<p>
</p>
<p>you don’t understand my argument, it’s not based on the majority of students in a major it’s based on a higher proportion of students in a major. </p>
<p>Illustrative example:</p>
<p>College A is 15% Asian, major X in College A has a 100% graduation rate, and the school’s overall graduation rate is 90%.
If Major X is 30% Asian, instead of the proportional 15%, then Asians would graduate at a higher rate than average, even though the school is doing absolutely nothing to affect graduation rate of a certain race. </p>
<p>The race just happens to self-select into the easier major at a higher rate. Given that self-selection is quite random, races choosing certain majors might by itself explain why some colleges have higher or lower minority vs. overall graduation rates.</p>