%PDF-1.6
%
1 0 obj<>
endobj
2 0 obj<>
endobj
3 0 obj<>
endobj
5 0 obj<>
endobj
7 0 obj[/Indexed/DeviceRGB 1 8 0 R]
endobj
8 0 obj<>stream
endstream
endobj
9 0 obj<>stream
Hbd(f0y 0 ( $
endstream
endobj
10 0 obj(http://www.esomar.org/upload/135730_7024_1089300248946-bulletLarge.gif)
endobj
11 0 obj(R$/"`\))
endobj
12 0 obj<>
endobj
13 0 obj<>
endobj
14 0 obj<>
endobj
15 0 obj<>
endobj
16 0 obj<>
endobj
17 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
18 0 obj<>
endobj
19 0 obj[17 0 R/XYZ 0 765.5 null]
endobj
20 0 obj<>
endobj
21 0 obj<>>>
endobj
22 0 obj<>
endobj
23 0 obj<><><><><><><><><><><><><><><><><><><><><>]/P 22 0 R/S/Article/T()/Pg 17 0 R>>
endobj
24 0 obj<>
endobj
25 0 obj[23 0 R]
endobj
26 0 obj[17 0 R/XYZ 0 751.155 null]
endobj
27 0 obj[17 0 R/XYZ 0 704.235 null]
endobj
28 0 obj<>/A<>/PA<>/Border[0 0 0]>>
endobj
29 0 obj[28 0 R 30 0 R 31 0 R 32 0 R 33 0 R 34 0 R 35 0 R]
endobj
30 0 obj<>/A<>/PA<>/Border[0 0 0]>>
endobj
31 0 obj<>/A<>/PA<>/Border[0 0 0]>>
endobj
32 0 obj<>/A<>/PA<>/Border[0 0 0]>>
endobj
33 0 obj<>/A<>/PA<>/Border[0 0 0]>>
endobj
34 0 obj<>/A<>/PA<>/Border[0 0 0]>>
endobj
35 0 obj<>/A<>/PA<>/Border[0 0 0]>>
endobj
36 0 obj[17 0 R/XYZ 0 546.027 null]
endobj
37 0 obj[17 0 R/XYZ 0 524.027 null]
endobj
38 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
39 0 obj[23 0 R]
endobj
40 0 obj[38 0 R/XYZ 0 605.4 null]
endobj
41 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
42 0 obj[23 0 R]
endobj
43 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
44 0 obj[23 0 R]
endobj
45 0 obj[43 0 R/XYZ 0 435.6 null]
endobj
46 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
47 0 obj[23 0 R]
endobj
48 0 obj[46 0 R/XYZ 0 108.6 null]
endobj
49 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
50 0 obj[23 0 R]
endobj
51 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
52 0 obj[23 0 R]
endobj
53 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
54 0 obj[23 0 R]
endobj
55 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
56 0 obj[23 0 R]
endobj
57 0 obj[55 0 R/XYZ 0 675.4 null]
endobj
58 0 obj<>/ColorSpace<>/Font<>/ProcSet[/PDF/Text/ImageC/ImageI]>>/Type/Page>>
endobj
59 0 obj[23 0 R]
endobj
60 0 obj[58 0 R/XYZ 0 701.8 null]
endobj
61 0 obj[58 0 R/XYZ 0 557.8 null]
endobj
62 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
63 0 obj<>
endobj
64 0 obj<>
endobj
65 0 obj<>
endobj
66 0 obj<>
endobj
67 0 obj<>
endobj
68 0 obj<>
endobj
69 0 obj<>
endobj
71 0 obj[23 0 R]
endobj
72 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
74 0 obj[23 0 R]
endobj
75 0 obj[72 0 R/XYZ 0 591.4 null]
endobj
76 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
78 0 obj[23 0 R]
endobj
79 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
81 0 obj[23 0 R]
endobj
82 0 obj<>/ColorSpace<>/Font<>/ProcSet[/PDF/Text/ImageC/ImageI]>>/Type/Page>>
endobj
84 0 obj[23 0 R]
endobj
85 0 obj[82 0 R/XYZ 0 68.2 null]
endobj
86 0 obj<>/ColorSpace<>/Font<>/ProcSet[/PDF/Text/ImageC/ImageI]>>/Type/Page>>
endobj
87 0 obj<>
endobj
89 0 obj[23 0 R]
endobj
90 0 obj[86 0 R/XYZ 0 741.6 null]
endobj
91 0 obj[86 0 R/XYZ 0 228.592 null]
endobj
92 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
94 0 obj[23 0 R]
endobj
95 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
97 0 obj[23 0 R]
endobj
98 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
99 0 obj<>
endobj
100 0 obj<>
endobj
102 0 obj[23 0 R]
endobj
103 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
105 0 obj[23 0 R]
endobj
106 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
107 0 obj<>
endobj
109 0 obj[23 0 R]
endobj
110 0 obj<>/ProcSet[/PDF/Text]>>/Type/Page>>
endobj
112 0 obj[23 0 R]
endobj
113 0 obj[110 0 R/XYZ 0 497.4 null]
endobj
114 0 obj<>/A<>/Border[0 0 0]>>
endobj
115 0 obj[114 0 R 116 0 R 117 0 R]
endobj
116 0 obj<>/A<>/Border[0 0 0]>>
endobj
117 0 obj<>/A<>/PA<>/Border[0 0 0]>>
endobj
118 0 obj<>
endobj
119 0 obj<>
endobj
122 0 obj<>
endobj
130 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
0.157 0.38 0.671 rg
BT
/TT0 1 Tf
16 0 0 16 15 751.155 Tm
(ESOMAR/WAPOR Guide to Opinion Polls)Tj
14 0 0 14 15 725.8106 Tm
(- including the ESOMAR International Code of Practice for the Publicatio\
n )Tj
0 -1.2 TD
(of Public Opinion Poll Results)Tj
0 0 0 rg
/T1_0 1 Tf
( )Tj
/TT0 1 Tf
12 0 0 12 15 683.093 Tm
(Contents)Tj
/TT1 1 Tf
( )Tj
0.157 0.38 0.671 rg
0 -2.784 TD
(Introduction to the Guide)Tj
0 -1.367 TD
(Opinion Polls and Democracy)Tj
0 -1.367 TD
(Representativeness of Opinion Polls)Tj
T*
(International Code of Practice for the Publication of Public Opinion Pol\
l Results )Tj
0 -1.367 TD
(Guideline to the Interpretation of the International Code of Practice fo\
r the Publication of Public )Tj
0 -1.367 TD
(Opinion Poll Results)Tj
T*
(Guidelines on Practical Aspects of Conducting Pre-Election Opinion Polls\
)Tj
0 0 0 rg
0 -2.1 TD
( )Tj
0.157 0.38 0.671 rg
14 0 0 14 15 511.0278 Tm
(Introduction to the Guide)Tj
0 0 0 rg
12 0 0 12 15 486.0848 Tm
(Public opinion polls are regularly conducted and published in many count\
ries. They measure )Tj
0 -1.2 TD
(not only support for political parties but also public opinion on a wide\
range of social and )Tj
T*
(political issues and are published frequently in a variety of newspapers\
and broadcast media. )Tj
0 -1.2 TD
(They are the subject of much discussion by the public, journalists and p\
oliticians, some of )Tj
0 -1.2 TD
(whom wish to limit or ban them completely. In a small number of European\
countries )Tj
T*
(legislation actually exists which restricts the publication of opinion p\
olls during the late stages )Tj
T*
(of election campaigns. )Tj
0 -1.1 TD
( )Tj
0 -1.1 TD
(The public discussion of opinion polls is not always well informed. The \
case for restriction on )Tj
0 -1.2 TD
(the publication of polls during election campaigns is hard to support wi\
th rational argument or )Tj
T*
(empirical evidence. ESOMAR has produced the present booklet in order to \
help those )Tj
T*
(interested in the subject of opinion polls to reach a more informed judg\
ement about the value )Tj
T*
(of such polls and the most appropriate ways of conducting and reporting \
them. )Tj
0 -1.1 TD
( )Tj
0 -1.1 TD
(WAPOR joins ESOMAR in the publication of this booklet. The WAPOR Council\
endorses all )Tj
0 -1.2 TD
(recommendations put forward in this document. WAPOR and ESOMAR believe t\
here is a need )Tj
T*
(for more consultation and coordination in issuing future guidelines and \
statements regarding )Tj
T*
(standards of practice in the industry, given the rapid growth in market \
and opinion research )Tj
T*
(around the world and the new technologies and developments. )Tj
0 -1.1 TD
( )Tj
T*
(An ESOMAR Code of Conduct for the Publication of Opinion Polls has exist\
ed since 1983. It is )Tj
0 -1.2 TD
(reproduced in this booklet together with guidance on the interpretation \
of opinion poll data. )Tj
T*
(There are also three additional sections. First, there is an ESOMAR posi\
tion statement on the )Tj
T*
(role of opinion polls in democratic systems. ESOMAR\222s position on thi\
s issue is quite clear. We )Tj
T*
(believe that there should be no restriction on the conduct or publicatio\
n of opinion polls which )Tj
T*
(have been carried out according to the ICC/ESOMAR International Code of \
Marketing and )Tj
T*
(Social Research Practice and published according to the ESOMAR Code for \
the Publication of )Tj
T*
(Opinion Poll Results. )Tj
0 -1.1 TD
( )Tj
0 -1.1 TD
(Second there is a section on the misuse of the term opinion poll to desc\
ribe unscientific and )Tj
0 -1.2 TD
(unrepresentative measurements of public opinion. Push polling, televotin\
g and certain Internet )Tj
T*
(polls are examples of the type of activity frequently, but incorrectly, \
presented as opinion polls. )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(1 of 22\)5/7/2005 8:58:55 AM)Tj
ET
EMC
endstream
endobj
131 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 756.058 Tm
( )Tj
0 -1.1 TD
(Third, there is a section providing guidelines on the conduct of pre-ele\
ction opinion polls. )Tj
0 -1.2 TD
(These guidelines are designed to help ensure that polls are carried out \
to high standards. They )Tj
0 -1.2 TD
(provide technical guidance to researchers and background information for\
journalists, )Tj
T*
(politicians, academics and other interested parties. They are not intend\
ed to be a "How to do )Tj
T*
(it" manual. The subject of public opinion research and the measurement o\
f voting intention )Tj
T*
(continues to evolve and every election may bring new circumstances to be\
addressed by the )Tj
T*
(researcher. The professional skills and previous experience of polling o\
rganisations are )Tj
0 -1.2 TD
(essential components of effective public opinion research. It is not pos\
sible to write this into )Tj
T*
(guidelines or codes of practice. )Tj
0 -1.1 TD
( )Tj
0.157 0.38 0.671 rg
14 0 0 14 15 592.401 Tm
(Opinion Polls and Democracy)Tj
0 0 0 rg
12 0 0 12 15 567.458 Tm
(For the last fifty years opinion polls have proved to be one of the perm\
anent operational )Tj
0 -1.2 TD
(constituents of modern democracies. )Tj
0 -1.1 TD
( )Tj
0 -1.1 TD
(Thanks to such polls, journalists can trace, step by step, the ups and d\
owns of election )Tj
0 -1.2 TD
(campaigns and the rise and fall of rulers' popularity. Political scienti\
sts obtain from them )Tj
0 -1.2 TD
(irreplaceable information on the electoral choices of different groups o\
f citizens and on the )Tj
T*
(motivations which explain those choices. They enable sociologists to fol\
