"尹's 40% approval rating, questions not biased." The real problem that pollsters analyzed is...

YTN Radio (FM 94.5) [YTN News FM Wise Radio Life]

□ Broadcast date and time: January 8, 2025 (Wednesday)
□ Host: announcer Park Gui-bin
□ Castor: Kim Heon-tae, CEO of Macy's Consulting (former head of the Korea Social Opinion Research Institute)

* The text below may differ from the actual broadcast content, so please check the broadcast for more accurate information.

◆ Announcer Park Gui-bin (hereinafter referred to as Park Gui-bin): There was a recent poll showing Yoon Suk Yeol's presidential approval rating at 40%. This is the result of the investigation that came out after the emergency martial law, with the president's duties currently suspended due to the passage of the impeachment bill in the National Assembly. The survey was conducted by Kopra, the Korea Public Opinion Research Institute, on 1,000 people aged 18 or older nationwide from the 3rd to the 4th at the request of Asia Today. The 95% confidence level and error range were plus or minus 3.1 percentage points at 4.7%. For more information, please visit the website of the Central Election Public Opinion Review Committee. Let's take a closer look with an expert on the content of this poll. I'll connect you to Dr. Kim Heon-tae, a pollster. Hello, doctor.

◇ Kim Heon-tae, CEO of Macy's Consulting (former head of the Korea Social Opinion Research Institute) (hereinafter referred to as Kim Heon-tae): Yes, hello.

◆ Park Gui-bin: A poll shows that the president's approval rating is 40%. There were also opinions that raised some issues with this investigation. Doctor, what do you think?

◇ Kim Heon-tae, CEO of Macy's Consulting (former head of the Korea Social Opinion Research Institute) (hereinafter, Kim Heon-tae): Before we talk about the approval rating itself, the political situation is at a very sensitive and critical point, such as martial law and impeachment. So, it's a pretty sensitive time for the numbers announced under the name of the president's approval rating, so I think it's controversial. And one thing is that if you look at the general public opinion, it seems that President Yoon is held accountable for martial law, and the figure itself that the president's approval rating is rising is quite resonating with people. I'm looking at it like this.

◆ Park Gui-bin: There's a bit of a controversy. As you said, so let's point out the issues one by one. First of all, it is pointed out that the question is biased. Is there anything wrong with the question?

◇ Kim Heon-tae: I've been going through the questionnaire all along. The presidential approval rating question itself is not very biased in terms of sentence composition, questioning method itself, or view. It's not inducing or anything like this, but the question about the president's approval rating itself is not a problem with the question itself, but other experts point out the question of the next question. The next question is what do you think about the forced arrest of the incumbent president by the Senior Civil Servant Corruption Investigations Unit despite the illegal controversy over the arrest warrant for President Yoon?

◆ Park Gui-bin: So you mentioned negative information in the question itself.

◇ Kim Heon-tae: Right, because if you ask what you think about forcibly taking a sitting president despite the illegal controversy, you can get the nuance of how you want to take a sitting president when you know that there's a problem. For some respondents, if you change the question, you will understand it quickly. If President Yoon Suk Yeol asks what he thinks about the forced arrest of the Senior Civil Servant Corruption Investigations Unit by refusing to attend the Senior Civil Servant Corruption Investigations Unit three times, many people of course did not, so it is not a good questionnaire to inform them that there is controversy in this way.

◆ Park Gui-bin: So if you look at the results of those respondents, can you judge that this question has had a little influence?

◇ Kim Heon-tae: In a sense, 57% of the respondents said that it is actually desirable to be forcibly taken, even though it seems a little biased, as controversial as it is now.

◆ Park Gui-bin: So that's the answer to that question, right?

◇ [Kim Hun-tae] That's right. So, or forced to do so is not desirable. In fact, if we accurately diagnose the current trend of public opinion through such surveys or results, public opinion is very negative about President Yoon's refusal to investigate by judicial institutions. Rather, it may be the result of showing this.

◆ Park Gui Bin: Oh, you're right. Then, no matter what positive information is contained in the question, it's not such a good question, is it?

◇ Kim Heon-tae: That's right. In fact, how much do you support the president in this cultural poll about the current president of Yoon Suk Yeol? I ended up asking like this. In fact, do you support or disagree with this rather than how much you support it? Do you support it again? Don't you support it? You have to maintain this level of neutrality quite mechanically, and in this way, despite the controversy, or by attaching an explanation, you refuse to be forced to perform. This is actually likely to create a bias against a particular response.

