Improved version of the robust-beautiful post

Author: cyberosa

nbs/blog/posts/robust-beautiful/robust-beautiful.ipynb

@@ -71,11 +71,11 @@
     "the same when visualized with a barplot on the right panel. (You can\n",
     "download the [dataset](_static/four_samples.csv) to see for yourself.)\n",
     "\n",
-    "We're not the first ones (see\n",
-    "[this](https://www.nature.com/articles/nmeth.2837),\n",
-    "[this](http://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.1002128),\n",
+    "We're not the first ones (see these articles:\n",
+    "[article 1](https://www.nature.com/articles/nmeth.2837),\n",
+    "[article 2](http://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.1002128),\n",
     "or\n",
-    "[that](https://onlinelibrary.wiley.com/doi/full/10.1111/ejn.13400))\n",
+    "[article 3](https://onlinelibrary.wiley.com/doi/full/10.1111/ejn.13400))\n",
     "to point out the barplot's fatal flaws. Indeed, it is both sobering and\n",
     "fascinating to realise that the barplot is a [17th century\n",
     "invention](https://en.wikipedia.org/wiki/Bar_chart#History) initially\n",
@@ -118,7 +118,7 @@
     "The figure above visualizes the same four samples as a swarmplot (left\n",
     "panel) and as a boxplot. If we did not label the x-axis with the sample\n",
     "size, it would be impossible to definitively distinguish the sample with\n",
-    "5 obesrvations from the sample with 50.\n",
+    "5 observations from the sample with 50.\n",
     "\n",
     "Even if the world gets rid of barplots and boxplots, the problems\n",
     "plaguing statistical practices will remain unsolved. Null-hypothesis\n",
@@ -148,24 +148,21 @@
    "id": "a7e3b1ad",
    "metadata": {},
    "source": [
-    "hown above is a Gardner-Altman estimation plot. (The plot draws its name from\n",
-    "[Martin J. Gardner]\n",
-    "(https://www.independent.co.uk/news/people/obituary-professor-martin-gardner-1470261.html)\n",
+    "This is a *Gardner-Altman* estimation plot. The plot draws its name from\n",
+    "[Martin J. Gardner](https://www.independent.co.uk/news/people/obituary-professor-martin-gardner-1470261.html)\n",
     "and [Douglas Altman](https://www.bmj.com/content/361/bmj.k2588), who are\n",
-    "credited with [creating the design]\n",
-    "(https://www.bmj.com/content/bmj/292/6522/746.full.pdf) in 1986).\n",
+    "credited with [creating the design](https://www.bmj.com/content/bmj/292/6522/746.full.pdf) in 1986.\n",
     "\n",
     "This plot has two key features:\n",
     "\n",
-    "  1. It presents all datapoints as a *swarmplot*, which orders each point to\n",
-    "  display the underlying distribution.\n",
+    "  1. It presents all data points as a swarmplot, ordering each point to display the underlying distribution.\n",
     "\n",
     "  2. It presents the effect size as a *bootstrap 95% confidence interval* (95% CI)\n",
-    "  on a separate but aligned axes. where the effect size is displayed to the right\n",
-    "  of the war data, and the mean of the test group is aligned with the effect size.\n",
+    "  on a separate but aligned axis. The effect size is displayed to the right of the raw data, and the mean of the test group is aligned with the effect size.\"\n",
+    "\n",
+    "<div style=\"padding: 15px; border: 1px solid transparent; border-color: transparent; margin-bottom: 20px; border-radius: 4px; color: #31708f; background-color: #d9edf7; border-color: #bce8f1;\"> Thus, estimation plots are robust, beautiful, and convey important statistical\n",
+    "information elegantly and efficiently. </div>\n",
     "\n",
-    "*Thus, estimation plots are robust, beautiful, and convey important statistical\n",
-    "information elegantly and efficiently.*\n",
     "\n",
     "An estimation plot obtains and displays the 95% CI through nonparametric\n",
     "bootstrap resampling. This enables visualization of the confidence interval as\n",
@@ -283,13 +280,11 @@
    "id": "b7b643f8",
    "metadata": {},
    "source": [
-    "For comparisons between 3 or more groups that typically employ analysis\n",
+    "For comparisons between three or more groups that typically employ analysis\n",
     "of variance (ANOVA) methods, one can use the [Cumming estimation\n",
     "plot](https://en.wikipedia.org/wiki/Estimation_statistics#Cumming_plot),\n",
-    "named after [Geoff\n",
-    "Cumming](https://www.youtube.com/watch?v=nDN-hcKR7j8), and draws its\n",
-    "design heavily from his 2012 textbook [Understanding the New\n",
-    "Statistics](https://www.routledge.com/Understanding-The-New-Statistics-Effect-Sizes-Confidence-Intervals-and/Cumming/p/book/9780415879682).\n",
+    "named after [Geoff Cumming](https://www.youtube.com/watch?v=nDN-hcKR7j8), and draws its\n",
+    "design heavily from his 2012 textbook [\"Understanding the New Statistics\"](https://www.routledge.com/Understanding-The-New-Statistics-Effect-Sizes-Confidence-Intervals-and/Cumming/p/book/9780415879682).\n",
     "This estimation plot design can be considered a variant of the\n",
     "Gardner-Altman plot.\n"
    ]
@@ -307,8 +302,8 @@
    "id": "b443b0a8",
    "metadata": {},
    "source": [
-    "The effect size and 95% CIs are still plotted a separate axes, but\n",
-    "unlike the Gardner-Altman plot, this axes is positioned beneath the raw\n",
+    "The effect size and 95% CIs are still plotted on a separate axis, but\n",
+    "unlike the Gardner-Altman plot, this axis is positioned beneath the raw\n",
     "data.\n",
     "\n",
     "Such a design frees up visual space in the upper panel, allowing the\n",