Studying the Effects of Natural Visual Scene Changes on Typical Adult Visual Perception: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Studying the Effects of Natural Visual Scene Changes on Typical Adult Visual Perception (NCT05004649), a na na study. The primary outcome is Psychophysical Measurements of Horizontal Discrimination Threshold.

2. Clinical Question

3. Trial Design

This completed study, NCT05004649, is titled “Studying the Effects of Natural Visual Scene Changes on Typical Adult Visual Perception.” Its official title is “The Effects of Stimulus Variability in Natural Visual Scenes.” The sponsor is the University of Pennsylvania. The record lists the study phase and allocation as not applicable, describes a single-group model, and lists no masking.

The study used a behavioral psychophysical task. Participants viewed a central object within background objects in a natural visual scene and made judgments about the central object’s horizontal position. The task is described as a two-interval forced-choice procedure: one stimulus was presented in each interval, and participants judged whether the second central object was to the left or right of the first. The registry describes measuring the discrimination threshold as a function of noise in the stimulus and examining within-trial variability in background features.

The registry lists enrollment of 19. The primary-outcome population is described as the 10 naive subjects who completed the study; one completed subject who was an experimenter was excluded from analysis. Participants were to be excluded before the experiment if best-corrected visual acuity was worse than 20/40 in either eye or if they made any errors on the Ishihara plate test. The record does not provide further details of recruitment or the number screened.

The study start date is 2021-08-09, and primary completion is listed as 2022-03-23. The primary outcome timeframe is approximately 3 weeks. The registry status is completed. It does not report a randomized allocation, a separate control arm, a protocol-based sample-size justification, or a detailed statistical analysis plan.

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: Prior to the experiment, participants will be interviewed and fill out a survey. They will be excluded prior to the experiment if their best-corrected visual acuity is worse than 20/40 in either eye or if they make any errors on the Ishihara plate test. Participants will be excluded after the conclusion of their first session if their horizontal position discrimination threshold in the control condition is higher than a maximum value of 0.6 degrees of visual angle.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Analysis is of the 10 naive subjects who completed the study, one completed subject was an experimenter for the study and thus is excluded from 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.

6. Primary Endpoint

The registry lists the primary outcome as Psychophysical Measurements of Horizontal Discrimination Threshold, measured over Approximately 3 weeks.

The primary endpoint is the horizontal discrimination threshold, measured in degrees of visual angle. It represents the minimum horizontal distance between two stimuli that allows them to be recognized as spatially separate. A lower threshold means that a smaller horizontal difference is sufficient for the stimuli to be perceived as distinct. This measure is used to quantify sensitivity to the position of the central object, rather than simply whether a participant gives a correct response on an individual trial.

The task presents a central object in the context of background objects. In a two-interval forced-choice judgment, participants decide whether the central object in the second interval is to the left or right of the object in the first interval. The study’s question concerns whether variability in background features affects this discrimination when those features are irrelevant to the position judgment. Measuring a threshold across stimulus noise conditions is intended to characterize perceptual performance under different levels or forms of scene variability; the exact conditions are not reported in the registry record.

The registry specifies that the endpoint is summarized by the median and full range, with a timeframe of approximately 3 weeks. These choices describe a planned summary and the period associated with the measurement. However, numerical endpoint values, the full range, and results by stimulus-noise condition are not reported in the registry record. Consequently, the endpoint’s definition can be interpreted, but the magnitude and precision of the observed thresholds cannot be assessed from this record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The endpoint is a psychophysical threshold rather than a clinical event count. Participants make repeated judgments about the relative horizontal position of a central object, and the threshold represents the minimum separation at which the stimuli are recognized as spatially distinct. As a general statistical concept, a threshold compresses performance across trials into a value that can be compared across stimulus conditions. Interpreting it depends on how the task and threshold-estimation procedure are implemented. The registry describes the task and the threshold definition, but the detailed estimation procedure is not reported in the registry record.

What does the median describe?

The planned summary is the median. In general, the median is the middle value after observations are ordered, and it is less affected by extreme observations than the arithmetic mean. It can be useful when measurements are asymmetric or when a few observations are unusually high or low. A median alone does not show the distribution’s shape, and it does not show how much individual participants vary. The registry also specifies the full range, which gives the lowest and highest observed values, but the actual median and range are not reported in the registry record.

Threshold measurements may vary between participants, across trials, and across stimulus conditions. In general, a summary of central tendency and a measure of spread answer different questions: the first describes a typical measurement, while the second describes variability. A full range communicates the observed extremes but can be sensitive to individual extreme values. The registry does not report participant-level values, other distribution summaries, or uncertainty intervals, so the degree of variation and precision cannot be evaluated from the record.

Why do stimulus conditions matter?

The study’s conceptual contrast concerns variability in background features that are irrelevant to the requested position judgment. The primary outcome is described as a function of stimulus noise. In general, comparing thresholds under different stimulus conditions can help assess whether added variability is associated with a change in discrimination performance. To interpret such a comparison, readers need to know how conditions were defined, how many observations contributed to each condition, whether the same participants experienced each condition, and how the resulting measurements were analyzed. Those specific details are not reported in the registry record.

Over-dispersion is a general statistical issue that arises when observed variation is greater than a model assumes. If a method assumes less variation than the data contain, its uncertainty estimates can be too small and inferential conclusions can appear more precise than warranted. This concept is relevant generally to repeated behavioral measurements, where observations from the same participant may be related. The registry record does not identify the statistical model, its assumptions, or whether repeated measurements were modeled, so it is not possible to assess over-dispersion or the handling of within-participant dependence in this study.

What analysis details are available?

The registry lists one primary outcome and three analyses. The primary outcome is summarized as a median with a full range, and the outcome timeframe is approximately 3 weeks. The record does not report the names of the analyses, their estimands, statistical tests, adjustment variables, model form, or results. It also does not report a confidence interval or a p-value for the primary question. The number of analyses alone does not establish what comparisons were made or how the analyses relate to one another.

In general, an analysis plan clarifies which outcome is primary, how measurements are grouped, which comparisons are intended, and how uncertainty is quantified. These details help distinguish a prespecified analysis from exploratory work and help readers understand how the reported estimates were obtained. A protocol-based sample-size justification, interim analyses, and a detailed statistical plan are not reported in the registry record. No assumptions about those features can therefore be made.

How should precision be assessed?

In general, an estimate describes the observed quantity, while a confidence interval conveys uncertainty under a specified statistical method. For repeated measures, the precision calculation should account for the structure of the data, including whether multiple observations from each participant are related. Without the numerical estimates, participant-level distribution, and analysis method, readers cannot assess the reliability of the threshold summaries or how precisely condition differences were estimated. The registry’s stated median and full range are descriptive choices, not a substitute for inferential uncertainty.

The study record also contains no numerical results that could support recalculation from posted summary statistics. Therefore, no recalculated comparison is available to set alongside a sponsor analysis. The statistical interpretation is limited to the described design and endpoint definitions, with the empirical estimates and the methods used to produce them not reported in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry identifies one primary outcome and notes three analyses. It does not report the analysis specifications or explain whether the analyses address separate conditions, alternative models, or other questions. It also does not state whether multiplicity was considered or whether any adjustment was planned. In general, multiplicity arises when several statistical tests or estimates are examined; the more analyses considered, the more important it is to explain which comparisons are confirmatory and how the interpretation accounts for multiple opportunities to find an apparent association. The record does not provide the information needed to evaluate that issue.

The outcome population is described as 10 naive subjects who completed the study, with one completed experimenter excluded from analysis. Enrollment is listed as 19. The registry does not give a complete participant flow, reasons for non-completion, the number with usable measurements for each analysis, or participant-level missingness. It does not report how missing outcome data were handled, whether any observations were excluded after the experiment, or whether an analysis included all available measurements. These points are not reported in the registry record.

In general, missing-data methods matter because conclusions may depend on why measurements are unavailable and on the assumptions made about the unobserved values. A complete-case analysis, an available-observation analysis, and an imputation approach can rely on different assumptions. The registry record does not identify which approach, if any, was used. It is therefore not possible to assess whether missingness could have influenced the estimated thresholds.

The record identifies the analysis population but does not provide a more detailed analysis-population definition or explain how the pre-experiment screening criteria were applied among those enrolled. No assumptions should be made about the study’s prespecification, multiplicity control, interim analyses, or missing-data strategy.

11. 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 registry lists 19 participants at risk for both serious and other adverse events. It records zero serious adverse events, zero other adverse events, and zero deaths. The serious-event list contains no terms. These are the safety counts reported in the registry record.

In general, a count of zero means that no events of that category were recorded among the participants at risk in the reported data. It does not establish that an event could never occur, and it should not be read as a precise estimate of risk beyond the observed study. The record does not provide further detail about event definitions, how safety information was collected, or the period of safety follow-up.

This was a behavioral psychophysical study, and the registry describes no randomized treatment comparison. The safety counts should be read as descriptive counts for the study group, not as a comparison of risks between interventions. They are separate from the primary endpoint, which measures horizontal discrimination threshold. The registry record does not report numerical primary-outcome results, so the adverse-event figures cannot be used to infer perceptual performance or the effect of stimulus variability.

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

The study addresses how visual perception of a task-relevant object may be affected by variability in other features of a natural scene. Its task asks participants to judge whether a central object in a second interval is positioned to the left or right of the central object in a first interval. The primary measure, horizontal discrimination threshold, is defined as the minimum distance at which two stimuli are recognized as spatially separate. The registry’s interpretation is that a lower threshold corresponds to a smaller horizontal difference being perceived as distinct.

The design is a single-group behavioral study, and the listed primary analysis population consists of 10 naive participants who completed the study. The record describes the intended outcome summary and the study’s conceptual question. It does not report the actual threshold values, their full range, or results across stimulus-noise conditions. Thus, the registry record does not establish whether background variability was associated with higher or lower thresholds, how large any difference was, or how precisely it was estimated.

The record also does not provide sufficient information to characterize the specific condition contrasts or statistical analyses. Three analyses are noted, but their methods and results are not reported. Without those details, no conclusion can be drawn about statistical significance, the consistency of participant responses, or the robustness of any result.

In general, a study’s finding applies most directly to the participants, task, and stimulus setup that were studied. The registry describes typical adults and a natural-scene psychophysical task, but it does not establish that the same pattern would apply to different populations, visual tasks, or real-world settings. The findings also cannot be interpreted as evidence about a clinical intervention: the study is not a randomized treatment comparison.

The record supports an account of the research question, task, endpoint definition, and analysis population. It does not support a claim that the study found a particular effect. The absence of numerical results in the registry record is a substantive limit on interpretation, not evidence that there was no effect.

13. Why This Trial Matters Statistically

This study is statistically interesting because it turns a broad question about perception in complex environments into a measurable task. Participants judge the horizontal position of a central object while background features may vary. The discrimination threshold offers a defined outcome that connects task performance to the smallest spatial difference perceived as distinct. That makes the endpoint conceptually aligned with the question about whether irrelevant scene variability can affect task-relevant perception.

The design also illustrates why the experimental structure and analysis population matter. The record describes a single-group study rather than a randomized comparison between groups. The primary outcome analysis is described for 10 naive participants who completed the study, while enrollment is listed as 19. The record’s population description excludes one completed participant who was an experimenter, but does not fully explain the remaining participant flow. Such details influence how readers assess the scope and precision of the evidence.

Statistically, a median and full range give a descriptive view of threshold measurements, but they do not by themselves show uncertainty around a condition contrast or explain repeated observations. A complete interpretation would also need the actual estimates, the stimulus-condition definitions, and the analysis approach. Those numerical findings and methodological details are not reported in the registry record. This makes the record useful for understanding the question and planned measurement, while limiting evaluation of the empirical result.

The study’s broader statistical lesson is that a well-defined endpoint and a compelling hypothesis are not enough to determine what the data show. Readers need the observed values and a transparent account of how comparisons were made. Without them, the magnitude, direction, and precision of an association between background variability and discrimination remain unknown.

14. 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 important limitation for interpreting the registry record is that it does not report numerical primary-outcome results. The median threshold and full range are specified as the summary, but their values are absent. No condition-specific estimates, effect estimate, confidence interval, or statistical test result is reported. The central empirical question therefore cannot be answered from the record alone.

The statistical analysis is also not described in enough detail for independent assessment. The record notes three analyses but does not explain their purposes, the comparisons they make, or the methods used. The stimulus-noise conditions and the amount of within-trial variability are not specified in detail. It is also not reported whether participants contributed repeated observations to each condition or how any dependence among measurements was handled.

The difference between enrollment of 19 and the primary analysis population of 10 naive participants who completed the study is not fully accounted for. The record notes that one completed experimenter was excluded, but does not describe the status or reasons for exclusion or non-completion for the other participants. It does not report missing-data methods or the number of usable measurements for each analysis.

The study is a single-group psychophysical experiment, not a randomized comparison. The registry does not provide evidence for a treatment effect or a comparison with a separate control group. Its results, if available, would need to be interpreted in the context of the task and participants studied. Finally, the record does not report a protocol-based sample-size justification, interim analyses, or a detailed statistical plan. These omissions limit assessment of prespecification and precision, and should not be filled by assumptions.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Studying the Effects of Natural Visual Scene Changes on Typical Adult Visual Perception — NCT05004649.
• Registry record last updated 2023-05-10; status last verified 2023-05.
• Results first posted 2023-05-10.
• 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.

18. Record Summary

NCT05004649 is a completed, single-group behavioral study sponsored by the University of Pennsylvania. It investigates whether variability in background features of natural visual scenes affects horizontal discrimination of a central object. The task uses two intervals and asks participants to judge whether the second central object is to the left or right of the first.

The primary outcome is horizontal discrimination threshold in degrees of visual angle, summarized by the median and full range, with a timeframe of approximately 3 weeks. Enrollment is listed as 19. The primary outcome description identifies 10 naive subjects who completed the study and excludes one completed experimenter from analysis. The record lists zero serious adverse events, zero other adverse events, and zero deaths among 19 participants at risk.

Numerical primary-outcome results, analysis details, and the complete participant flow are not reported in the registry record. The record therefore describes the question, design, and endpoint, but does not establish the observed direction, magnitude, or precision of any relationship between scene variability and discrimination.