low shifts of opinion on )Tj
0 -1.2 TD
(the major social problems and chart the evolution of values. To those in\
power and their )Tj
T*
(opponents they trace the movements in their relative support between ele\
ctions, as well as the )Tj
T*
(public impact of important national or international events. And they al\
low citizens to make )Tj
T*
(themselves heard at all times and to see where their own views stand com\
pared with those of )Tj
T*
(other people. )Tj
0 -1.1 TD
( )Tj
T*
(The more direct the link established by the institutions of a country be\
tween its citizens and )Tj
0 -1.2 TD
(the nation\222s political direction, the more opinion polls are in deman\
d. Wherever the choice of )Tj
T*
(the chief executive rests on the whole electorate, the media resort more\
extensively to opinion )Tj
T*
(polls. )Tj
0 -1.1 TD
( )Tj
T*
(Conversely, those countries whose institutions set up a screen between t\
he citizens\222 vote and )Tj
0 -1.2 TD
(the designation of rulers make far more sparing use of election polls. )Tj
0 -1.1 TD
( )Tj
T*
(When new nations join the ranks of democracies, polls soon appear. That \
is the case today in )Tj
0 -1.2 TD
(Eastern Europe, as it was yesterday in Latin America. Totalitarian regim\
es, even when they )Tj
T*
(claim to have won 99% of votes in single list elections, have never risk\
ed allowing opinion )Tj
T*
(polls to be conducted in which their citizens might be able to say wheth\
er they would prefer a )Tj
T*
(freely chosen regime. )Tj
0 -1.1 TD
( )Tj
T*
(Indeed opinion polls and freedom cannot be separated. The relationship b\
etween the pollsters )Tj
0 -1.2 TD
(and the polled presupposes a climate of confidence, and published result\
s are credible only in )Tj
T*
(so far as all opinions can be freely expressed, unpleasant as they might\
be to those in power. )Tj
0 -1.1 TD
( )Tj
T*
(Yet, however important their place in modern democracies, the right to c\
onduct and publish )Tj
0 -1.2 TD
(polls freely is occasionally contested by the political establishment. T\
hose same leaders who )Tj
T*
(eagerly scrutinize the shifts in their popularity rating are sometimes u\
neasy about what they )Tj
T*
(perceive as the dangerous \223tyranny of polls\224 and are concerned to \
protect voters from any )Tj
T*
(risks of manipulation that might be attributable to opinion polls. )Tj
0 -1.1 TD
( )Tj
T*
(Thus in the last fifteen years some European countries have passed laws \
intended to regulate )Tj
0 -1.2 TD
(the practice of election polls. These laws generally lay down a pre-elec\
tion period during which )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(2 of 22\)5/7/2005 8:58:55 AM)Tj
ET
EMC
endstream
endobj
132 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 752.658 Tm
(the publication of opinion polls is forbidden. The justification advance\
d for such action is a )Tj
0 -1.2 TD
(supposed need to protect the citizen against any excesses which might co\
nfuse him or )Tj
T*
(interfere with his freedom of choice. )Tj
0 -1.1 TD
( )Tj
0 -1.1 TD
(Now that we have been able to observe the impact of such laws in practic\
e, it is possible to )Tj
0 -1.2 TD
(make a preliminary assessment of the impact of this policy. )Tj
0 -1.1 TD
( )Tj
0 -1.1 TD
(From the French experience, the longest in Europe, we note that the cont\
rolling body, )Tj
0 -1.2 TD
(although it has succeeded in bringing into line certain questionable ope\
rators, still had to admit )Tj
T*
(it was unable to prevent attempts at disinformation by another public in\
stitution, the Service )Tj
T*
(des Renseignements G\351n\351raux of the Interior ministry. This was the\
source in France of the )Tj
T*
(clearest proven effort to manipulate opinion through the publication of \
misleading polls. )Tj
0 -1.1 TD
( )Tj
T*
(Belgian legislation is the most extreme concerning the duration of the c\
ensorship period, )Tj
0 -1.2 TD
(initially fixed at four weeks. Experience in this country suggests that \
the silence created by )Tj
T*
(censorship seems to encourage manipulation by special interest groups mu\
ch more than free )Tj
T*
(and potentially contradictory information would have done. During the 19\
85 legislative )Tj
T*
(election, when this law was first applied, there was lively speculation \
on the stock exchange )Tj
T*
(during the black-out period when those with access to unpublished polls,\
secured an )Tj
T*
(advantage over those kept in ignorance by the effects of the law. )Tj
0 -1.1 TD
( )Tj
T*
(Despite the good intentions underlying the introduction of legislation o\
f this kind, these and )Tj
0 -1.2 TD
(other unfortunate results have stemmed from reliance on two equally deba\
table assumptions. )Tj
0 -1.1 TD
( )Tj
T*
(The first of these is the assumption that it is actually possible to man\
ipulate opinion through )Tj
0 -1.2 TD
(the publication of pre-election polls. For this to work the manipulator \
needs to secure the )Tj
T*
(complicity of the numerous competing polling organisations whose reputat\
ions are involved, )Tj
T*
(and of the many competing newspapers whose credibility is at stake. In p\
ractice attempts to )Tj
T*
(manipulate polls in this way have been generally unsuccessful not least \
because nobody knows )Tj
0 -1.2 TD
(the answer to the crucial question: "Which polls - or which series of po\
lls - should I publish in )Tj
T*
(order to favour the candidate of my choice?" And the reason no one knows\
the answer to this )Tj
T*
(crucial question is because there is no answer. )Tj
0 -1.1 TD
( )Tj
T*
(Electoral choices are not a mechanical affair; the voter\222s mind is no\
t a computer into which you )Tj
0 -1.2 TD
(can input percentages and be sure to receive a given vote as output. Few\
electors vote merely )Tj
T*
(on the strength of their knowledge of other people\222s choice. And even\
those electors who )Tj
T*
(operate in that way have diverse reactions to a given opinion poll. Thei\
r conclusion is )Tj
T*
(determined by the credibility of the source, each individual\222s politi\
cal sympathies, )Tj
T*
(temperament, and many other elements which combine to create a unique pe\
rsonal decision. )Tj
T*
(Legislation to "protect" adult citizens is an insult to their ability to\
make their own decisions. )Tj
0 -1.1 TD
( )Tj
T*
(The second debatable assumption is the idea that the citizen's freedom o\
f choice is better )Tj
0 -1.2 TD
(protected by some form of regulation rather than by free and competitive\
information. )Tj
T*
(However, it is silence, not freedom, which lends itself to rumour and ma\
nipulation. Censorship )Tj
T*
(creates two categories of citizens, those who are entitled to full infor\
mation, \(in this case )Tj
T*
(through private polls conducted by those with the resources to do so - o\
ften including the )Tj
T*
(legislators themselves\), and those who are held to be too easily duped \
and who must be kept )Tj
T*
(unaware of any changes in the public opinion of candidates toward the en\
d of a campaign. )Tj
0 -1.1 TD
( )Tj
T*
(The French referendum on the Maastricht Treaty highlighted the dangers o\
f two speed access )Tj
0 -1.2 TD
(to information in this way. Small investors were denied the right to mon\
itor and consider the )Tj
T*
(evolution of the views of the electorate, while large financial organisa\
tions were daily )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(3 of 22\)5/7/2005 8:58:55 AM)Tj
ET
EMC
endstream
endobj
133 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 753.058 Tm
(commissioning private polls which enabled them to foresee the ups and do\
wns of the European )Tj
0 -1.2 TD
(monetary system - an unforeseen result of a law whose declared aim was "\
to protect the )Tj
T*
(citizen against abuses and manipulations". )Tj
0 -1.1 TD
( )Tj
0 -1.1 TD
(The Council of Europe had no misconceptions on this point. In September \
1985 it approved the )Tj
0 -1.2 TD
(conclusions of a report on opinion polls prepared by Sir John Page on be\
half of the Committee )Tj
T*
(on Parliamentary and Public Relations, which stated that "all the eviden\
ce of the influence of )Tj
0 -1.2 TD
(opinion polls on election results is subjective ...\224; \223the objecti\
ve publication of genuine public )Tj
T*
(opinion polls did not have a strong and discernable influence on the res\
ult ..."; "the Committee )Tj
T*
(are not of the opinion that stronger controls are shown to be desirable \
or necessary ...". )Tj
0 -1.1 TD
( )Tj
T*
(By explicit reference to the ICC/ESOMAR International Code, the Committe\
e recommended )Tj
0 -1.2 TD
(that polling institutes should follow the existing Codes and that furthe\
r restrictions on public )Tj
T*
(opinion polls were both unnecessary and undesirable. )Tj
0 -1.1 TD
( )Tj
T*
(Not only are further restrictions unnecessary and undesirable, but also \
the viability of existing )Tj
0 -1.2 TD
(restrictions is under severe threat. The explosive growth of the Interne\
t is the main factor )Tj
T*
(making it almost impossible to prevent the widespread publication of opi\
nion polls. During the )Tj
T*
(period when legislation prevents print and broadcast media from local pu\
blication, opinion polls )Tj
T*
(can be \(and have been\) conducted and published on the Internet to a wo\
rldwide audience. If )Tj
T*
(the force of reason and logic does not persuade legislators that poll re\
strictions are )Tj
T*
(unnecessary, the world wide web will probably make restrictions impossib\
le to impose. )Tj
0 -1.1 TD
( )Tj
0.157 0.38 0.671 rg
14 0 0 14 15 422.601 Tm
(Representativeness of Opinion Polls)Tj
