<$BlogRSDUrl$>

Maintaining Optimism in the Face of Reality. Occasional observations on the state of the world, society, business and politics. Usually anchored by facts, always augmented by opinion.


Digging Into Ignorance  | e-mail post

This is a followup from my prior post discussing political ignorance. At the end, I promised to dig into a decent-sized dataset from a survey, primarily to measure media consumption, conducted by the Pew Center for the People and the Press in 2002. This all relates to a post from a couple of days ago about the great "political divide" in America, which I tend to question. Read the first post on that subject for what political ignorance has to do with it.

To assess the ignorance rampant in this great country of ours, I started to dig into the Pew survey results. The survey, with 3,002 respondents, asked three questions that I decided to examine, as they seemed like an obviously interrelated set:
My assumption is that the first question provides the respondent's self-assessment of how closely they follow a relatively specific news story, while the second two questions are indicators of how much knowledge they have about the current situation and background directly related to that specific news story. While this is a very small set of questions on which to base any determination, I felt that the specific interrelatedness of the questions provided some counterbalance to the narrow focus. In addition, I liked this question set as it doesn't have an inherent political bias to it. That is, one would think that Democrats and Republicans would have equal interest in the topic.

People Overrate The Extent to Which They Follow The News

The results were, shall we say, unfortunate. Recoding the data to compare whether the respondent got neither, either, or both of 39(e) and 39(f) correct, and comparing it to their self-assessment of how closely they followed news of the Israeli-Palestinian conflict result in this table:

How closely Mideast news followed vs. Related Knowledge

Neither Correct
One Correct
Both Correct
Total Count
Don't know/Refused 65 26 8.7 100 23
Not at all closely 83 11 5.4 100 258
Not too closely 57 29 14 100 395
Fairly closely 43 29 28 100 1131
Very closely 25 31 45 100 1195

What!? Only 45% of people who honestly think they are following Palestinian/Israeli violence very closely can answer both of those questions? Even more astounding is that 43% of those who think they are following it "fairly closely" couldn't answer either of them. Bear in mind that with Israel's establishment, it was a multiple choice question. But maybe I'm being too harsh, suppose we just consider at the incredibly simple question of who Yasser Arafat is. Certainly everyone should nail that one?

How closely Mideast news followed vs. Who Is Arafat?

Don't Know
Incorrect Correct Total Count
Don't know/Refused 39 43 17 100 23
Not at all closely 57 35 8.5 100 258
Not too closely 28 44 28 100 395
Fairly closely 18 38 44 100 1131
Very closely 9.9 30 60 100 1195

Still, only 60% of these self-assessed very close followers of Mideast news can say who Yasser Arafat is. C'mon, in 2002, he had been head of the PLO for 33 years. This is an embarassment that had me ready to put the "for sale" sign in my yard and get a one-way international plane ticket. I figured the Thanksgiving travel crowds would be too much, though. So I'm staying put for now.

Well, It's Probably Because of Those Stupid [Insert Opposing Party Here]

I am sure that partisans on both sides of the political spectrum are assuming that the "other guys" are weighing those numbers down. Not as much as you might think, with respect to those who claim to follow things "very closely" however there is a significant partisan difference between people who think they follow things "fairly closely."

Party Affiliation vs Knowledge For "Very Close" News Followers

Neither Correct
One Correct
Both Correct
Total Count
Don't know 29 39 32 100 28
Other 18 18 64 100 11
No Preference 40 33 27 100 52
Independent 23 29 49 100 308
Democrat 26 30 44 100 406
Republican 22 32 46 100 390


Party Affiliation vs. Knowledge for "Fairly Close" News Followers

Neither Correct
One Correct
Both Correct
Total Count
Don't know 47 39 13 100 38
Other 42 33 25 100 12
No Preference 57 28 15 100 67
Independent 39 33 28 100 291
Democrat 50 27 22 100 352
Republican 36 27 37 100 371


Among those who follow the news "very closely" there is a statistically insignificant difference Democrats and Republicans, with Independents slightly ahead, and all three within a margin of error for the subsample sizes of 308-406. However, it appears both Democrats and independents have a different standard for what watching the news "fairly closely" means. That is, while the "both correct" score for Republicans went down 9 points between the very closely to fairly closely group, Democrats who follow the news "fairly closely" went down by 22 points, and independents by 20. compared to their similarly-aligned "very closely" group. Half of Democrats who "fairly closely" followed Mideast news missed both questions, while slightly more than a third of similarly self-assessing Republicans could got neither correct.

Well It's Probably the [Stupid Kids/Senile Old People] Then

I thought I would look at ages as well, given that maybe the numbers for the Dems were being skewed by the presumably larger proportion of young people, my assumption being that the natural hubris of youth combined with a lack of perspective might have young people even more likely to incorrectly regard themselves as following the news when they really aren't. The first quick check I did was to see how party affiliation lined up by age within the sample:



REF 65+ 55-64 45-54 35-44 25-34 18-24
Don't know 34 2.8 4.3 1.3 1.9 1.9 2.0
Other 4 0.5 0.25 1.3 1.2 1.1 0.58
No Preference 2 4.7 5 6 7.2 8 7.9
Independent 24 20 24 28 25 29 36
Democrat 18 38 34 34 30 29 27
Republican 18 35 32 29 34 31 26
Total 100 100 100 100 100 100 100
Count 50 599 397 552 584 477 343
Really not too much of a skew for party affiliation, but I thought I would still look at the results of knowledge versus age for the very close and fairly close news followers.

