Comment Re:Republicans? (Score 1) 899
I am in a school psychology graduate program and the issue of cultural biases on tests is discussed often. In the field of psychometrics, bias means that the test predicts performance on some criterion differently for one group than for another. If this happens then a test is statistically biased. All of the modern common cognitive tests currently in use are not statistically biased [Brown, R.T., Reynolds, C.R. and Whitaker, J.S. (1999). Bias in mental testing since Bias in mental testing. School psychology quarterly, 14(3), 208-238.]
While tests are not biased many of them are unfair in that they contain culturally loaded questions. Even with the unfairness they do not show statistical bias. It is a distinction that seems to be only semantics for those outside psychology, but they are really two different things. It is good that bias has been removed, but the current question is how to remove culturally loaded questions while still measuring the desired construct.
While tests are not biased many of them are unfair in that they contain culturally loaded questions. Even with the unfairness they do not show statistical bias. It is a distinction that seems to be only semantics for those outside psychology, but they are really two different things. It is good that bias has been removed, but the current question is how to remove culturally loaded questions while still measuring the desired construct.