Just hours after La Nación reported the results of another survey by Unimer Costa Rica, the security ministry was issuing a press release lauding the numbers.
The problem was that the ministry did not get the details correctly.
The survey showed that insecurity no longer was the top concern of the majority of the residents. The ministry correctly noted that those listing insecurity as their top concern amounts to just 18 percent of those surveyed. That is a big drop from the 48 percent who said the same thing in 2009.
The ministry said that concern about insecurity had dropped to third position among Costa Ricans. Actually the news story in La Nación showed it to be in second place.
Some 21 percent of the respondents said the employment was their main concern. In third place was the cost of living, cited as the primary concern by 17 percent of the respondents.
Lesser primary concerns were corruption, the economy and poverty, according to La Nación.
The Laura Chinchilla administration certainly will be remembered for the way it put a favorable spin on daily events. But it appears that at least the security ministry will get a failing grade in statistics.
What the Unimer firm did was survey about 1,200 Costa Ricans. Then they projected the results from these respondents to the Costa Rica population. This is inferential statistics, and it is not a good idea to consider the results as highly accurate.
La Nación itself reported a margin of error of plus or minus 2.8 percent for the results. That means employment, cost of living and insecurity are basically tied for the top concern of all Costa Ricans.
If the Unimer survey is accurate and not one of those soundings that are incorrect one out of 20 times, the security ministry has less to cheer about.
Such surveys give broad answers, as those who follow election polls know. The closer the election race the less useful are the polls.