Community Survey Blog

In our previous posts in this series, we covered demographic data and analyzed the community’s mean responses to the survey’s questions.  But, we suspected that answers would vary across locales.  As we continued to look at the data, we started to examine how different communities responded to the same question.  In this post, we will examine the differences in responses across locales to the first question about the overall health of the community.

First question: The overall health of the community
Here are locale specific answers to that familiar first question: how do community members feel about the community?  You can see that the responses seem pretty positive from country to country.

We then decided to go a bit further with that data we collected.

  One interesting thing we tried is called factor analysis.  When a survey has many response variables, you can run a factor analysis to see if any set of variables can be condensed into a new variable — one that is more easily explained by the common characteristic of all the original responses.  We did the factor analysis and were able to create two new variables called “Community Health and Dynamics” and “Feeling close to Mozilla”.

The above graph is a visualization of how we determined the two factors.  Notice how three of the variables gravitate to the vertical axis that we describe as “Feeling Close to Mozilla”?  Recognizing that, we decided to create a new variable to test, naming that variable “feeling close to Mozilla” since we felt that this was a common description of each of the variables.  The same goes for the horizontal axis.  Three variables gravitated toward what we have called “Community Health and Dynamics”.

With these two new variables from the factor analysis, we can test against our locales.

  In the next graph, you can see responses to the two new variables across locales.  But before anyone draws, remember that this analysis on the two new variables is directly related to the mean response variables we saw in the original question.  And, remember that all those answers were above the midpoint response of three.  Mean responses were above three and we felt pretty good about that.

From the graph, we can start to see that some locales have a slightly lower feeling of closeness to Mozilla (new variable one) and the health and dynamics (new variable two).  It is also important to note that some countries have much higher response rates than others.  So, drawing very concrete conclusions is hard to do.  Like all statistical analysis, it’s really imperative to understand the data before jumping to conclusions

Let’s focus first on Italy and Spain.  What might we conclude by these responses?  It seems that both Italy and Spain have responded with more exuberance than other countries about community health and dynamics and feeling close to Mozilla.  Remember though that this does not mean that Italy and Spain are “better” than others or more supported, it simply means that their communities responded with higher answers on the scale when compared to other locales.  

It also appears that the Russian community, on average, feels a bit less “connected” to Mozilla when comparing their answers to other countries.  Perhaps people from the Mozilla community can do better outreach in Russia to raise this response.  Good news: this may happen! We recently hired some new Mozilla staff to do direct outreach and community building in Eastern Europe.  In the future, it will be interesting to see if responses change after Mozilla takes action.

Interestingly, countries like Hungary and Poland had responses that were more toward the middle of the pack (less so for Poland).  They seem to have a pretty balanced opinion compared to other countries, feeling connected to Mozilla and that the community dynamics are good.

Time for some speculation.  Do cultural differences play into the results we see?  Is it unreasonable to think that different cultures would answer more or less enthusiastically?  We could ask the Spanish what they think.  But, we are not sociologists and are not geared to answer those questions — but, it is not hard to believe that they might.

—————————-

In the next post, we will look at the responses to our second question where we asked about the resources that Mozilla could provide to the community. We will apply again the factor analysis in order to reduce the number of variables and try to learn about the *real needs* of the communities.

Trackback URI | Comments RSS