August 2008 Poll - Discussion

Welcome to the discussion of results from the August 2008 Type & Politics Poll. The discussion presented here provides a more in-depth look at the results from the poll.

This section presents the following:

  • Responses to the Type & Politics Poll are compared with the national averages calculated by realclearpolitics.com.
  • Summary results from statistical analyses of poll responses are presented and discussed.

If you would like to view the "raw results" from the poll, simply click on the August 2008 Poll Results in the 2008 Elections Studies menu above.

Click on "Next" (below) to start viewing the discussion.

  


Poll Overview - August 2008

 

Methods

The fourth round of the Type & Politics Poll was conducted during the first full week of August 2008. Approximately 270 e-mails were sent to people who registered to participate in the Poll.  The Poll also was open to any visitor to the website during this time.


Participants

Eighty-nine people completed the poll. The majority of these respondents indicated preferences for E (55%), N (87%), F (63%), and P (58%). The average age of respondents was 56 years and the majority (59%) were female.

Eighty-nine percent of the responses came from U.S. citizens. The political affiliations of those who indicated this information was: 44% Democrat, 28% Independent, and 15% Republican.

With respect to MBTI experience, 83% indicated they were members (or had been members) of an APT association and 84% indicated they were MBTI qualified. One-hundred percent indicated they were at least moderately confident that their reported type was their best-fit type.


Approval Ratings for U.S.

As show in the following three graphs, trends in approval ratings were fairly consistent with other polls to date: poll respondents gave significantly lower approval ratings to President Bush and Congress than were given by national samples. The poll respondents also were less likely to indicate the country was headed in the right direction.

(Note: Simply refresh page if a chart does not load completely, especially if you are using Safari as you browser.)

BUSH APPROVAL RATINGS

CONGRESS APPROVAL RATINGS

 

RIGHT  DIRECTION 

 Note: August 08 results shown are for all 89  poll respondents with 79 U.S. citizens.

Preferences for Presidential Candidates:  Poll Responses Compared with National Averages (Democrats)

Participants in the August 08 Poll showed marginal support for Obama over Clinton (Democrats).

 

November 2007 Poll participants who indicated they were registered democrats (n = 197) expressed very similar preferences for presidential candidates when compared with national averages.

 

February 2008 Poll participants who indicated they were registered democrats (n = 54) expressed very similar preferences for presidential candidates when compared with national averages. Note: Results for Edwards are no longer reported in national polls.

 

May 2008 Poll participants who indicated they were democrats (n = 28) expressed slightly more support for Obama than Clinton when compared with national averages. 

 

August 2008 Poll participants who indicated they were democrats (n = 48) show strong support for Obama over Clinton.  However, results also show some support for Clinton. Note: National averages are no longer reported.

Note: Of the 84 participants in the November 2007 Poll who indicated they were registered as "Other," or were not registered but planned to register, the following candidates were preferred: Obama (26%), Clinton (14%), Edwards (10%), and Don't Know (23%). No other candidates received more than 9% preference.

Note: Of the 27 participants in the February 2008 Poll who indicated they were planning to register or were registered as "Other," the following candidates were preferred: Obama (44%), Clinton (15%), McCain (13%), and Don't Know (25%). No other candidates received more than 9% preference.

Note: Of the 21 participants in the May 2008 Poll who indicated they were "Independents" or  "Other," the following candidates were preferred: Obama (53%), Clinton (5%), and McCain (16%). No other candidate received more than one vote.

Note: Of the 14 participants in the August 2008 Poll who indicated they were "Independents" or  "Other," only Obama (50%) received more than 1 vote.


Preferences for Presidential Candidates: Poll Responses Compared with National Averages (Republicans)

Participants in the August 08 Poll showed general support for McCain (Republicans).

 

November Poll participants who indicated they were registered republicans (n = 64) expressed very similar preferences for presidential candidates when compared with national averages. Preferences for Romney, however, seem somewhat lower than the national average.

 

February Poll participants who indicated they were registered republicans (n = 22) expressed very similar preferences for presidential candidates when compared with national averages. The 3 votes for Obama, however, indicate a slight emergence of "Obamicans".

 

May Poll participants who indicated they were registered republicans (n = 12) generally expressed support for McCain.  Note: National results for the Republican primary are no long reported at realclearpolitics.com

August Poll participants who indicated they were registered republicans (n = 16) generally expressed support for McCain. The slight support for Romeny seems to reflect speculation about McCain's VP nominee.  Note: National results for the Republican primary are no long reported at realclearpolitics.com

 Note: Of the 84 participants in the November 2007 Poll who indicated they were registered as "Other," or were not registered but planned to register, the following candidates were preferred: Obama (26%), Clinton (14%), Edwards (10%), and Don't Know (23%). No other candidates received more than 9% preference.

Note: Of the 27 participants in the February 2008 Poll who indicated they were planning to register or were registered as "Other," the following candidates were preferred: Obama (44%), Clinton (15%), McCain (13%), and Don't Know (25%). No other candidates received more than 9% preference.

