PRESS RELEASE
FOR IMMEDIATE RELEASE TO ALL MEDIA HOUSES

THE UNZA PF THINK TANK‘S POSITION ON THE MUVI TV OPINION POLLS

Section 14, Sub-section 4 of the Electoral Code of Conduct permits broadcasters to conduct opinion polls as long as they indicated the source of the public opinion poll and indicate the margin of error. By source, we understand this to mean methodology. In our view, the MUVI TV opinion poll in terms of methodology does not fulfil the requirements of the Electoral Code of Conduct.

The validity of the polls being conducted by MUVI TV and other media houses on the internet is highly questionable because of the seriously flawed methodology being used. Because of this, a false impression has been created that these are scientific polls and are accurately reflective of what the Zambian people feel about the candidates. But this is far from the truth.

For one thing, MUVI TV’s approach in conducting the poll is not sufficiently scientific. For an opinion poll to be scientific, there must be some important preconditions it must meet. One of these preconditions is that the selection of respondents must not discriminate against any member of the target population. Related to this, the distribution of the characteristics of the sample must closely resemble those of the population.

This requires therefore that each member of the target population must have the same chance of being part of the sample as any other member of that population.

For example, if as MUVI TV reported they had a sample of 1,800 respondents, given a population of 5,166,088 registered voters, then each registered voter will have 1,800/5,166,088 or 3 chances out 10,000 of participating in their opinion poll. In other words, their sample must be representative of all the 5,166,088 registered voters in Zambia.

Given below is the breakdown, in percentage terms, of registered voters by age province and gender by November 2014 obtained from the Electoral Commission of Zambia website.

 

Province Total Female Male
Central

9.4

51.0

49.0

Copperbelt

16.4

54.1

45.9

Eastern

11.6

46.3

53.7

Luapula

7.9

48.0

52.0

Lusaka

15.4

54.4

45.6

Muchinga

5.2

48.5

51.5

Northern

8.4

48.7

51.3

North western

6.1

49.3

50.7

Southern

12.0

47.9

52.1

Western

7.6

43.4

56.6

Total

100.0

49.9

50.1

 

If the sample of 1,800 MUVI TV obtained was representative of all registered voters in Zambia, the distribution of the respondents, broken down by gender and province, would, in absolute numbers, resemble the one below. But is this the case with MUVI TV sample?

Province Female Male Total
Central

86

83

169

Copperbelt

160

135

295

Eastern

97

112

209

Luapula

68

74

142

Lusaka

151

127

278

Muchinga

46

48

94

Northern

74

77

151

North western

54

56

110

Southern

103

112

215

Western

59

78

137

Total

898

902

1,800

To be sufficiently representative, a poll must ideally use probability sampling in the selection of people from the population into the sample.

The most basic form this takes is random sampling. With random sampling, one must have a sampling frame of the population based on the target population. For an opinion poll on political preferences, for example, one must have the most up to date voters register of obtained from the ECZ as a sampling frame. To ensure that every registered voter has an equal chance of being part of the sample, names must be selected from the voters register at random. Then the selected voters can be contacted through face-to-face interviews or by phone if such information is available on the register.

The MUVI TV opinion poll is unscientific as it does not meet any of the basic criteria for a scientific opinion poll described in the preceding sections; it is thus likely to give highly misleading results.

To begin with, their sample of respondents does not give all registered voters in Zambia an equal chance of participating in the opinion poll. Even among mobile phone users, it favours Airtel and Zamtel subscribers and discriminates against MTN subscribers. Furthermore, it gives no chance of participation to those with no access to, or without interest in MUVI TV. In a word, it is a travesty of what a scientifically derived sample is expected to be.

As is the case with polls based on the phone – in approach or internet based, the sample used is self-selecting, attracting participants who are overzealous or passionate about a topic or candidate. Consequently, these people will resort to any means to ensure that their views prevail. MUVI TV claims that each participant was restricted to one vote from one cell phone number only.

But how does MUVI TV deal with situations where one person has more than one number? And how do they control for the possibility of some people using the phones from less zealous friends and relatives to cast their votes many times over? If the restriction is on the votes coming from the same IP address on a computer, some respondents may simply use different computers in different locations in order to cast more votes.

