Table 3. Bonferroni intervals to test for categories percentage cheetah prey with observed frequencies of livestock attacks by cheetahs that differed significantly from the expected (k = 4, α = 0.05, Z = 2.4977).

A chi square test was used to determine if the number of grid cells (N = 403) (expected values) and the number of livestock attacks by cheetahs (N = 188) recorded (observed values) differed significantly for the four categories of percentage cheetah prey (χ² = 52.42, df = 1, P < 0.001). From these results, 20% cheetah prey appears to be the critical point to differentiate between low and high probable levels of human-cheetah conflict.

Percentage cheetah prey	Expected proportion Pio	Observed proportion Pi	Bonferonni intervals for Pi	Use index Pi/Pio	Significant
0	0.106700	0.180851	0.1107 ≤Pi ≤0.2510	1.69	+
>0 to ≤20	0.476427	0.526596	0.4357 ≤Pi ≤0.6175	1.11	0
>20 to ≤80	0.079404	0.042553	0.0058 ≤Pi ≤0.0793	0.54	−
>80 to ≤100	0.337469	0.250000	0.1711 ≤Pi ≤0.3289	0.74	−