Reply to thread

Message

<blockquote data-quote="Umbran" data-source="post: 9397753" data-attributes="member: 177">From several responses here, it seems like folks could use a bit more in the way of statistical literacy.In short: Statistically, that twitter poll is garbage, and should not be taken as evidence of anything.&nbsp; Same goes for other twitter polls in general, or polling threads here on EN World - they don't even reliably represent the opinion of EN Worlders.At slightly more length: Proper polling calls for a random sample of the population of interest.&nbsp; When you allow a broad population to self-select whether they will be respondents to the poll, the result isn't random.&nbsp; It is generally skewed in several ways.That people who aren't actually rpg fans might join in is probably the least concerning possibility.&nbsp; You'll gets a skew towards people with strong opinions, or who have a point to make or axe to grind, and an under-representation of people who feel less strongly.&nbsp; Then, there'll be bias coming from the network connection of people, and the kind of responses twitter tends to favor.&nbsp; Not to mention biases brought on by twitter's algorithms for whether or not they show you an item.And that's just a few obvious things.&nbsp; Overall, this is not how you determine how people broadly feel about something.&nbsp; Even calling it an &quot;informational&quot; poll is giving it too much credit.&nbsp; There is no reliable information there.</blockquote>

[QUOTE="Umbran, post: 9397753, member: 177"] From several responses here, it seems like folks could use a bit more in the way of statistical literacy. In short: Statistically, that twitter poll is [I]garbage[/I], and should not be taken as evidence of anything. Same goes for other twitter polls in general, or polling threads here on EN World - they don't even reliably represent the opinion of EN Worlders. At slightly more length: Proper polling calls for a [I]random[/I] sample of the population of interest. When you allow a broad population to self-select whether they will be respondents to the poll, the result [I]isn't random[/I]. It is generally skewed in several ways. That people who aren't actually rpg fans might join in is probably the [I]least[/I] concerning possibility. You'll gets a skew towards people with strong opinions, or who have a point to make or axe to grind, and an under-representation of people who feel less strongly. Then, there'll be bias coming from the network connection of people, and the kind of responses twitter tends to favor. Not to mention biases brought on by twitter's algorithms for whether or not they show you an item. And that's just a few obvious things. Overall, this is not how you determine how people broadly feel about something. Even calling it an "informational" poll is giving it too much credit. There is no reliable information there. [/QUOTE]

Verification