Problem 10.4. We continue with the situation in Problem 8.8. Assume that the two sample sizes are nį = 19 and n2 = 12 and the two sample variances are s = 0.81 and s = 0.49. Is there enough evidence...

Extracted text: Problem 10.4. We continue with the situation in Problem 8.8. Assume that the two sample sizes are nį = 19 and n2 = 12 and the two sample variances are s = 0.81 and s = 0.49. Is there enough evidence that fam- ilies from culled populations have a lower bunching intensity than families from non-culled populations? Use a test of hypothesis at level a = 0.005. Suppose that the two populations are normally distributed with equal vari- %3D ances.


Extracted text: Problem 8.8. Between 1967 and 1995, South Africa controlled its elephant populations through “culling", i.e. killing older animals. Scientists believe that in some populations, the surviving young elephants who experienced culling have symptoms similar to the post-traumatic stress disorder in hu- mans. The authors of article [60] investigated the effects of culling, using a variable called "bunching intensity", which gives the response to threat for a family of adult female elephants. This variable has values between 0 and 4, with 0 = "no response" and 4 = "very fast response". We consider two populations of elephants, one of which had experienced culling and the other had not. A sample of n1 families from the culled population has a mean bunching intensity of 1.2, whereas a sample of n2 families from the non-culled population has a mean bunching intensity of 2.5. The mean bunching intensity for the combined two samples is 1.7. What is the pro- portion of families who experienced culling in the combined two samples?