0 0 0 rg
12 0 0 12 15 397.658 Tm
(All opinion polls should be based on scientific and representative measu\
rements of public )Tj
0 -1.2 TD
(opinion. Far too often the term opinion poll is misused to describe unsc\
ientific and )Tj
T*
(unrepresentative measurements of public opinion. Representativeness mean\
s the obtaining of )Tj
0 -1.2 TD
(measurements which can be generalised to apply without any statistical b\
ias to the whole )Tj
0 -1.2 TD
(population under consideration. ESOMAR is determined to oppose the misus\
e of the term )Tj
T*
(Opinion Poll to describe activity in contravention of this Code. Members\
are asked to bring any )Tj
T*
(instance of misuse to the attention of their ESOMAR Representative. ESOM\
AR\222s Professional )Tj
T*
(Standards Committee has prepared letters and support material for Repres\
entatives to use. )Tj
0 -1.2 TD
(Particular example of current concern are push polling, televoting, Inte\
rnet polls, frugging and )Tj
T*
(mega databases. These are explained below. )Tj
0 -1.1 TD
( )Tj
T*
(1. Push polling has become a feature of politics in some countries. Push\
polls use deliberately )Tj
0 -1.2 TD
(biased questionnaires or samples. Some seek to produce false poll findin\
gs in support of a )Tj
T*
(particular issue. Others use biased questions in an attempt to convince \
those being )Tj
0 -1.2 TD
(interviewed to support a particular point of view. In both cases the \223\
poll\224 is a deliberate )Tj
0 -1.2 TD
(attempt to manipulate public opinion. This clearly contravenes the funda\
mental principles of )Tj
T*
(the ICC/ESOMAR Code which prohibits researchers from participating in su\
ch exercises. Rule )Tj
T*
(15 of the Code makes it clear that such non research activities as push \
polling must not be )Tj
T*
(carried out by research organisations. )Tj
0 -1.1 TD
( )Tj
T*
(2. Televoting involves inviting viewers or readers to phone special numb\
ers in order to register )Tj
0 -1.2 TD
(their vote on a particular issue. The results are then published or broa\
dcast on TV or radio as )Tj
T*
(editorial content. With advances in technology, these televotes can now \
be live, on screen as )Tj
T*
(viewers phone in. While this is a perfectly legitimate media exercise fo\
r entertainment )Tj
T*
(purposes, it can not be described as an opinion poll which satisfies the\
requirement of the )Tj
T*
(ESOMAR Code. Televoting polls cannot deliver reliable and representative\
samples. They can )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(4 of 22\)5/7/2005 8:58:55 AM)Tj
ET
EMC
endstream
endobj
134 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 754.858 Tm
(represent the views only of those viewing or listening, who chose to res\
pond. Even then, )Tj
0 -1.2 TD
(because there is no control of the number of times each person can vote,\
there is no )Tj
T*
(guarantee that the televote is representative of those who voted. The re\
sults from televotes )Tj
0 -1.2 TD
(cannot be generalised to apply to the whole population. Researchers shou\
ld not carry out such )Tj
0 -1.2 TD
(projects. )Tj
0 -1.1 TD
( )Tj
0 -1.1 TD
(3. Internet polls, conducted by placing questions on a web site and invi\
ting visitors to the site )Tj
0 -1.2 TD
(to give their opinion, have recently started to become common. It is at \
present impossible to )Tj
T*
(get a representative sample of public opinion using the Internet in this\
way. Only a small and )Tj
T*
(unrepresentative minority has access to the Internet. Further, it is ver\
y difficult to obtain a )Tj
T*
(representative sample of Internet users using this method as it will alm\
ost certainly over )Tj
T*
(represent the heavy users who will have the greatest chance of coming ac\
ross the )Tj
T*
(questionnaire. Researchers should be cautious when creating web sites co\
ntaining opinion poll )Tj
T*
(type questions. While this may be a valid activity for surveys of certai\
n computer users, or for )Tj
T*
(experimental purposes, researchers should be careful not to set the wron\
g example by )Tj
T*
(publicising the findings of Internet surveys unless they are sure that t\
he sample is )Tj
T*
(representative. )Tj
0 -1.1 TD
( )Tj
T*
(4. Frugging stands for fund raising under the guise of polling. Some pol\
itical parties, charities )Tj
0 -1.2 TD
(and pressure groups have started to use false opinion poll approaches in\
order to request )Tj
T*
(donations from respondents. While understanding the need of these organi\
sations to collect )Tj
T*
(funds and establish the views of their supporters, it is in the interest\
s of all concerned that the )Tj
T*
(public should not be misled into co-operation by making false claims abo\
ut the intention of the )Tj
T*
(enquiry. Again, researchers should not carry out fund raising under the \
guise of polling and )Tj
T*
(should bring any examples to the attention of their ESOMAR Representativ\
e. )Tj
0 -1.1 TD
( )Tj
T*
(5. One final area of concern is the creation of mega databases for the d\
istribution of millions of )Tj
0 -1.2 TD
(questionnaires. There is no scientific support for the notion that just \
because millions of people )Tj
T*
(have answered some questions, the results produced will be valid and rel\
iable. Some direct )Tj
0 -1.2 TD
(marketing databases now claim to have answers from tens of millions of p\
eople; one UK utility )Tj
T*
(company recently sent a questionnaire to all 17 million of its customers\
; a French transport )Tj
T*
(company distributes 1.5 million self completion questionnaires to young \
people using its )Tj
T*
(service; these are all examples of massive but unscientific sampling lik\
ely to deliver )Tj
T*
(unrepresentative results. The accuracy and reliability of the results of\
an opinion poll depend )Tj
T*
(not merely on the number of people interviewed, but more importantly on \
the scientific )Tj
T*
(representativeness of the sample questioned and responding. )Tj
0 -1.1 TD
( )Tj
T*
(The 1936 US presidential election campaign provided the first, and still\
the best, )Tj
0 -1.2 TD
(demonstration of the inaccuracy of the \223ask millions approach\224. Th\
e Literary Digest poll )Tj
T*
(dispatched questionnaires to 10,000,000 Americans. 2,376,523 replied and\
the analysis )Tj
T*
(pointed to a Landon victory. \(If you asked your self \223Who is Landon?\
\224, you already know they )Tj
T*
(got it wrong!\). Among others, George Gallup using scientifically constr\
ucted samples of only )Tj
T*
(3,000 respondents predicted a Roosevelt win. This was the real birth of \
modern opinion polling )Tj
T*
(and the lesson still applies today. The Literary Digest \223poll\224 of \
millions was in error by some )Tj
T*
(19%. )Tj
0 -1.1 TD
( )Tj
0.157 0.38 0.671 rg
14 0 0 14 15 95.601 Tm
(International Code of Practice for the Publication of Public Opinion Pol\
l Results )Tj
0 0 0 rg
/TT1 1 Tf
12 0 0 12 15 70.6662 Tm
(4.1 Introduction to the Code)Tj
/TT0 1 Tf
( )Tj
0 -1.101 TD
( )Tj
0 -1.1 TD
(1. Public opinion research - the study of people's attitudes and beliefs\
about political, social )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(5 of 22\)5/7/2005 8:58:55 AM)Tj
ET
EMC
endstream
endobj
135 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 754.858 Tm
(and other issues - forms a part of the total marketing and social resear\
ch field. It is subject to )Tj
0 -1.2 TD
(exactly the same professional and ethical requirements as other forms of\
survey research. )Tj
T*
(These requirements are set out in the ICC/ESOMAR International Code of M\
arketing and Social )Tj
0 -1.2 TD
(Research Practice. )Tj
0 -1.1 TD
( )Tj
T*
(2. However, public opinion research tends to be a specially `sensitive' \
area. It deals with )Tj
0 -1.2 TD
(issues which arouse greater public interest and emotion than do most com\
mercial market )Tj
T*
(research projects. In addition, its findings are much more widely publis\
hed and debated, and )Tj
0 -1.2 TD
(may sometimes be presented in a provocative or even tendentious way. ESO\
MAR has )Tj
T*
(therefore set out specific recommendations about the publication of such\
research. )Tj
0 -1.1 TD
( )Tj
T*
(3. Opinion polls have a valuable role to play in present-day society. It\
is desirable that the )Tj
0 -1.2 TD
(general public, politicians, the media and other interested groups shoul\
d through research )Tj
T*
(have access to accurate and unbiased measures of public attitudes and in\
tentions. We )Tj
T*
(recognise that there are some sincerely-held worries about possible \(bu\
t mostly unproven\) )Tj
T*
(effects which some polls could in theory have upon voting or other behav\
iour. However, the )Tj
T*
(alternative is that the public is exposed only to unscientific and proba\
bly inaccurate assertions )Tj
T*
(about the situation, in many cases presented by individuals or organisat\
ions who have an )Tj
T*
(insufficient understanding of the nature of the information they are usi\
ng or who take an )Tj
T*
(extremely partisan approach to presenting the facts. The objective of th\
is Code is to reduce )Tj
T*
(the risk of the public being misled by research which is inadequate or b\
adly presented. )Tj
0 -1.1 TD
( )Tj
T*
(4. The Parliamentary Assembly of the Council of Europe has examined this\
ESOMAR Code for )Tj
0 -1.2 TD
(the Publication of Opinion Polls and has given the Code its blessing. Th\
e Council of Europe has )Tj
T*
(recommended the widespread application of this Code to govern the public\
ation of polls. )Tj
0 -1.1 TD
( )Tj
T*
(5. The validity and value of public opinion polls depend on three main c\
