NEWS12 August 2009

Data points to cultural difference in community participation

Features North America Trends

US— Researchers at Communispace are investigating whether culture affects the types of activities people are willing to take part in within online communities.

The Boston-based agency recently analysed one of its multinational online communities and discovered that Japanese and Thai members were less likely to participate in discussion activities, but more likely to complete traditional survey tasks.

Analysis of other communities are now under way to see if this pattern is repeated. Results aren’t due back until September, but Communispace’s director of research Manila Austin told Research: “Having now looked at seven or eight more international communities and all the data collected so far, I think that when we get out next results we will find that south-east Asian cultures are under-represented [in discussions].”

Possible reasons for this are twofold, says Austin. Perhaps the simplest explanation is language differences. “An open-ended format where there’s a burden of being proficient in a second language [usually English] is perhaps one reason to think that some of these countries are under-represented,” she says.

But cultural differences may also come into play. As part of its research Communispace surveyed 1,300 of its community members from 37 different countries using a method developed by Maastricht University professor Geert Hofstede, which ascribes cultural values to several elements, including acceptance of power hierarchies, individualism versus collectivism, and willingness to accept uncertainty.

Hofstede found – as did Communispace – that Asian countries are in general more collective-minded. And according to senior vice president Julie Wittes Schlack, “those cultures that are at the individualist end of the spectrum are over-represented in discussions, whereas cultures at the collectivist end of the spectrum are over-represented in surveys”.

Why this should be is not clear. Wittes Schlack says: “Some cultures might see it as immodest to say, ‘Well, this is me, this is what I think, my opinion matters’”, preferring the anonymity afforded by surveys. But Austin suggests the problem may also lie in how people are invited to take part in discussions. Perhaps participation could be improved if invitations focused less on encouraging people to have their say and instead put more emphasis on helping achieve community goals.

“What we do know for sure,” says Austin, “is that it is important to provide people with multiple ways to participate so that we are ‘meeting people where they’re at’ regardless of culture.”

  • You can read more thoughts on culture and its impact on multinational communities in the ‘Breakthroughs without Borders’ white paper, available online here.

Brian Tarran

Features North America Trends

@RESEARCH LIVE

2 Comments

Esteban Kolsky

15 years ago

This is nothing new. We have known about culture issues when completing surveys, participating in focus groups, and becoming member of communities for a long, long time. This is one of the reasons multi-language surveys must be produced separately as opposed to translate it. Different cultures are not only more or less inclined to participate in surveys, they also react different to words and phrases we find common in our surveys (certain groups don't believe a transaction can be satisfying - thus asking for customer satisfaction will yield crazy low scores, for example). Also, companies that aggregate results from similar surveys in different countries find that the scores vary dramatically in some cases. It is never a good idea to tackle global surveys as a single entity. Finally, there are several cases I read about where communities that were deployed failed miserably in one place and succeeded in another. Again, an issue of culture adopting a model over another. While it is easy to collect a country-by-country list of problems (I have been asked about it repeatedly over the years), it would be counter-productive. Even stereotypes sometimes don't come true. It is better to do pilots for the feedback events you are trying to deploy and see the results that assume that something does not work in a specific situation.

Like Report

Manila Austin

15 years ago

Thank you, Esteban, for your thoughtful comments. Your example about customer satisfaction surveys is instructive, and I agree that stereotypes are uncertain predictors of behavior—on- or off-line; certainly this research underscores the importance of providing multiple modes of engaging and expressing one’s self…and not jumping to conclusions about what or won’t work across cultures. As to the newness of these findings, I do think that studying cultural differences as they manifest in online settings is a valuable line of inquiry; in designing and conducting this study we did not see, for example, other studies that applied Hofstede’s measure to Internet communities or other social media forms. (And I would be delighted to know of any, for as Brian points out, I am currently engaged in a second round of research that looks at participation and language across 7 or 8 of Communispace’s multi-country communities.) I think the phenomenon you mention—where a community is adopted in one country but fails in another—is critically important to understand; companies will need to rely more and more on online, social media methods for listening to and connecting with global consumer and customer groups as Internet penetration and usage continues to grow. Finding ways to respectfully and successfully engage diverse groups online will continue to be a key for corporate learning and competitive advantage.

Like Report

Featured company