Some overzealous voters may even open several Facebook accounts in order to increase the number of votes. And how does MUVI prevent some people from mobilizing others who share their political views and preferences to cast their votes as many times as they wish? It is clear that the MUVI opinion poll is what is called a push opinion poll. This is a method of denigrating or promoting a viewpoint or a candidate under the guise of a public-opinion poll in a biased fashion in an effort to influence voters.

The other glaring weakness of the MUVI TV opinion poll is that does not collect any additional information on the respondents to give a clear picture of who is in the sample. It is impossible to know, for example, if the people participating in the poll are registered voters. There is also no information on characteristics of respondents such as age, gender, political affiliation, province of residence or origin, ethnicity, race, etc. which would give us further insights in understanding what underlies the political preferences of the registered voters.

If an opinion poll is conducted scientifically and is representative of the population, it is possible to estimate how accurately the information we have based on a sample reflects what actually obtains in the true population of voters. This is possible through the computation of the margin of error.

In simple terms, the margin of error is a measure of how accurately the information based on the sample estimates what is in the true population. For example, if we wanted to know the percentage of registered voters who are likely to vote for Edgar Lungu or Hakainde Hichilema, it is possible to devise a sample design to ensure that our sample estimate will not differ from the true population value by more than, say, 5 percent. With the MUVI TV opinion poll, it is not possible to compute the margin of error because their sample was not scientifically obtained and therefore not representative of the population of registered voters.

The deficiencies noted in the MUVI TV opinion poll call into question their motives. It is very evident that those conducting the poll have severely limited or no knowledge of sampling theory and basic statistics. Why then do they want to conduct such an activity when they could have contracted a number of credible organizations and experts to do a more credible opinion poll?

The danger of such polls as they may excite the emotions of people who have limited understanding of sampling theory and statistics and may thus believe in such misleading polls predicting victory of their preferred candidate. What if such a candidate loses? Won’t the results of such polls be used as grounds for disputing the election results? Isn’t this potentially a recipe for civil strife?

Going by what we have pointed out, the MUVI TV opinion polls have not conformed to the basic principles scientific sampling; they have also violated the basic requirements of Section 14, Sub-section 4 of the Electoral Code of Conduct. ECZ should invoke penal proceedings against MUVI TV.

These polls must therefore be stopped before they mislead the nation on the probable outcome of the 20th January 2014 by-elections. The polls should only be allowed if they demonstrate strict adherence to the principles of scientific sampling and abide by the dictates of Section 14, Sub-section 4 of the Electoral Code of Conduct.

RELEASED BY THE UNZA PF THINK SECRETARIAT

THE VIEWS OF THIS PRESS STATEMENT DO NOT REFLECT THE VIEWS OF THE UNIVERSITY OF ZAMBIA OR ANY OTHER ORGANIZATIONS TO WHICH MEMBERS OF THE THINK TANK BELONG INDIVIDUALLY OR SEVERALLY.

[Read 97 times, 1 reads today]
Loading...

57 COMMENTS

    • One wanders if they would have bothered to give this long analysis had the opinion poll showed ECL winning

      0

      0
    • PF has a lot of work to clean up. Now Emmanuel Mwamba has shifted office and gone to UNZA as a “Think Tank”. Yesterday they had a rejected
      – FORMER bishop Chomba
      – Excommunicated Fr. Bwalya (Who has since decided to withdraw his PF support)
      – They have taken RB, a total embarrassment of his parties UNIP & MMD.

      There is no point for PF to continue campaigning, people have decided.
      Its HH.

      0

      0
    • Muvi Tv being a pay tv cant have a proper anlsysis. First anyone who watches Muvi T.v is a frustrated citizen who wants to hear the other version of news considering how ZNBC IS baised towards P.F. By that u expect that kind of opinion though not conclusive

      0

      0
    • PF, PF, PF! Why did you believe and advertise the opinion poll that Muvi TV conducted before the PF extra-ordinary general conference which indicated Mr. Lungu would win in Kabwe? Why didn’t you question the methodology then? PF just knows how to use pangas to analyse issues and I forgive you Pathetic Failures. PF wawa!

