Indikator Politik Questions Sample Bias in IndexMundi's Cross-Country Data
Founder and Senior Researcher of Indikator Politik Indonesia, Burhanuddin Muhtadi, has issued a critique regarding the methodological validity of research released by IndexMundi Global Surveys. The focus of his concern is the standardisation and reliability of the sampling system used by the international platform to map public perception.
According to Burhanuddin, a research conclusion published in the public domain should ideally be supported by rigorous methods to avoid biased or inaccurate results. “The main weakness of the IndexMundi Global Surveys methodology lies in the use of open web-based polling without strict sample controls,” Burhanuddin stated in a written release on Monday (6/7/2026).
The Professor of Political Science at UIN Syarif Hidayatullah Jakarta assessed that the mechanism of collecting data through open online polls has systemic limitations in fully representing the real population conditions of a country. Consequently, the data gathered only reflects the accumulated views of a segment of internet users who happened to choose to participate in the questionnaire on that platform. “The data produced reflects the subjective perceptions of internet users, not empirically verified statistical data,” he said.
Furthermore, Burhanuddin explained that the fundamental problem with such open digital research models is the emergence of sampling bias. This occurs because the characteristics of respondents tend to be homogeneous, covering only individuals with stable internet access, technological literacy, and proficiency in the language used on the website. This condition means the sample is not drawn through pure random sampling, resulting in a respondent composition that fails to represent the macro demographic structure of a country, including age distribution, gender proportion, income levels, or proportional regional representation.
Beyond sampling bias, the voluntary nature of the data collection system also opens the door to self-selection bias. In an open website scheme, anyone visiting the page can freely fill out the survey. This pattern risks the recorded opinions being dominated by specific groups with extreme sentiments, whether highly positive or highly negative, towards a particular issue. “Without identity verification, the loose filling system creates a loophole for an individual or group to fill out the survey multiple times using VPNs or different devices to manipulate a country’s ranking,” Burhanuddin noted.
He also highlighted the issue of research transparency, particularly the absence of public information regarding the exact sample size used to calculate the final scores for each country. He argued that without a minimum sample threshold, a country’s ranking fluctuation becomes vulnerable to being influenced by very small changes in responses. The institution is also considered less than transparent about its data cleaning mechanisms to filter out interference from bots, spam accounts, or invalid answers.
Academically, Burhanuddin concluded that the data output from IndexMundi is more appropriately positioned as an early indicator or a general overview of sentiment dynamics in the digital space, rather than a reflection of factual conditions on the ground. He suggested that such data should be verified using results from international research institutions that apply rigorous sampling methodologies, such as Transparency International or Gallup Poll. “Therefore, international organisations and academics reject using IndexMundi Global Surveys as a primary scientific reference,” he concluded.
Source: jawawa.id (https://jawawa.id/newsitem/indikator-politik-questions-sample-bias-in-indexmundis-cross-country-data-1783439471)
Originally published by Media Indonesia (https://mediaindonesia.com/politik-dan-hukum/908561/potensi-bias-sampel-indikator-politik-soroti-validitas-data-komparatif-antarnegara), translated from Indonesian