RE: STEDMAN POLL?

My good Dr. Otieno,

Good evening. I do not think the statistical population is the problem. 2000 is a very generous sample for a smallish voting block like that of Kenya (they use the same number, or less, in polling in the United States of America with a far much larger voting population). Maybe, you should have addressed the frequency with which the said polls were taken, and the political atmosphere during the polling period. I know you wont ask this: Given that we talk in English or Kiswahili in Kenya, but vote in LUO, Kikuyu, Kamba, ... is it not right that Pollster should have stated how many Luos, Kikuyu's etc participated in the poll? Claiming to have sampled Kamau in Kiambi is not enough; I know of My Cousins who are Kamaus but I am a Luo in every way, except the dentist in my village thought that, as a scholar, changing my dental formula could have interfered with my speech (sorry for this rare liberty of the tongue in the middle of a serious discussion)

Maybe, I am too old; but if I were running for ANY National Office in Kenya, and I saw this poll (irrespective of its source); I should be asking if political marriages between Warriors from various Houses have offered me any advantages?

I am not ruling out the manipulation of opinion polls. That happens all the time. But there is always some truth in a poll, however twisted it is.

Joseph Ramogi Alila
Author: SUNSET ON POLYGAMY (Tragedy: Cultural Practices and Disease Epidemic) _www.ramogialilablog at yahoo dot co dot uk_