|
Post by kamalet on Jan 14, 2013 12:47:05 GMT 3
The following is a response by Infotrak to allegations of favoritism issued in a letter to newsrooms by its CEO and Founder. Dear esteemed Presidential Candidate, With approximately 50 days to the General Election, I write to share my two cents as a political pollster. As you may be aware, we recently released our maiden popularity poll of the year on Friday. Whilst some of you may be over the moon with the findings, I know that many of you are probably seething in anger at the fact that the numbers did not treat you too favourably. Please believe that I empathise with you fully because like you, I hate to lose. However, I think it imperative that you all keep your eyes on the ball and appreciate that winning requires strategy and strategy is ineffective void of the pertinent tools. Allow me to limit myself to the tool I know best which is research or poll statistics. It is often quipped that numbers are tools not rules and the skills in interpreting them require talent, not evidence in divine intervention. Indeed, those of you who have used numbers as tools tend to excel in their political endeavours. As your campaigns hot up and pollsters continue to release opinion polls my first prayer is that you will desist from attacking the numbers and instead deem them as facts that seldom lie. I hope you will not belabour on sideshows and awful propaganda aimed at shooting the messenger simply because you don’t like the message. I urge you to remember that the poll “message” emanates from the very electorate you are so keen to govern. In shunning their views you are literally saying you are not ready to listen to the people and are as such inept to lead them. Because the numbers are tools, they are not an end unto themselves but rather a useful means to the end. That being the case, I implore you to understand that the statistics we release periodically are not static; they operate in a continuum and are liable to changes depending all your efforts. Those of you who use the numbers to address your weaknesses and bolster your strengths often see positive traction in the numbers. However, those of you who tend to listen to your sycophants who advise you to ignore the opinion polls as incorrect often find yourselves overtaken by events. Thirdly, it may be prudent to look at best practice on the optimisation of opinion polls. It is my belief President Barack Obama and his team perhaps possess the most superior skills in bringing numbers to life. Studying how they were able to recapture the presidency might provide useful tips on how to garner a first round win in the elections. My mentor and friend, the Harris poll chairman states: you need to be alive to the fact that many things are important to ensure free and fair elections. These include but are not limited to: freedom of speech and freedom to run campaigns; freedom of the Press; impartiality of the media to all candidates; government non-interference in media affairs; transparent electoral processes including registering of voters, announcing election dates; transparent vote tallying etc. Because it is unlikely all of these criteria are fully met, we need to add a ninth: that there be reliable, independent pre-election and exit polls. It is not lost to me that opinion polls are by no means infallible. However, they are the only reliable way to measure public opinion. Knowledge of public opinion, whether well-used or misused, is far better than ignorance, and those who attack, censor, corrupt or intimidate the polls are enemies of democracy. In the coming weeks, my final prayer is that you will have confidence in the electorate’s ability to decipher their own opinion. I hope you will not insinuate that knowledge of public perception instigates violence; I trust you will not try to manipulate your voters into thinking that ignorance is bliss. As former British PM James Callaghan once wrote “If you cannot trust the public with polls, you should not trust them with the vote”. May the best team win! The writer is founder and CEO of Infotrak Research and Consulting ....terrible commedy if the intention was that we laugh!
|
|
|
Post by Omwenga on Jan 14, 2013 15:47:14 GMT 3
The Standard, www.standardmedia.co.ke/?articleID=2000075030&pageNo=1&story_title=Kenya-Cord-denies-swaying-opinion-pollsNAIROBI,KENYA: Cord presidential hopeful Prime Minister Raila Odinga has refuted claims that his party has been swaying opinion polls in their favor. The PM denied claims by some of their opponents that the Cord alliance was behind last week’s Infotrak opinion polls which placed his candidature at 51 per cent against Jubilee’s 39 per cent as Musalia Mudavadi’s Amani coalition was rated at 3 per cent. Raila stated that Cord does not control opinion polls. Speaking at the Orange house as he received Richard Onyonka and Joseph Lekuton who formally joined with the orange party, Raila said that added that he agrees with Infotrak’s findings. “People are now being asked whom they are voting for and those were findings of a scientific research and we have no reason at all to manufacture them,” Raila said adding that Cord was a national alliance and not a tribal grouping. Jubilee presidential nominee Uhuru Kenyatta and his running mate William Ruto have since dismissed the findings saying the true opinion poll will be proven by Kenyans on election day. Raila at the same time predicted a landslide victory for his side in the March 4 general election. The PM said he is confident the Cord alliance will trounce their rivals by securing an emphatic 75 per cent victory in the first round of the polls. He said in spite the jubilee coalition distributing what he termed the as misleading tribal figures to show that there strongholds hold number, the cord is more popular across the country. “Its now coming out clear who the winners will be. After the launch of Cord in Uhuru Park and subsequent rallies in western the country is ready for a cord victory. I am seeing nothing less than 75 percent,” the PM said. Raila’s comments were directed at some data of voters being circulated in the country to suggest that the Gema and Kalenjin communities total to about 6.4 million of the 14 million registered voters. The Cord leader said that he was not a tribal leader and that he doesn’t consider himself a regional kingpin having represented a cosmopolitan constituency for the 20 years. ODM chairman Henry Kosgey said the contest for leadership should not be tribal competition but competitions of ideas and policies. Onyonka said although he had worked close with ODM for the last five years, he was now joining the Orange outfit.
|
|
|
Post by mwalimumkuu on Jan 14, 2013 15:56:15 GMT 3
Ambitho, if she cares about her reputation and that of her company need not engage in such cheap jabs but use such opportunities to address herself to moot issues and questions being raised about her firm and these polls.
|
|
|
Post by nok on Jan 14, 2013 15:59:04 GMT 3
As they say " if you can't stand the heat leave the kitchen". If you can't stand the polls, quit
|
|
|
Post by mwalimumkuu on Jan 14, 2013 16:20:13 GMT 3
As they say " if you can't stand the heat leave the kitchen". If you can't stand the polls, quit Why quit when the real poll that will prove these misguided elements wrong is coming? We have our eyes firmly on the prize, no amount of destruction will change this. Hesabu imefanywa kila kitu kiko sawa, this will be a blowout for jubilee, watch this space.
|
|
|
Post by Luol Deng on Jan 14, 2013 17:16:55 GMT 3
mwalimumkuu,
What issues have been raised? She is Luo, so should she change her ethnicity to please your likes. Infotrack released their numbers & the factual basis behind it. The fact that you have conveniently overlooked that and gone for non issues just shows how shallow you are. Having a comment every now and then on all matters political doesn't mean that you have anything meaningful to say.
