NasDem's Summons to LSI Denny JA: A Reminder of How to Evaluate Survey Results

If survey institutions are employed truly independently and honestly, they accurately depict people's political participation; nevertheless, if they are presented mostly from surveys paid by specific sponsors, they will undoubtedly convey a different image, and tend to be unfavorable for opposing politicians.

So are pollsters political communicators or merely image builders for their clients?

In particular, on Monday, October 9, 2023, the National Democratic political party, commonly abbreviated as NasDem, in North Sumatra, held an audience.

Iskandar, the chairman of the NasDem chapter North Sumatra -- stated that the survey result seemed odd in light of the information disclosed by LSI (Indonesian Survey Circle) Denny JA. One of them is the substantial decline in Anies' electability in North Sumatra.

Iskandar and his team objected to the survey results and wished to have them reevaluated because they discovered anomalies.

According to Iskandar, this anomaly was observed in May 2023, when the same survey organization - LSI Denny JA - published the results of a presidential electability poll. At the time, Anies had achieved 32.6%. This implies that Anies' electability decreased by 28% in just four months, or an average of nearly 7% per month. And this has never occurred in historical research.

Regarding the survey results, Iskandar stated that he questioned LSI Denny JA on the methodology, quantity, and distribution of respondents, as well as the survey's funder.

LSI Denny JA's Reaction

Adjie Alfaraby, a researcher at LSI Denny JA, determined that NasDem's response was excessive. According to him, the summons was disproportionate because his agency's survey results met standards and could be explained.

Adjie stated that his institution bore academic responsibility for the released survey results. He considered the responses of several parties to the survey results to be typical. He acknowledged that he frequently receives criticism from parties who are dissatisfied with the survey results they publish.

Similarly, LSI Denny JA was praised by those who benefited from the results they published. Consequently, inreasing polls are typically praised, whereas declining polls are criticized.

As the institute's founder, Denny JA himself expressed his opinions.

In many other situations, like county election cases, the losing side even thought there was high-level interference. They even claimed that the results of this survey were manipulated to justify election fraud or regional elections in the future.

This legal subpoena does have a silver lining. At the very least, he compels us to offer some standard advice or direction. There are three essential guidelines for evaluating survey results.

1. Consider the survey firm's track record.

This epoch is dominated by Google and the internet. Digital traces exist for everything.

LSI Denny JA is the oldest survey institution in Indonesia that is presently conducting public opinion surveys. Since 2004, this institution has provided a publication a month before election day detailing the winner of the previous presidential election.

In 2004, LSI declared that SBY would be the winner. Even more so, SBY would prevail again in just one round in 2009.

LSI Denny JA predicted that Jokowi would triumph in 2014 or 2019 as well. A month later, the survey's findings and projections for four presidential elections (2004, 2009, 2014, and 2019) were confirmed. Everything is accessible via the internet.

However, there are also more survey institutes whose names were only recently brought to light. Those without a verifiable track record in previous presidential elections have every right to be viewed with scepticism.

2. Consider the survey organization's credibility as well.

Institutions that conduct research and surveys can be evaluated by their development and achievements. For instance, how many institutions receive recognition from international or national institutions?

TIME magazine has awarded LSI Denny JA and Denny JA personally (for his work in survey institutions), for instance. In addition, the Guinness Book of World Records awarded them for surpassing the world record for political education. Additionally, they received an award from Twitter and the journalist organization PWI Jaya.

Awards and accomplishments from national institutions, particularly significant international institutions, demonstrate that the general public values his work.

Moreover, a self-destructive act by an institution that has laboriously built its reputation would be confounding.

3. Surveys should also be viewed within a specific time frame

Standard and typical survey question: "If the presidential/regional elections were held today, who would you vote for?"

This indicates that the survey reflects the attitudes of respondents on the day the survey was conducted.

Obviously, presidential candidates' popularity can fluctuate. For example, a presidential candidate who did very well in a poll in June could do badly in a poll in September. That's because he made a blunder, for instance. However, in the next two months, the presidential contenders who are towering in December may get much taller. This might due to the presidential candidate's exceptional influence.

Such as the 2017 DKI Regional Elections. In January of 2017, LSI Denny took a picture of Anies with his low current poll. In April 2017, however, this institution announced that Anies would defeat Ahok in the DKI regional elections.

Why did LSI Denny JA declare Anies' different position between January and April for the 2017 regional elections? This is because Anies' own electability in the sphere has changed. Credible surveys can document these alterations.

Just pay attention to a few LSI Denny JA publications about the presidential election. Even though Anies is consistently losing support, there is always this caveat: a lesson from the DKI regional elections -- that the final vote is always subject to change.

Knowledgeable Perspectives on Evaluating Political Opinion Surveys

The bandwagon effect refers to the phenomena of survey findings that motivates individuals to prefer persons that are superior in the survey. While the opposite of this phenomenon is known as the underdog effect, which is an increase in sympathy for figures who tend to lose in a survey. The survey results can also be used to determine the strategy for the next campaign.

Pollsters can conduct surveys to gauge public support, as well as conduct quick counts after a general election has been conducted. This is permissible under Law No. 19 of 1997 pertaining to Statistics, which basically permits guaranteed surveys of results summaries submitted to the Central Statistics Agency (BPS).

Article 14 (1) of KPU Decree No. 701 of 2003 legalized opinion surveys for the general election, provided that they included an explanation of the superiority and inferiority of the methodology employed.

This is coupled with Article 186 of the 2008 Election Law Decree, which states that surveys or opinion polls can be conducted to increase political participation in society at large.

In conclusion, if NasDem wishes to outlaw research results, their efforts will be in futile -- as research institutions and their research results are legally protected in Indonesia. There is no indication of fraud, embezzlement, or illegal activity.

And to broaden our perspective, in a democratic nation, it is acceptable to be disappointed or furious with a research study. This is a valid personal opinion. But if you want to speak of the findings of a scientific study, you should first learn the basics of that discipline.

It is apparent that the findings of the poll conducted between September 4 and September 12, 2023 were included in the research that NasDem North Sumatra objected to. There it was plainly written and also announced publicly by LSI Denny JA that the study was based on face-to-face interviews with 1,200 respondents selected using multistage random sampling, with a 2.9% margin of error.

This highlights the significance of the national margin of error of 2.9%.

The margin of error decreases with increasing population size. Nevertheless, if we delve deeply into the data, let alone shrink the national population area to the provincial population area (in this instance, North Sumatra), the margin of error is bound to rise.

So, can we be more intelligent and prudent in our opinions?

***

Monica JR, 2023