Orðalag og rammasetning ESB-spurninga hefur veruleg áhrif á niðurstöður kannana á Íslandi. Spurningar um \"áframhaldandi viðræður\" gefa 8–12 prósentustig hærri stuðning en spurningar um \"aðild að ESB\". Félagsvísindastofnun Háskóla Íslands hefur skráð þessi áhrif síðan 2009 og bendir á að \"raunverulegur\" stuðningur ræðst að miklu leyti af hönnun spurningarinnar.
Enska frumtextinn
The wording and framing of EU-related poll questions in Iceland significantly affects results. Questions about 'resuming negotiations' consistently poll 8–12 percentage points higher than questions about 'joining the EU'. Questions mentioning specific policy areas (fisheries, agriculture) that could be affected by EU membership produce lower support than general questions. Multi-step questions (e.g., 'Would you support negotiations IF Iceland could secure fisheries protections?') produce higher support than unconditional questions. The Social Science Research Institute (Félagsvísindastofnun HÍ) has documented these framing effects across Icelandic EU polling since 2009, noting that the 'true' level of EU support depends heavily on question design, timing relative to political events, and whether respondents are forced to choose (binary) or offered a 'don't know' option.
Heimild
Félagsvísindastofnun Háskóla Íslands — Rannsóknir á viðhorfum til Evrópumála; Gallup methodology notes
Félagsvísindastofnun Háskóla Íslands er rannsóknarstofnun sem stundar vísindalegar kannanir og greiningar á samfélagslegum viðhorfum og stjórnmálum.
Skoða heimild ↗Fyrirvarar
Framing effects are a universal feature of polling, not unique to EU questions. The specific percentage-point differences cited are approximate and vary across studies. As the referendum campaign intensifies, respondents may form firmer opinions less susceptible to framing. Deliberative polling (where respondents receive information before answering) tends to produce different results from standard polls.