Regarding Infotrak opinion poll

Angela Ambitho (Infotrak) said that they conducted interviews between July 16 and July 22 (yesterday) in 100 constituencies in 31 counties.
My question to the statisticians here is whether it is possible to process and analyse your data and present it for public consumption the following day (today)?

Sample size?

she said 2000

This is what the good lady was saying on 19th July:

http://www.nation.co.ke/news/politics/Jubilee-Infotrak-o-release-fake-poll/1064-4021434-gjgrovz/index.html

Just read for yourself and make a decision.

Ok. Methodology of collecting data? Were questionnaires given to the sample size with questions to tick? Or someone came to people and asked them questions and tick himself?

With good machines and software, automated analysis can be done in seconds. Data collection and entry on the other hand…

This is how The Nation put it:

A sample of 2,000 respondents drawn from 31 counties and 100 constituencies took part in the Infotrak survey that had a 2.19 per cent margin of error.

Data, according to Ms Ambitho, was collected through household computer-assisted personal interviews.

In 2007 they used to call themselves Infotrak harris. Is it the company that was cited by Kriegler report for ‘cooking’ opinion polls?

The Infotrak poll is part of a well orchestrated plan to lay groundwork for rejection of election results. The Zogby poll was part of this plan. Release two polls days apart alluding to the same outcome. Thank God for Ipsos. Their poll results have been consistent.

2 Likes

Kwa hivyo ipsos ikitoa next poll and shows Raila leading/has closed gap it will be correct?

Another one for you guka Kipotato…
published in 2014
http://nairobiwire.com/2014/09/list-hated-women-kenya.html

Everything is computerized. It is possible to analyze and present results even in an hour.
I think there is a whole misunderstanding of polls of course poisoned by politicians. Polls are just meant to be a guide. And they are a useful tool. Without which candidates just guess their position. To us, they show the standing of your candidate. He may win or he may not. The poll doesn’t determine that. Its just a standing at a particular day which may change in the blink of an eye.
Right now, the polls just point to a very tight race and nothing more. They are not predicting an outcome on 8/8, even if taken on 8/7.
I see people here crying Angela Ambitho this, Ambitho that…blah blah…She didnt even carry out the poll. She just presented the results.

do you know anybody who has been questioned in those polls, and how do they get their phone numbers anyway!?

1 Like

Its strange how the results of those polls are fairly accurate. Think 2013 the polls had is 46%-42%. Final results 50%-44%. Fairly close. You see polls are conducted by private corporations. Their integrity is on the line. I haven’t heard any complaints about those companies. And to date, their results (predictions) have been very accurate.

Ambitho was supposed to release her polls days ago. They were to show Raila leading with even a bigger margin. Akina Itumbi and Mutahi made her shy away. Only an odd hour after Ipsos released their polls showing Kenyatta on the lead, she had no option but to unleash an edited version of her earlier polls.
[MEDIA=twitter]887354412305440768[/MEDIA]

1 Like

:D:D:D:D
I’m tempted to believe this,hehehe!
[MEDIA=twitter]889090500178956289[/MEDIA]

Watu wababaika.

stop making your own facts. In February 2013, Infotrak predicted that Raila 46% and Kenyatta 45%. They said their survey had been done between feb 24- feb 26 2012

3 Likes

A usual, you woke up with defensive factoids straight out of your exhaust pipe. In the 2016 US elections, practically ALL pollsters missed the mark, except the L.A. Times.

1 Like

I am giving Infotrak the benefit of the doubt, but I would want them to come out and explain.
What big thing has Raila done in the last 2 weeks to gain 4%?
What big thing has Uhuru done in the last 2 weeks to lose 5%?

This kind of inconsistency is too much to ignore.

3 Likes