dispersed individual scores are around that average. Neither summary conveys the full distribution of scores. This matters for an outcome whose theoretical range is 0-infinity and whose practical scores were described as mostly between 0-4: the group average does not show whether most scores were similar or whether a smaller number of higher scores influenced it.

What does the mean difference represent?

The comparison is expressed as Transportation Safety minus Dog Safety. The posted means were 1.6 and 1.4, respectively, and the mean difference was 0.2. In general, a mean difference expresses the difference between group averages on the outcome's scale. A positive value in this contrast means the Transportation Safety mean is higher. Because higher scores indicate higher risk-taking, the direction of the estimate corresponds to a higher average risk-taking score in the comparator group. This is a scale-based contrast; it is not a risk ratio, a difference in bite counts, or a percentage change.

The comparison is unadjusted and recalculated from the posted summary statistics. An unadjusted comparison uses the group summaries without accounting for other participant characteristics or covariates. The registry record does not report an adjusted estimate or say that adjustment was part of the original statistical plan. The record also does not explain the outcome-specific restriction to children with known noncompliance in the Dog Safety group, so that analysis-population detail should remain visible when interpreting the contrast.

How should the p-value be read?

The unadjusted p-value was 0.315. In general, a p-value describes how compatible the observed data are with a specified statistical null model, under the assumptions of the analysis. It is not the probability that the null hypothesis is true, and it does not measure the size or practical importance of an effect. A p-value alone also does not distinguish a small effect from an imprecisely estimated one. Here, it should be read alongside the mean difference and the reported interval, which crosses zero.

The registry record does not report whether this p-value matches a sponsor-conducted analysis or identify the exact test used in an original analysis plan. It is therefore most useful as part of the unadjusted summary-statistic comparison, not as evidence that every design or analysis choice has been reconstructed. Random assignment is reported, but allocation details and the treatment of the outcome-specific analysis population are not described.

What does the interval add?

A confidence interval expresses uncertainty around an estimated quantity. In general, an interval gives a set of values reasonably compatible with the data and the method used to construct it. A narrower interval indicates greater precision on the outcome's scale; a wider interval indicates less precision. The reported interval for this mean difference includes values on both sides of zero. It therefore does not isolate one direction of group difference. The interval's exact confidence level is not reported in the registry record, so no level should be assumed.

For a continuous outcome, the standard deviations help describe within-group spread, while the standard error helps describe uncertainty in the estimated mean. The registry comparison includes standard errors and a pooled standard deviation, but these are not substitutes for understanding the individual-score distribution or the way the sample was selected. General statistical reasoning also cautions that a pooled spread can summarize variability across groups under particular assumptions; it does not prove that group variances are identical. No diagnostic information about distributional assumptions is reported.

Why does the construction of the scale matter?

The score combines seven standardized behaviors selected through factor analysis from 15 observed tasks, activities, or decisions. Standardizing items places them on a common scale before averaging, and the linear transformation makes all scale values positive. These construction choices define what a one-unit score difference means. The registry does not report the factor-analysis details, the transformation formula, or evidence supporting a clinical interpretation of particular score values. The reported average intercorrelation of .50 and Cronbach's alpha of .65 offer information about the item set, but they do not establish test-retest reliability, predictive validity, or responsiveness to change.

In general, over-dispersion means that observed variability is greater than a statistical model expects. It can affect uncertainty estimates when a method assumes less variability than the data contain. The registry record does not report an over-dispersion assessment, and the endpoint is summarized as a continuous measure with standard deviations; therefore no claim about over-dispersion in this analysis is warranted. More broadly, outcome distributions, outlier checks, model diagnostics, missing-data methods, and sensitivity analyses are not reported in the registry record.

The trial is described as randomized and parallel-group, but masking was none. General statistical interpretation recognizes that unmasked assessments can be vulnerable to differences in participant behavior or outcome coding if expectations influence what is done or recorded. The registry does not say whether coders were independent or whether coding reliability was assessed. These are limits on what can be established from the posted methods, not proof that bias occurred.

