{
  "alt": "A dumbbell chart of thirteen institutions with two dots each: the in-person gap and the web gap, with intervals, around a line at zero. Science is at the top with the in-person dot well to the right of zero and the web dot farther right. The military, organized religion and the Supreme Court are on the Republican side of zero for both modes, and the intervals for banks, companies and labor include zero in person.",
  "caption": "The web gap is larger than the in-person gap for {fact:n_web_larger} of the thirteen institutions and the two point estimates have the same sign for {fact:n_same_sign}, and before the correction for thirteen comparisons science is the only difference whose interval excludes zero. Each row joins the in-person gap (ring) to the web gap (solid dot), and the right-hand column, headed Change, is the web gap minus the in-person gap. Intervals are wide for most rows, so the chart cannot rule out mode differences as large as the science difference for the other institutions. Rows are sorted as in the heatmap.",
  "data": [
    {
      "diff_hi": 24.7,
      "diff_lo": 4.2,
      "difference": 14.4592,
      "gap_inperson": 19.7951,
      "gap_web": 34.2543,
      "hi_inperson": 27.8,
      "hi_web": 40.7,
      "institution": "Scientific community",
      "lo_inperson": 11.8,
      "lo_web": 27.8,
      "n1": 1259,
      "n2": 1630
    },
    {
      "diff_hi": 9.4,
      "diff_lo": -4.2,
      "difference": 2.5901,
      "gap_inperson": 14.6455,
      "gap_web": 17.2356,
      "hi_inperson": 19.9,
      "hi_web": 21.5,
      "institution": "Executive branch",
      "lo_inperson": 9.4,
      "lo_web": 13,
      "n1": 1295,
      "n2": 1628
    },
    {
      "diff_hi": 6.4,
      "diff_lo": -10.1,
      "difference": -1.8181,
      "gap_inperson": 14.3855,
      "gap_web": 12.5674,
      "hi_inperson": 21.2,
      "hi_web": 17.2,
      "institution": "Education",
      "lo_inperson": 7.6,
      "lo_web": 8,
      "n1": 1305,
      "n2": 1622
    },
    {
      "diff_hi": 18.6,
      "diff_lo": -1.4,
      "difference": 8.5953,
      "gap_inperson": 10.4127,
      "gap_web": 19.008,
      "hi_inperson": 18.3,
      "hi_web": 25.2,
      "institution": "Medicine",
      "lo_inperson": 2.5,
      "lo_web": 12.8,
      "n1": 1309,
      "n2": 1626
    },
    {
      "diff_hi": 9.5,
      "diff_lo": -1.5,
      "difference": 3.9813,
      "gap_inperson": 7.8166,
      "gap_web": 11.7978,
      "hi_inperson": 11.8,
      "hi_web": 15.6,
      "institution": "Press",
      "lo_inperson": 3.8,
      "lo_web": 8,
      "n1": 1297,
      "n2": 1628
    },
    {
      "diff_hi": 2.5,
      "diff_lo": -8.5,
      "difference": -2.9942,
      "gap_inperson": 7.0623,
      "gap_web": 4.0681,
      "hi_inperson": 11.6,
      "hi_web": 7.2,
      "institution": "Television",
      "lo_inperson": 2.5,
      "lo_web": 1,
      "n1": 1291,
      "n2": 1628
    },
    {
      "diff_hi": 1.5,
      "diff_lo": -7.5,
      "difference": -2.9811,
      "gap_inperson": 5.0088,
      "gap_web": 2.0277,
      "hi_inperson": 8.8,
      "hi_web": 4.5,
      "institution": "Congress",
      "lo_inperson": 1.3,
      "lo_web": -0.4,
      "n1": 1296,
      "n2": 1629
    },
    {
      "diff_hi": 17.6,
      "diff_lo": -1.3,
      "difference": 8.1512,
      "gap_inperson": 0.8029,
      "gap_web": 8.9541,
      "hi_inperson": 8.6,
      "hi_web": 14.4,
      "institution": "Organized labor",
      "lo_inperson": -7,
      "lo_web": 3.5,
      "n1": 964,
      "n2": 1223
    },
    {
      "diff_hi": 9,
      "diff_lo": -7.7,
      "difference": 0.6277,
      "gap_inperson": 0.3472,
      "gap_web": 0.9748,
      "hi_inperson": 7,
      "hi_web": 6,
      "institution": "Banks and finance",
      "lo_inperson": -6.3,
      "lo_web": -4,
      "n1": 1307,
      "n2": 1632
    },
    {
      "diff_hi": 2.7,
      "diff_lo": -13.5,
      "difference": -5.4162,
      "gap_inperson": -0.8068,
      "gap_web": -6.223,
      "hi_inperson": 5.5,
      "hi_web": -1.2,
      "institution": "Major companies",
      "lo_inperson": -7.1,
      "lo_web": -11.2,
      "n1": 1265,
      "n2": 1629
