{
  "creator": [
    "Höfler, Michael",
    "Pronizius, Ekaterina",
    "Buchanan, Erin"
  ],
  "date": [
    "2024-12-23"
  ],
  "description": [
    "An observational study might support a causal claim if the association found cannot be explained by bias due to unconsidered confounders. This bias depends on how strongly the common predisposition, a summary of unconsidered confounders, is related to the factor and the outcome. For a positive effect to be supported, the product of these two relations must be smaller than the left boundary of the confidence interval for, e.g., a standardised mean difference (d). We suggest means to derive heuristics for how large this product must be to serve as a confirmatory threshold. We also provide non-technical, visual means to express researchers’ assumptions on the two relations to assess whether a finding on d is explainable by omitted confounders. The ViSe tool, available as an R package and Shiny application, allows users to choose between various effect sizes and apply it to their own data or published summary results."
  ],
  "format": [
    "application/pdf",
    "text/html",
    "text/xml"
  ],
  "identifier": [
    "https://meth.psychopen.eu/index.php/meth/article/view/14579",
    "10.5964/meth.14579"
  ],
  "language": [
    "eng"
  ],
  "publisher": [
    "PsychOpen GOLD / Leibniz Institut for Psychology (ZPID)"
  ],
  "relation": [
    "https://meth.psychopen.eu/index.php/meth/article/view/14579/14579.pdf",
    "https://meth.psychopen.eu/index.php/meth/article/view/14579/14579.html",
    "https://meth.psychopen.eu/index.php/meth/article/view/14579/14579.xml"
  ],
  "rights": [
    "Copyright (c) 2024 Michael Höfler, Ekaterina Pronizius, Erin Buchanan",
    "https://creativecommons.org/licenses/by/4.0"
  ],
  "source": [
    "Methodology; Vol. 20 No. 4 (2024); 318-335",
    "1614-2241",
    "1614-1881",
    "10.5964/meth.v20i4"
  ],
  "subject": [
    "causality",
    "confirmation",
    "observational studies",
    "effect size",
    "visualisation",
    "software"
  ],
  "title": [
    "How Large Must an Associational Mean Difference Be to Support a Causal Effect?"
  ],
  "type": [
    "info:eu-repo/semantics/article",
    "info:eu-repo/semantics/publishedVersion",
    "Peer-reviewed article"
  ]
}