◆ Park Gui-bin: The first question, as you just said, pointed out the problem because you asked me in one direction whether I supported it in a straightforward way, unlike other investigative agencies.

◇ [Kim Hun-tae] That's right. In this part, we need to discuss whether it is reasonable to ask for support for a president whose job has been suspended. So that's a bit of a problem, and if you look at the approval rating question, you don't even ask about the president's state administration. So I'm simply asking if I support President Yoon. So this part can also be a little controversial. In fact, then these parts are actually ambiguous whether this is asking the president's personal support or political support or support for state administration. So I think that's something that needs a little discussion.

◆ Park Gui Bin: That's right. You said that it is different from the usual method of investigation.

◇ Kim Heon-tae: That's right. During this copra-like survey, I saw a poll for Chuseok last year. I asked how he asked about the president's approval rating, and normally there, the teacher asks me how I evaluate President Yoon Suk Yeol's overall state administration. So it's not the usual or usual way to ask President Yoon Suk Yeol directly whether he supports President Yoon Suk Yeol. For now, it may be because the job is currently suspended, but I think controversy related to this part can inevitably arise.

◆ Park Gui-bin: Where should we usually look at the main body of the place where these questions are made?

◇ Kim Heon-tae: That's actually because the answer is through consultation. However, in the field, there are cases where the person in charge of a poll company or poll agency reflects more opinions, or in some cases, the media company's specific media carries out more arguments and opinions. So it depends on the situation, but from that point of view, it will be a little difficult to know whose opinions this survey reflected more.

◆ Park Gui-bin: Yes, you mentioned earlier that major media research institutes are not currently investigating the president's approval rating for state administration. Come to think of it, haven't you usually done regular research on places like Gallup and Realmeter? I'm not doing it now, but is it because I'm not doing my job?

◇ Kim Heon-tae: That's right. Anyway, the meaning of asking for support for the president's Yoon Suk Yeol is actually ambiguous when the president is not running state affairs. That's why other investigative agencies have not asked for presidential support recently, but what I want to tell you today is that when I went to the NEC's website today, various investigative companies asked for presidential support. So, I think I asked the question yesterday, so when the results of the survey come out, some say that the president's approval rating is rising due to the gathering of conservatives. Or this is a really ridiculous thing about the results of the investigation. I think there will be more information on this part.

◆ Park Gui-bin: Did you conduct an investigation asking for support for the president? At another organization recently?

◇ Kim Heon-tae: Yes, it's a poll review committee on the NEC's website this morning. When I went to the website, I think some research companies did research to verify this because this was a problem. However, the results of the investigation have not yet been announced.

◆ Dr. Park Gui-bin: You said this in an interview before. Since there are so many polling companies, it's necessary to make a judgment based on the company's reputation and brand. When you mentioned this, what is the Korea Public Opinion Research Institute called Kopra, the institution that conducted this survey that we are referring to now?

◇ Kim Heon-tae: I don't know. I think I can tell you two things first. One was not so active. According to this poll, it's the first time I've done it with a media company since September last year. As far as I can see, I haven't done public opinion polls at all for almost three or four months. So, it doesn't seem that much recognition or reliability has been accumulated because it wasn't that active. Second, this survey method combines a kind of number called ARS and RDD and performs dialing. As I said in the last interview, I've mentioned that there are usually high-end and low-end types depending on the survey method, but in fact, the ARS method is a low-end type among experts. Therefore, the reliability tends to be slightly lower than the polls conducted through the mobile carrier list. So I think we can say that this survey is also difficult for us to see the reliability very high.

◆ Park Gui-bin: So in the case of ARS surveys, is it a little unreasonable to actually reflect public sentiment?

◇ Kim Heon-tae: It's hard to say that, but I think it can be a reference point for how much data you believe. And this time, the response rate that many people talk about is also relatively low. On the other hand, in the case of random rotation through the mobile carrier's list and the interviewer asks, of course, the quality is high and expensive. So what I always insist on is that in the case of ARS, it's better to put the name ARS poll or ARS poll in front of it than the name of poll.

◆ Park Gui-bin: He is an expert among the poll experts in Korea who served as the head of the Korea Social Opinion Research Institute. We are conducting an interview with Dr. Kim Heon-tae. Another issue is the response rate. I'm sure the supporters of the Democratic Party would have cut it off in the middle. There are comments like this. How do you see this?