10. Confidence Intervals and Precision

The reported mean difference is 0.2, with the contrast defined as Transportation Safety minus Dog Safety. Its reported interval extends from below zero to above zero. This crossing of zero means that the interval includes no mean difference as well as differences in either direction. The available estimate therefore does not pin down the direction of the group contrast with precision.

Precision concerns the range of values compatible with the data and the estimation method; it is distinct from whether an estimate is important in practical terms. The interval is not confined to a single side of zero, so it allows more than one interpretation of the direction. The registry record does not provide a clinical or behavioral threshold for a meaningful difference on this transformed scale. It is therefore not possible to use the interval to classify all compatible values as important or unimportant for real-world dog safety.

The reported group summaries are based on 12 in Dog Safety and 34 in Transportation Safety for this endpoint. Those are the analysis populations listed for the outcome, not a reconciliation of the registry's overall enrollment. The analysis restriction in the Dog Safety group is particularly relevant: only children with known noncompliance were analyzed for this measure. The registry record does not report how this selection affected uncertainty or whether missing outcome information contributed to the analyzed group counts.

Standard deviations of 0.5 and 0.8 describe spread in the two groups, but they do not by themselves communicate the precision of the mean difference. The reported interval is the direct summary of uncertainty for that contrast. Its confidence level and the original sponsor's inferential method are not reported in the registry record. The unadjusted p-value of 0.315 is another summary of the comparison, but it should not replace consideration of the interval's range and the limitations of the analysis population.

Finally, precision for this short-term behavioral score should not be confused with precision about dog-bite prevention. The study's primary measure is observed behavior during a structured encounter, and the registry record does not report an injury endpoint. The interval therefore describes uncertainty about the posted scale comparison, not about the effect on bite incidence or longer-term safety.

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 a report of the trial's original design assumptions. The registry record does not report the original sample-size justification. The listed assumptions for the illustration 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 at the stated fraction of the observed effect. For this continuous outcome, the figures mean participants per arm, not total events.

The rows consider planning for 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect. The corresponding figures are 213, 379, and 852 participants per arm for 80% power, and 286, 507, and 1,141 participants per arm for 90% power. The 80% power column represents the planned chance of detecting the specified effect under the assumptions if that effect and model were true; the 90% column uses a higher target chance of detection. These are design targets, not guarantees about what a future study would find.

Smaller effects need more participants to detect under the same planning assumptions. That pattern is reflected in the larger figures for planning at 75% and 50% of the observed effect. The curve presents the same general planning relationship across effect sizes: as the effect to be detected becomes smaller, the required participants per arm increase. The displayed values are not observed power for the completed study; they are prospective planning illustrations based on the observed effect and the stated assumptions.

The illustration should not be read as evidence that the completed trial was designed for these sample sizes. It does not establish whether the original study used the same significance level, allocation, approximation, effect target, or power target. The registry record does not report those original design choices.

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 = 0.200; pooled standard deviation σ = 0.737.
2. n = 2 × (1.96 + 0.842)² × 0.737² / 0.200² = 212.9, rounded up to 213 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 <- 0.2                      # observed difference in means
sigma <- 0.736545993132812                     # 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(2299427)
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 = 213 (80% power), 286 (90% power)
 75% of observed effect: n per arm = 379 (80% power), 507 (90% power)
 50% of observed effect: n per arm = 852 (80% power), 1,141 (90% power)
power.t.test: 213.9 
Simulated power at n = 213 per arm: 0.800 (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 lists 2 outcome measures and identifies the Children's Behavior With Dogs on Standardized Objective Scale as the primary outcome. It reports 0 secondary outcomes posted and 0 analysis entries. The information available therefore does not provide a complete set of secondary outcome results or a full account of analyses conducted. Another registered outcome is not described in the posted endpoint information in enough detail to interpret its measurement or result.

