Updating a quick question

Last Friday I noted that jobless claims had shot up in the previous week but that economists said to avoid reading too much into a single week's figures, urging instead looking at a four-week average to smooth out natural volatility.

True enough, I said, "but why is that only mentioned when it makes things look better?"

Well, this week's figures are out.

Washington (AP, April 22) - Fewer people signed up for jobless benefits last week, a sign that companies are feeling less inclined to slash their work forces now that the economy is rebounding.

The Labor Department reported Thursday that new applications filed for unemployment insurance dropped by a seasonally adjusted 9,000 to 353,000 for the week ending April 17. The decline, which came after new applications rose sharply in the previous week, left claims at a level that was higher than the 340,000 analysts were expecting.

See anything about a four-week average in there? Neither do I. Interestingly, my very rough calculation indicates that a four-week average would show a 4,000-5,000 increase rather than the one-week 9,000 decrease. But that couldn't have been the reason for not mentioning a four-week average.

Could it?