PS. Can you please give us the basis of the jubilee projections?
|
|
|
Post by mwalimumkuu on Jan 14, 2013 17:38:16 GMT 3
mwalimumkuu, What issues have been raised? She is Luo, so should she change her ethnicity to please your likes. Infotrack released their numbers & the factual basis behind it. The fact that you have conveniently overlooked that and gone for non issues just shows how shallow you are. Having a comment every now and then on all matters political doesn't mean that you have anything meaningful to say. PS. Can you please give us the basis of the jubilee projections? Mos mos nani. Will you please show me where I have said that the lady is Luo and therefore needs to change that. If you go back to my contribution to this issue you will see the specific issues I raised about this poll on day one. Stop confusing this board, it has a memory of its own. Kenya as a country has a history in matters such as this. There is nothing so suggesting that this has changed. That is why for instance, no one is wasting their time and money campaigning in Luo Nyanza.
|
|
|
Post by phil on Jan 14, 2013 17:38:53 GMT 3
|
|
|
Post by mwalimumkuu on Jan 14, 2013 17:41:20 GMT 3
Ahaa!! Very telling. My fears confirmed.
|
|
|
Post by phil on Jan 14, 2013 17:58:26 GMT 3
Can we publish a similar management structure for Synovate - which incidentally published similar ratings for presidential candidates?
Mwalimumkuu, calm down and start debating issues instead of desperate moves to try to be seen to having proven something while in essence you are just defacating on threads with nonsense.
Uhuru, Duale and Ruto were categorical that Raila is related to the top management (all luos) of the Strategic Research. This document goes to prove that they were wrong. I need not tell you how much damage these baseless allegation have impacted on the corporate image perhaps even their firms bottom line. This is company that employs many young Kenyans. Does Jubilee want to create employment for youth or not?
Here is the company trying to discount that Saturday and Sunday Jubilee rally allegations and as it has done for all the other insinuation about cooking numbers.
And here is Angela Ambitho, the firms CEO on Kenyatta-24 where again she proves that Jubilee Coalition do not have the numbers to win this election. Again all the issues raised in this board and at the rallies are adequately addressed.
|
|
|
Post by nowayhaha on Jan 14, 2013 23:44:57 GMT 3
Polling firms on the spot over findings www.nation.co.ke/News/politics/Polling-firms--on-the-spot-over-findings/-/1064/1620362/-/15kusra/-/index.htmlA team set up to monitor the media has criticised research companies for conducting what it termed as shallow and unscientific opinion polls on Kenya’s elections. In its weekly briefing, the National Steering Committee on Media Monitoring said the pollsters were being used by some unnamed politicians. Ms Mary Ombara, the committee secretary, specifically took issues with a recent opinion poll by Infotrak, which she claimed had only served to belittle certain presidential aspirants by only releasing figures on the popularity of certain candidates. Infotrack’s response “The design and analysis of the polling survey tools fell short of a scientific exercise. We think that the polls propositions... were an attempt at make-believe and push for non-existent power axes,” she said in Nairobi. They argued that by releasing a poll only centred on a contest between Prime Minister Raila Odinga and Deputy Prime Minister Uhuru Kenyatta, the pollster had made the two feel “they have an upper hand” and presented the rest of the contenders as losers. When contacted, Infotrak chief executive Angela Ambitho accused the team of targeting her firm when other research firms had issued “curious” findings in the past. “Why wasn’t it an issue when it was revealed that Uhuru was catching up with Raila yet there was no explanation for that sudden rise? she posed.
|
|
|
Post by nowayhaha on Jan 14, 2013 23:59:38 GMT 3
www.the-star.co.ke/news/article-89629/opinion-polls-fooling-some-people-all-timeOpinion polls: Fooling some of the people all of the time It was the American novelist Mark Twain who once wrote “There are three kinds of lies: Lies, damned lies, and statistics”. And Abraham Lincoln, the 16th President of the United States, once said: “You can fool some of the people all of the time, and all of the people some of the time, but you cannot fool all of the people all of the time”. There are messages and lessons in these great quotes for Kenya’s present-day, deeply flawed opinion-polling sector, particularly for the firm called Infotrak Harris. It is high time Kenyan opinion pollsters were professionally and publicly vetted and certified. Indeed, the process of professionalizing opinion pollsters would flush out all manner of quacks and would-be fortune tellers. Secondly, no political opinion poll should be published and disseminated to the public through the media of mass communication, or even directly through the Internet or social media networks, without first being required to make full disclosure of its financiers and the political leanings and, or affiliations of the firm’s Board and senior polling personnel. If these parameters and paradigms were scrupulously observed, Kenyans would be shocked at what hidden factors would rise to the surface like scum in a tank of dirty water. These eye-opening factors would explain how one presidential candidate has been artificially kept at the top of the contenders’ list for almost five years straight, through thick and thin. For instance, Infotrak Harris should always disclose that it is chaired by Mr. Jerry Okingu, one of the most diehard Odingaists who has ever lived. The CEO is Ms. Angela Ambitho, another great Odinga serial booster. Despite hemorrhaging key allies and strategic supporters, some of them in command of large swing-vote ethnic blocs, Prime Minister Raila Odinga is constantly portrayed by Infotrak Harris as enjoying a commanding lead over all other contenders, including Vice President Kalonzo Musyoka Deputy Prime Ministers Uhuru Kenyatta and Musalia Mudavadi. In spite of suffering a massive loss of face and credibility – including among the Diplomatic Corps and the Diaspora – following the publication of his estranged former aide Miguna Miguna’s book Peeling Back the Mask, A Quest for Justice in Kenya, Raila is still portrayed as the man to beat at the next presidential poll by the rest of the field by opinion pollsters. The diplomatic cables sent to Washington, London, Paris, Bonn, Tokyo and other world capitals following the launch of the Miguna expose will doubtless make very interesting reading one fine day when they reach the public domain. The Orange Democratic Movement (ODM), despite having imploded and lost