onsiderations: )Tj
T*
( )Tj
0 -1.1 TD
(\(i\) the nature of the research techniques used and the efficiency with\
which they are applied, )Tj
0 -1.2 TD
(\(ii\) the honesty and objectivity of the research organisation carrying\
out the study, )Tj
T*
(\(iii\) the way in which the findings are presented and the uses to whic\
h they are put. )Tj
0 -1.1 TD
( )Tj
T*
(This Code concentrates primarily on the second and third of these issues\
. Guidelines on )Tj
0 -1.2 TD
(techniques and the conduct of opinion polls, and pre-election in particu\
lar, are given in the )Tj
T*
(next section. )Tj
0 -1.1 TD
( )Tj
T*
(6. Major problems can arise when opinion poll findings are published and\
debated. It would )Tj
0 -1.2 TD
(clearly be unrealistic, and unreasonable, to expect the media to quote t\
he full technical )Tj
T*
(background of a survey when presenting its findings: they have limitatio\
ns of space and must )Tj
T*
(also hold the interest of their audience. However, there is certain basi\
c information which must )Tj
T*
(be provided if that audience is to have the opportunity of judging for i\
tself the evidence )Tj
T*
(presented and deciding whether or not it agrees with any conclusions dra\
wn from the )Tj
T*
(research. This Code is primarily concerned with trying to ensure that th\
e public has reasonable )Tj
T*
(access to this key information about the survey, and that published repo\
rts of the findings are )Tj
T*
(not misleading. The Code tries to strike a realistic balance between wha\
t would be theoretically )Tj
T*
(desirable and what is practicable. )Tj
0 -1.1 TD
( )Tj
T*
(7. All reputable research organisations apply the appropriate scientific\
methods and operate )Tj
0 -1.2 TD
(with professional objectivity. In doing so they conform to the ICC/ESOMA\
R International Code )Tj
T*
(of Marketing and Social Research Practice. There is also general agreeme\
nt among them on )Tj
T*
(the principles which should underlie the publication of research results\
. However, normal )Tj
T*
(professional practice varies between countries in some respects and in c\
ertain countries )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(6 of 22\)5/7/2005 8:58:55 AM)Tj
ET
EMC
endstream
endobj
136 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 754.858 Tm
(additional information to that specified in this Code will also customar\
ily be provided as part of )Tj
0 -1.2 TD
(the standard key material. )Tj
0 -1.1 TD
( )Tj
0 -1.1 TD
(8. Research organisations have a particular responsibility in the field \
of public opinion polls, to )Tj
0 -1.2 TD
(make sure that both the client and the public have a reasonable understa\
nding of the special )Tj
0 -1.2 TD
(problems and limitations involved in measuring attitudes and beliefs as \
distinct from )Tj
T*
(behaviour. Such research frequently deals with complex and sensitive iss\
ues about which )Tj
0 -1.2 TD
(respondents have varying degrees of knowledge and interest, and where th\
eir views may often )Tj
T*
(be half-formed, confused and inconsistent. High professional integrity a\
nd skill is essential if )Tj
T*
(the research itself is to be unbiased and meaningful, and if the finding\
s are to be presented )Tj
T*
(and interpreted clearly and accurately. It is important also that the re\
search budget available )Tj
T*
(is sufficient to carry out a valid study. ESOMAR fully recognises that s\
uch considerations are )Tj
T*
(vital if public opinion polls are to merit public confidence and support\
. )Tj
0 -1.1 TD
( )Tj
T*
(9. Finally, if as a result of past experience, a research organisation h\
as reason to believe that a )Tj
0 -1.2 TD
(particular client will not fairly present opinion poll results in his pu\
blished version of the )Tj
T*
(findings, the research organisation has a responsibility to stop carryin\
g out polls for publication )Tj
T*
(by that client. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(4.2 The Code )Tj
0 -1.1 TD
( )Tj
T*
(A. Basic Requirements of the ICC/ESOMAR International Code of Marketing \
and )Tj
0 -1.2 TD
(Social Research Practice)Tj
/TT0 1 Tf
( )Tj
0 -1.101 TD
( )Tj
0 -1.1 TD
(1. All research organisations which conduct public opinion polls must co\
nform to the ICC/)Tj
0 -1.2 TD
(ESOMAR International Code of Marketing and Social Research Practice. Par\
ticular attention is )Tj
T*
(drawn to the requirements of Rule 15 \(concerning the clear separation o\
f research from non-)Tj
T*
(research activities\), Rules 14 and 27 \(concerning misleading reporting\
\), Rules 25 and 26 )Tj
T*
(\(concerning preparation of reports\) and Rule 29 \(concerning making th\
e client aware of the )Tj
0 -1.2 TD
(ESOMAR Code\). These Rules together with relevant extracts from the Note\
s on How the Code )Tj
T*
(should be applied are reproduced in Appendix 1 to this document. )Tj
0 -1.1 TD
( )Tj
T*
(2. It is important to distinguish between the requirements which apply t\
o the reporting of )Tj
0 -1.2 TD
(public opinion poll results by a research organisation to its original c\
lient, and those which )Tj
T*
(apply to the subsequent publishing of any poll findings by that client t\
o a wider audience. The )Tj
T*
(first of these situations is largely covered by the Notes on the applica\
tion of Rule 25 of the )Tj
T*
(existing International Code which specifies reporting requirements in de\
tail. This )Tj
T*
(supplementary Code is intended to clarify certain additional requirement\
s which arise in )Tj
T*
(connection with the wider publication of the findings, and therefore app\
lies especially to the )Tj
T*
(second situation. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(B. Additional Requirements)Tj
/TT0 1 Tf
( )Tj
0 -1.101 TD
( )Tj
0 -1.1 TD
(3. When any public opinion poll findings are published in print media th\
ese should always be )Tj
0 -1.2 TD
(accompanied by a clear statement of: )Tj
0 -1.1 TD
( )Tj
T*
(\(a\) the name of the research organisation carrying out the survey; )Tj
0 -1.2 TD
(\(b\) the universe effectively represented \(i.e. who was interviewed\);\
)Tj
T*
(\(c\) the achieved sample size and its geographical coverage; )Tj
T*
(\(d\) the dates of fieldwork; )Tj
T*
(\(e\) the sampling method used \(and in the case of random samples, the \
success rate achieved\); )Tj
T*
(\(f\) the method by which the information was collected \(personal or te\
lephone interview, etc.\); )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(7 of 22\)5/7/2005 8:58:55 AM)Tj
ET
EMC
endstream
endobj
137 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 754.858 Tm
(\(g\) the relevant questions asked. In order to avoid possible ambiguity\
the actual wording of )Tj
0 -1.2 TD
(the question should be given unless this is a standard question already \
familiar to the audience )Tj
T*
(or it is given in a previous published report to which reference is made\
. )Tj
0 -1.1 TD
( )Tj
0 -1.1 TD
(4. In the case of broadcast media it may not be possible always to give \
information on all )Tj
0 -1.2 TD
(these points. As a minimum, points \(a\) - \(d\) above should normally b\
e covered in any )Tj
T*
(broadcast reference to the findings of a public opinion poll, preferably\
in visual \(written\) form )Tj
0 -1.2 TD
(where practical. )Tj
0 -1.1 TD
( )Tj
T*
(5. The percentages of respondents who give 'don't know' answers \(and in\
the case of voting-)Tj
0 -1.2 TD
(intention studies, of those who say they will not vote\) must always be \
given where they are )Tj
T*
(likely to significantly affect the interpretation of the findings. When \
comparing the findings )Tj
T*
(from different surveys, any changes \(other than minor ones\) in these p\
ercentages must also )Tj
T*
(be indicated. )Tj
0 -1.1 TD
( )Tj
T*
(6. In the case of voting-intention surveys, it must always be made clear\
if voting-intention )Tj
0 -1.2 TD
(percentages quoted include any of these respondents who answered `don't \
know' or `may not/)Tj
T*
(will not vote' in reply to the voting questions asked. )Tj
0 -1.1 TD
( )Tj
T*
(7. Whatever information may be given in the published report of the surv\
ey, the publisher and/)Tj
0 -1.2 TD
(or the research organisation involved must be prepared on request to sup\
ply the other )Tj
T*
(information about the survey methods described in the Notes on the appli\
cation of Rule 25 of )Tj
T*
(the International Code. Where the questions reported on have formed part\
of a more extensive )Tj
T*
(or `omnibus' survey, this must be made clear to any enquirer. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
T*
(C. Arrangements between the Research Organisation and its Client)Tj
/TT0 1 Tf
( )Tj
0 -1.101 TD
( )Tj
0 -1.1 TD
(8. In order to ensure that these Code requirements are followed, and to \
avoid possible )Tj
0 -1.2 TD
(misunderstandings, the research organisation must make clear in advance \
to its client: )Tj
0 -1.1 TD
( )Tj
T*
(\(i\) that the research organisation itself is bound by the requirements\
of the general )Tj
0 -1.2 TD
(International Code. )Tj
T*
(\(ii\) that subsequent wider publication of the research findings should\
be in accordance with )Tj
T*
(this supplementary Code. )Tj
0 -1.1 TD
( )Tj
T*
(It is therefore the responsibility of the research organisation to draw \
its client's attention to )Tj
0 -1.2 TD
(the present Code on Publication of Results and to use its best endeavour\
s to persuade the )Tj
T*
(client to follow the Code's requirements. )Tj
0 -1.1 TD
( )Tj
T*
(9. The research organisation and the client each have a responsibility i\
