Online crowdsourcing platforms like Witmart have become common tools for subject recruitment. This study examines whether samples drawn from these sites are representative of the general population or just internet users. By benchmarking Witmart data against national surveys, we find online respondents differ significantly from those in traditional polls but align closely with internet users specifically. Importantly, when weighted appropriately using post-stratification techniques, our findings show that crowdsourced samples can produce statistically similar coefficients to those of standard internet-based studies. This suggests these platforms may effectively represent Chinese netizens for researchers focused on digital populations. The study also investigates potential political and social biases in online responses, revealing notable desirability effects among subjects recruited through Witmart-like mechanisms.