Multiplicity arises when several outcomes, time points, subgroups, or statistical comparisons are examined. In general, examining multiple results can increase the chance that some appear noteworthy by chance, and a prespecified plan can explain how that issue is handled. The registry record does not report a multiplicity adjustment, a hierarchy among outcomes beyond naming the primary outcome, or a statistical analysis plan. It is not possible to determine whether any adjustment was planned or used.

Missing-data methods are also not reported in the registry record. The endpoint population is described as including only children with known noncompliance in the Dog Safety group, but the record does not say how noncompliance was determined, whether the comparator group had a corresponding criterion, or why other participants were not included in the endpoint analysis. No missingness pattern, imputation method, or sensitivity analysis is reported. The distinction between overall enrollment and the endpoint analysis populations should therefore be retained rather than treating the posted endpoint summaries as a complete analysis of everyone enrolled.

The record does not report an intention-to-treat analysis, a per-protocol analysis, or another named analysis population. Nor does it report interim analyses or stopping rules. Those details are not reported in the registry record and should not be inferred from the completed status or from the randomized design. The reported comparison remains an unadjusted summary-statistic analysis of the listed endpoint groups, with the limitations that follow from the available population and analysis information.

14. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: adverse medical or social situations (0.0%).

For serious adverse events, the registry lists 35 participants at risk in the Dog Safety group and 34 in the Transportation Safety group, with 0 serious events reported as affecting participants in each group. For other adverse events, it lists the same at-risk counts and 0 events reported as affecting participants in either group. The listed top serious-event term is “adverse medical or social situations,” with 0.0%.

These figures describe the adverse-event reporting fields in the registry. They do not show that no unwanted experiences of any kind occurred; the record specifically reports zero events in the listed serious and other adverse-event categories. The at-risk counts for safety reporting are distinct from the endpoint analysis populations and should not be treated as the primary-outcome group sizes.

Death counts are not reported in the registry record. A blank or unreported death value should not be interpreted as a reported zero. The registry also does not provide event descriptions, severity details beyond the categories, ascertainment procedures, or an explanation of how adverse events were collected. The safety entries should therefore be read narrowly as the reported counts and category information, rather than as a complete characterization of all possible harms.

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

The primary result describes short-term behavior during a semi-structured interaction with a live therapy dog. The mean score was 1.4 for Dog Safety and 1.6 for Transportation Safety, and the unadjusted mean difference was 0.2 in the direction of a higher average risk-taking score in the Transportation Safety group. The reported interval crosses zero, and the unadjusted p-value was 0.315. These results do not establish a clear difference between the groups on this measure, and they do not demonstrate equivalence.

The outcome is a constructed scale based on seven standardized behaviors, selected from observation of 15 tasks, activities, or decisions. Higher values indicate higher risk-taking, but the registry does not report a threshold that distinguishes safe from unsafe behavior or a minimum meaningful difference. The scale's reported internal consistency information does not make the score interchangeable with actual dog-bite risk.

The comparison also has a specific scope. Both groups used websites for 2 weeks, and the comparator website addressed transportation safety. The result therefore contrasts two educational website experiences, rather than a dog-safety website with no intervention. The assessment was conducted post-intervention, about 2 weeks after the pre-intervention assessment. It does not establish what happened after a longer period or whether participants behaved similarly in other settings.

The record's stated analysis population further limits interpretation: only children with known noncompliance in the Dog Safety group were analyzed for the primary measure. The record does not explain that restriction or provide enough participant-flow information to establish how these endpoint groups relate to overall enrollment. Caution is needed when extending the result to all children who enrolled or to the intended target ages generally.

The registry describes a broader goal of preventing child dog bites and supporting children's and parents' learning. However, the posted primary outcome is behavior on a scale, not bite incidence, injury, hospitalization, or death. The result cannot by itself establish whether the educational websites changed real-world bite risk. It is a limited statistical comparison of the reported short-term behavioral measure in the analyzed groups.

