SSC Survey Results: Sexual Harassment Levels By Field

Original research filling a real gap: existing harassment surveys can't be compared (rates swing 12-85% on question wording; EEOC just counts raw claims), so Scott uses the 8,077-respondent SSC survey with consistent methodology. Findings: 21% of women / 6% of men harassed at work; at-work and out-of-work harassment correlate 0.81 across fields (implying it tracks the PEOPLE in a field, not field culture); STEM/nerdy/male-dominated fields show the LEAST harassment, balanced/verbal fields the most (preregistered: 12% of women in STEM vs 25% in non-STEM, p=0.002). Handled with exemplary care -- preregistration, the in-work/out-of-work control, a clever 'prefer not to answer' debiasing instruction, the honest 'my prior was confirmed, so be skeptical' caveat, and a call for replication with the raw data attached. Argues the 'techbro' media narrative may simply be wrong, like the immigrant-crime myth.

Quality 74 · Strong. Strong original, preregistered amateur social science on a charged topic where good research is nearly absent; the 0.81 people-not-culture correlation and the counter-narrative STEM-is-low result are genuinely interesting, and the methodological self-awareness is exemplary. Top of Strong, held below Excellent by the self-selected blog sample (which Scott stresses) and the single-dataset scope.

Claude’s paradigm shift 60 · Notable shift. Fresh data on a question with almost no comparable research, plus two counter-narrative findings; the analysis is original even if the methods are standard. Notable.

Real-world impact 2 · Minor. Contributed real data to a contested public debate; influence stayed within-discourse, no material change. Within-blog/community influence.