◇ Kim Heon-tae: Well, I also have that point, but it's not easy to say that there's no basis at all. Because when you compare the response rate itself, some people say that the response rate is 4% too low, but most of the ARS polls have a response rate of 3 to 4%. That's why the response rate is always lower than the random sampling through the mobile carrier list I mentioned earlier. So the Copra survey doesn't have a particularly low response rate.

◆ Park Gui-bin: Yes, 4.7%.

◇ Kim Heon-tae: Yes, so if you look at other companies that use similar ARS polls, there's 3%, so it doesn't seem to be low in itself. Also, if this is the case, if many Democratic Party supporters or people who have resistance to President Yoon hung up, the response rate should be lower, but it wasn't necessarily the case. So if we want to see this clearly, there is originally an example of the actual response rate or response rejection rate, can survey companies print data by age and region? In fact, if so, we can verify that those who have relatively high support for President Yoon by age showed more response rates. Experts say, but it is difficult to affirm this now because the response rate is not particularly low compared to other ARSs.

◆ PARK GUEBIN: Yes, the Democratic Party's base was a question that led to a break in the middle. I asked him because he was making this claim, but the Democratic Party is now saying that it will file another complaint against him for violating the Public Official Election Act because it is a biased poll.

◇ Kim Heon-tae: If there is some clear data, of course, there should be a complaint or a criminal charge. In fact, it is not easy to say that there is a real problem with the data released so far, and the other thing is that some media companies asked about the president's approval rating, so as the ruling party claims, the president's approval rating has risen slightly recently due to conservative assembly. Also, opposition support was falling. In line with this trend, I think it will be a little clearer whether this is the result or whether there is actually a problem with a result that is too different from other research companies.

◆ Park Gui-bin: Yes, the agency that conducted this investigation. Kopra, the Korea Public Opinion Research Institute, is arguing that this is not the part we are talking about, this biased opinion poll. Then how about this one? Among some experts, the conservative group may be over-represented. There were also comments like this.

◇ Kim Heon-tae: There are not many questions that reflect the opposition's position or ask the opposition's position, but it is true that there are more questions that reflect the ruling party's argument. However, there is currently no evidence to suggest that the Democratic Party's supporters continued to hang up while listening to this question, and then conservative respondents actively participated in the poll and that the response rate was oversampled. It's not easy to say something like this because someone did it like this.

◆ Park Gui-bin: Yes, with the announcement of the results of the poll this time, in fact, not only the president's approval rating came out. The next presidential suit and political party approval ratings came out. Were there any other points that we should pay close attention to in the results of those polls?

◇ Kim Heon-tae: As I looked at that part, there was a little bit of a mystery, but maybe this part needs to be explained more actively by the investigative agency. Because on the same day, other media companies also polled. There was a poll for other media, and despite the survey on the same day, the president's approval rating was similar, but when looking at the approval rating of one party in the corporate company, the ruling party of people's power was higher than the Democratic Party of Korea. In fact, in this current situation, there are no cases in which the support of the ruling party of people's power is higher than that of the opposition party. This is where any other research company, and even though the same company did the same day, this is where we can actually raise a significant question of whether the results of this research company's poll are reliable.

◇ Kim Heon-tae: So I think that the investigative agency itself needs some clarification or explanation on this part.

◆ Park Gui-bin: When you conduct a normal poll, you see a similar trend or trend. Usually?

◇ Kim Heon-tae: Yes. So maybe that's probably the biggest reason why it's hard to manipulate polls is because other research companies do it together. So if they get weird results, it's quite a problem. So that's probably one of the most important parts of maintaining the public confidence of the research company.

◆ Park Gui-bin: It's very difficult to say that a true poll is how to read it and interpret it whenever it's released. What should we do to not misunderstand or be deceived by the results of this poll? So you have to be aware of this. If you know this and look at the polls, can you point it out?

◇ Kim Heon-tae: First of all, we have to compare the questionnaire. In fact, I think it is a desirable phenomenon for people to look at the questionnaire or the composition of the questionnaire like this after this controversy. So by looking at the questionnaire questions, this research company asked like this and that research company asked like that, so I can see how I should look at the results like this. So one thing is important to look at the questionnaire like that, and the other thing is to look at the survey design and compare it with the previous survey of the same survey company or other survey companies. So, it's most important to look at this trend and judge how it came out last time, whether it's going up or down. And when comparing the results of the poll, it is always important to compare the same survey designs to see if this is the ARS survey, the mobile carrier list method, or the interviewer method.

◆ Park Gwi-bin: Yes, I see. So far, this has been pollster Dr. Kim Heon-tae. Thank you, doctor.

◇ Kim Heon-tae: Yes, thank you.