 |
High-Quality vs. High-Brow: Can Soft News Ever be the Best News?
| |
| | Unformatted Document Text:
8
on the opinions of NES respondents whose political knowledge, based on a battery of factualknowledge questions, exceed the overall mean in the survey. For two reasons, we elected toemploy alternative means calculating the candidates’ actual policy positions. First, because ourmodels interact media consumption with respondents’ political awareness, employing factualknowledge as a basis to determine where the candidates stood on the issues would produce afundamental endogeneity problem. In other words, highly politically aware individuals would, bydefinition, tend to vote more correctly according to the algorithm. After all, in the Lau andRedlawsk formulation, their opinions form much of the basis for determining what constitutes acorrect vote. Second, we believe that defining as “experts” all respondents who exceed theoverall survey mean is too lenient a definition of “expert,” particularly given that the meanrespondent answered only about 1/3 of the political knowledge questions correctly.Consequently, we employed two alternative measures of the candidates’ “true” policy positions.
First, we surveyed a group of actual “experts.” We sent an email request to about 80
political scientists with particular expertise in American politics, asking them to rate thecandidates on the identical issue scales as those employed by the NES survey. We received atotal of 40 replies. We employed the mean ratings of these 40 issue experts as our first estimateof the candidates’ “true” issue positions.
Notably, the standard deviation on our experts’ assessments is far smaller than that for
the estimates of NES “expert” respondents, as defined by Lau and Redlawsk: .64 for our experts,versus 1.02 for NES “expert” respondents (i.e., those whose factual political knowledgeexceeded the survey mean). This suggests that our experts were in far greater agreement on thecandidates’ issue positions than were the NES “expert” respondents.
3
For our second alternative means of assessing the candidates’ true issue positions, we
relied on the judgments of fully informed researchers. Specifically, we asked five undergraduatesand one graduate student to literally read everything the candidates said during the 2000campaign. This was possible because all candidate speeches, advertisements, press releases, andother statements are available on a single CD-ROM, entitled “In their Own Words: Sourcebookfor the 2000 Presidential Election.” After our researchers read all of the CD-ROM’s contents, weasked them to rate the candidates’ positions on each pertinent issue area.
Figure 1 summarizes the estimates of our American politics experts and student research
assistants on each issue dimension included in the “correct vote” algorithm. Each candidate issue
3
The mean ideological score across our 40 experts was 4.75 on a 1-7 scale, where 7 represented
“extremely liberal” and 1 represented “extremely conservative.” This compares to an averageideological score among NES “expert” respondents of 3.55. Hence, our experts were somewhatmore liberal than Lau and Redlawsk’s NES “expert” respondents. It is possible that our liberaland conservative experts might fundamentally disagree on the candidates’ issue positions. If so,this difference might have important consequences for our estimates of the candidates’ “true”issue positions. In fact, among our experts, the overall difference – across all of the issuedimensions we investigate -- between self-described liberals and conservatives was quite small,about seven percent. And the standard deviations are nearly identical. Moreover, if one limits theNES “experts” to those identifying themselves as liberals, the standard deviation remainsvirtually unchanged, at 1.0. This suggests that even after controlling for systematic ideologicaldifferences, our experts converged in their ratings far more than the NES “expert” respondents.This further suggests that our experts’ estimated issue positions are, in all likelihood, morereliable than those of the NES “expert” respondents, as defined by Lau and Redlawsk.
|
| |
| |
|
|
8
on the opinions of NES respondents whose political knowledge, based on a battery of factual
knowledge questions, exceed the overall mean in the survey. For two reasons, we elected to
employ alternative means calculating the candidates’ actual policy positions. First, because our
models interact media consumption with respondents’ political awareness, employing factual
knowledge as a basis to determine where the candidates stood on the issues would produce a
fundamental endogeneity problem. In other words, highly politically aware individuals would, by
definition, tend to vote more correctly according to the algorithm. After all, in the Lau and
Redlawsk formulation, their opinions form much of the basis for determining what constitutes a
correct vote. Second, we believe that defining as “experts” all respondents who exceed the
overall survey mean is too lenient a definition of “expert,” particularly given that the mean
respondent answered only about 1/3 of the political knowledge questions correctly.
Consequently, we employed two alternative measures of the candidates’ “true” policy positions.
First, we surveyed a group of actual “experts.” We sent an email request to about 80
political scientists with particular expertise in American politics, asking them to rate the
candidates on the identical issue scales as those employed by the NES survey. We received a
total of 40 replies. We employed the mean ratings of these 40 issue experts as our first estimate
of the candidates’ “true” issue positions.
Notably, the standard deviation on our experts’ assessments is far smaller than that for
the estimates of NES “expert” respondents, as defined by Lau and Redlawsk: .64 for our experts,
versus 1.02 for NES “expert” respondents (i.e., those whose factual political knowledge
exceeded the survey mean). This suggests that our experts were in far greater agreement on the
candidates’ issue positions than were the NES “expert” respondents.
3
For our second alternative means of assessing the candidates’ true issue positions, we
relied on the judgments of fully informed researchers. Specifically, we asked five undergraduates
and one graduate student to literally read everything the candidates said during the 2000
campaign. This was possible because all candidate speeches, advertisements, press releases, and
other statements are available on a single CD-ROM, entitled “In their Own Words: Sourcebook
for the 2000 Presidential Election.” After our researchers read all of the CD-ROM’s contents, we
asked them to rate the candidates’ positions on each pertinent issue area.
Figure 1 summarizes the estimates of our American politics experts and student research
assistants on each issue dimension included in the “correct vote” algorithm. Each candidate issue
3
The mean ideological score across our 40 experts was 4.75 on a 1-7 scale, where 7 represented
“extremely liberal” and 1 represented “extremely conservative.” This compares to an average
ideological score among NES “expert” respondents of 3.55. Hence, our experts were somewhat
more liberal than Lau and Redlawsk’s NES “expert” respondents. It is possible that our liberal
and conservative experts might fundamentally disagree on the candidates’ issue positions. If so,
this difference might have important consequences for our estimates of the candidates’ “true”
issue positions. In fact, among our experts, the overall difference – across all of the issue
dimensions we investigate -- between self-described liberals and conservatives was quite small,
about seven percent. And the standard deviations are nearly identical. Moreover, if one limits the
NES “experts” to those identifying themselves as liberals, the standard deviation remains
virtually unchanged, at 1.0. This suggests that even after controlling for systematic ideological
differences, our experts converged in their ratings far more than the NES “expert” respondents.
This further suggests that our experts’ estimated issue positions are, in all likelihood, more
reliable than those of the NES “expert” respondents, as defined by Lau and Redlawsk.
|
|
 Convention | | Convention is an application service for managing large or small academic conferences, annual meetings, and other types of events! | | Submission - Custom fields, multiple submission types, tracks, audio visual, multiple upload formats, automatic conversion to pdf. | | Review - Peer Review, Bulk reviewer assignment, bulk emails, ranking, z-score statistics, and multiple worksheets! | | Reports - Many standard and custom reports generated while you wait. Print programs with participant indexes, event grids, and more! | | Scheduling - Flexible and convenient grid scheduling within rooms and buildings. Conflict checking and advanced filtering. | | Communication - Bulk email tools to help your administrators send reminders and responses. Use form letters, a message center, and much more! | | Management - Search tools, duplicate people management, editing tools, submission transfers, many tools to manage a variety of conference management headaches! | | Click here for more information. |
|
|
|
| |
|
|
|