n the public interest to )Tj
0 -1.2 TD
(ensure that the published report on a public opinion poll does not misre\
present or distort the )Tj
T*
(survey data. For example, misleading comments based on non-significant d\
ifferences must be )Tj
T*
(avoided. Special care must be taken to ensure that any graphs or charts \
used do not convey a )Tj
T*
(misleading impression of the current survey's results or of trends over \
time. It is also )Tj
T*
(important that the reader or listener should be able clearly to distingu\
ish between the survey )Tj
T*
(findings as such and any editorial or other comments based upon them. Pa\
rticularly in the case )Tj
T*
(of print reports, the research organisation must wherever feasible appro\
ve in advance the )Tj
T*
(exact form and content of publication as required in Rule 27 of the orig\
inal International Code. )Tj
0 -1.1 TD
( )Tj
T*
(10. The research organisation cannot normally be held responsible for an\
y subsequent use )Tj
0 -1.2 TD
(made of public opinion poll results by people other than the original cl\
ient. It should however )Tj
T*
(be ready to issue immediately such comments or information as may be nec\
essary to correct )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(8 of 22\)5/7/2005 8:58:55 AM)Tj
ET
EMC
endstream
endobj
138 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 754.858 Tm
(any cases of misreporting or misuse of results when these are brought to\
its attention. )Tj
0 -1.1 TD
( )Tj
T*
(11. In the event that a client releases data from a survey which was not\
originally intended for )Tj
0 -1.2 TD
(publication, this Code of Conduct will apply to it as if it had original\
ly been commissioned for )Tj
0 -1.257 TD
(publication.)Tj
/T1_0 1 Tf
14 0 0 14 84.852 698.9756 Tm
( )Tj
0.157 0.38 0.671 rg
/TT0 1 Tf
-4.989 -2.612 Td
(Appendix 1)Tj
0 0 0 rg
/TT1 1 Tf
12 0 0 12 15 637.4662 Tm
(ICC/ESOMAR International Code of Marketing and Social Research Practice)Tj
/TT0 1 Tf
( )Tj
0 -1.101 TD
( )Tj
0 -1.1 TD
(While all Rules of the International Code apply to Public Opinion Polls,\
the following Rules have )Tj
0 -1.2 TD
(special significance in this connection: )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(The Professional Responsibilities of Researchers)Tj
/TT0 1 Tf
( )Tj
0 -1.101 TD
( )Tj
0 -1.1 TD
(14. Researchers must not knowingly allow the dissemination of conclusion\
s from a marketing )Tj
0 -1.2 TD
(research project which are not adequately supported by the data. They mu\
st always be )Tj
T*
(prepared to make available the technical information necessary to assess\
the validity of any )Tj
T*
(published findings. )Tj
0 -1.1 TD
( )Tj
T*
(15. When acting in their capacity as Researchers the latter must not und\
ertake any non-)Tj
0 -1.2 TD
(research activities, for example database marketing involving data about\
individuals which will )Tj
0 -1.2 TD
(be used for direct marketing and promotional activities. Any such non-re\
search activities must )Tj
0 -1.2 TD
(always, in the way they are organised and carried out, be clearly differ\
entiated from marketing )Tj
T*
(research activities. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(The Mutual Rights and Responsibilities of Researchers and Clients)Tj
/TT0 1 Tf
( )Tj
0 -1.101 TD
( )Tj
0 -1.1 TD
(25. The Researcher must provide the Client with all appropriate technica\
l details of any )Tj
0 -1.2 TD
(research project carried out for that Client. )Tj
0 -1.1 TD
( )Tj
T*
(26. When reporting on the results of a marketing research project the Re\
searcher must make )Tj
0 -1.2 TD
(a clear distinction between the findings as such, the Researcher's inter\
pretation of these and )Tj
T*
(any recommendations based on them. )Tj
0 -1.1 TD
( )Tj
T*
(27. Where any of the findings of a research project are published by the\
Client the latter has a )Tj
0 -1.2 TD
(responsibility to ensure that these are not misleading. The Researcher m\
ust be consulted and )Tj
T*
(agree in advance the form and content of publication, and must take acti\
on to correct any )Tj
T*
(misleading statements about the research and its findings. )Tj
0 -1.1 TD
( )Tj
T*
(29. Researchers must ensure that Clients are aware of the existence of t\
his Code and of the )Tj
0 -1.2 TD
(need to comply with its requirements. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(Notes on how to apply the ICC/ESOMAR International Code of Marketing and\
Social )Tj
0 -1.2 TD
(Research Practice )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
T*
(\(Rule 14\))Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(The kinds of technical information which should on request be made avail\
able include those )Tj
0 -1.2 TD
(listed in the Notes to Rule 25. The Researcher must not however disclose\
information which is )Tj
T*
(confidential to the Client's business, nor need he disclose information \
relating to parts of the )Tj
T*
(survey which were not published. )Tj
0 -1.1 TD
( )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(9 of 22\)5/7/2005 8:58:55 AM)Tj
ET
EMC
endstream
endobj
139 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 754.8662 Tm
(\(Rule 15\))Tj
/TT1 1 Tf
( )Tj
0 -1.201 TD
(The kinds of "non-research activity" that must not be associated in any \
way with the carrying )Tj
0 -1.257 TD
(out of marketing research include:)Tj
/T1_0 1 Tf
14 0 0 14 224.328 725.3755 Tm
( )Tj
ET
q
8 0 0 7 15 694.8000031 cm
/Im0 Do
Q
BT
/TT1 1 Tf
12 0 0 12 23 691.858 Tm
( )Tj
0 -0.1 TD
(Enquiries whose objectives are to obtain personal information about priv\
ate individuals per )Tj
0 -1.2 TD
(se, whether for legal, political, supervisory \(e.g., job performance\),\
private or other )Tj
T*
(purposes; )Tj
ET
q
8 0 0 7 15 651.6000061 cm
/Im0 Do
Q
BT
/TT1 1 Tf
12 0 0 12 23 648.658 Tm
( )Tj
0 -0.1 TD
(The acquisition of information for use for credit-rating or similar purp\
oses; )Tj
ET
q
8 0 0 7 15 637.1999969 cm
/Im0 Do
Q
BT
/TT1 1 Tf
12 0 0 12 23 634.258 Tm
( )Tj
T*
(The compilation, updating or enhancement of lists, registers or database\
s that are not for )Tj
0 -1.2 TD
(scientific research purposes \(e.g., those that may be used for direct m\
arketing\); )Tj
ET
q
8 0 0 7 15 608.3999939 cm
/Im0 Do
Q
BT
/TT1 1 Tf
12 0 0 12 23 605.458 Tm
( )Tj
0 -0.1 TD
(Industrial, commercial or any other form of espionage; )Tj
ET
q
8 0 0 7 15 594 cm
/Im0 Do
Q
BT
/TT1 1 Tf
12 0 0 12 23 591.058 Tm
( )Tj
T*
(Sales or promotional approaches to individual Respondents; )Tj
ET
q
8 0 0 7 15 579.6000061 cm
/Im0 Do
Q
BT
/TT1 1 Tf
12 0 0 12 23 576.658 Tm
( )Tj
T*
(The collection of debts; and )Tj
ET
q
8 0 0 7 15 565.1999969 cm
/Im0 Do
Q
BT
/TT1 1 Tf
12 0 0 12 23 562.258 Tm
( )Tj
T*
(Fund-raising.)Tj
-0.667 -2.033 Td
(Certain of these activities - in particular, the collection of informati\
on for databases for )Tj
0 -1.2 TD
(subsequent use in direct marketing and similar operations - are legitima\
te activities in their )Tj
T*
(own right. Researchers \(e.g. those working within a client company\) ma\
y be involved with )Tj
T*
(such activities either directly or indirectly. In such cases it is essen\
tial that a very clear )Tj
T*
(distinction be made between such activities and marketing research. Any \
work that involves )Tj
T*
(the collection and use of personal data for non-research purposes \(such\
as those listed above\) )Tj
0 -1.2 TD
(must not be carried out under the name of marketing research or of a mar\
keting research )Tj
T*
(organisation as such, or be incorporated into a marketing research surve\
y. Personal data )Tj
T*
(collected for marketing research purposes must never be used in connecti\
on with non-research )Tj
T*
(activities such as direct marketing. )Tj
0 -1.1 TD
( )Tj
T*
(However, the use of marketing research data in connection with non-resea\
rch databases is )Tj
0 -1.2 TD
(permissible where Researchers have first ensured that such research info\
rmation has been )Tj
0 -1.2 TD
(fully depersonalised \(i.e., anonymised\). A common way of achieving thi\
s is by "modelling" the )Tj
T*
(research data before its fusion with any other data. This is permissible\
only if there is no risk )Tj
T*
(that any data in the database which is derived from marketing research c\
ould be linked to )Tj
T*
(individual Respondents or Data Subjects. )Tj
0 -1.1 TD
( )Tj
T*
(These issues are considered at greater length in the ESOMAR Guideline on\
Maintaining the )Tj
0 -1.2 TD
(Distinctions between Marketing Research and Direct Marketing. )Tj
0 -1.1 TD
( )Tj
T*
(There are no additional requirements which apply to customer satisfactio\
n research in cases )Tj
0 -1.2 TD
(where no personal data are disclosed outside the research organisation\(\
s\) responsible for the )Tj
T*
(project. Furthermore, if a survey sample or mailing list has been provid\
ed for the project by an )Tj
T*
(outside company \(eg client or other research organisation\) it is also \
reasonable for the )Tj
T*
(Researcher to notify that company of any names and addresses which have \
been found during )Tj
T*
(the course of the survey to be no longer operational \(eg because a Resp\
ondent has died or )Tj
0 -1.2 TD
(moved away from the address given\). The situation is more complex in th\
e case of a study )Tj
T*
(which may involve other data about identified Respondents \(eg specific \
queries or comments\) )Tj
T*
(being disclosed outside the research organisation: these issues are addr\
essed in more detail in )Tj
T*
(a separate ESOMAR Guideline on Customer Satisfaction Research. )Tj
0 -1.1 TD
( )Tj
/TT0 1 Tf
0 -1.099 TD
