Noah Carl found limited midwit support in GSS for declining liberalism at low IQ and conservatism at average IQ. GSS used low-quality WORDSUM and interviewer ratings. ANES vocab test and interviewer ratings correlate strongly at r=.69. Author and Emil found no midwit effect on government vs individual responsibility using ASVAB. NLSY79 showed no midwit on gender roles or politics except flat correlation past 115 IQ for conservative gender roles. Emil's datasets showed zilch for midwit. WLS Henmon-Nelson IQ found strong midwit effect with low and high IQ trending Democrat or left-wing. Midwit effect passes significance in WLS women strongly and men barely. Midwit specific to conservative self-ID, absent in Republican residuals independent of conservatism. No midwit for gender roles despite conservative ID effect. Raw results mostly reject midwit but WLS too strong to ignore. Author plans formal WLS publication. Libertarianism as high-IQ ideology explains no paradox in Republican residuals. Midwit used as heuristic for dumb ideas, not universal theory.