Polling Peculiarities

Here is an interesting set of polls from a while back, with an obvious discrepancy between the initial findings and the follow up:

Do you personally think you would watch the televised coverage of
the US Senate if it were available to you, or not?Question: Yes 72% No 26% Don’t know %2
(Source: Survey by the Associated Press/Media General, April 3-11, 1986.)

A few months later...
As you may know, the US Senate began televising its sessions on June 2.
Have you yourself watched part of any of those sessions on TV since then?

No, did not
watch 79%
Yes, watched
20%
Can’t get them on my TV (vol.) 1%
(Source: Survey by the ABC News/Washington Post, June 19-24, 1986.)

Although the initial findings suggested that there was an overwhelming interest in watching the Senate on TV, this does not seem to have corresponded to the actual viewership once the Senate was actually televised. Many factors may have contributed to this inconsistency:

1. Respondents may have interpreted the question as a vote on whether Senate sessions should be televised. The question may have been understood as "Should one be able to watch the senate in session?" to which many responded positively.

2.Respondents may have felt that saying they would watch the senate was the "right answer". Since citizens should want to watch the Senate, they reasoned, we should say we want to watch the senate.

3. The subject of televised senate sessions may have been receiving much attention in the media at the time, artificially creating an interest in watching the Senate. Since this did not reflect actual interest, the percentage actually watching was significantly lower.

4. The question presented the allure of something which is out of reach. Human psychology dictates that one desires something more if he cannot have it. Once Senate viewing was actually available, interest dropped because of the very fact that it was now available.

This poll shows the difficulty of accurately determining public interest and should serve as a reminder against over-reliance on public opinion surveys, without proper analysis.