(\(Rule 25\))Tj
/TT1 1 Tf
( )Tj
0 -1.201 TD
(The Client is entitled to the following information about any marketing \
research project to )Tj
0 -1.2 TD
(which he has subscribed: )Tj
0 -1.1 TD
( )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(10 of 22\)5/7/2005 8:58:55 AM\
)Tj
ET
EMC
endstream
endobj
140 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 754.858 Tm
(1. Background )Tj
0 -1.2 TD
(- for whom the study was conducted )Tj
T*
(- the purpose of the study )Tj
0 -1.2 TD
(- names of subcontractors and consultants performing any substantial par\
t of the work )Tj
0 -1.1 TD
( )Tj
T*
(2. Sample )Tj
0 -1.2 TD
(- a description of the intended and actual universe covered )Tj
T*
(- the size, nature and geographical distribution of the sample \(both pl\
anned and achieved\);and )Tj
0 -1.2 TD
(where relevant, the extent to which any of the data collected were obtai\
ned from only part of )Tj
T*
(the sample )Tj
T*
(- details of the sampling method and any weighting methods used )Tj
T*
(- where technically relevant, a statement of response rates and a discus\
sion of any possible )Tj
T*
(bias due to non-response )Tj
0 -1.1 TD
( )Tj
T*
(3. Data collection )Tj
0 -1.2 TD
(- a description of the method by which the information was collected )Tj
T*
(- a description of the field staff, briefing and field quality control m\
ethods used )Tj
T*
(- the method of recruiting Respondents; and the general nature of any in\
centives offered to )Tj
T*
(secure their co-operation )Tj
T*
(- when the fieldwork was carried out )Tj
T*
(- \(in the case of \221desk research'\) a clear statement of the sources\
of the information and their )Tj
T*
(likely reliability )Tj
0 -1.1 TD
( )Tj
T*
(4. Presentation of results )Tj
0 -1.2 TD
(- the relevant factual findings obtained )Tj
T*
(- bases of percentages \(both weighted and unweighted\) )Tj
T*
(- general indications of the probable statistical margins of error to be\
attached to the main )Tj
T*
(findings, and of the levels of statistical significance of differences b\
etween key figures )Tj
0 -1.2 TD
(- the questionnaire and other relevant documents and materials used \(or\
, in the case of a )Tj
T*
(shared project, that portion relating to the matter reported on\). )Tj
0 -1.1 TD
( )Tj
T*
(The Report on a project should normally cover the above points or provid\
e a reference to a )Tj
0 -1.2 TD
(readily available separate document which contains the information. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(\(Rule 27\))Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(It is clearly impossible for a Researcher fully to control the ways in w\
hich research findings are )Tj
0 -1.2 TD
(interpreted or applied once these are in the public domain. However, Res\
earchers should use )Tj
T*
(their best endeavours to prevent any misinterpretation or misuse of rese\
arch findings, and \(as )Tj
T*
(far as is practicable\) to correct any such misinterpretation or misuse \
once they become aware )Tj
T*
(that this has happened. )Tj
0 -1.1 TD
( )Tj
T*
(The publication of research findings may sometimes prove to be misleadin\
g because certain of )Tj
0 -1.2 TD
(the technical aspects or limitations of the research have not been fully\
appreciated and/or )Tj
T*
(because the public presentation, explanation and discussion of the findi\
ngs \(eg in the media\) )Tj
T*
(have not clearly spelt out all the relevant considerations. This can hap\
pen accidentally, or as a )Tj
T*
(result of the pressures on media time and space, rather than for any mor\
e undesirable reason. )Tj
T*
(Researchers can reduce the danger of such problems arising by making sur\
e \(eg in their )Tj
T*
(contract for a research project\) that they are consulted in advance by \
the Client about the )Tj
T*
(form in which any research findings will be published. If following publ\
ication it becomes clear )Tj
T*
(that serious misinterpretation of the research and its findings has occu\
rred, leading to )Tj
T*
(misleading discussion of the implications of the research, the Researche\
r should endeavour to )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(11 of 22\)5/7/2005 8:58:55 AM\
)Tj
ET
EMC
endstream
endobj
141 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 754.858 Tm
(correct such misinterpretation by any available and appropriate means. )Tj
0 -1.2 TD
(In a case where the Client does not consult and agree in advance the for\
m of publication with )Tj
T*
(the researcher, the latter is entitled to: )Tj
0 -1.1 TD
( )Tj
0 -1.1 TD
(\(i\) refuse permission for his name to be used in connection with the p\
ublished findings and )Tj
0 -1.2 TD
(\(ii\) publish the appropriate technical details of the project \(as lis\
ted in the Notes to Rule 25\). )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(\(Rule 29\))Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(It is recommended that Researchers specify in their research proposals t\
hat they follow the )Tj
0 -1.2 TD
(requirements of this ICC/ESOMAR International Code and that they make a \
copy available to )Tj
0 -1.257 TD
(the Client if the latter does not already have one.)Tj
/T1_0 1 Tf
14 0 0 14 313.944 614.9756 Tm
( )Tj
0.157 0.38 0.671 rg
/TT0 1 Tf
-21.353 -2.612 Td
(Guideline to the Interpretation of the International Code of Practice fo\
r the )Tj
0 -1.2 TD
(Publication of Public Opinion Poll Results)Tj
0 0 0 rg
12 0 0 12 15 536.658 Tm
(These Guidelines to the Interpretation of the Code for the Publication o\
f Opinion Polls are )Tj
0 -1.2 TD
(designed to help answer some of the questions and problems which inevita\
bly arise when )Tj
T*
(putting the Code's recommendations into practice. )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
0 -1.099 TD
(5.1 General principles)Tj
/TT0 1 Tf
( )Tj
0 -1.101 TD
( )Tj
0 -1.1 TD
(1. The key consideration must be to hold firmly to the objectives and sp\
irit of the Code. )Tj
0 -1.2 TD
(ESOMAR will look for steady progress in the general standards reached. )Tj
0 -1.1 TD
( )Tj
T*
(2. ESOMAR appreciates that it may not always be easy to persuade some cl\
ients that certain )Tj
0 -1.2 TD
(recommendations should be adhered to in the public interest. Also it is \
essential to recognise )Tj
T*
(and support the important principle of \(responsible\) editorial freedom\
. In the words of the )Tj
T*
(Code itself, members are expected to )Tj
/TT2 1 Tf
(use their "best endeavours")Tj
/TT0 1 Tf
( to achieve the objectives )Tj
0 -1.201 TD
(laid down in the interests both of their profession and of the public in\
general. )Tj
0 -1.1 TD
( )Tj
T*
(3. ESOMAR will regularly review any problems which individuals and organ\
isations encounter )Tj
0 -1.2 TD
(with the Code in practice. If experience suggests that it would be sensi\
ble and justified to do )Tj
T*
(so, some of the recommendations may then be amended in the light of this\
experience. )Tj
0 -1.1 TD
( )Tj
T*
(4. Membership of ESOMAR - and therefore the primary responsibility for f\
ollowing the Code - is )Tj
0 -1.2 TD
(an individual one. However, ESOMAR would expect organisations associated\
with the Society )Tj
T*
(through individual membership equally to do their best to ensure that th\
e Code is followed, )Tj
T*
(and will fully support such efforts. )Tj
0 -1.1 TD
( )Tj
T*
(5. Many research organisations already exercise strong control over the \
way in which their )Tj
0 -1.2 TD
(research findings are publicly reported on through the form of contract \
they have with their )Tj
T*
(clients. This is a desirable safeguard which is referred to again later \
in the Guidelines. )Tj
0 -1.1 TD
( )Tj
0 -1.1 TD
(6. Where individual members - and their organisation - are involved with\
a study purely as a )Tj
0 -1.2 TD
(sub-contractor \(for example, for the fieldwork\) their responsibilities\
must relate primarily to )Tj
T*
(this more limited aspect of the total study. The key requirements here a\
re covered by the )Tj
T*
(main International Code. At the same time, members would naturally be ex\
pected to use their )Tj
T*
(influence as far as possible to ensure that the total study is handled i\
n a way which conforms )Tj
T*
(with the recommendations in this supplementary Code - for example, by ob\
taining an )Tj
T*
(assurance beforehand to this effect. If the sponsoring organisation for \
whom they carry out )Tj
T*
(such sub-contracting work repeatedly breaches the Code, they must very s\
eriously consider )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(12 of 22\)5/7/2005 8:58:55 AM\
)Tj
ET
EMC
endstream
endobj
142 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 754.858 Tm
(of exit polls. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(14. Local Laws)Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(All pre-election polls must be carried out in conformance with local law\
s. If any guides in this )Tj
0 -1.2 TD
(document contravene local legislation, the local laws must take priority\
. )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
T*
(ESOMAR\222s mission is to promote the use of opinion and market research\
for improving decision )Tj
0 -1.2 TD
(making in business and society worldwide. Founded in 1948, ESOMAR unites\
4000 members in )Tj
T*
(100 countries. The Society facilitates the exchange of experiences betwe\
en suppliers and users )Tj
T*
(of research in order to optimise the integration of research results int\
o the decision making )Tj
T*
(process. )Tj
0 -1.1 TD
( )Tj
T*
(Founded in 1947, the World Association for Public Opinion Research - WAP\
OR aims to further )Tj