      0

      0
    • @Zedman, and nothing would have stopped UPND from rebating such a poll like what this PF UNZA Think Tank has done.

      And at least this rebuttal makes a lot of sense compared to the gibberish explanation Costa Mwansa gave yesterday. Clearly Costa and his team at Muvi TV have no clue as to how opinion polls are done. Next time they should ask, there is no shame in it!!!!!

      0

      0
    • PF think-tank?! My foot!! Better call themselves “PF Disorganized-Tank”. Why cant this group come up with their own opinion poll?

      If this group was sensible, they should`ve talked about the qualities of their candidate and what PF has achieved. Of course even themselves have nothing good to talk about their PF.

      0

      0
    • I thought you were going to give us your opinion poll?

      Carry out yours and show as the results. Don’t Christianize others without offering an alternative.

      Give us yours we compare.

      0

      0
    • Clearly, the biaseness in the author shows that he is alien to the intellectual work of think tanks. Then he wants readers to read his article like the bible because of UNZA belonging. If the Author was an intellectual, he could also have explained why the Poll which favored Lungu for PF presidency was good. PF advertised it in the papers. Secondly, he does not explain why he thinks only UPND supporters would vote multiple times and not PF cadres? Why did PF supporters show so much interest in that Poll before the convention? If MUVI is biased, why did they not rig that Poll against Lungu for PF presidency?

      This is what it means to have a dirty mind- thinking with emotions when things dont favor you.

      0

      0
    • The UNZA monks have spoken..UNZA remains the moral conscious of Zambia WHERE AND WHEN UNZA SAYS YES JUST KNOW ZAMBIA HAS ALSO SAID YES!!!

      January 21, it will be announced…Edgar Lungu is leading in the elections

      January 23/24 it will be announced…Edgar Lungu has won the elections

      By the first week of February it will be announced…HH has resigned as UPND president..

      ****VIVA EDGAR LUNGU 2015 AND 2016****

      0

      0
  1. Is the UNZA PF think tank a recognized structure within the PF framework or a loose conglomeration of PF sympathizers? In other universities elsewhere, the think tank would have been more proactive on matters of policy and engage the UNZA community including those with contrary views.

    0

    0
    • Who said a think tank has to be a “recognized structure” of a political party? By the way, even here in America there are political Party groups in universities. They may not called themselves Think Tanks, but definitely there are both Republican and Democratic student/faculty discussion groups in universities here. Think Tanks come in all shapes, sizes, philosophies, agendas, and everything else you can think of. What is a ‘think tank’ anyway? Answer that question and you may find out that your comment is misplaced.

      And it looks like you simply chose to skip over the disclaimer at the bottom of the article!

      0

      0
    • I have evaluated the candidates. I am voting HH. It’s my vote, and he is my choice.

      No more PF lies. God save us from the PF plunder, corruption, pangas, high borrowings, high prices of cement, mealie meal, fuel, foreign exchange.

      0

      0
  2. That’s well and good on the face of it – sampling with known proportions; better still would be stratified random sampling based on gender, age, urban-rural, jobless-employed, office-factory, education etc! NEVER over-simplify a priori probabilty models bwana – JUST GO VOTE on 20-01-2015 😉

    0

    0
  3. It has to bear the name PF! Anyway, whether the MUVI TV opinion polls are true or not people ve already made up their mind, it’s forward chabe. Even Scott has been saying ‘Zambia has to move forward’.

    0

    0
  4. How I wish we could also have an analysis from Harvard. I would like to hear the views of universities which produce engineers who can design an airplane or a staple for sticking papers together.

    0

    0
    • WISDOM, THE ANALYSIS FROM HARVARD WOULD HAVE BEEN THE SAME AS THE THINK TANK. TO START WITH THE SAMPLE SIZE WAS ABOUT .0003% OF THE REGISTERED VOTER POPULATION SIZE. TELL MUVI TV THAT THIS IS NOT BIG BROTHERS HOUSE WERE YOU CAN VOTE AS MANY TIMES AS YOU WANT

      0

      0
  5. HH all the way. may 20th January come quickly and results will show to shame the think tanks.

    0

    0
  6. Makes aaaaa loooooottttt….. of sense!

