FLAWED poll results due to incorrect methodology may affect the credibility of the studies that are being carried out by research centres, said two research bodies.

The comments from Ilham Centre and Merdeka Center come after a survey conducted by O2 Research Malaysia Sdn Bhd was criticised for allegedly being methodologically distorted and only reflected the opinions of media groups’ readership rather than the Malaysian electorate.

The survey by 02 had gauged the popularity of Prime Minister Anwar Ibrahim and his new unity government.

Ilham Centre executive director Hisomuddin Bakar said the research commissioned by five news organisations only involved readers from said outlets.

He said the research body must have a proper sampling technique and must also include respondents from various demographics.

“The sampling itself is one of the most important aspects when you carry out a survey. Even when you have 1,200 respondents, if you carry out the correct methodology your result will be accurate.

“There are clear flaws in the study carried out by O2. One is that they interviewed 81% male respondents but the census shows 52.3% of people in Malaysia are males.

“This is the first time poll methodology has been criticised by academics. The impact will be felt by other research bodies, it affects the confidence of the people on studies carried out by pollsters.

“It does not matter how many respondents they have as long as they follow the right methodology and know their objective,” said Hisomuddin.

Earlier this week, the head of Universiti Utara Malaysia’s (UUM) Malaysian Institute of Political Analysis (Mapan) Mashitah Mohd Udin said the survey carried out by O2 Research Centre was not representative of the electorate.

This was after the body suggested that the public was generally unhappy with Anwar’s administration.

“There were 35,077 respondents but it only reflected the audiences of the five news organisations.

“Normally, scientific surveys have a high threshold to be considered legitimate, through face-to-face or telephone surveys among respondents chosen through random stratified samples. Or, at the very least, simple random sampling encompassing gender, race, age, economic background, education level, income levels and locality.

“This did not happen for the survey (in question),” said Mashitah.

The survey in question was commissioned by Sinar Harian, Awani, The Star, Sin Chew Daily and Malaysia Nanban in conjunction with Anwar’s 100 days as prime minister.

Among others, the survey suggested that 71% of Malays were “not satisfied” with the government’s effort and ability to fulfil their “GE manifesto”.

Other ethnic groups were generally more confident in the government to deliver on those promises.

The survey results paint a negative public perception of the government regarding political stability, ethnic relations, education, inflation, jobs, wages, natural disasters, economic growth, food supplies, welfare, health and economic resilience.

Mashitah argued that apart from the need for proper survey methods, it would take time before the people could properly assess the government’s performance.

She said the government only started rolling out new policies, such as Malaysia Madani, on January 19 while Budget 2023 was only tabled on February 24.

Reflecting reality

Ibrahim Suffian of Merdeka Center said opinion polls must be conducted properly in order to yield results that reflect reality even if a survey consists of just 1,200 respondents, it can represent Malaysia’s 21 million voter base.

“Online surveys based on individuals answering questions without being selected through proper methods are usually unscientific and their results do not represent the population of a country.

“In summary, conducting surveys is not an easy task. Recently, even surveys conducted by independent bodies in developed countries sometimes miss their mark and fail to predict outcomes such as election results.

“This usually involves the accuracy of obtaining respondents that represent all groups.

“Therefore, it is important for pollsters to carry out surveys that meet scientific standards so that the results obtained reflect the actual situation.

“This not only promotes a culture of constructive criticism in the context of democracy in Malaysia, but it also helps to increase public trust in the country’s surveys,” Ibrahim said in a statement.

Following the release of the survey, Communications and Digital Minister Fahmi Fadzil has cast doubt on the findings saying the findings were unreliable and not reflective of the people’s views.

He said the ministry will meet with the media agencies concerned to get clarification on the survey. –  March 4, 2023.