#samplingerror — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #samplingerror, aggregated by home.social.
-
Every #Mastodon / #Fediverse poll that includes the text "boost for larger sample size" should be required to have "Dewey Defeats Truman" as one of the poll answers.
#peeve #PetPeeve #statistics #poll #sample #bias #DeweyDefeatsTruman #SamplingError #SamplingBias
-
Every single one of those suffers from severe self-selection biases - and even if a representative sample could be correctly taken, they're not actually closely related to how programming languages are matched to various problems.
It's lazy, click-baity #journalism at best, and #listicle fodder in actual fact most of the time.
[1] #CTOs may like to pick "popular" languages so they have more programmers to pick from, lowering what they need to #pay. See "Java", 1998-2015.
2/2
-
Every single one of those suffers from severe self-selection biases - and even if a representative sample could be correctly taken, they're not actually closely related to how programming languages are matched to various problems.
It's lazy, click-baity #journalism at best, and #listicle fodder in actual fact most of the time.
[1] #CTOs may like to pick "popular" languages so they have more programmers to pick from, lowering what they need to #pay. See "Java", 1998-2015.
2/2
-
Every single one of those suffers from severe self-selection biases - and even if a representative sample could be correctly taken, they're not actually closely related to how programming languages are matched to various problems.
It's lazy, click-baity #journalism at best, and #listicle fodder in actual fact most of the time.
[1] #CTOs may like to pick "popular" languages so they have more programmers to pick from, lowering what they need to #pay. See "Java", 1998-2015.
2/2
-
Every single one of those suffers from severe self-selection biases - and even if a representative sample could be correctly taken, they're not actually closely related to how programming languages are matched to various problems.
It's lazy, click-baity #journalism at best, and #listicle fodder in actual fact most of the time.
[1] #CTOs may like to pick "popular" languages so they have more programmers to pick from, lowering what they need to #pay. See "Java", 1998-2015.
2/2
-
Every single one of those suffers from severe self-selection biases - and even if a representative sample could be correctly taken, they're not actually closely related to how programming languages are matched to various problems.
It's lazy, click-baity #journalism at best, and #listicle fodder in actual fact most of the time.
[1] #CTOs may like to pick "popular" languages so they have more programmers to pick from, lowering what they need to #pay. See "Java", 1998-2015.
2/2