    It is pretty obvious Movi TV’s “journalistic integrity” has been compromised. Somebody must have oiled their hands to produce such bogus polls and influence public opinion. Even to many of us who are uninitiated in statistics and scientific sampling are able to tell that a spread of almost 70 percentage points between the “winner” (HH) and the second placed person (Edigar Lungu) is simply [email protected]” to say the least.

    This is the kind of intellectual laziness on the part of our Zambian media that will always keep this industry (Television/Radio) in Zambia a relic of the third world. It is like everybody in this industry, from media Executives to Journalists, are comfortable with mediocrity and no one wants to raise the bar.

    0

    0
    • @Yambayamba
      It is they themselves who claim to be PF UNZA think tank. In a posting below, I have described what a think tank should ideally do. I am NOT persuaded that the reasoning given in the article is by university minds my friend Yambayamba. The level of reasoning in the article is below the caliber of many of the guys I interact with at UNZA. As for your USA example, I find it amusing since I have no clue what the hell goes on in your USA; I am here, waiting for 20th January which is coming too slow for my liking.

      0

      0
  7. Am enjoying this.Edgar wapya munzi.Watchout mwaice icalo nacisala superman HH.Kuyabebele Jan 20.Start emptying your in trays at the office.Viva HH.

    0

    0
  8. Ba Monko, that data is too much for these overzealous political amatures whose aim is only to propagate propaganda. Some of us who had an opportunity to study Demography and Statistics are alive to the fact that Sampling has certain benchmarks for it to be credible and there are several variables that need to be taken into consideration before a representative sample is selected! All the so called opinion polls that have been conducted recently lack the basic tenets of statistical analysis and can not be empirically verified as true representation of the Voter Population. Moreover, all demographic variables such as age, sex as well as the regions have not been properly qualified! Therefore, such useless opinions should be dismissed with the contempt they deserve!

    0

    0
  9. The problem with UPND bloggers is that they think they are the only intelligent imbeciles. The PF think tank has put everything to perspective and all the blind followers of UPND have to do is to ask whether the PF think tank is an organ of PF. The best the above bloggers can do is to state whether the PF think tank is right or wrong. He has educated you on statistics and all you is harden your hearts like steel.

    0

    0
    • @BBC
      No sir/madam, a Think Tank as the name says, puts forward ideas proactively and not retroactively. If PF has a think tank based at UNZA, what do they have to say on the carbon tax which is so high? on the paltry drop in the fuel pump price in Zambia against the massive fall in world crude oil prices? on the plummeting of the Kwacha value against major currencies? on the refusal by PF to acknowledge that Sata was ill to his last breath? on matters of science and technology? on the VAT controversy? These are think tank issues NOT arguments on grade 12 statistics which they waste so much time on in the article above.
      I repeat, at other universities think tanks deal with real issues not statistical trivialities.

      0

      0
    • @permanent link. we know you are purging wherever you are because of the wind of change which is in hh’s favour. when you think of panga fighters going into political extinction on the 20th of jan 2015, especially at the hands of hh, you cant stomach it. in the meantime continue hallucinating. 12 days to go is the countdown.

      0

      0
  10. Fair analysis but has failed to explain how the issue of multiple fone numbers and emails can not be used by all parties involved. To claim that they are only being used by HH supporters is incorrect. What these polls have clearly indicated is that pf has lost urban areas. Zambian eye, mwebantu and muvi are all exposed to these urban electorates and farmers are likely to cast a protest vote. You dont need rocket science to know that the playing ground has changed.

    Thank you

    0

    0
  11. There is no such thing as a perfect opinion poll.
    I conducted an opinion poll several years ago when I was a student at a very low college.the results of the poll where confirmed by the actual results when elections were held to choose college leaders.
    You don’t need a professional standard to obtain an accurate result of a poll.

    0

    0
  12. In a normal society, it should have been the so-called PF UNZA ‘Think Tank’ that should have provided us with opinion poll results NOT this rubbish….

    Think Tank, my foot……….

