GAMAAN, the Group for Analyzing and Measuring Attitudes in Iran, is an independent, non-profit research foundation registered in the Netherlands. We study Iranians’ attitudes towards different social and political topics. 

Surveying in an authoritarian context

The rationale for GAMAAN’s innovative approach, spreading surveys on a large variety of digital channels and collaborating with VPN tool providers, in particular Psiphon VPN, is the fact that conventional survey modes like face-to-face and telephone interviewing cannot yield valid results in the existing Iranian context. In an environment of state repression, individuals often censor their true views or even actively alter them to avoid scrutiny by authorities—a phenomenon known as “preference falsification.” While sampling bias is a challenge in any country, Iran presents an additional obstacle: collecting data untainted by the “fear-of-the-state bias.”

We believe that evidence matters and offer results that consider scientific standards and the Iranian context. We do not assume that the same survey modes and methods used in free societies produce valid results in authoritarian contexts, but examine similarities and differences by engaging with the state-of-the-art in scientific analyses of survey methodologies in general and Iran in particular.

Online surveys

In 2019, we started testing and working with multiple chain referral sampling, a combination of river sampling (online opt-in samples gathered through social media), network sampling (starting “seeds” to distribute the survey in specific networks), and quota sampling (real-time monitoring of samples’ socio-demographic distribution, starting new seeds to improve sample quality). The surveys repeatedly went viral, which helped to sample tens of thousands of respondents, including from groups such as women without higher education, people with low income, in rural areas, marginalized groups, and also regime supporters.

Since the 2022 nationwide protests, we collaborated with satellite television channels (once, in December 2022) and in all surveys with VPN providers (Psiphon VPN, once also with Lantern VPN) to innovate existing sampling methods tailored for the Iranian context. We assessed the extent to which random VPN-sampling in Iran can approximate probability surveying. This question became more pressing when the regime filtered commonly used apps such as WhatsApp and Instagram, while on the other hand Internet penetration grew to a percentage similar to Western countries. According to the latest statistics from the International Telecommunication Union, around 80% of Iran’s population are internet users, and research within Iran indicates that around 90% of these internet users utilize Internet censorship circumvention tools. Distributing the survey through random sampling among users connected to VPN tools mitigates network bias (i.e., the likelihood of participants sharing similar views with the organizers). 

Extracting representative samples 

In addition to the main questions of each survey, we ask respondents about their demographic characteristics (gender, age group, level of education, province of residence, urban or rural area, employment status, household income level, spoken language at home, type of health insurance); we also ask about their political orientation and past electoral behavior. Since respondents participate anonymously without questions about personal characteristics, they feel more secure in expressing their genuine opinions. 

To extract representative samples, we use matching and weighting methods; we then compare our results with external data and other survey institutes’ results for non-sensitive questions such as about employment, household income levels, languages spoken at home, and health insurance types. In this way, we strive to understand the extent to which our results can be said to be representative. GAMAAN’s survey reports present and discuss the target population and sample demographics, weighted results, and comparisons with external data and other survey institutes’ results.

Academic analysis and ethics

We present our work at international academic conferences, write journal articles, and collaborate with researchers based in survey institutes, universities, and think tanks. Our approach has gone through ethical review at Utrecht University and together with staff specialized in data management, privacy, and ethics, and with human rights experts working on Iran, GAMAAN’s approach was approved after conducting a Data Protection Impact Assessment.