statesmen from key regions like the Rift Valley, Western and Coast, is still portrayed as Kenyans’ political party of choice. The portrayal of such key operatives as the Vice President and Eldoret North MP William Ruto as being of barely any consequence in the estimate of the Kenyan public is such a barefaced lie that it is almost sick-making. The crowds of ordinary Kenyans who welcome the VP wherever he goes in Kenya – with the exception of Luo Nyanza for reasons we shall not linger over in this analysis – are proof positive that he rates much more than the insulting sub-10% permanently reserved for him by patently fraudulent opinion pollsters. Ditto Ruto, the undisputed king of the disposition of the Rift Valley vote bloc, the self-described political “hustler” who has totally diminished the Daniel and Gideon arap Moi factor among the Kalenjin communities, to say nothing of his demolition in that region of the Raila factor.The fool-the-people opinion pollsters have made Raila break opinion-polling world records – where are those Guinness Book of World Records chaps? They should be told of this politician who leads the polls for five years running! These fraudsters long ago chose DPM Uhuru to be Raila’s permanent but distant Number Two challenger for a very good reason: the reactivation, at a strategically appointed time, of the 41 against 1 genocidal ideology agenda of the 2007 elections. The tragedy of defective, even fatally flawed, opinion polling in Kenya is that it conceals the real picture, expels the truth and sells and imposes a poisoned political fiction. It interferes with Kenyans’ own deepest convictions and self-knowledge. When you know that you will vote for the VP based on his track record of professionalism, diplomacy and peacemaking, but you are constantly told by apparently professional pollsters that you belong to a minority of less than 10% likeminded Kenyans, you might begin to doubt your choice.This is psychological advance rigging of the next presidential poll in Kenya. The narrative of the invincibility of the Odinga political brand consistently imposed by politicized and hopelessly partisan opinion pollsters must be explicitly challenged, once and for all, ahead of the next General Election, Kenya’s biggest and most expensive electoral event ever. If the opinion-polling sector will not clean up its act, professionalize as well as make full disclosure not only of its methods but also of its funding, ideologies and strategic political alliances, then it must be forced to do so by Kenyans themselves. Otherwise we are being fed pure political poison and confusion. These mind games must cease. Kenyans must be left alone to make enlightened and well-informed choices. The truth of the matter is that the Odinga political brand counts for nothing in vast swathes of Kenya outside of Luo Nyanza and Nairobi’s sprawling Kibera slum. This is a fact you will never see reflected by an opinion-polling firm chaired by Mr. Okungu and fronted by Ms. Ambitho! Instead, you will hear only of how the VP brings up the rear and the DPMs struggle far behind Raila and Ruto is of no consequence. The latest numbers from the Okungu-Ambitho opinion-polling operation purporting to show how Raila would fare in a runoff contest with each one of the big contenders, including the VP, contrives to imply a 50-50 dead-heat result if the PM were to run against Uhuru. This is pesa nane deception of the worst kind, a hoodwinking of village idiots. The only reason the Odinga strategists want a Kenyatta for a runoff foe is that they hope to play the anti-Kikuyu card by pointing out emotively, at the last minute, that it isn’t exactly equitable for Uhuru, a Kikuyu to succeed President Mwai Kibaki , another Kikuyu. This is what almost brought Kenya onto its knees after 2007 General election The Odinga clique has no deadlier weapon in its electoral arsenal than this emotional anti-Kikuyu card. This is the device through which they contrive to keep genuine contenders like the VP permanently in the shadows, while promoting the foes of their choice. And then there is the question of professionalism and methods. How many Kenyan opinion polling firms enjoy world-class sampling, data collection and data analytic capacity for survey research? How many reputable research organizations, including university departments, go to Kenyan pollsters for their sampling and data collection needs? How many manufacturers and other corporates are professionally and satisfactorily served by Kenya’s pollster sector? Is there a professional body that vets and certifies Kenyan pollsters’ methodological excellence and sampling expertise? How many of them employ the services of sampling scientists? Do they have multimedia monitoring capacity in an era of digital information communications technologies (ICTs) and social networks? All these questions require urgent and comprehensive answers. A probe of Kenya’s opinion polling, data collection and data analysis sector might reveal a snake-pit of non-professionalism and quackery that would shock even laymen. Indeed, it might turn out to be the case that some so-called research firms exist only to periodically “survey” the race to State House and to position some candidates and their political brands ahead of all others. Going forward, I believe I can without fear of contradiction state categorically that the results of the next presidential contest in Kenya, whether or not it goes into a runoff, are going to be anything but what the highly partisan Infotrak keeps harping on. The margin of error will be breathtaking and the Fourth President of Kenya will be a person whom Infotrak has belittled and ignored for almost five years. Why do I say this? Quite simply because, to all intents and purposes, in all probability, no truly professional and impartial opinion poll on the race for State House has yet been conducted in Kenya. All opinion polls so far have had a stealth agenda and a political axe to grind. What’s more, they have been mishandled by non-professionals and non-scientists. One of the most succinct and to-the-point definitions of opinion polling is to be found in Wikipedia: “An opinion poll, sometimes simply referred to as a poll, is a survey of public opinion from a particular sample. Opinion polls are usually designed to represent the opinions of a population by conducting a series of questions and then extrapolating generalities in ratio or within confidence intervals”. The process of extrapolating generalities in Kenyan opinion polling very clearly has no ratio and nor does it fall within any empirical confidence levels. Instead, it is weighted to perpetrate lies, damned lies and tainted statistics. The author is the Media and Communications Advisor to the Vice President. The views expressed here-in are his own.