    0

    0
  13. This is what happens when people are desperate!UPND and HH are very desperate and they will do anything to win this election,including buying media like MUV TV.What MUV TV does not realise is that it will lose face and the confidence that people who watch it have in it.To be honest there is noway a rulling party can get 16% in an election like this,noways!!These polls are designed to influence people to vote for HH. Based on this HH now claims to have won the election prematurely.This is unprofessional and unacceptable in this time and era. MUV TV might be digging its own grave because these actions will definitely backfire and the editor will be ashamed!!HH know he will lose and he cannot stand the shame of losing again that is why he is trying by all means to create lies through MUV TV.

    0

    0
  14. Great scientific analysis. Fully agreed. Unfortunately MUVI TV has no capacity to think before action!

    0

    0
  15. Even the writer confirms, the views are not from UNZA, but to just another individual, who cannot explain, which organization he belongs to, another fake one. HH forward

    0

    0
  16. Who are these idi.ots calling themselves Think Tank? The tank is full of maggots. The opinion poll has scared the shi.t out of PF and now they are seeking some fake intellectuals to add some credibility to the rebuttals. No way! The opinion poll stays and we encourage Muvi TV not to be shaken by these starved pseudo academicians without substance. Hunger can really drive you to do very useless things just for you to eat. This is what these chaps are doing. They have been promised some food. We don’t need your academic jargon and we are not threatened by it. They think by presenting figures and methodologies people will be scared. Not at all. Don’t forget that we are entitled to carry out convenience research. Nobody can stop us. So far HH is leading and they re scared shi.t. Its PF OUT.

    0

    0
  17. The think tank has a point…..the MUVI TV poll is a push opinion poll aimed at swaying the minds of the gullible. it is a good strategy by UPND coz to some extent they have managed to sway some minds with no basic statistical background…its a dirty game they r playing and since politics is dirty game …play on guys! Final whistle on the 20th of jan

    0

    0
  18. A “Think Tank” that is partisan is no think tank, it is more of a party propaganda machinery. So what is what the “PF Think Tank” is worth.

    0

    0
  19. PF think-tank?! My foot!! Better call themselves “PF Disorganized-Tank”. Why cant this group come up with their own opinion poll?

    If this group was sensible, they should`ve talked about the qualities of their candidate and what PF has achieved. Of course even themselves have nothing good to talk about their PF.

    0

    0
  20. So what’s the lesson here? That no one who supports UPND should take victory for granted and must instead get up and vote on election day?

    Sounds like a good idea – thanks PF think tank!

    0

    0
  21. Why hasn’t this Think Tank conducted a scientifically valid opinion poll instead of complaining MUVI TV’s opinion poll? The voters would have a chance to compare different opinion polis. PF has enough financial resources to fund such a poll.

    0

    0
  22. The analysis in the story lacks credibility so to speak. In the modern world there are software which can analyze the data without going through what you read from books. MUVI TV used the same software to analyze the data that were collected from respondents. It appears like the opinion poll conducted by MUVI has sent shivers in PF. It is common understanding that the PF is loosing this election. Opinion poll or not people have made decision. You don’t need to go through the technicalities to understand who is going to win. For instance Copperbelt which was previously a stronghold of PF most likely will go to HH. I interviewed most of the people on the Copperbelt and it appears they have lost interest in PF. EL is not liked so to speak and a new marriage with RB has messed him up.

    0

    0
  23. First of all the methodology is wrong and biased to people who have access to muvi and that viewership given coverage of areas

    Edgar is winning with huge margins we will be adding satas votes plus those that should have gone to lupia had he stood and taking away from HH and others for his Own

    We will post confirmation of numbers with s confidence internal and skewness of Edgar lungu HH votes distribution with a degree of confidence of 95%. and an acceptable standard error tested on all scenarios worst case or near or extream positive

    Edgar will win northern central Lusaka eastern western northwestern and two marginally southern

    Unless estreem event happens team Lungu should be cerebrating from now to. 21

    Its confirmed scenarion

    0

    0
  24. While these students have elaborately described one way of sampling, it is not the only method of sampling, so the analysis is flawed. Go back to your text books and resrarch on tv polling. With all the energy that they used in trashing the Muvi tv poll, they could have easily conducted their own.

    0

    0
  25. The so called PF Think Tank is a group of PF sympathizers who merely want to wipe out the tears from Members. The so called PF Think Tank should have conducted own Opinion Poll.

    Country men and women the adverts for Edgar Lungu on Radio Phoenix and ZNBC are just overdone and irritating so to speak. PF lacks proper public relations. In a space of 15 minutes a song is played. Worse is that the message is not communicating something to the electorates. Most people talked have complained. It is better to have balanced campaign messages so that the voters make decision who to vote for than hearing from one political party.

    0

    0
  26. So that katuni insulting street boys katondo boys is the best advert by HH

    What advert can we see from HH depicting his promises can we see the Zambian scenario and show us the actual against of delivered

    HH the kaponya advert is an insult to intelligent minds of struggling street children’s and men

    Edgar lungu is a better president to lead Zambia with good team maintaining ba guy

    0

    0
  27. Wapya minzi just run mudala the Titanic has hit the rock and there is confusion everywhere trying to justify a wrong…..a wrong is just a wrong ba Paya Farmer ( condolences to the Serenje mans family..dying while waiting for his sweat..PF you will be paid by No Vote…..mwailisha bamumbwe?

    HH 2015 and beyond

    0

    0
  28. AFTER FREE MASON STORY, I HAVE NET OVER 21 GOD FEARING TONGAS THAT HAVE REJECTED SATANISIM AND ARE NOW GOING TO VOTE FOR LUNGU.
    THEY KNOW GOD DOES NOT CONDON SATANISM, ASSOCIATED WITH BLOOD MONEY.

    0

    0
  29. PF Think tanks have lost it by attacking MUVI TV,they should have also condemned the likes of Mwebantu,QFM,COMET RADIO,MILLENIUM RADIO,UNZA RADIO and the POST because it is saying an opposition party can win,so i would advise the think tanks to go back and complete the work they still have time.

    0

    0
  30. can some one lecture me when it was during the PF convention in Kabwe all was well for EL right ???

    0

    0
    • NO, ALL YOU UpDn *****S WANT TO DWELL ON THE FACT THAT OUR BELOVED PRESIDENT ECL WAS VOTED BY THE SHOW OF HANDS. HOW WAS HH VOTED FOLLOWING THE DEATH OF MAZOKA(MHSRIEP)?
      EVEN IN THE USA IF THE PRESIDENTIAL CANDIDATE FOR PARTY NOMINATIONS IS UNOPPOSED VOTING IS BY SHOW OF HANDS.
      HH IS LOSING THE NEXT ELECTIONS, THE REST OF ZAMBIA IS WAITING TO SEE HOW UpDn WILL CONDUCT THEIR CONVENTION.

      0

      0
  31. HH IS A FREE MASON . HH IS NOT A GOD FEARING SDA AS HE WANTED TO POTRAY HIMSELF.
    HOW DO YOU EXPLAIN THIS, A SATAUNCH SDA ELDER WHO CAMPAINGS ON SATURDAY AND GOES TO A ROMAN CATHOLIC CHURCH ON SUNDAY IN KASAMA WITH GBM.
    PEOPLE I HAVE FRIENDS WHO ARE REAL SDA, THEY DO NOT EVEN COOK ON SATURDAY.
    REPEAT HH IS A FREE MASON
    DO NOT MAKE A MISTAKE OF VOTING FOR HH

    0

    0
  32. Let HH be cheated since he is not in touch with us people on the ground.any honest UPND cadre can easily tell that PF will win after talking to many people in compounds,streets,markets,offices,etc who want PF to continue in power. majority people want PF to continue due to good projects which our beloved Sata initiated! Plus so many voters will still be voting for Sata! Honestly only a miracle would make HH win this election! OTHERWISE,PF IS RETAINING POWER COME WHAT MAY AS THIS PARTY IS STILL LOVED BY MASS ZAMBIANS!EDGAR LUNGU IS JUST LUCK TO BE ANOINTED BY PEOPLE’S FAVOURITE SATA! GO PF GO!

    0

    0
  33. This think Tank is very dull. An opinion poll is not a survey where all conditions of research are met. I am sure he is aware of quasi studies also. Unlike a survey where you randomly choose who to participate in your study, in an opinion poll especially one conducted via SMS and the internet the researcher has no power to choose who participates. The only thing they can do is encourage people to participate. What a shame on UNZA to produce such a dull chap…

    0

    0

Comments are closed.