0 -1.2 TD
(the use of scientific survey research in national and international affa\
irs. WAPOR is officially )Tj
T*
(recognised as a member of the International Social Science Council \(ISS\
C\) and is supported by )Tj
T*
(UNESCO. There are over 500 members in more than 60 countries. )Tj
0 -1.1 TD
( )Tj
T*
(Copyright ESOMAR/WAPOR 1998. 1st reprint 2003)Tj
/TT0 1 Tf
( )Tj
/TT1 1 Tf
0 -2.783 TD
(ESOMAR)Tj
15.503 0 Td
(The World Association for Public Opinion Research)Tj
/TT0 1 Tf
-15.503 -1.201 Td
(Vondelstraat 172)Tj
15.503 0 Td
(UNL Gallup Research Center)Tj
-15.503 -1.2 Td
(1054 GV Amsterdam)Tj
15.503 0 Td
(200 N. 11th Street)Tj
-15.503 -1.2 Td
(The Netherlands)Tj
15.503 0 Td
(Lincoln, NE 68588-0241)Tj
-15.503 -1.2 Td
(Tel: +31-20-664.2141)Tj
15.503 0 Td
(Tel: +1-402-458.2030)Tj
-15.503 -1.2 Td
(Fax: +31-20-664.2922)Tj
15.503 0 Td
(Fax: +1-402.458.2038)Tj
-15.503 -1.2 Td
(Email: )Tj
0.157 0.38 0.671 rg
(email@esomar.org)Tj
0 0 0 rg
15.503 0 Td
(Email: )Tj
0.157 0.38 0.671 rg
(wapor@unl.edu)Tj
0 0 0 rg
-15.503 -1.45 Td
(Website: www.esomar.org)Tj
15.503 0.083 Td
(Internet: )Tj
0.157 0.38 0.671 rg
(www.wapor.org )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(22 of 22\)5/7/2005 8:58:55 AM\
)Tj
ET
EMC
endstream
endobj
143 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 754.858 Tm
(whether or not they are justified in carrying out further work for such \
a sponsor \(please note )Tj
0 -1.2 TD
(the comments on the Code\222s Introduction Point 9 later in these Guidel\
ines\). )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(5.2 Specific issues )Tj
0 -1.1 TD
( )Tj
0 -1.1 TD
(Introduction to point 9)Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(There are liable to be occasions on which, despite the researcher\222s \223\
best endeavours\224, the )Tj
0 -1.2 TD
(detailed recommendations of this Code are not completely met by a given \
report on a public )Tj
0 -1.2 TD
(opinion poll. ESOMAR continues to strive for full conformity with the re\
commendations; but the )Tj
T*
(sanction recommended in Point 9 of the Introduction is intended to apply\
mainly to the case of )Tj
/TT1 1 Tf
T*
(repeated deliberate misinterpretation)Tj
/TT0 1 Tf
(, not to more limited \223technical\224 short-comings in the )Tj
0 -1.201 TD
(published reports. )Tj
0 -1.1 TD
( )Tj
T*
(Cases where recommendations included in this Code appear to be contraven\
ed may )Tj
0 -1.2 TD
(sometimes be handled at national level under a local Code or legislation\
. Whether or not this )Tj
T*
(happens, the ESOMAR Secretariat should be informed, especially if the pr\
oblem involves a )Tj
T*
(member of the Society. In the latter cases the Society may offer advice \
about avoiding such )Tj
T*
(difficulties in the future. If it appears that a member has seriously fa\
iled to uphold the spirit of )Tj
T*
(this Code, the ESOMAR Professional Standards Committee will consider whe\
ther the case )Tj
T*
(merits some form of disciplinary action. Any action would be taken only \
after the member )Tj
T*
(involved has been given a full opportunity of showing whether or not he \
had in practice fully )Tj
T*
(used his best endeavours to follow the Code \(see ESOMAR\222s Disciplina\
ry Procedures\). )Tj
T*
(ESOMAR\222s primary concern is to encourage and support its members in t\
rying to establish )Tj
T*
(more general conformity to the Code\222s recommendations, and trusts tha\
t this will normally be )Tj
T*
(achieved by discussion and agreement with the parties concerned. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(Article B3)Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(Although the wording of this article is not completely mandatory, its re\
commendations are )Tj
0 -1.2 TD
(strong ones and should be adhered to as far as possible. Experience in c\
ertain countries )Tj
T*
(demonstrates that, even without the backing of legislation \(as applies \
for example to certain )Tj
T*
(aspects of polling in France\), it is quite possible in practice for pub\
lished reports on polls to )Tj
T*
(include all, or virtually all, of the information listed. )Tj
0 -1.1 TD
( )Tj
T*
(Any recommendations for a standard format must take account of the diffe\
rent styles, layouts, )Tj
0 -1.2 TD
(etc., of widely varying types of publication. One example of a suitable \
form of wording would )Tj
T*
(be: )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
T*
(\223This survey was carried out by ABC Research, on behalf of Intergalac\
tic News Inc., using a )Tj
0 -1.2 TD
(national quota sample of 1111 adults of voting age personally interviewe\
d in 102 locations )Tj
T*
(between 1st-5th March 1995.\224)Tj
/TT0 1 Tf
( )Tj
0 -1.1 TD
( )Tj
T*
(Another alternative is to use a \221data box\222 of the kind: )Tj
T*
( )Tj
/TT2 1 Tf
T*
(\223Survey carried out by XYZ Research, on behalf of Intergalactic News \
Inc. National survey of )Tj
0 -1.2 TD
(1234 adults aged 18 and above, personally interviewed between 25th-28th \
February 1996 in )Tj
T*
(86 locations. Random sample \(effective interviews = 76% of those eligib\
le for interview\)\224)Tj
/TT0 1 Tf
( )Tj
0 -1.1 TD
( )Tj
T*
(There are certain specific situations in which it is clearly difficult t\
o follow all the )Tj
0 -1.2 TD
(recommendations listed: )Tj
0 -1.1 TD
( )Tj
T*
(\(i\) where the survey reported on is )Tj
/TT1 1 Tf
(very extensive and complex )Tj
/TT0 1 Tf
(and where the media )Tj
0 -1.201 TD
(report can therefore provide only a relatively brief overview of the tot\
al survey )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(13 of 22\)5/7/2005 8:58:55 AM\
)Tj
ET
EMC
endstream
endobj
144 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 756.058 Tm
( )Tj
0 -1.1 TD
(\(ii\) where an article summarises the results of a )Tj
/TT1 1 Tf
(number)Tj
/TT0 1 Tf
( of surveys, when again it would be )Tj
0 -1.201 TD
(too complicated to give all the key information for each of the surveys \
referred to. )Tj
0 -1.1 TD
( )Tj
T*
(Also, where a given survey is reported on `serially' \(for example in th\
e course of several )Tj
0 -1.2 TD
(consecutive issues of a newspaper\) it will frequently be unnecessary to\
repeat all the technical )Tj
T*
(details in every issue. )Tj
0 -1.1 TD
( )Tj
0 -1.1 TD
(These situations are the exceptions. Most published reports on public op\
inion polls refer to )Tj
0 -1.2 TD
(much more limited studies than these. Even in the more complex cases it \
should frequently be )Tj
T*
(possible to give much of the key information asked for in Article B3 oth\
er than the detailed )Tj
T*
(question wordings. In all cases where the key information cannot be full\
y provided the basic )Tj
T*
(principle of fair and informative reporting must be followed, and it sho\
uld be made clear how )Tj
T*
(and where the serious enquirer can obtain fuller details. )Tj
0 -1.1 TD
( )Tj
0 -1.1 TD
(On specific points in this Article: )Tj
T*
( )Tj
T*
(\(3c\) \223achieved\224 sample size is the number of interviews actually\
reported on. \223Geographical )Tj
0 -1.2 TD
(coverage\224 should state which broad regions of the country \(national \
or other\) were )Tj
T*
(represented. In addition the number of sampling locations used should be\
given as an )Tj
T*
(indication of the adequacy of the sample design. In referring to the num\
ber of \223locations\224 the )Tj
T*
(objective is to provide a realistic picture of the extent to which the s\
ample is widely distributed )Tj
T*
(geographically. The best terms to use would vary by country - for exampl\
e \223D\351partements\224 )Tj
T*
(might be best in France, \223Parliamentary Constituencies\224 in the U.K\
. )Tj
0 -1.1 TD
( )Tj
T*
(\(3e\) it is important for the reader to be given some general indicatio\
n of the sampling )Tj
0 -1.2 TD
(approach used since this may in certain cases have very strong implicati\
ons for the likely )Tj
T*
(representativeness of the sample. The information it is possible to incl\
ude in a published report )Tj
T*
(cannot hope to give all the relevant data for a technical assessment; bu\
t even a limited )Tj
0 -1.2 TD
(reference can be helpful. In the case of random sampling, the main objec\
tive is to identify )Tj
T*
(those studies where an )Tj
/TT1 1 Tf
(unusually)Tj
/TT0 1 Tf
( low success rate has been achieved, for whatever reasons. )Tj
0 -1.201 TD
(Although it is preferable wherever possible to quote the actual success \
rate, the main )Tj
0 -1.2 TD
(requirement is therefore to indicate if the success rate is below that r\
egarded in the profession )Tj
T*
(as \223normal\224 for the type of study \(this is a matter for experienc\
ed professional judgement\). )Tj
0 -1.1 TD
( )Tj
T*
(\(3g\) the guiding principle is the need to avoid possible ambiguity and\
misunderstanding. This )Tj
0 -1.2 TD
(is particularly important where the actual wording of the question is cr\
itical to the )Tj
T*
(interpretation of the findings, and where the answers reported on can be\
affected by the )Tj
T*
(precise form of the question - especially on issues which are politicall\
y or socially )Tj
T*
(\221sensitive\224 \(for example, attitudes towards abortion\). The reade\
r should therefore be helped to )Tj
T*
(understand exactly what was asked. In some cases this will be sufficient\
ly clear from the text )Tj
T*
(itself and the actual answers reported; but in any case of possible doub\
t it is much more )Tj
T*
(preferable to include the question wording used. Certainly where tabular\
data is given it is )Tj
T*
(good practice to include the full question wording. Experience shows tha\