|
|
|
Post by reporter911 on Jan 15, 2013 0:46:08 GMT 3
Jubilee poised for poll victory, say Uhuru and Ruto Jubilee presidential candidate Uhuru Kenyatta and his running mate William Ruto on Saturday expressed confidence their alliance will win the elections. Mr Kenyatta and Mr Ruto said they had put in place a clear plan to clinch victory. The leaders dismissed the latest opinion poll that projected an election win for their rival, the Coalition for Reforms and Democracy (Cord). The DPM said that contrary to the results released by the pollsters his coalition was also popular in Nairobi and urged his supporters to turn out in large number to disapprove the polls. “If they are saying we do not have the numbers then what will they say when they see this huge crowd here,” Mr Kenyatta told his supporters during a rally at Uhuru Parkl, Nairobi Saturday. Mr Ruto claimed that top managers at Infotrak Research and Consulting that conducted the poll were related to Cord presidential candidate Raila Odinga. ‘On March 4 we are asking that if they lose elections, that if Kenyans decide that it is Jubilee they should be ready to accept defeat,” Mr Ruto said. elections.nation.co.ke/news/-/1631868/1663948/-/format/xhtml/-/pww0u2/-/index.html ;D ;D desperation has many faces... yarn!!! Uhuruto are going to disappear in central province.. curtsy of robotic voters... they are on the way to oblivious.. 'CORD TSUNAMI IS RIDING THE WAVE.. Kenyans with clear vision who have a clear reminder of 2007-2008 are not fools.. even money won't buy their votes.. I wonder whether Uhuruto have learned the lession.. Mudavadi their bank door protoge as rumored backed by statehouse is a none-starter in Western province.. yani the guy needs to seek votes in Nyeri
|
|
|
Post by reporter911 on Jan 15, 2013 1:08:41 GMT 3
Here is the polls report ... just as a reminder of the "CORD" Tsunami waving through Kenya currently.. Tribal nonsensical spewed out diarrhea won't stop the "CORD" Tsunami
|
|
|
Post by mangai on Jan 15, 2013 14:30:23 GMT 3
I found this interesting, from Macharia Gaitho. www.nation.co.ke/blogs/Dismissing-poll-numbers-wont-help-candidates/-/446672/1665472/-/view/asBlogPost/-/8nae0sz/-/index.htmlDismissing poll numbers won’t help any of the presidential candidatesPosted Monday, January 14, 2013 | By MACHARIA GAITHO I was both bemused and outraged by the varied reactions to the latest opinion poll on the presidential and General Elections. Prime Minister Raila Odinga’s supporters purred like contented kittens with the news that the Cord Alliance ticket would capture State House ain the first round. However, it was gloom and anger within the rival Jubilee Coalition. Deputy Prime Minister Uhuru Kenyatta went on the assault against pollsters Infotrak Harris at his weekend campaign events. His supporters launched on social media a vicious propaganda blitz, based mostly on falsehoods, designed to show that the polling company is associated with Mr Odinga, even if only by the ethnic origins of its owners and key managers. At least this time Mr Kenyatta did not go to the ridiculous extent of threatening to commission his own ‘friendly’ pollster. But the germane point is that had Infotrak Harris found him leading, the Jubilee Coalition would be out on the stumps telling voters how independent polls confirmed the certainty of victory. And by the same token, of course, Mr Odinga’s supporters would be apoplectic all over the place on how Infotrak had been ‘bought’ to produce fake figures. In this case, it was not just the Kenyatta camp that dismissed the Infotrak poll, but also the minor party candidates including Mr Musalia Mudavadi, Mr Peter Kenneth, Ms Martha Karua and Prof James Kiyiapi. I would advise against those knee-jerk reactions. Kenyan politicians should by now have wised up to the reality that you ignore opinion polls at your own peril. Number-crunching is an integral tool of democratic elections anywhere. Dismissing numbers because they do not favour you is as self-defeating as can be. Even more foolish would be paying some pliable pollsters to cook the numbers according to your dictates. Right now the country is awash with briefcase pollsters and researchers offering themselves to the highest bidders, so there will be no shortage of extremely dodgy surveys being hawked around. But they will serve absolutely no purpose other than massaging the egos of the paymasters. I do not hold brief for Infotrak and neither do I endorse their findings, but I restate that it is foolish and self-delusional for any candidate or party to dismiss the results because they don’t come out tops. The owners and managers of all the other reputable polling outfits in Kenya – I would mention Synovate, Consumer Insight and Strategic – all have names that must start with ‘M’, ‘O’, ‘K’, ‘A’ or any other letter of the alphabet. It would be more than foolish if Mr Odinga launched a slander campaign against Consumer Insight, for instance, on the simplistic premise that the name of the owner starts with the letter ‘U’ and therefore he or she must be a relative or village-mate of Uhuru Kenyatta. This type of ethnic-profiling we are seeing on Facebook and Twitter in reaction to the Infotrak poll is a stark reminder of the dangerous politics that we engage in. I think no self-respecting leader, especially with the memory of the 2007-08 violence still so fresh in mind, would want to be associated with such propaganda based on nothing but fuelling ethnic hatred. Candidates would profit much more from learning from the numbers rather than dismissing them without basis. The Infotrak numbers may well be inaccurate. They could also be spot-on, but we all know that things could change dramatically come March. If anyone has a dispute with the numbers, however, then the intelligent thing would be to professionally scrutinise them rather than just react out of outrage. If will be interesting to see how the numbers compare with studies from other reputable pollsters in the coming weeks. Wide discrepancies between the results from different pollsters is what might reveal whether some are deliberately skewing their figures to favoured candidates, or employing faulty methodology. We have learnt since the 2002 elections that the science of political opinion polling has come of age in Kenya. There have been variations, as everywhere else on the world, but the final results have always been strikingly accurate within the margins of error. mgaitho@ke.nationmedia.com
|
|
|
Post by Luol Deng on Jan 15, 2013 17:33:33 GMT 3
mwalimumkuu, What issues have been raised? She is Luo, so should she change her ethnicity to please your likes. Infotrack released their numbers & the factual basis behind it. The fact that you have conveniently overlooked that and gone for non issues just shows how shallow you are. Having a comment every now and then on all matters political doesn't mean that you have anything meaningful to say. PS. Can you please give us the basis of the jubilee projections? Mos mos nani. Will you please show me where I have said that the lady is Luo and therefore needs to change that. If you go back to my contribution to this issue you will see the specific issues I raised about this poll on day one. Stop confusing this board, it has a memory of its own. Kenya as a country has a history in matters such as this. There is nothing so suggesting that this has changed. That is why for instance, no one is wasting their time and money campaigning in Luo Nyanza. You can do better than playing dumb mwalimumkuu. Considering you have not raised any worthwhile queries, I'll leave it at that.
