Surveys on the French presidential election of 2007

Opinion polls on French presidential Election of 2007:

Results of the principal surveys

Results of principal the surveys carried out during the pre-countryside and the presidential campaign:

1st turn

Surveys TNS Sofres


	17- May 18th, 2006	14-15 jun.	17 jul.	4-5 sep.	4-5 oct.	12-13 oct.	8-9 nov.	6-7 déc.	17-18 Jan. 2007	31 jan. - fév.	7-8 fév.	14-15 fév.	28 fév. - mar.	7-8 mar.	14-15 mar.	21-22 mar.	28-29 mar.	4-5 avr.	11-12 avr.
Arlette Laguiller	3.5	3.5	4	3	3	3	2.5	3.5	2	3	3	2	2	2.5	2	2	1	2.5	2.5
Olivier Besancenot	5	6	4	3.5	5	4	4	3.5	3	2.5	3	3.5	3	4	2	3.5	3.5	4	4
Gerard Schivardi	-	-	-	-	-	-	-	-	-	-	-	-	-	-	<0.5	<0.5	0.5	<0.5	<0.5
Jose Bove	-	-	-	-	-	-	-	-	-	4	3.5	3	2	1.5	2	2.5	2.5	2	2
Marie-George Buffet	3	2	3	3	3	2	2	2	3	2.5	2	2.5	3.5	1.5	2.5	2	2.5	3	2.5
Ségolène Royal	30	32	32	34	29.5	34	34	33	31	26	26	26	25.5	25.5	24	26.5	27	23.5	26
Dominique Voynet	2.5	2	1.5	1.5	2	1	1.5	2	2	1.5	1	1	1	1	1	1	1	1	1.5
Francois Bayrou	8	8	6	7	7	7	7	8	9	13	14	12	18.5	23	22	21.5	18	20	17
Nicolas Sarkozy	34	31	35	36	38	36	34	33	35	32	33	33	31	27	31	28	30	28	30
Frederic Nihous	-	-	-	-	-	-	-	0.5	<0.5	<0.5	<0.5	<0.5	<0.5	<0.5	<0.5	1	1	1.5	1.5
Philippe de Villiers	4	3	3	2	3	2	2	2	1	2	1	2.5	1	1.5	0.5	1	1	1.5	1
Jean-Marie Le Pen	10	12.5	11.5	10	9.5	11	13	11.5	13	12.5	13	13	12	12	12	11	12	13	12

To note that these figures do not give the number of people who answered the question of their vote, because these figures are rectified to correspond to the analyzes of the institutes, and take account neither of the abstentionnists, nor of the people who refused to answer or were not there at the time of the phone call. Moreover, these figures are published without their margin of uncertainty, which is an absolute statistical data, which depends on the number of questioned people, and which exists whatever the quality of the sample and the relevance of the software of rectification. Are not quoted either in the table, the people who are not candidates officially, which explains a sum which can be lower than 100%. The selected order is that of survey institute TNS Sofres.

turn

This list of surveys concerns possible second turns between two candidates; the two candidates are chosen by the survey institutes.

Divergence according to the institutes on a given date

Here comparative enters the surveys of several institutes carried out at the same period.

The error made on a result can be estimated by multiplying the standard deviation by 3. If one is interested in the “third man” for example, the surveys put Francois Bayrou in front of Jean-Marie Le Pen, but with the margin of error, one can have 11% voters for Mr. Bayrou (minimum: 15 - 3×1,4) and 16% for Mr. Le Pen (maximum: 12 + 3×1,4). This of course is to be balanced by other factors, in particular the stability of the electorate or his volatility:

one does not know how the vote of the 40% the undecided one will be distributed;
in a general way, according to François Miquet-Marty (LH2), more half (56%) of probed which quotes a candidate think of being able to change opinion and between 9,5 and 13% between on May 18th, 2006 and on February 15th, 2007 according to TNS Sofres);
Mr. Bayrou on the other hand knew a clear progression in the surveys in 2007 (between 11 and 19% according to Ifop enter mid-January and mid-February, between 8 and 14% according to TNS Sofres between mid-May 2006 and mid-February 2007), and records only 38% of “firm” votes (62% of the people stating to vote for him with the first turn state to provide to change opinion). However, the institutes are denied this criticism while alleging that the investigations comparing the answers of probed contacted by fixed telephone and other methods different step

Another point of method was discussed many in the media. The sounders use indeed methods of Redressement to correct the answers given by probed sometimes which do not dare to give their voting intention. This technique is however necessary: for example, by comparing what probed say to have voted in 2002 and the real results of the vote, one observes indeed a difference that the sounder must correct to give an image more representative of the real intentions of vote. This behavior is indicated under the term of Désirabilité social in social Psychologie.

See the article: