Short-form travel videos correlate with destination image and travel intention

A meta-analysis of nine independent samples involving 3,200 participants found that evaluations of short-form travel videos were positively associated with both travelers’ impressions of a destination and their stated intention to visit. Using random-effects two-stage meta-analytic structural equation modeling, the researchers estimated correlations of 0.514 between video characteristics and destination image, 0.492 between video characteristics and travel intention, and 0.541 between destination image and travel intention. The model found an indirect statistical association of 0.201 linking video characteristics to travel intention through destination image, with a 95% confidence interval from 0.133 to 0.290. A residual direct association between video characteristics and travel intention also remained, indicating that the image pathway did not account for the entire pooled relationship.

The synthesis, published in Frontiers in Communication, focuses on short-form destination videos circulated through platforms designed for rapid viewing and algorithmic recommendation. The analyzed characteristics were divided into three theoretical families: content, such as information quality, informativeness, entertainment and authenticity; source, such as creator credibility, expertise and trustworthiness; and media-interaction features, including interactivity, presence and immersion. However, seven of the nine strict-model samples used content measures, one used a source-related definition and one used an author-defined overall communication construct. The estimates therefore primarily summarize relationships involving video content rather than establishing equivalent patterns for every type of short-video characteristic.

Destination image served as the proposed intermediate variable in the analysis. The concept includes cognitive beliefs about a place, affective feelings about it, or broader holistic impressions, although the included studies did not always measure these components in the same way. Travel intention likewise covered several operational definitions, including willingness to visit, revisit, book or select a destination. That variation matters because revisit intention presupposes previous travel, while booking intention is closer to a commercial decision. Most importantly, the outcome was a stated intention rather than observed travel behavior, and all nine strict-model studies used cross-sectional self-report surveys.

The structural model separated the association from video characteristics to destination image, labeled a, from the association between destination image and travel intention after accounting for video characteristics, labeled b. The pooled estimates were 0.514 for the first path and 0.391 for the second. Once destination image was included in the model, the residual path from video characteristics to travel intention was 0.291, with a 95% confidence interval of 0.136 to 0.435. The indirect estimate of 0.201 and residual direct estimate of 0.291 summed to the total pooled association of 0.492. A likelihood-ratio comparison also favored retaining the residual path: fixing it to zero worsened model fit, with a change in chi-square of 12.10 on one degree of freedom and p below 0.001.

Although the pattern was consistent in several checks, it was not uniform across research settings. Heterogeneity was substantial, with I-squared values of 91.02% for destination image and travel intention, 92.56% for video characteristics and travel intention, and 93.01% for video characteristics and destination image. Approximate 95% prediction intervals were wide, ranging from 0.208 to 0.820 for the video–image correlation and from 0.286 to 0.796 for the image–intention correlation. This variation means that the average relationship may not transfer evenly to particular platforms, destinations, creator types or cultural settings. Too few samples occupied the major subgroup categories to support reliable inferential comparisons among them.

Sensitivity analyses produced broadly similar results. Adding a broader media-use exposure measure from a tenth peer-reviewed journal sample, for a total of 3,587 participants, yielded an indirect association of 0.188 with a 95% confidence interval of 0.131 to 0.264. Removing a sample whose matrix required reconstruction produced an indirect estimate of 0.206, with a confidence interval of 0.128 to 0.312; the primary estimate was 0.201 with a confidence interval of 0.133 to 0.290. Leave-one-out refits also remained positive, with the indirect estimate ranging from 0.186 to 0.218. These checks show that the principal pattern did not depend on a single included sample or on the expanded definition of exposure. They do not resolve the underlying differences among the studies.

Several features of the evidence limit causal interpretation. The researchers described the results as consistent with partial statistical mediation, not causal mediation. Adjusting for video characteristics within the image-to-intention path, and for destination image within the residual video-to-intention path, did not account for external variables such as prior familiarity, previous travel experience, involvement or pre-existing attitudes. Shared self-report methods may also inflate relationships when respondents evaluate videos, destinations and their own intentions in the same survey. Although one study performed a common-method-factor test and another reported a Harman single-factor result below its chosen threshold, the aggregate matrices could not distinguish method covariance from substantive association.

The review’s search and selection process also introduces potential coverage and publication-status biases. The researchers screened 469 records, excluded 369 at the title-and-abstract stage and assessed 49 full-text reports. Ten peer-reviewed journal reports entered at least one model, nine of them contributing independent samples to the strict analysis. Following a transparently reported postregistration amendment, 171 dissertations, theses, conference papers and proceedings were classified as publication-type exclusions, even though some had entered earlier exploratory analyses. The restriction created a consistent publication-status rule but narrowed the available evidence base. In addition, 51 of 100 reports sought for full-text review could not be retrieved, and several planned information sources were unavailable or not completed, leaving the possibility of availability and source-coverage bias.

For destination communication research, the results offer a more specific conclusion than the idea that popular videos automatically drive tourism. Positive evaluations of short-form travel-video content were associated with more favorable destination images and stronger stated travel intentions, and destination image accounted for part, but not all, of that relationship. The evidence does not establish that any particular platform, creator type, format or video feature performs best, nor does it show that changing a destination image would cause a change in travel behavior. It also cannot establish whether enjoyment, inspiration, usefulness, creator trust, parasocial interaction or other unmeasured processes produce the residual association. The most defensible takeaway is therefore statistical rather than prescriptive: across a small, content-dominated and geographically concentrated evidence base, destination image carried a measurable but incomplete share of the association between short-form travel-video characteristics and stated travel intention.

Subject of Research: Meta-analysis of short-form travel-video characteristics, destination image and travel intention

Article Title: Short-form travel video characteristics and travel intention: a meta-analytic structural equation modeling study of the indirect role of destination image

Article References: Fu, S., Liu, H., & Hashim, M. H. (2026). Short-form travel video characteristics and travel intention: a meta-analytic structural equation modeling study of the indirect role of destination image. Frontiers in Communication, 11, 1961179. https://doi.org/10.3389/fcomm.2026.1961179

Image Credits: AI Generated

DOI: 10.3389/fcomm.2026.1961179

Keywords: short-form video, travel intention, destination image, tourism marketing, meta-analysis, structural equation modeling, destination communication, cross-sectional research, statistical mediation, Short-form, travel, video