|
|
|
Post by reporter911 on Jan 16, 2013 5:45:55 GMT 3
Mos mos nani. Will you please show me where I have said that the lady is Luo and therefore needs to change that. If you go back to my contribution to this issue you will see the specific issues I raised about this poll on day one. Stop confusing this board, it has a memory of its own. Kenya as a country has a history in matters such as this. There is nothing so suggesting that this has changed. That is why for instance, no one is wasting their time and money campaigning in Luo Nyanza. You can do better than playing dumb mwalimumkuu. Considering you have not raised any worthwhile queries, I'll leave it at that. ;D ;D hysterical is what Mwalimumkuu has become, looking for a needle in a haystack which was never there in the first place.. ' CORD" TSUNAMI can do that to people ;D ;D
|
|
|
Post by kamalet on Jan 16, 2013 9:28:41 GMT 3
More criticism of the Ambitho poll....
Opinion polls data should be taken with a heavy pinch of salt in Kenya
By NJERU WA NJAGI Posted Tuesday, January 15 2013 at 20:00
There is considerable evidence, from a statistician’s assessment, that pollsters do not well understand the theoretical basis of what they say they do.
This casts aspersions on the so-called “numbers” that they so love to present in front of mesmerised journalists, and provides sufficient grounds for a statistician to dismiss these “numbers” as a sham.
It arguably also lends credence to suspicions that unscrupulous pollsters may be masquerading behind the veil of “the science of opinion polling” to peddle personal, and/or sponsored agenda.
Opinion polling, in its proper application, is a concept hinged on survey sampling theory and an active research domain within the academic and professional field of statistics.
The crucial role that statistics and statistical concepts play in society cannot be over-emphasised. For instance, statistical concepts are applied by medical statisticians in all the phases of drug development to ensure that only safe drugs are approved for public use.
Medical statisticians also use statistical concepts in clinical trials to test whether new treatments provide better prognosis or fewer side-effects.
It is incumbent on anyone who even purports to evoke concepts of statistics to do so in a manner that will not cast unnecessary mud on the noble practice of statistics.
On to my first point. The opinion poll results whose presentation I have watched via online TV, and whose reports I read online, leave a great deal to be desired.
A poll report should be a statistical document, and should, as such, be written with both the non-statistical and the statistical audiences in mind. Therefore, as much as technical details must be avoided, certain aspects that are definitely of interest to a statistician must be stipulated. This makes the results verifiable.
However, none of these polls specify whether or not the survey design was taken into account in the analysis, and, if so, how this was done. Maintaining silence on such a crucial aspect leaves a statistician guessing between the following:
1.That the pollster does not appreciate that there are theoretical ramifications that stem from the design, and which need to be duly taken into account in the analysis, if the results are to be reliable.
2.That the pollster only has a slight clue that the design needs to be somehow accounted for, but, due to limitations in the understanding of the underlying theory, chooses to remain mum, in order not to open a can of worms.
It is highly unlikely in a survey that the problem of non-response will not surface. None of these polls specify whether or not there were non-respondents, and if yes, how these were dealt with.
Due to the prominence of this issue, the US Food and Drug Administration had to convene a panel of eminent statisticians and medical statisticians, to provide guidelines to the agency on how the problem needs to be handled, in the context of clinical trials.
Missing responses, and how they are handled, can invalidate the entire process. If respondents who would have answered in support of political candidate X systematically abstain from the poll, it is possible that a fallaciously underestimated popularity will be peddled, unless the pollster appreciates and invokes missing data methods. Silence on this issue should ring the alarm bells.
Thirdly, I have not seen a single pollster attempt to interpret the confidence intervals associated with their estimates. This is very important because the public needs to know that it is possible that, for instance, the true political popularity percentage of candidate X may not even be contained in the interval provided.
It is, therefore, very misleading for pollsters to, by omission or commission, create an impression to the public that the true popularity of candidate X is even anywhere near what they present as the estimate.
In plain language, candidate X’s popularity may not be even anywhere near the percentage provided. Failure to provide guidelines on how to construe these intervals is outrageous.
Fourthly, in utter disbelief, I have also seen pollsters draw conclusions to the effect that candidate X’s popularity has surged, or plummeted, based on a comparison of results from surveys conducted at different periods.
They never, in addition, specify whether it is the same group of respondents who were followed over time. This is important because any statistician who has taken a basic course in longitudinal data analysis knows that you cannot draw such conclusions unless you have been following up the same group of respondents.
That’s a basic difference between longitudinal and cross-sectional studies!
In conclusion, I would like to call upon fellow professionally trained statisticians, especially during this Year of Statistics, to remain vigilant, to ensure that concepts that touch on our noble profession are not dragged into disrepute by pretenders.
