GodIndex by country: what the old comparisons reveal

Six country views, not six national surveys

The first report discussed France, Germany, Poland, Russia, the United Kingdom and the United States. A label on a chart establishes that the project published that view. It does not establish the demographic composition of the people behind it, how their country was determined, or whether the participating group resembled the national population.

A low score in a country view should therefore be described as a low score among the responses feeding that view. It should not become a statement that a whole nation disliked god or supported a political position.

Three more countries named in the early commentary

A recovered 14 July 2008 post mentions increased participation from China, Japan and South Korea. Together with the six country views in the report, this provides evidence that the project discussed participation across at least these nine named countries.

The three additional countries are mentioned in commentary about changing index composition. No counts or independently verified participant locations accompany that claim. We do not equate the existence of a named country in a post with a recovered national index series.

Read the July composition note.

A conflict seen through different curves

The original authors described contrasting movements during the Georgia war. Their political explanation is an interpretation of those curves, not direct survey evidence about why each participant voted. A rating of god is not the same question as approval of a military action.

For researchers, the contrast suggests a question about information environments and response composition. To test it, one would need comparable time windows, response counts, participant characteristics and independent measures of news exposure. Those are not supplied by the images alone.

Where are the voter counts?

We have not recovered a complete table of unique voters or votes by country. The global participation total cannot be divided across countries using chart height, line position or a country’s population. Nor can a reported contribution to the composite be treated as a head count without a verified formula.

A 2009 eKAI story mentions approximately 65,000 Polish users and approximately 700,000 worldwide, using language about daily activity. Those are historical press claims with ambiguous definitions, not a verified country breakdown. The press notes explain why we do not combine them with the 2008 report totals.

How a future comparison could be useful

A responsible country comparison should publish the time period, response volume, definition of location and treatment of small groups. Readers should be able to distinguish a difference between countries from a shift caused by a different mix of respondents.

Translation also matters. An apparently identical question may carry different religious or cultural assumptions across languages. Before ranking results, collaborators could test whether people understand the prompt and scale similarly. A pilot with that goal would be more informative than adding more flags to a map.

We welcome cross-cultural research proposals and additional dated country-level source material.