The good news is that we finally have updated detailed demographic data for Vashon-Maury Island. And the bad news: It is of very questionable quality.

For those of you who follow the figures that tell us how our Island population changes, you are probably aware that the Census Bureau has altered the way it collects information, such as income and employment. It formerly used the “long form” as part of the survey we complete every 10 years. Now such information is collected by the Census Bureau through a monthly sample called the American Community Survey (ACS), then reported as accumulated data.

Sounds good because we do not have to wait a full 10 years to get new characteristics statistics. However, if your population is small, like ours, there is cause to question the accuracy.

Here is what I mean. In 2000, 1,354 (or 13 percent) of our Island residents completed that long survey form. The ACS information just released came from a sample of only 653 (or 6 percent) of those who live here — less than half.

Here’s the other major difference. The people questioned in the year 2000 were asked to report data as of April of that year. Those who contributed information for our most recent ACS statistics were surveyed throughout a five-year period, from 2005 through 2009, and their answers were combined. Technically, this means you cannot compare one-point-in-time data with figures from a five-year period, because they are measuring two different things. But of course we will, since we have no other update from 2000 information.

When one does that comparison, some things, such as average household size and percent of people in the labor force, do not appear to change. But when measuring other characteristics, the ACS says we are dramatically different. That begs the question: Can you trust the information?

One nice thing that comes with ACS figures is that the good statisticians at the Census Bureau give us some idea of how wrong they might be. For every piece of information, they offer a margin of error, a statistical measure that says the reported figure can be off by so much in either direction. The real number lies within a certain range. Unfortunately, however, for Vashon-Maury estimates, the margin of error can be very large.

For example, ACS reports there are an estimated 174 Hispanic/Latinos on Vashon, with a margin of error of 106. In other words, the true Hispanic/Latino population is between 78 and 280 — quite a spread.

The ACS indicates that the population has only grown by 302 (give or take a margin of error of 620 people). Almost all of those must be, as we like to say, gray hairs, because our median age has jumped from 43 in the year 2000 to 50 for 2005-09. That is a pretty extreme increase, leaving us to ask our local real estate agents the true ages of the people moving on and off the Island in the last 10 years.

Or let’s say the Vashon Maury Community Food Bank is completing a grant application that asks for median household income. The ACS 2005-09 estimate is $79,446 — an increase of 36 percent over the $58,261 median income reported in the 2000 Census. However, the application will have to include a lot of qualifiers, since that 2005-09 number blends some economically strong years with the recession that hit in 2008. What’s more, there’s that pesky problem of a margin of error again: This time, it’s nearly $10,000, placing the real median income somewhere between $69,449 and $89,443.

Those reviewing the application will no doubt compare those Vashon numbers to the median income for all of King County. In 2000, there was only a 10 percent difference between the $53,157 median income for the county and what Vashon reported. But by 2009, the ACS says there is an 18 percent difference. In King County, the median income is $67,806, with a $1,200 margin of error — or a range between $66,521 and $69,091.

Conclusion? Vashon seems to be doing pretty well, according to ACS statistics, so why give funds to our Food Bank?

Here is the frustrating thing. I wish I could tell you that our data will improve, that we will get more accurate information in the future. For some things we might.

When the 2010 census figures come out, we will have a more accurate picture of our total population, what percent are Hispanic/Latino and the median age of those who live here. But for other important topics, such as what kind of jobs we hold, what percent of people live below the poverty level, how we get to work and what our housing looks like, we will only have the ACS to rely upon.

I know many Islanders will have reason to consider and even use these new ACS statistics. But when you do, think hard about their accuracy and present these new figures with care.

­— Alice Larson, Ph.D., a social service researcher, teaches the Vashon Demographics class for Vashon College and often makes presentations on this subject to Island organizations.