16. Why This Trial Matters Statistically

This trial illustrates why an outcome's definition and analysis population are central to statistical interpretation. The primary measure is not simply a count of safe actions: it is an average of seven standardized behaviors selected from observed tasks and transformed so all values are positive. The direction is defined clearly—higher values indicate higher risk-taking—but the scale has no registry-reported threshold for a clinically meaningful difference. A mean difference therefore needs to be interpreted on the scale's terms rather than translated into a claim about injuries.

The result also shows the value of reading an estimate together with its uncertainty. The group means and unadjusted mean difference give a direction and scale for the observed contrast, while the interval crossing zero indicates that the direction is not established precisely. The p-value alone would not convey the full range of compatible differences. Neither statistical summary substitutes for a clear description of how the endpoint groups were selected.

In this record, the primary outcome's analysis population has a notable restriction: only children with known noncompliance in the Dog Safety group were analyzed. The record lists different endpoint group sizes and does not explain how they relate to overall enrollment. Such information affects what population the estimate describes. Randomized assignment is valuable for comparisons, but an outcome analysis restricted after assignment still requires a transparent account of eligibility, missingness, and participant flow.

The active website comparator is another important design feature. The contrast concerns two educational experiences, not a website versus no activity. That distinction shapes the scientific question and prevents an overbroad interpretation. Finally, planning calculations based on an observed effect can help illustrate the scale of a future study, but they do not reveal the rationale used to design the completed study. Statistical reading depends on keeping these design, measurement, population, and precision limits together.

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.

Several trial-specific details limit what can be concluded from the posted result. The most prominent is the outcome-specific analysis rule: only children with known noncompliance in the Dog Safety group were analyzed for the primary measure. The registry record does not define noncompliance, explain the reason for restricting the group, or describe a corresponding restriction in the Transportation Safety group. It also does not provide enough participant-flow information to reconcile overall enrollment with the endpoint analysis populations. The estimate should therefore be understood as applying to the groups included in the posted outcome summary.

The endpoint is a transformed behavioral scale, not a direct measure of dog bites or injuries. Although the registry describes the scale construction and reports item intercorrelation and Cronbach's alpha, it does not report a clinically meaningful threshold, predictive validity for future bites, or a formula for the transformation. The registry does not report whether the observed interaction captures children's behavior in familiar or everyday settings.

The intervention period was 2 weeks, and the endpoint was assessed post-intervention about 2 weeks after the pre-intervention assessment. The record does not report longer-term follow-up for this outcome. Masking was none, and the registry does not say whether outcome coders were blinded to group assignment or how coding consistency was checked. These omissions do not establish that bias occurred, but they leave those aspects of measurement quality unclear.

The active comparator was a publicly available transportation-safety website, so the contrast does not isolate the effect of receiving a website versus receiving no website. The registry record does not report detailed allocation procedures, missing-data methods, the original sample-size rationale, a full statistical analysis plan, or a complete set of secondary results. These limits leave important design and analysis details unreported.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: A Website to Teach Children Safety With Dogs — NCT02299427.
• Registry record last updated 2017-02-06; status last verified 2016-12.
• Results first posted 2017-02-06.
• 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 registry identifies a completed randomized, parallel-group Phase 1 study sponsored by the University of Alabama at Birmingham. Its enrollment is 69, and masking is listed as none. The interventions were 2 weeks of regular use of a research-developed child dog-safety website or a publicly available child transportation-safety website.

The primary posted outcome is a mean score for observed behavior with a live therapy dog, with higher scores indicating higher risk-taking. The posted means were 1.4 for Dog Safety and 1.6 for Transportation Safety; the unadjusted mean difference was 0.2, with an interval crossing zero and an unadjusted p-value of 0.315. For this outcome, only children with known noncompliance in the Dog Safety group were analyzed.

The registry reports zero serious and other adverse events affecting participants in the listed safety groups. Death counts are not reported in the registry record. Original design assumptions, detailed analysis methods, missing-data handling, and the broader outcome results are not reported in the registry record.