Mr Njagi is a doctoral research Fellow in Biostatistics at the Interuniversity Institute for Biostatistics and Statistical Bioinformatics of Hasselt University in Belgium
|
|
|
Post by reporter911 on Jan 16, 2013 9:45:55 GMT 3
More criticism of the Ambitho poll.... Opinion polls data should be taken with a heavy pinch of salt in KenyaBy NJERU WA NJAGI Posted Tuesday, January 15 2013 at 20:00 There is considerable evidence, from a statistician’s assessment, that pollsters do not well understand the theoretical basis of what they say they do. This casts aspersions on the so-called “numbers” that they so love to present in front of mesmerised journalists, and provides sufficient grounds for a statistician to dismiss these “numbers” as a sham. It arguably also lends credence to suspicions that unscrupulous pollsters may be masquerading behind the veil of “the science of opinion polling” to peddle personal, and/or sponsored agenda. Opinion polling, in its proper application, is a concept hinged on survey sampling theory and an active research domain within the academic and professional field of statistics. The crucial role that statistics and statistical concepts play in society cannot be over-emphasised. For instance, statistical concepts are applied by medical statisticians in all the phases of drug development to ensure that only safe drugs are approved for public use. Medical statisticians also use statistical concepts in clinical trials to test whether new treatments provide better prognosis or fewer side-effects. It is incumbent on anyone who even purports to evoke concepts of statistics to do so in a manner that will not cast unnecessary mud on the noble practice of statistics. On to my first point. The opinion poll results whose presentation I have watched via online TV, and whose reports I read online, leave a great deal to be desired. A poll report should be a statistical document, and should, as such, be written with both the non-statistical and the statistical audiences in mind. Therefore, as much as technical details must be avoided, certain aspects that are definitely of interest to a statistician must be stipulated. This makes the results verifiable. However, none of these polls specify whether or not the survey design was taken into account in the analysis, and, if so, how this was done. Maintaining silence on such a crucial aspect leaves a statistician guessing between the following: 1.That the pollster does not appreciate that there are theoretical ramifications that stem from the design, and which need to be duly taken into account in the analysis, if the results are to be reliable. 2.That the pollster only has a slight clue that the design needs to be somehow accounted for, but, due to limitations in the understanding of the underlying theory, chooses to remain mum, in order not to open a can of worms. It is highly unlikely in a survey that the problem of non-response will not surface. None of these polls specify whether or not there were non-respondents, and if yes, how these were dealt with. Due to the prominence of this issue, the US Food and Drug Administration had to convene a panel of eminent statisticians and medical statisticians, to provide guidelines to the agency on how the problem needs to be handled, in the context of clinical trials. Missing responses, and how they are handled, can invalidate the entire process. If respondents who would have answered in support of political candidate X systematically abstain from the poll, it is possible that a fallaciously underestimated popularity will be peddled, unless the pollster appreciates and invokes missing data methods. Silence on this issue should ring the alarm bells. Thirdly, I have not seen a single pollster attempt to interpret the confidence intervals associated with their estimates. This is very important because the public needs to know that it is possible that, for instance, the true political popularity percentage of candidate X may not even be contained in the interval provided. It is, therefore, very misleading for pollsters to, by omission or commission, create an impression to the public that the true popularity of candidate X is even anywhere near what they present as the estimate. In plain language, candidate X’s popularity may not be even anywhere near the percentage provided. Failure to provide guidelines on how to construe these intervals is outrageous. Fourthly, in utter disbelief, I have also seen pollsters draw conclusions to the effect that candidate X’s popularity has surged, or plummeted, based on a comparison of results from surveys conducted at different periods. They never, in addition, specify whether it is the same group of respondents who were followed over time. This is important because any statistician who has taken a basic course in longitudinal data analysis knows that you cannot draw such conclusions unless you have been following up the same group of respondents. That’s a basic difference between longitudinal and cross-sectional studies! In conclusion, I would like to call upon fellow professionally trained statisticians, especially during this Year of Statistics, to remain vigilant, to ensure that concepts that touch on our noble profession are not dragged into disrepute by pretenders. Mr Njagi is a doctoral research Fellow in Biostatistics at the Interuniversity Institute for Biostatistics and Statistical Bioinformatics of Hasselt University in Belgium By central province Kijana ati Njagi ;D ;D ;D ;D what a joke!! now this s sher desperation going as far as looking for a tribesman that can ease the pain from the polls facts ..poleni sana Next.... get one more tribesman with 10 PHD's.. it ain't going to change the facts of the polls and the real "CORD" gigantic wave on the ground..
Meanwhile check this polls closely, currently "CORD" has surpassed the 49%.. it is at 58% heading toward 100%
|
|
|
Post by mangai on Jan 16, 2013 9:48:02 GMT 3
More criticism of the Ambitho poll.... Opinion polls data should be taken with a heavy pinch of salt in KenyaBy NJERU WA NJAGI Posted Tuesday, January 15 2013 at 20:00 There is considerable evidence, from a statistician’s assessment, that pollsters do not well understand the theoretical basis of what they say they do. This casts aspersions on the so-called “numbers” that they so love to present in front of mesmerised journalists, and provides sufficient grounds for a statistician to dismiss these “numbers” as a sham. It arguably also lends credence to suspicions that unscrupulous pollsters may be masquerading behind the veil of “the science of opinion polling” to peddle personal, and/or sponsored agenda. Opinion polling, in its proper application, is a concept hinged on survey sampling theory and an active research domain within the academic and professional field of statistics. The crucial role that statistics and statistical concepts play in society cannot be over-emphasised. For instance, statistical concepts are applied by medical statisticians in all the phases of drug development to ensure that only safe drugs are approved for public use. Medical statisticians also use statistical concepts in clinical trials to test whether new treatments provide better prognosis or fewer side-effects. It is incumbent on anyone who even purports to evoke concepts of statistics to do so in a manner that will not cast unnecessary mud on the noble practice of statistics. On to my first point. The opinion poll results whose presentation I have watched via online TV, and whose reports I read online, leave a great deal to be desired. A poll report should be a statistical document, and should, as such, be written with both the non-statistical and the statistical audiences in mind. Therefore, as much as technical details must be avoided, certain aspects that are definitely of interest to a statistician must be stipulated. This makes the results verifiable. However, none of these polls specify whether or not the survey design was taken into account in the analysis, and, if so, how this was done. Maintaining silence on such a crucial aspect leaves a statistician guessing between the following: 1.That the pollster does not appreciate that there are theoretical ramifications that stem from the design, and which need to be duly taken into account in the analysis, if the results are to be reliable. 2.That the pollster only has a slight clue that the design needs to be somehow accounted for, but, due to limitations in the understanding of the underlying theory, chooses to remain mum, in order not to open a can of worms. It is highly unlikely in a survey that the problem of non-response will not surface. None of these polls specify whether or not there were non-respondents, and if yes, how these were dealt with. Due to the prominence of this issue, the US Food and Drug Administration had to convene a panel of eminent statisticians and medical statisticians, to provide guidelines to the agency on how the problem needs to be handled, in the context of clinical trials. Missing responses, and how they are handled, can invalidate the entire process. If respondents who would have answered in support of political candidate X systematically abstain from the poll, it is possible that a fallaciously underestimated popularity will be peddled, unless the pollster appreciates and invokes missing data methods. Silence on this issue should ring the alarm bells. Thirdly, I have not seen a single pollster attempt to interpret the confidence intervals associated with their estimates. This is very important because the public needs to know that it is possible that, for instance, the true political popularity percentage of candidate X may not even be contained in the interval provided. It is, therefore, very misleading for pollsters to, by omission or commission, create an impression to the public that the true popularity of candidate X is even anywhere near what they present as the estimate. In plain language, candidate X’s popularity may not be even anywhere near the percentage provided. Failure to provide guidelines on how to construe these intervals is outrageous. Fourthly, in utter disbelief, I have also seen pollsters draw conclusions to the effect that candidate X’s popularity has surged, or plummeted, based on a comparison of results from surveys conducted at different periods. They never, in addition, specify whether it is the same group of respondents who were followed over time. This is important because any statistician who has taken a basic course in longitudinal data analysis knows that you cannot draw such conclusions unless you have been following up the same group of respondents. That’s a basic difference between longitudinal and cross-sectional studies! In conclusion, I would like to call upon fellow professionally trained statisticians, especially during this Year of Statistics, to remain vigilant, to ensure that concepts that touch on our noble profession are not dragged into disrepute by pretenders. Mr Njagi is a doctoral research Fellow in Biostatistics at the Interuniversity Institute for Biostatistics and Statistical Bioinformatics of Hasselt University in BelgiumThe title of this thread is trying to suggest that since Angela Ambitho is a Luo, she might have doctored the poll to suit Raila, a fellow Luo. What would you have expected from this Njeru wa Njagi guy? Isn't this article meant to favour his 'cousin' Uhuru Kenyatta?