    },
    {
      "diff_hi": 8.8,
      "diff_lo": -7.7,
      "difference": 0.5573,
      "gap_inperson": -11.6318,
      "gap_web": -11.0745,
      "hi_inperson": -4.8,
      "hi_web": -6.3,
      "institution": "Organized religion",
      "lo_inperson": -18.4,
      "lo_web": -15.8,
      "n1": 1264,
      "n2": 1622
    },
    {
      "diff_hi": 18,
      "diff_lo": -3.6,
      "difference": 7.2016,
      "gap_inperson": -13.7458,
      "gap_web": -6.5441,
      "hi_inperson": -5.4,
      "hi_web": 0.4,
      "institution": "Military",
      "lo_inperson": -22.1,
      "lo_web": -13.4,
      "n1": 1302,
      "n2": 1630
    },
    {
      "diff_hi": 3,
      "diff_lo": -14,
      "difference": -5.5183,
      "gap_inperson": -16.1518,
      "gap_web": -21.6701,
      "hi_inperson": -9.5,
      "hi_web": -16.4,
      "institution": "Supreme Court",
      "lo_inperson": -22.8,
      "lo_web": -27,
      "n1": 1289,
      "n2": 1632
    }
  ],
  "id": "gap_by_mode",
  "notes": [
    "Telephone and other-mode interviews of 2022 and 2024 are left out of this figure; they are included in the all-modes estimates elsewhere in the chapter. The interval of each web-minus-in-person difference is in the data table.",
    "Intervals assume a simple random sample of the weighted respondents (the Kish effective sample size); they do not use the survey's sampling design."
  ],
  "source": "General Social Survey, 1973-2024; weighted (weight_use); Democrats and Republicans include leaners (2022 and 2024 pooled); intervals on the Kish effective sample size",
  "spec": {
    "hi1": "hi_inperson",
    "hi2": "hi_web",
    "label1": "In person",
    "label2": "Web",
    "lo1": "lo_inperson",
    "lo2": "lo_web",
    "reference": 0,
    "sort": "data",
    "x1": "gap_inperson",
    "x2": "gap_web",
    "xDomain": [
      -40,
      50
    ],
    "xFormat": {
      "digits": 1,
      "signed": true
    },
    "xLabel": "Democratic minus Republican gap (percentage points)",
    "y": "institution"
  },
  "subtitle": "Democratic minus Republican share with a great deal of confidence, percentage points with 95% intervals, respondents interviewed in person and on the web, 2022 and 2024 pooled",
  "table": {
    "columns": [
      {
        "key": "institution",
        "label": "Institution"
      },
      {
        "format": {
          "digits": 1,
          "signed": true
        },
        "key": "gap_inperson",
        "label": "In-person gap (points)"
      },
      {
        "format": {
          "digits": 1,
          "signed": true
        },
        "key": "lo_inperson",
        "label": "In-person gap, lower end"
      },
      {
        "format": {
          "digits": 1,
          "signed": true
        },
        "key": "hi_inperson",
        "label": "In-person gap, upper end"
      },
      {
        "format": {
          "digits": 1,
          "signed": true
        },
        "key": "gap_web",
        "label": "Web gap (points)"
      },
      {
        "format": {
          "digits": 1,
          "signed": true
        },
        "key": "lo_web",
        "label": "Web gap, lower end"
      },
      {
        "format": {
          "digits": 1,
          "signed": true
        },
        "key": "hi_web",
        "label": "Web gap, upper end"
      },
      {
        "format": {
          "digits": 1,
          "signed": true
        },
        "key": "difference",
        "label": "Web minus in person (points)"
      },
      {
        "format": {
          "digits": 1,
          "signed": true
        },
        "key": "diff_lo",
        "label": "Difference, lower end"
      },
      {
        "format": {
          "digits": 1,
          "signed": true
        },
        "key": "diff_hi",
        "label": "Difference, upper end"
      },
      {
        "key": "n1",
        "label": "In-person respondents"
      },
      {
        "key": "n2",
        "label": "Web respondents"
      }
    ]
  },
  "title": "The web science gap differs from the in-person gap by {fact:diff_science} points, before correcting for thirteen comparisons",
  "type": "dumbbell",
  "width": "wide"
}