{
  "title": "Academic evidence for instructional video: attributed findings",
  "checkedAt": "2026-10-20",
  "scope": "Publicly accessible academic evidence only. Peer-reviewed papers and meta-analyses, with DOI and full bibliographic attribution. No vendor surveys or Demobro data.",
  "reuseNote": "Cite the original authors and preserve study context. This is a reference list, not participant-level data. Standardised effect sizes are not percentages or sales lifts. Rows with a calculation field are DemoBro arithmetic derived from published rounded group means, not new experiments or independent studies.",
  "highlights": [
    "mooc-observation-scale",
    "calculated-segment-recall-gap",
    "calculated-whole-lecture-gap"
  ],
  "sources": {
    "guo-2014": {
      "publisher": "ACM",
      "title": "How video production affects student engagement: An empirical study of MOOC videos",
      "reportYear": 2014,
      "observationPeriod": "Four edX courses first offered in autumn 2012",
      "authors": [
        "Guo, P. J.",
        "Kim, J.",
        "Rubin, R."
      ],
      "doi": "10.1145/2556325.2566239",
      "url": "https://doi.org/10.1145/2556325.2566239",
      "venue": "Proceedings of the First ACM Conference on Learning at Scale, pp. 41–50",
      "accessUrl": "https://pure.kaist.ac.kr/en/publications/how-video-production-affects-student-engagement-an-empirical-stud/",
      "methodology": "Peer-reviewed observational conference paper. 6.9 million viewing sessions across four courses; sessions are not unique learners. Engagement proxies do not establish causal effects or SaaS conversions."
    },
    "noetel-2021": {
      "publisher": "SAGE / AERA",
      "title": "Video Improves Learning in Higher Education: A Systematic Review",
      "reportYear": 2021,
      "observationPeriod": "Published synthesis of earlier randomised higher-education trials; not 2026 observations",
      "authors": [
        "Noetel, M.",
        "Griffith, S.",
        "Delaney, O.",
        "Sanders, T.",
        "Parker, P.",
        "del Pozo Cruz, B.",
        "Lonsdale, C."
      ],
      "doi": "10.3102/0034654321990713",
      "url": "https://doi.org/10.3102/0034654321990713",
      "venue": "Review of Educational Research, 91(2)",
      "accessUrl": "https://journals.sagepub.com/doi/abs/10.3102/0034654321990713",
      "methodology": "Peer-reviewed systematic review of 105 studies and 7,776 students; multilevel random-effects meta-analysis. Numerical claims here were verified in the public publisher abstract. Hedges’ g is a standardised effect, not a percentage or revenue uplift."
    },
    "rey-2019": {
      "publisher": "Springer Nature",
      "title": "A Meta-analysis of the Segmenting Effect",
      "reportYear": 2019,
      "observationPeriod": "Synthesis of earlier multimedia-learning investigations; published January 2019",
      "authors": [
        "Rey, G. D.",
        "Beege, M.",
        "Nebel, S.",
        "Wirzberger, M.",
        "Schmitt, T. H.",
        "Schneider, S."
      ],
      "doi": "10.1007/s10648-018-9456-4",
      "url": "https://doi.org/10.1007/s10648-018-9456-4",
      "venue": "Educational Psychology Review, 31, 389–419",
      "accessUrl": "https://maria-wirzberger.de/wp-content/uploads/2019/01/Rey2019_Article_AMeta-analysisOfTheSegmentingE.pdf",
      "methodology": "Peer-reviewed meta-analysis of 56 investigations and 88 pairwise comparisons. Table 3 reports retention d=0.32 (95% CI 0.20–0.43; k=67, n=6,100) and transfer d=0.36 (95% CI 0.24–0.48; k=56, n=4,754). Outcome-specific counts are not independent participant totals to add together."
    },
    "szpunar-2013": {
      "publisher": "National Academy of Sciences",
      "title": "Interpolated memory tests reduce mind wandering and improve learning of online lectures",
      "reportYear": 2013,
      "observationPeriod": "Two experiments reported in 2013; not a current industry survey",
      "authors": [
        "Szpunar, K. K.",
        "Khan, N. Y.",
        "Schacter, D. L."