Age vs Knowledge for "Very Close" Mideast news followers

Neither Correct
One Correct
Both Correct
Total Count
Refused 23 32 45 100 22
65+ 21 37 42 100 298
55-64 17 26 57 100 192
45-54 24 27 49 100 236
35-44 32 30 38 100 213
25-34 24 33 43 100 146
18-24 36 28 35 100 88


Age vs Knowledge for "Fairly Close" Mideast news Followers

Neither Correct
One Correct
Both Correct
Total Count
Refused 26 35 39 100 23
65+ 41 30 30 100 179
55-64 31 30 39 100 152
45-54 39 22 39 100 205
35-44 40 34 26 100 233
25-34 50 29 22 100 208
18-24 64 30 6.1 100 131


So it looks like those from 18-44 have a slightly different perspective of what "very close" means compared to 45-64. Again though, this "fairly close" segment blew me away. Fewer than 25% of those under 44 who think they are fairly close followers of Mideast news could answer both questions, and among the 18-24 crowd, it's a particularly sad 6.1%. I hesitate to read too much into this question because both of the questions are things that have been known facts for decades, and thus could arguably favor older people, although I have a tough time accepting that argument.

But, given that age and party affiliation were pretty well-distributed, I wanted to look a little more closely. The next thing I did was to look at the percentage of individuals who got both questions correct within each age and party affiliation segment for the groups that reported following the news very closely or fairly closely. Remember that this is a fairly large total sample (n=3002), so these are still decent-sized slices, with n=1195 for very close news followers and n=1131 for fairly close news followers. I then computed the percentage decline with each cross-sectional group.

% of Cross-section of "Very Close" Mideast News Followers Answering Both Mideast News Questions Correctly

REF 65+ 55-64 45-54 35-44 25-34 18-24
Don't know 33 50 25 0 0 67 0
Other 0 100
67 50 100 100
No Preference 0 33 0 50 18 13 33
Independent 80 53 70 52 39 47 16
Democrat 25 36 51 47 42 46 48
Republican 60 41 64 48 37 39 45

% of Cross-section of "Fairly Close" Mideast News Followers Answering Both Mideast News Questions Correctly

REF 65+ 55-64 45-54 35-44 25-34 18-24
Don't know 30 14 0 0 20 0 0
Other 0 0
25 33 50
No Preference
8.3 33 43 6.3 13 0
Independent 17 40 45 42 18 25 4.3
Democrat 100 20 33 30 19 19 3
Republican 75 41 48 49 41 23 12

% Decline of Cross-section's ability to answer Both Mideast News Questions Correctly Between Very Close and Fairly Close news followers

REF 65+ 55-64 45-54 35-44 25-34 18-24
Don't know 10 71 100
-Inf 100
Other
100
63 33 50
No Preference
75 -Inf 14 66 -6.7 100
Independent 79 25 35 20 55 47 73
Democrat -300 44 36 36 54 58 94
Republican -25 0.41 25 -3.2 -10 41 73

Looking at the numbers, the one thing that I think is visible is that while everyone seems to think they are better informed than their knowledge suggests they are, younger people seem to have a lower standard of being informed on the news, but Democrats, even more than young people, are more likely to overstate their familiarity with the news compared to their actual knowledge of that news. To look at this, I examined just the 25 and older crowed (throwing out those who refused to provide and age as well as the 18-25 group).

Among this group (25 and up) Democrats who considered themselves "very close" followers of Mideast news (n=375) scored about the same as Republicans who only rated themselves as "fairly close" news followers (n=326) (41% VCD to 40% FCR). In contrast, self-described Republicans (n=363) and Independents (n=272) who were very close news followers did much better than similarly self-reporting Democrats, with 48% and 52%, respectively, able to answer both questions accurately, versus the 41% for Dems. (Note that this contrasts some from the early observation that Dems and Republicans seemed to be quite similar in the "very close" news follower segment, the 18-24 group was hurting Republicans and helping Democrats.)

Education and Income Do Have An Effect, Obviously

Since the Republican numbers would tend to be natural skewed by education and income, I thought it would be worthwhile to just take a quick skim at those numbers as well:

Income vs Knowledge for self-reported "Very close" followers of Mideast news

Neither Correct
One Correct
BothCorrectTotalCount
Don't know/Refused 27 33 39 100 218
$100,000 or more 12 26 62 100 137
$75,000 - $100,000 17 23 60 100 110
$50,000 - $75,000 20 31 49 100 197
$40,000 - $50,000 26 31 43 100 120
$30,000 - $40,000 32 32 36 100 133
$20,000 - $30,000 24 32 45 100 127
$10,000 - $20,000 37 32 31 100 94
Less than $10,000 39 36 25 100 59


Education vs. Knowledge for self-reported "Very close" followers of Mideast news

Neither Correct
One Correct
Both Correct
TotalCount
Refused 80 20 0 100 5
At Least 4-year Degree
12 26 62 100 444
Some College/2 Year Deg
21 33 46 100 282
High School Graduate
36 34 30 100 374
Less Than High School
48 34 18 100 90

The numbers for fairly close watchers fall off about as one would expect. That is, while the more educated and the weatlhier drop off about 10 points, the lower categories fall off at a faster rate.

General Observations

Here are some quick general observations from the data.


I will add a comic relief section to this later that lists some of the verbatim responses captured when asking who Yasser Arafat was. ("Bad ass" is my favorite so far.)
e-mail post | Link Cosmos | [Permalink]  |  | Friday, November 26, 2004
Comments: Post a Comment

This page is powered by Blogger. Isn't yours?