Header Ads

Adakah kajian Ilham Center boleh dipercayai?


Ini bukan setakat untuk membezakan analisa pusat kajian dengan sembang kedai kopi, malah lebih daripada itu, ia berkaitan dengan rasa takut kepada Tuhan.
Incorporating Reliability and Credibility into Public Opinion Research: Where Ilham Center is failing us.

In God we trust, everybody else must bring data to the table. This saying is famously attributed to W. Edwards Deming half a century ago. Dubbed as among the fathers of industrial revolution for his monumental contribution in modern quality management, his tenacity and quest for the truth and incessant improvement has influenced many facets of our life. In 1987, he was awarded the National Medal of Technology for his forceful promotion of statistical methodology which form the basis of objective quality measurement and improvement.

In contrast, sentiment sells. Many times this land has witnessed how easily one can pave his access to fame in the mainstream media with intangible allegations and data-less assumptions; provided the sentiments are marketable. Advancement of information technology act as catalyst in its propagation; spreading half-baked ideas and premature conclusion or even fallacies and deceits. As a consequence, never-ending political turmoil, lack of substantiate discourse and wild prejudice are hampering this nation prospect to move forward.

Public opinion is important because it matters. Thus, research that looks into the subject has to be approached empirically and with great caution. The knowledge and skill on how information should be collected, measured, digested and presented; is critical. More importantly when it comes to researching public opinion where ideological clashes may expose and mislead researchers to bias and fabricated conclusion. Though for obvious reasons, research methodology and analysis should always be independent from one political ambition or vendetta.

In academia, it is a norm for a researcher to declare his or her conflict of interest at the beginning of any research presentation or documentation. In cases where there is clashes between personal interest and the research subject, the researcher has to follow that up with mitigation steps taken to ensure impartial analysis and eliminate favoritism. While it is only humane to lean to one side of any argument, a researcher has to demonstrate an unwavering commitment and depth to uphold the truth.

In any research, the result is only as important as the method. For a research outcome to be reliable and credible, the methodology has to be solid, invincible from fallacy. Construct validity is a concept which determine the ability of a research tool to measure what it claims to measure. For example, if CEO of a research center make a claim that PAS is not serious in fighting Barisan National in upcoming election. He has to back up that claim with evidence. There has to be a procedure to arrive at that conclusion. The procedure has to be proven effective and consistent in determining the level of seriousness of any political party in any election in Malaysia. There has to be parameters that can be measured numerically or evaluated subjectively. Or at least, there has to be an operational definition or benchmark to outline the seriousness in fighting Barisan National.  

It goes the same if the CEO make a claim that PAS will only garner 18% of Malay votes in the next local election. He needs to support that on-air claim with on-land proof. There has to be a statistical model to arrive at that percentage. A reliable and credible research center should not have blurted numbers out of thin air. To be valid, the statement must be reinforced with details on the variance and regression involved in the analysis, the confidence interval and the margin of error. If a survey has been done, introduce the conceptual framework, explain the power of the study and accuracy of the sample size calculation.  

Ultimately, if we really want public opinion to matters to politicians, policy makers and public themselves, research or finding related to it has to carry a significant weightage. For that to happen, research has to be built upon a sound methodology and acceptable degree of accountability and transparency. It should endure critical appraisal. This is not just to differentiate between a research center analysis and a kopitiam rumbling, but more than that, this is a question of taqwa.    

Dr Abdullah Fahmi,
Profesional Istiqamah