Note: Of the 21 participants in the May 2008 Poll who indicated they were "Independents" or  "Other," the following candidates were preferred: Obama (53%), Clinton (5%), and McCain (16%). No other candidate received more than one vote.

Note: Of the 14 participants in the August 2008 Poll who indicated they were "Independents" or  "Other," only Obama (50%) received more than 1 vote.


Type and Liberal-Conservative Orientation

As discussed elsewhere on this site , some research suggests a relationship between S-N preferences and self-identified liberal-conservative political orientation. That is, to the extent that conservative orientations involve preferences for sustaining traditional institutions and a cautious approach to change, then a conservative orientation appears to reflect S preferences. To the extent that liberal orientations involve a preference for change and new possibilities, then a liberal orientation appears to reflect N preferences.

We tested statistically for this possible relationship in the November 2007 Poll, the February 2008 Poll, and the May 2008 Poll.  Results from these polls generally support for the S-N effect in both polls. We also found near-significant effects for the T-F preferences (with Ts more conservative and Fs more liberal).

 

Liberal

Intuition (& Feeling?)

Conservative

Sensing (& Thinking?)

 

For the results from the August 2008 Poll, we used multiple regression analysis to examine the relationship between liberal-conservative political orientation and the following: (1) the four type preference sets (E-I, S-N, T-F, J-P), (2) party identification (Democratic, Republican, Independent, etc.), and (3) age and sex. We used a step-wise procedure whereby the four type preference sets were entered first followed by the other variables. (This procedure often is referred to as hiearchical multiple regession).

Results indicated a significant effect for the set of type preferences alone and a significant effect when the party identification scores also were added.  The first set of results indicated a significant effect for the S-N preferences (p = .02) and a significant effect (p = .04) for the T-F preferences. E-I preferences also produced a near-sgnificant effect (p = .06). 

When party identification was added, results indicated significant effects for Democrats and Republicans with Democrats reporting a more liberal orientation and Republicans reporting a more conservative orientation. (Interestingly, the variable for Independents showed no relationship.) The results for the psychological type preferences were quite different, however, when assessed in combination with party identification. In this case, neither the S-N preferences nor the T-F preferences remained significant predictors of liberal-conservative political self-identity. Rather, the E-I and J-P preferences were significant with E and J preferences relating to a more conservative orientation whereas I and P preferences related more to a liberal orientation.  These results did not change when age and sex were added in the third step of the analysis. (And, neither age nor sex indicated any signficant relationships with liberal-conservative identity.)

These results continue to support the hypothesis that S-N and T-F preferences are in some degree related to liberal-conservative political self-identity.  However, the influence of the mental functions seems better represented through their associated party identifications. The results for the attitudes/orientations, however, is something of a new finding in the polls.  That Js might express a more conservative orientation seems reasonable to the extent that Js prefer an external world where the theme is "order," a conservative trait.  That Ps might express a more liberal  orientation also seems reasonable to the extent that Ps prefer an external world that is more "flexible" and "open," both liberal traits.  As for the E-I findings, we're at something of a loss to interpret these results. If you have any ideas, please send them along.


Political Affection or "Feeling Thermometer" Ratings

Research into how people self-identify as "liberal," "moderate," "conservative" and so on suggests both cognitive and emotional factors are involved in the formation of one's political identity.  The cognitive approach explores how one's beliefs about certain topics influence one's political identity. For example, someone who believes the practive of abortion is immoral is likely to self-identify as a conservative. On the other hand, someone who belives that a woman has an inabliable individual right to make the decision about whether or not to have an abortion is likely to self-identify as a liberal.  The emotional or affective approach explores how one's "feelings" about certain topics influence one's political identity. Thus, someone who believes abortion is immoral--yet has little emotional or affective connection with this belief--may be less likely to self-identify as a conservative than someone who holds the belief and has a strong affective response to the issue. Thus, the cognitions (beliefs, opinions, etc.) and emotions (affects, preferences, etc.) that one experiences relative to a political "object" both may influence political self-identity.

{mosimage}In our previous polls we have examined how such factors as psychological type, party identification, age, sex, and so on correlate with one's political identity as "liberal", "conservative," and "moderate".  The August 08 Poll included items to help us explore the role of political affections in addition to these previously examined factors.  Thus, the August 08 Poll included eight items called "feeling thermometers" to assess respondents political affection for eight different political groups: liberals, republicans, libertarians, conservatives, politicians, democrats, independents and moderates. ("Feeling Themometers" have been used for some time in national surveys of U.S. political dynamics such as those conducted by the ANES ). Participants in the August Poll were asked to:

Please indicate how you feel toward each of the following groups using a "feeling thermometer" ranging from 0 degrees to 100 degrees.

    • Ratings between 50 degrees and 100 degrees mean that you feel favorable and warm toward the group.
    • Ratings between 0 degrees and 50 degrees mean that you don't feel favorable toward the group and that you don't care too much for that group.
    • You would rate the group at the 50 degree mark if you don't feel particularly warm or cold toward the group.

