1[2 {"id":"C1","text":"The classical one-way ANOVA does not reject equality of the five weekday mean changes (F=0.263, p=0.90).","results_keys":["anova_f_stat","anova_p_value","n_analysed"],"status":"supported","caveat":"Equal-variance assumption violated (Levene); rank-based and robust checks agree."},3 {"id":"C2","text":"The Monday contrast is +0.05 bp with a 95% Welch CI of [-0.32, 0.42] bp, which contains zero.","results_keys":["effect_estimate","effect_ci_low","effect_ci_high","effect_se"],"status":"supported","caveat":"iid assumption; volatility clustering not modelled."},4 {"id":"C3","text":"Out of sample (2010-2019), weekday-specific means estimated on 2000-2009 do not improve on the zero-change forecast (RMSE 4.884 vs 4.876 bp).","results_keys":["model_metric","baseline_metric","model_minus_baseline_rmse","n_train","n_test"],"status":"supported","caveat":"No formal forecast-comparison test; difference is tiny."},5 {"id":"C4","text":"Overall conclusion: no detectable day-of-week effect on the mean daily change of the 10-year yield.","results_keys":["conclusion"],"status":"supported","caveat":"Effects below ~0.4 bp/day cannot be excluded."},6 {"id":"C5","text":"The null result holds under HC1, pooled, Kruskal-Wallis, post-holiday exclusion and per-period analyses.","results_keys":["robust_mon_hc1_ci","robust_mon_pooled_ci","robust_kruskal_p","robust_anova_p_excl_post_holiday","robust_anova_p_train","robust_anova_p_test"],"status":"supported","caveat":"Welch-ANOVA joint test not computed."},7 {"id":"C6","text":"Weekday variances of the daily change differ significantly, with Mondays least volatile.","results_keys":["diag_levene_p","diag_sd_by_weekday_bp"],"status":"supported","caveat":"Side observation; outside the mean-change estimand."}8]9