| Region | States | Percentage owning some firearm |
|---|---|---|
| NEW_ENGLAND | ME VT NH MA CT RI | 29.41 |
| MIDDLE ATLANTIC | NY NJ PA | 27.72 |
| E. NOR. CENTRAL | WI IL IN MI OH | 42.19 |
| W. NOR. CENTRAL | MN IA MO ND SD NE KS | 44.59 |
| SOUTH ATLANTIC | DE MD WV VA NC SC GA FL DC | 45.36 |
| E. SOU. CENTRAL | KY TN AL MS | 61.83 |
| W. SOU. CENTRAL | AR OK LA TX | 53.33 |
| MOUNTAIN | MT ID WY NV UT CO AZ NM | 46.23 |
| PACIFIC | WA OR CA AK HA | 39.01
(Originally from GSS data 1988-1993) |
| Region | States | Percentage owning rifle |
|---|---|---|
| NEW_ENGLAND | ME VT NH MA CT RI | 15.95 |
| MIDDLE ATLANTIC | NY NJ PA | 17.28 |
| E. NOR. CENTRAL | WI IL IN MI OH | 24.97 |
| W. NOR. CENTRAL | MN IA MO ND SD NE KS | 32.32 |
| SOUTH ATLANTIC | DE MD WV VA NC SC GA FL DC | 23.85 |
| E. SOU. CENTRAL | KY TN AL MS | 34.74 |
| W. SOU. CENTRAL | AR OK LA TX | 31.13 |
| MOUNTAIN | MT ID WY NV UT CO AZ NM | 33.33 |
| PACIFIC | WA OR CA AK HA | 22.95
(Originally from GSS data 1988-1993) |
| Region | States | Percentage owning pistol |
|---|---|---|
| NEW_ENGLAND | ME VT NH MA CT RI | 13.85 |
| MIDDLE ATLANTIC | NY NJ PA | 12.48 |
| E. NOR. CENTRAL | WI IL IN MI OH | 21.14 |
| W. NOR. CENTRAL | MN IA MO ND SD NE KS | 18.69 |
| SOUTH ATLANTIC | DE MD WV VA NC SC GA FL DC | 27.84 |
| E. SOU. CENTRAL | KY TN AL MS | 38.37 |
| W. SOU. CENTRAL | AR OK LA TX | 32.98 |
| MOUNTAIN | MT ID WY NV UT CO AZ NM | 25.79 |
| PACIFIC | WA OR CA AK HA | 21.97
(Originally from GSS data 1988-1993) |
| Region | States | Percentage owning shotgun |
|---|---|---|
| NEW_ENGLAND | ME VT NH MA CT RI | 16.45 |
| MIDDLE ATLANTIC | NY NJ PA | 17.44 |
| E. NOR. CENTRAL | WI IL IN MI OH | 27.81 |
| W. NOR. CENTRAL | MN IA MO ND SD NE KS | 33.08 |
| SOUTH ATLANTIC | DE MD WV VA NC SC GA FL DC | 25.93 |
| E. SOU. CENTRAL | KY TN AL MS | 42.90 |
| W. SOU. CENTRAL | AR OK LA TX | 34.30 |
| MOUNTAIN | MT ID WY NV UT CO AZ NM | 25.79 |
| PACIFIC | WA OR CA AK HA | 18.69
(Originally from GSS data 1988-1993) |
y = 0.269795*x + 7.64939 varx 53.3817 se 12.9738 r 0.184419 2 sided test at 90.0% CI tc = 1.68957 at 35 d.f. beta in 0.269795 +- 0.41063 = [-0.140836,0.680425] test H0 beta == 0.000000 against H1 beta != 0.000000 t = 1.11009 at 35 d.f. |t| <= tc (1.68957); accept H0 test H0 beta == 0.000000 against H1 beta > 0.000000 t <= tc (1.30621); accept H0 P(beta!=0.000000) = 0.725471 P(beta>0.000000) = 0.862735 2 sided test at 90.0% CI tc = 1.68957 at 35 d.f. alpha in 7.64939 +- 18.1576 = [-10.5082,25.807] r2 = 0.0340109 tag x y yp SD 44.5902 3.66388 19.6796 MO 44.5902 33.3957 19.6796 MN 44.5902 17.2704 19.6796 KS 44.5902 9.08783 19.6796 IA 44.5902 5.04224 19.6796 VA 45.356 17.4408 19.8862 SC 45.356 13.4485 19.8862 NC 45.356 17.3617 19.8862 MD 45.356 45.8051 19.8862 GA 45.356 31.719 19.8862 FL 45.356 15.6488 19.8862 DC 45.356 66.2148 19.8862* WA 39.0144 11.2067 18.1753 OR 39.0144 8.5425 18.1753 CA 39.0144 14.5853 18.1753 AK 39.0144 10.0533 18.1753 RI 29.4118 14.9364 15.5845 MA 29.4118 13.5108 15.5845 CT 29.4118 