|
|
|
Post by mank on Jan 16, 2013 15:38:18 GMT 3
More criticism of the Ambitho poll....
Opinion polls data should be taken with a heavy pinch of salt in Kenya
By NJERU WA NJAGI Posted Tuesday, January 15 2013 at 20:00
.... However, none of these polls specify whether or not the survey design was taken into account in the analysis, and, if so, how this was done. Maintaining silence on such a crucial aspect leaves a statistician guessing between the following:
1.That the pollster does not appreciate that there are theoretical ramifications that stem from the design, and which need to be duly taken into account in the analysis, if the results are to be reliable.
2.That the pollster only has a slight clue that the design needs to be somehow accounted for, but, due to limitations in the understanding of the underlying theory, chooses to remain mum, in order not to open a can of worms.
..... Thirdly, I have not seen a single pollster attempt to interpret the confidence intervals associated with their estimates. This is very important because the public needs to know that it is possible that, for instance, the true political popularity percentage of candidate X may not even be contained in the interval provided. ..... Mr Njagi is a doctoral research Fellow in Biostatistics at the Interuniversity Institute for Biostatistics and Statistical Bioinformatics of Hasselt University in Belgium The most significant flaw I have heard about with regard to the topical study is that it selectively sampled CORD leaning constituencies (counties close to the homelands of the CORD candidates, and those that have shown favour for CORD). There is nothing you are going to do to such results to make them speak about the vote of the Kenyan voter. They will just speak about the vote of the CORD-leaning voter - how such information is useful to the pollsters must be the devil in the detail. But what Njagi alludes to here, the confidence intervals, is also a concept worth scrutiny. In statistics you cannot conclude there is a difference between two comparative characteristics if the confidence interval of parameter estimates for the comparative characteristics are "too large", or, alternatively stated, if the difference in parameter estimates is statistically insignificant. So it would be helpful to actually look at the full report of this study and see what levels of significance they report for the differences they cite. Not that such scrutiny will change the first and fatal flaw of the study.
|
|
|
Post by reporter911 on Jan 16, 2013 22:21:26 GMT 3
|
|
|
Post by nowayhaha on Jan 17, 2013 11:30:01 GMT 3
So after joining Railas which has been accused by several quarters as the one sponsoring the Infrotrack polls Kalonzo sees the light . Please Kalonzo dont go that route you are an exception in CORD as you are well respected across the political divide and btw you have been doing well talking of peace , reconciliation and policies which will change the country please stick to that leave siasa ya opinion polls to Raila and ODM Yes we have given him the" OPINION POLLS PRESIDENT " title but his doesnt necessarily mean he is or will be the PRESIDENT OF KENYA. Here Kalonzos opinion on opinion polls before he went to the political group bankrolling the opinion polls kenya.news24.com/MyNews24/Kalonzo-Dismisses-Latest-Opinion-Poll-20121112Kalonzo dismisses latest opinion poll 12 November 2012, 10:40Steve Wambugu Vice President Kalonzo Musyoka on Saturday took his campaigns to Nakuru town where he dismissed a recent opinion poll that rated him behind other presidential hopefuls. Kalonzo was addressing a Muslims congregation before opening a Wiper Democratic Movement office at Shabab estate when he said that the opinion poll results were crafted by well paid individuals who are out to malign him. He said that he will be exposing those behind crafted opinion polls soon, adding that he is firmly in the presidential race. He blamed other presidential hopefuls for rumor mongering after sensing defeat and promised to carry out robust campaigns immediately he officially launches his campaigns. He also called on all Kenyans to register as voters in big numbers once the exercise kicks off. Kalonzo was accompanied by nominated MP Shakila Abdallah and former MP Francis Nyenze. www.nation.co.ke/News/politics/Kalonzo-faults-opinion-polls-/-/1064/1285096/-/3us3v5z/-/index.htmlPosted Tuesday, December 6 2011 at 16:12 SHARE THIS STORY Tweet 0 inShare Vice President Kalonzo Musyoka has said the conduct of opinion polls in Kenya is wanting and urged pollsters to embrace professionalism, integrity and accuracy. In developed countries, such surveys observe strict guidelines and that is why results are acceptable, the VP noted. “In this country opinion polling is a subject of serious under hand manipulation which is a cause for worry,” he said during the Kenya National Dialogue and Reconciliation conference at Crowne Plaza Hotel, Nairobi Tuesday. Mr Musyoka said MPs were of the view that the country should have an Act of Parliament to regulate the opinion polls since skewed polling may contribute to post election chaos. He assured participants that Kenya will not witness a repeat of violence as happened after the 2007 elections, since the Constitution has ensured that the electoral environment will be regulated. He said the Constitution will not allow candidates who have lost in an election to agitate and incite the population to violence. "Any candidate who believes that he or she cannot accept electoral defeat must withdraw from the race and allow Kenyans the opportunity to run their affairs peacefully and responsibly,” he said. The VP said the Constitution will eliminate corruption, tribalism and violence. Mr Musyoka said the Judiciary has undergone obvious rejuvenation under Chief Justice Willy Mutunga and also pointed to reform commissions including the Independent Boundaries and Electoral Commission, National Cohesion and Integration Commission, the Constitutional Implementation Commission and the Commission on Revenue Allocation. “The Commission for the Administration of Justice, National Police Service Commission and the Salaries and Remuneration Commission are among bodies awaiting their turn for redesigning or reestablishment respectively,” said the Vice President. Mr Musyoka said these institutions are at the core of rebuilding a new Kenya. “If these bodies are set up and managed as the Constitution dictates, then we shall leave a meaningful and enduring legacy that future generations will find hard to tarnish. And that must be our intention, our guiding spirit," he said. He said that the Constitution entrenches very progressive rights and freedoms particularly the right to food and shelter. The VP said this aspect of the Constitution compels the government to rededicate resources and commitment to improvement of the standards of living of Kenyans and progressively end poverty. He said the onus to communicate that Kenya is reborn belongs to the leadership. “We leaders bear the burden of inspiring a sense of confidence among Kenyans in the fact that our rejuvenated country is adequately prepared to reap the full benefits borne by the new Constitution”, noted the Vice President. Among those present at the KNDR meeting were former UN Secretary-General Kofi Annan, former Tanzanian President Benjamin Mkapa, former Ghanaian President John Kufuor, former South African First Lady Graca Machel and the head of a team that investigated the Kenya’s disputed 2007 elections Johann Kriegler.
