From McClatchy.com:

Not only were the October job losses higher than anticipated, Labor Department statisticians sharply revised September's jobless figures.

Statisticians acknowledged that the preliminary estimate of 159,000 jobs lost that month was far off, and that there were instead 284,000 workers tossed into the ranks of the unemployed. They also revised August jobless numbers from the estimate of 73,000 to 127,000.

These revisions are significant because they suggest that the economy was suffering a steep drop even before the financial crisis exploded into a global problem in mid-September. Many of the September job losses preceded the financial meltdown.

A good chunk of the revisions reflected changes in jobs for teachers and school employees, a statistic that state and municipal governments compile and report with a lag. Another part of the revision reflected greater layoffs from small employers, who weren't captured in the broad employer sample. These small-company jobs often show up in revisions once statisticians have seen payroll and unemployment insurance information from state governments.

"It suggests that small firms did worse than expected," Wyss said.

My point with all the talk of revisions comes from the fact that I teach in China.  The Chinese only give their data once, no revisions.

I am just writing up a PowerPoint presentation to put on the Intermediate Macroeconomics website for the class so that they can see how the data are calculated and understand more of the nuances of research as well as data collection. 

Many wish to come to the U.S. for graduate school - the next year will be especially tough for them as MANY U.S. graduates will be heading to Graduate school to escape paying student loans as there may not be many jobs.