21.3638 15.5845 UT 46.2312 13.6748 20.1223 NV 46.2312 23.0473 20.1223 NM 46.2312 11.8934 20.1223 ID 46.2312 2.39747 20.1223 CO 46.2312 9.86592 20.1223 AZ 46.2312 13.3307 20.1223 PA 27.7207 23.6066 15.1283 NY 27.7207 18.4852 15.1283 NJ 27.7207 21.6628 15.1283 TN 61.8321 21.939 24.3314 MS 61.8321 47.3855 24.3314 KY 61.8321 13.6833 24.3314 AL 61.8321 27.4608 24.3314 WI 42.1875 16.0772 19.0314 OH 42.1875 16.6971 19.0314 MI 42.1875 31.3097 19.0314 IN 42.1875 4.63237 19.0314 IL 42.1875 28.2103 19.0314 Spearman corr = 0.092461 H0: vars are independent Critical r at 5% = 0.306000; accept H0Note that Washington, DC, is a 2-sigma outlier for its region in terms of homicide rates. The OLS shows that at up to 86% confidence, the beta found is > 0. Although the Spearman coeff shows there is little correlation between ownership and homicide rankings, none was expected as the region ownership rates were applied to each state in that region. (A later study will cull the GSS data for state-by-state ownership rates). Note, also, the "explanation power" -- the r^2 -- is also small. In this OLS the "ownership" variate explains less than 3% of the state-by-state variation in homicide rates. But there is, nevertheless, "some" statistical evidence that ownership rates and homicide rates are linked (this could be for several reasons). For each 1% of households reporting they own at least 1 firearm, there is an associated additional 0.27 homicides per 100K population.
y = 0.147781*x + 12.5426 varx 28.2928 se 4.44047 r 0.358806 2 sided test at 90.0% CI tc = 1.94317 at 6 d.f. beta in 0.147781 +- 0.304975 = [-0.157194,0.452756] test H0 beta == 0.000000 against H1 beta != 0.000000 t = 0.941595 at 6 d.f. |t| <= tc (1.94317); accept H0 test H0 beta == 0.000000 against H1 beta > 0.000000 t <= tc (1.43975); accept H0 P(beta!=0.000000) = 0.617252 P(beta>0.000000) = 0.808626 2 sided test at 90.0% CI tc = 1.94317 at 6 d.f. alpha in 12.5426 +- 13.18 = [-0.637336,25.7226] r2 = 0.128742 tag x y yp W_NOR_CENTRAL 44.5902 16.3408 19.1322 SOUTH_ATLANTIC 45.356 24.7476 19.2454 PACIFIC 39.0144 14.0431 18.3082 NEW_ENGLAND 29.4118 16.5275 16.8891 MOUNTAIN 46.2312 13.129 19.3747 MIDDLE_ATLANTIC 27.7207 19.656 16.6392 E_SOU_CENTRAL 61.8321 23.7173 21.6802 E_NOR_CENTRAL 42.1875 21.885 18.7771 Spearman corr = 0.166667 H0: vars are independent Critical r at 5% = 0.643000; accept H0The OLS shows we would accept the hypothesis that beta > 0 at up to 80% confidence. The Spearman, OTOH, shows no significant correlation at 95% confidence. The r^2 shows the "ownership" variate accounts for somewhat more than 10% of variation in region-to-region homicide rates. (A small r^2 is consistent with a small Spearman corr). We have, however, "some" statistical reason (i.e. the T-test on the OLS beta) to accept there is a linkage between ownership report rates and homicide rates. This could be for several reasons. For each 1% of households reporting they own at least 1 firearm, there's an additional 0.15 homicides per 100K population.