|
|
|
Post by reporter911 on Jan 17, 2013 18:27:22 GMT 3
So after joining Railas which has been accused by several quarters as the one sponsoring the Infrotrack polls Kalonzo sees the light . Please Kalonzo dont go that route you are an exception in CORD as you are well respected across the political divide and btw you have been doing well talking of peace , reconciliation and policies which will change the country please stick to that leave siasa ya opinion polls to Raila and ODM Yes we have given him the" OPINION POLLS PRESIDENT " title but his doesnt necessarily mean he is or will be the PRESIDENT OF KENYA. Here Kalonzos opinion on opinion polls before he went to the political group bankrolling the opinion polls kenya.news24.com/MyNews24/Kalonzo-Dismisses-Latest-Opinion-Poll-20121112Kalonzo dismisses latest opinion poll 12 November 2012, 10:40Steve Wambugu Vice President Kalonzo Musyoka on Saturday took his campaigns to Nakuru town where he dismissed a recent opinion poll that rated him behind other presidential hopefuls. Kalonzo was addressing a Muslims congregation before opening a Wiper Democratic Movement office at Shabab estate when he said that the opinion poll results were crafted by well paid individuals who are out to malign him. He said that he will be exposing those behind crafted opinion polls soon, adding that he is firmly in the presidential race. He blamed other presidential hopefuls for rumor mongering after sensing defeat and promised to carry out robust campaigns immediately he officially launches his campaigns. He also called on all Kenyans to register as voters in big numbers once the exercise kicks off. Kalonzo was accompanied by nominated MP Shakila Abdallah and former MP Francis Nyenze. www.nation.co.ke/News/politics/Kalonzo-faults-opinion-polls-/-/1064/1285096/-/3us3v5z/-/index.htmlPosted Tuesday, December 6 2011 at 16:12 SHARE THIS STORY Tweet 0 inShare Vice President Kalonzo Musyoka has said the conduct of opinion polls in Kenya is wanting and urged pollsters to embrace professionalism, integrity and accuracy. In developed countries, such surveys observe strict guidelines and that is why results are acceptable, the VP noted. “In this country opinion polling is a subject of serious under hand manipulation which is a cause for worry,” he said during the Kenya National Dialogue and Reconciliation conference at Crowne Plaza Hotel, Nairobi Tuesday. Mr Musyoka said MPs were of the view that the country should have an Act of Parliament to regulate the opinion polls since skewed polling may contribute to post election chaos. He assured participants that Kenya will not witness a repeat of violence as happened after the 2007 elections, since the Constitution has ensured that the electoral environment will be regulated. He said the Constitution will not allow candidates who have lost in an election to agitate and incite the population to violence. "Any candidate who believes that he or she cannot accept electoral defeat must withdraw from the race and allow Kenyans the opportunity to run their affairs peacefully and responsibly,” he said. The VP said the Constitution will eliminate corruption, tribalism and violence. Mr Musyoka said the Judiciary has undergone obvious rejuvenation under Chief Justice Willy Mutunga and also pointed to reform commissions including the Independent Boundaries and Electoral Commission, National Cohesion and Integration Commission, the Constitutional Implementation Commission and the Commission on Revenue Allocation. “The Commission for the Administration of Justice, National Police Service Commission and the Salaries and Remuneration Commission are among bodies awaiting their turn for redesigning or reestablishment respectively,” said the Vice President. Mr Musyoka said these institutions are at the core of rebuilding a new Kenya. “If these bodies are set up and managed as the Constitution dictates, then we shall leave a meaningful and enduring legacy that future generations will find hard to tarnish. And that must be our intention, our guiding spirit," he said. He said that the Constitution entrenches very progressive rights and freedoms particularly the right to food and shelter. The VP said this aspect of the Constitution compels the government to rededicate resources and commitment to improvement of the standards of living of Kenyans and progressively end poverty. He said the onus to communicate that Kenya is reborn belongs to the leadership. “We leaders bear the burden of inspiring a sense of confidence among Kenyans in the fact that our rejuvenated country is adequately prepared to reap the full benefits borne by the new Constitution”, noted the Vice President. Among those present at the KNDR meeting were former UN Secretary-General Kofi Annan, former Tanzanian President Benjamin Mkapa, former Ghanaian President John Kufuor, former South African First Lady Graca Machel and the head of a team that investigated the Kenya’s disputed 2007 elections Johann Kriegler. The Polls are right on the money, Kalonzo knows it so does most people in Kenya except for central province who are jumping up and down crying hysterically, but that won't change the "CORD" Tsunami swiping the ground.."CORD is now at 59% polling strong against Jubilee which is down to 28% Mambo bado..
|
|