      ],
      "doi": "10.1073/pnas.1221764110",
      "url": "https://doi.org/10.1073/pnas.1221764110",
      "venue": "Proceedings of the National Academy of Sciences, 110(16), 6313–6317",
      "accessUrl": "https://www.pnas.org/doi/10.1073/pnas.1221764110",
      "methodology": "Peer-reviewed experiment paper, 80 students across two experiments with a 21-minute statistics lecture. The cited combined analysis concerns final-test recall of the fourth lecture segment, not the entire course; repeated-test, restudy and nontested groups are distinct."
    }
  },
  "statistics": [
    {
      "id": "mooc-observation-scale",
      "value": 6900000,
      "displayValue": "6.9M",
      "unit": "viewing sessions",
      "audience": "Four edX MOOC courses",
      "metric": "Viewing sessions analysed; observational study scale",
      "sourceId": "guo-2014",
      "locator": "Abstract; methods"
    },
    {
      "id": "video-replacement-learning",
      "value": 0.28,
      "displayValue": "g = 0.28",
      "unit": "Hedges g",
      "audience": "Higher-education students; video replaces existing instruction",
      "metric": "Standardised learning effect when video replaces instruction",
      "sourceId": "noetel-2021",
      "locator": "Public abstract"
    },
    {
      "id": "video-addition-learning",
      "value": 0.8,
      "displayValue": "g = 0.80",
      "unit": "Hedges g",
      "audience": "Higher-education students; video supplements existing instruction",
      "metric": "Standardised learning effect when video supplements instruction",
      "sourceId": "noetel-2021",
      "locator": "Public abstract"
    },
    {
      "id": "segmentation-retention",
      "value": 0.32,
      "displayValue": "d = 0.32",
      "unit": "standardised effect size d",
      "audience": "Multimedia learning; retention outcomes",
      "metric": "Pooled retention effect for segmented instruction",
      "sourceId": "rey-2019",
      "locator": "Table 3; 95% CI [0.20, 0.43], k=67, n=6,100"
    },
    {
      "id": "segmentation-transfer",
      "value": 0.36,
      "displayValue": "d = 0.36",
      "unit": "standardised effect size d",
      "audience": "Multimedia learning; transfer outcomes",
      "metric": "Pooled transfer effect for segmented instruction",
      "sourceId": "rey-2019",
      "locator": "Table 3; 95% CI [0.24, 0.48], k=56, n=4,754"
    },
    {
      "id": "interpolated-testing",
      "value": 86,
      "displayValue": "86% vs 72%",
      "unit": "percent correct",
      "audience": "Final-test recall of fourth lecture segment; tested vs nontested groups",
      "metric": "Mean percentage correct in the combined two-experiment analysis",
      "sourceId": "szpunar-2013",
      "locator": "Discussion, combined final cumulative test analysis; restudy comparison is 66%"
    },
    {
      "id": "calculated-segment-recall-gap",
      "value": 14,
      "displayValue": "+14 pp",
      "unit": "percentage points",
      "audience": "Final-test recall of fourth lecture segment; tested vs nontested groups",
      "metric": "Calculated mean recall-score gap: 86% minus 72%",
      "sourceId": "szpunar-2013",
      "locator": "Discussion, combined final-test analysis of fourth lecture segment",
      "calculation": "86 - 72 = 14 percentage points; (86 - 72) / 72 × 100 = 19.4% relative mean-score difference (rounded)."
    },
    {
      "id": "calculated-whole-lecture-gap",
      "value": 22,
      "displayValue": "+22 pp",
      "unit": "percentage points",
      "audience": "Entire final cumulative test, all four lecture segments; tested vs nontested groups",
      "metric": "Calculated mean final-test score gap: 90% minus 68%",
      "sourceId": "szpunar-2013",
      "locator": "Discussion, combined entire final cumulative test across both experiments",
      "calculation": "90 - 68 = 22 percentage points; (90 - 68) / 68 × 100 = 32.4% relative mean-score difference (rounded)."
    }
  ]
}