<?xml version="1.0" encoding="UTF-8"?><rss version="2.0" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>statistics &amp;mdash; leastaction</title>
    <link>https://write.as/leastaction/tag:statistics</link>
    <description></description>
    <pubDate>Mon, 24 Aug 2026 12:20:23 +0000</pubDate>
    <image>
      <url>https://i.snap.as/Sw9WqHns.jpg</url>
      <title>statistics &amp;mdash; leastaction</title>
      <link>https://write.as/leastaction/tag:statistics</link>
    </image>
    <item>
      <title>On polling and propaganda</title>
      <link>https://write.as/leastaction/on-polling-and-propaganda?pk_campaign=rss-feed</link>
      <description>&lt;![CDATA[For almost two years now polls in Canada have shown that the Conservative Party of Canada, led by Pierre Poilievre, Canada&#39;s answer to Donald Trump, have a commanding lead over the governing Liberal Party led until recently by Justin Trudeau. There is no doubt that this was a major factor leading to Trudeau&#39;s resignation, in the hope that the Liberal Party might recover lost ground under a new leader.&#xA;&#xA;At the same time, I can&#39;t help feeling that the polls have a lot of influence on how Canadians will vote.!--more--&#xA;&#xA;I wrote a thread on Mastodon about how this might work. The Conservative Party&#39;s messaging consists simply of repeating &#34;Justin Trudeau is a horrible person and he should resign&#34;. This is spread both by conventional media and corporate social media as news and commentary.&#xA;&#xA;Then one day you get a phone call from a polling company asking whether you agree that Justin Trudeau is a horrible person and he should resign&#34;. What are the odds that you will say yes, if you&#39;ve been exposed to the statement repeatedly?&#xA;&#xA;Polls are viewed as representing the most probable outcome of an election &#34;if it were held today&#34;. But that probability is only conditional. It is the probability that you will vote one way or another given that you&#39;ve been exposed to a certain propaganda. I am sure that if in a parallel universe where the Conservative Party of Canada has not been hammering the message about Trudeau for years a similar poll was conducted, the result would be quite different.&#xA;&#xA;But there&#39;s more. The poll itself has an influence on the election outcome. One reason is what is known as &#34;social proof&#34;. Gil Duran has a great post about this in The Framelab. He writes&#xA;&#xA;  People tend to copy what other people do. If something is popular, we tend to trust it more. When many other people buy a product or follow a trend, we tend to assume it’s good. Our brains are wired to look for clues from other people’s choices. It’s a basic human instinct to follow the crowd. &#34;Monkey see, monkey do.&#34;&#xA;&#xA;If you see that a majority of your fellow voters will vote Conservative, you are more likely to do so yourself. Marketers know this very well.&#xA;&#xA;The poll may also have another effect on the outcome of the election: If you are not a Conservative voter, but you are convinced that the Conservatives will win, that makes you less likely to take the trouble to vote. The poll becomes a self-fulfilling prophecy.&#xA;&#xA;I see propaganda as a process, where a certain political message is spread to as many people as possible and then polling is a feedback mechanism that confirms that the message is true. Polling skeptics will scrutinize the methodology and the statistical methods, and I have been one of them, but I believe now that that is missing the point. The polls can be quite accurate in that they reflect popular opinion, and yet be a key component of the propaganda process.&#xA;&#xA;#Democracy #Elections #Statistics #Propaganda]]&gt;</description>
      <content:encoded><![CDATA[<p>For almost two years now polls in Canada have shown that the Conservative Party of Canada, led by Pierre Poilievre, Canada&#39;s answer to Donald Trump, have a commanding lead over the governing Liberal Party led until recently by Justin Trudeau. There is no doubt that this was a major factor leading to Trudeau&#39;s resignation, in the hope that the Liberal Party might recover lost ground under a new leader.</p>

<p>At the same time, I can&#39;t help feeling that the polls have a lot of influence on how Canadians will vote.</p>

<p>I wrote a <a href="https://indieweb.social/@alfredo/113818517829931793">thread on Mastodon</a> about how this might work. The Conservative Party&#39;s messaging consists simply of repeating “Justin Trudeau is a horrible person and he should resign”. This is spread both by conventional media and corporate social media as news and commentary.</p>

<p>Then one day you get a phone call from a polling company asking whether you agree that Justin Trudeau is a horrible person and he should resign”. What are the odds that you will say yes, <em>if you&#39;ve been exposed to the statement repeatedly</em>?</p>

<p>Polls are viewed as representing the most probable outcome of an election “if it were held today”. <em>But that probability is only conditional</em>. It is the probability that you will vote one way or another <em>given that you&#39;ve been exposed to a certain propaganda</em>. I am sure that if in a parallel universe where the Conservative Party of Canada has not been hammering the message about Trudeau for years a similar poll was conducted, the result would be quite different.</p>

<p>But there&#39;s more. The poll itself has an influence on the election outcome. One reason is what is known as “social proof”. Gil Duran has a <a href="https://www.theframelab.org/90-percent-will-open-this-message-heres-why-social-proof/">great post about this in The Framelab</a>. He writes</p>

<blockquote><p>People tend to copy what other people do. If something is popular, we tend to trust it more. When many other people buy a product or follow a trend, we tend to assume it’s good. Our brains are wired to look for clues from other people’s choices. It’s a basic human instinct to follow the crowd. “Monkey see, monkey do.”</p></blockquote>

<p>If you see that a majority of your fellow voters will vote Conservative, you are more likely to do so yourself. Marketers know this very well.</p>

<p>The poll may also have another effect on the outcome of the election: If you are not a Conservative voter, but you are convinced that the Conservatives will win, that makes you less likely to take the trouble to vote. The poll becomes a self-fulfilling prophecy.</p>

<p>I see propaganda as a process, where a certain political message is spread to as many people as possible and then polling is a feedback mechanism that confirms that the message is true. Polling skeptics will scrutinize the methodology and the statistical methods, and I have been one of them, but I believe now that that is missing the point. The polls can be quite accurate in that they reflect popular opinion, and yet be a key component of the propaganda process.</p>

<p><a href="https://write.as/leastaction/tag:Democracy" class="hashtag"><span>#</span><span class="p-category">Democracy</span></a> <a href="https://write.as/leastaction/tag:Elections" class="hashtag"><span>#</span><span class="p-category">Elections</span></a> <a href="https://write.as/leastaction/tag:Statistics" class="hashtag"><span>#</span><span class="p-category">Statistics</span></a> <a href="https://write.as/leastaction/tag:Propaganda" class="hashtag"><span>#</span><span class="p-category">Propaganda</span></a></p>
]]></content:encoded>
      <guid>https://write.as/leastaction/on-polling-and-propaganda</guid>
      <pubDate>Mon, 13 Jan 2025 02:49:40 +0000</pubDate>
    </item>
  </channel>
</rss>