Before examining the relationship between feeling thermometer responses, type preferences, and other factors we first analyzed the reponses using factor analysis.  Results indicated that respondents appeared to discriminate between three different political groups in their ratings:

  • Group 1: Liberals, Democrats, Conservatives, Republicans
  • Group 2: Independents, Moderates, Libertarians
  • Group 3: Politicians

We further analyzed the results for groups 1 & 2 using reliability analysis. Results showed that responses to Group 1 were highly consistent (coefficient alpha = .90). Results for Group 2 were only slightly consistent (coefficient alpha = .61).

These results suggest at least three impresssions. First, in terms of political affection, respondents tended to associate closely political party with political ideology.  For example, for Group 1, respondents who expressed high political affection for Democrats were also likely to express high political affection for liberals (and less political affection for Republicans and conservatives). Similarly, in Group 2, Libertarians (a political party) and Independents (a growing political party) are associated with moderates (a political ideology).  This association of ideologies with parties seems reasonable given the increased political polarization in the U.S. over the past 20 or so years and the tendency to assign liberal-conservative ideology to Democrat-Republican party categories. (However, as we've noted here , the association has been stronger for the Republican-Conservative category than the Democrat-Liberal category. This trend may be changing, however, with John McCain as the Republican candidate in that his somewhat "maverick" persona seems to be leading some conservatives (and particularly social conservatives) to distinguish their ideology from what the Republican party stands for.  We've noticed this trend especially with some of the political commentators such as Glenn Beck and Rush Limbaugh.  The Bush Administration's less-than-conservative "big government" policies also may be contributing to this movement.)

The second impresssion is that the results suggest respondents differentiated somewhat between the primary political parties in the U.S. and the secondary political groups.  This interpretation seems valid given extant political dynamics in the U.S. For example, analysts of the Presidential primaries generally note that candidates often target themselves to their base during the primaries and, once proclaimed the presumptive nominee, move more to the "center" in an attempt to attract the independents, moderates and other "swing voters". (See this article for an example of this consideration.) The low reliability rating for this group also suggests that respondents viewed them as a somewhat loosely connected "other" group of politicial entities in comparison to the more tightly connected entities in Group 1.

A third impression is that respondents viewed "politicians" as separate from political parties and groups. Future studies using the approach used here might want to include "feeling themometer" ratings for particular politicians such as state govenors, House of Representative members and so on to bolster this interpretation. 


Psychological Type and Political Affection Ratings

"Birds of a feather flock together." "Likes attract." These cliches reflect the findings of years of research showing people are more positively predisposed to other people who are similar to them. Thus, it seems reasonable to assume that political affection ratings likely will correlate with factors that suggest similarity between the rater and the object rated. For example, we might expect that liberals and Democrats will rate liberals and Democrates more "warmly" than they would conservatives and republicans.

{mosimage}We examined the relationship between political affection ratings and the following variables: type preferences (EI, SN, TF, JP), self-identified ratings of liberal-conservative orientation, party membership (Republican, Democratic, Independent), sex, and age.  We used heirarchical multiple regression with political affection ratings as the dependent variable.  Groups of variables were entered in the following order: (1) type preferences, (2) liberal-conservative self-ratings along with party membership, and (3) age and sex.

Results indicated a significant effect for the SN and TF preferences on the first step. Respondents with S and T preferences rated Rebublicans and conservatives more warmly whereas respondents with N and F preferences rated Democrats and liberals more warmly. No significant relationships were found for the EI and JP preferences.

The SN and TF effects became non-significant, however, when the political orientation variables were included on the second step. Rather, the key (and only) significant predictor of political affections ratings was the liberal-conservative self-identification variable.  Moreover, the partial correlation between political affection ratings and liberal-conservative ratings was .91, a very strong relationship.  As might be expected, higher ratings of oneself as conservative were correlated with respondents rating  conservatives and Republicans more warmly whereas higher ratings of oneself as liberal were correlated with respondents rating liberals and Democrats more warmly.

Age and sex were included in step three.  Results indicated that males, moreso than females, rated conservatives and Republicans more warmly.  Conversely, females, moreso than males, rated liberals and Democrats more warmly.  Results for liberal-conservative self ratings remained strongly significant. (The results for male and females is interesting and accords somewhat with political commentator Chris Matthew's labeling of the Democrat pary as the "mommy" party and the Republican party as the "daddy" party. These characterizations and their relationship to  type and politics are discussed further here. )

In summary, these results generally support results from the previous polls. That is, certain psychological type preferences (SN preferences especially and TF preferences also) seem related to a variety of political orientations when considered in isolation. However, when other variables (other political variables as well as demographic variables) are added to the analyses, the role of psychological type preferences are slight or negligible.  The emerging picture from these polls is that type preferences may influence the development of certain values which lead in varying degrees to various political beliefs, identities, and affections which then influence other political variables such as voting behavior. 

Note: We did not explore the relationship of type preferences and other variables on political affection ratings for the "other" group given the low reliability for the scale. 

  

Print Email

 
 
 
 
 © Politicaltypes.com 2017. All Rights Reserved.

Share This