One should always beware of over-reading or over-interpreting polls.
Polls are useful indicators at a moment of time and may be accurate to a certain number of percentage points.
Or not.
It is particularly hard to poll a primary many, many months ahead of that primary.
Take the potential Republican primary between long-time Republican incumbent Sen. Arlen Specter and a likely challenger, ex-Representative and 2004 primary challenger Pat Toomey.
Toomey came very close to winning in 2004 and is expected to launch a new challenge for 2010.
Two polls came out this week on this potential Republican primary. One showed Toomey 14% ahead of Specter while the other showed Specter 15% ahead.
One of these polls is wrong; on the other hand, perhaps both polls are wrong. The variance is too great.
The bottom line is the Specter is in a world of hurt and could lose a Republican primary.
The other bottom line is to take all polls, particularly primary polls long before the election, with a few grains of salt.
Read on for the poll results
A Quinnipiac Univ. poll conducted March 19-23 and released March 25 – surveyed 1,056 registered voters. Subsample of 423 Republicans
.
Republican Primary Matchup
41% – Pat Toomey
27% – Sen. Arlen Specter
32% – undecided, other
A Philadelphia Daily News/WGAL-TV/Pittsburgh Tribune-Review/WTAE-TV/WPVI-TV/Times-Shamrock Newspapers poll conducted March 17-22 by Franklin & Marshall College – surveyed 662 adults. Subsample of 211 Republican registered voters.
Republican Primary Matchup
33% – Sen. Arlen Specter
18% – Pat Toomey
2% – Margaret Luksik
47% – undecided, other
