4 ms·

Just because a test might have some component you can vaguely articulate as having some cultural influence does not make the test itself invalid for comparison

• learc83 7y ago
  >does not make the test itself invalid for comparison across cultures If you can't accurately quantify the cultural component, then yes it does. >Claims that basic pattern recognition is somehow not a valid means of assessment across cultures are dubious at best. Basic pattern recognition isn't the only thing that's being measured--that's the problem. Even something as simple as how often a student has been exposed to a matrix of rows and columns has an impact on performance. Repeated exposure to similar tests also increases performance.
  • georgeburdell 7y ago
    If it’s the best thing we have, then it’s worth using. You are proposing no alternative.
    • learc83 7y ago
      If you have only one poll with an unknown margin of error and it shows Candidate A at 51 and Candidate B at 49, then assuming Candidate A is winning just because it's the best thing we have is worse than just admitting we don't know. Wrong information is often worse than no information. That's one of the problems with PISA results. Policy makers look at other countries with higher scores and try to emulate them without understanding why their scores are higher. In that case it's just cargo culting education. This doesn't mean that standardized testing is completely useless, but you have to know the limitations, and comparing results between cultures is one of the limitations.