t it is in practice often )Tj
T*
(quite possible to include the questions without overloading the publishe\
d report. )Tj
0 -1.1 TD
( )Tj
T*
(As a general principle it is also good practice to indicate if the resul\
ts quoted have been )Tj
0 -1.2 TD
(adjusted to take account of weighting procedures or other statistical ca\
lculations, wherever )Tj
T*
(these are likely to mean that the findings reported differ substantially\
from the raw data )Tj
T*
(collected in the field. \(This recommendation is especially relevant in \
the case of )Tj
/TT1 1 Tf
(non-standard)Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(weightings - i.e. other than conventional sample design weighting proced\
ures such as normal )Tj
0 -1.2 TD
(weighting by area and similar generally-accepted procedures.\) )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(14 of 22\)5/7/2005 8:58:55 AM\
)Tj
ET
EMC
endstream
endobj
145 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 756.058 Tm
( )Tj
/TT1 1 Tf
0 -1.099 TD
(Article B4)Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(In the case of broadcast media, where the scope for providing basic info\
rmation about a )Tj
0 -1.2 TD
(survey is clearly more restricted, some research organisations currently\
arrange with their )Tj
T*
(client to provide a press release at the same time as the broadcast repo\
rt on a survey. Such a )Tj
T*
(press release can more easily include some kind of fact sheet covering a\
ll the basic pieces of )Tj
T*
(information referred to in Article B3. The publication of this type of i\
nformation on the Internet )Tj
T*
(is also a way of making fuller detail available. These types of practice\
s are strongly )Tj
0 -1.2 TD
(recommended. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(Article B5There are many occasions on which the interpretation of partic\
ular findings )Tj
0 -1.2 TD
(will be quite different if the level of \223don\222t know\224 answers is\
5% or 50%. In the case )Tj
T*
(of voting-intention studies the same consideration also applies to \223w\
ill not vote\224 )Tj
T*
(answers. A research organisation must apply its experience and professio\
nal )Tj
0 -1.2 TD
(judgement in deciding when such situations arise. It may not be necessar\
y to include )Tj
0 -1.2 TD
(all the \223don\222t know\224 percentages in any tables given, although \
where this is possible )Tj
T*
(it is frequently the best way of dealing with the issue. It may be quite\
sufficient, for )Tj
T*
(example, to make a general comment such as: \223the proportion of \221do\
n\222t knows\222 was )Tj
T*
(never higher than 5%\224 - or to comment specifically on those instances\
where the )Tj
T*
(proportion was much higher. \(In the case of voting-intention studies, i\
t is not )Tj
T*
(necessary to quote \223will not vote\224 percentages separately from \223\
don\222t know\224 )Tj
T*
(answers, if by quoting them separately a false idea of likely voting tur\
nout may be )Tj
T*
(given.\) )Tj
0 -1.1 TD
( )Tj
T*
(Article B6 )Tj
0 -1.2 TD
(Where a voting-intention study is one of a series carried out by a resea\
rch )Tj
T*
(organisation, and follows the normal standard calculation practices for \
that )Tj
T*
(organisation, it may not be necessary to refer to this point in every re\
port. However, )Tj
0 -1.2 TD
(confusion has been caused on occasions because of the differing practice\
s of )Tj
T*
(different organisations and it is desirable to avoid this problem arisin\
g. It must in )Tj
T*
(any case be made easy for enquirers to check what is the practice involv\
ed. )Tj
0 -1.1 TD
( )Tj
T*
(Article B7 )Tj
0 -1.2 TD
(Research organisations must be prepared to supply the basic information \
about the )Tj
T*
(survey methods used according to the Notes on the application of Rule 25\
of the )Tj
T*
(main International Code. There is no obligation under either Code for fu\
rther )Tj
T*
(information beyond this to be supplied - although organisations will nor\
mally be )Tj
T*
(prepared to discuss their research methods in more detail with bona fide\
enquirers. )Tj
0 -1.1 TD
( )Tj
T*
(Article C9 )Tj
0 -1.2 TD
(In preparing material for publication, journalists and others connected \
with the )Tj
T*
(media themselves normally follow professional codes of practice and ethi\
cs )Tj
T*
(concerned to uphold the public interests. The present Code is not intend\
ed in any )Tj
T*
(way to substitute for these but rather to support them. \(In this contex\
t, \223published )Tj
T*
(report\224 covers non-print as well as print media.\) )Tj
0 -1.1 TD
( )Tj
T*
(The research institute should reserve the right to publish the total stu\
dy and not only )Tj
0 -1.258 TD
(the technical specifications in the event of:)Tj
/T1_1 1 Tf
14 0 0 14 305.34 91.7756 Tm
( )Tj
ET
q
8 0 0 7 15 61.1999969 cm
/Im0 Do
Q
BT
/TT0 1 Tf
12 0 0 12 23 58.258 Tm
( )Tj
0 -0.1 TD
(A shortened version of the publication prevaricating the analysis of the\
results )Tj
ET
q
8 0 0 7 15 46.8000031 cm
/Im0 Do
Q
BT
/TT0 1 Tf
12 0 0 12 23 43.858 Tm
( )Tj
T*
(An unforeseen and abridged version of the publication )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(15 of 22\)5/7/2005 8:58:55 AM\
)Tj
ET
EMC
endstream
endobj
146 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
q
8 0 0 7 15 759 cm
/Im0 Do
Q
BT
/TT0 1 Tf
12 0 0 12 23 756.058 Tm
( )Tj
0 -0.1 TD
(A publication which does not conform to the prior agreements)Tj
-0.667 -2.033 Td
(The fourth sentence of this Article emphasises the importance of disting\
uishing as far as )Tj
0 -1.2 TD
(possible between the figures which emerge directly from the questions as\
ked, and any )Tj
T*
(commentary/interpretation based on these. Although the dividing line is \
not always a simple )Tj
T*
(one to define, in most cases the distinction between \223fact\224 and \223\
comment\224 is in practice a )Tj
T*
(workable one. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(5.3 Contractual arrangements)Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(ESOMAR encourages the use of contracts between research organisations an\
d their clients to )Tj
0 -1.2 TD
(cover certain of the points dealt with in this Code. For example, some c\
ontracts stipulate that )Tj
T*
(the agency has the right to examine and approve copy based on its resear\
ch. Where the )Tj
T*
(agency reserves the copyright of the findings this can also help to redu\
ce some of the )Tj
T*
(problems involved in unscrupulous \223secondary reporting\224 of the fin\
dings by other people. In )Tj
T*
(addition to any other requirements it is suggested that such a contract \
could usefully cover: )Tj
0 -1.1 TD
( )Tj
T*
(1. clarification of the point that the contract binds )Tj
/TT1 1 Tf
(both)Tj
/TT0 1 Tf
( the fund-supplier and the media )Tj
0 -1.201 TD
(involved, where these are different parties )Tj
0 -1.2 TD
(2. some measure of control by the research organisation over the publish\
ed form of the results )Tj
T*
(including figures and graphs )Tj
0 -1.1 TD
( )Tj
T*
(Certain contracts also provide that if research findings commissioned fo\
r publication are not in )Tj
0 -1.2 TD
(fact published, such findings can subsequently \(after a specified perio\
d of time\) be released by )Tj
T*
(the research organisation itself; or alternatively the organisation is f\
ree to repeat the survey )Tj
T*
(for another client. It is also increasingly common practice in certain c\
ountries for data tapes )Tj
T*
(from public opinion surveys to be lodged with appropriate archives for s\
ubsequent secondary )Tj
T*
(research by academic researchers and others. Such steps can help to redu\
ce the danger that )Tj
T*
(polls may be thought sometimes to be used in a \221manipulative\222 way \
by less scrupulous clients. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(5.4 Summary)Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(Any code of practice in this area must have obvious limitations, in that\
researchers can )Tj
0 -1.2 TD
(exercise only restricted control over how their results are presented in\
the media, and still less )Tj
T*
(influence over any comments and interpretations \(sometimes misguided an\
d tendentious\) )Tj
T*
(based on the findings. A code must therefore depend on trying to spread \
the use of \221best )Tj
T*
(practice\222 and to influence media clients to avoid misleading presenta\
tion of survey results. )Tj
0 -1.1 TD
( )Tj
0 -1.157 TD
(ESOMAR expects its members to follow the Code with this objective firmly\
in mind.)Tj
/T1_1 1 Tf
14 0 0 14 517.512 252.1673 Tm
( )Tj
0.157 0.38 0.671 rg
/TT1 1 Tf
-35.894 -2.612 Td
(Guidelines on Practical Aspects of Conducting Pre-Election Opinion Polls\
)Tj
0 0 0 rg
12 0 0 12 15 190.658 Tm
(6.1 Introduction)Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(The following guidelines concentrate on the conduct of the pre-election \
polls. At first it may )Tj
0 -1.2 TD
(seem strange to concentrate the guidelines on pre-election polls since t\
hey are just one )Tj
0 -1.2 TD
(particular type of political poll. However, while it is true that all op\
inion polls require high )Tj
T*
(technical standards, it is pre-election polls that feature most frequent\
ly in the debate about )Tj
T*
(polls, and which are restricted in some European countries. These guidel\
ines have two main )Tj
T*
(objectives - to protect the interests of the voter in a democracy and to\
protect the credibility of )Tj
0 -1.2 TD
(market and opinion research. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(1. Protecting the interests of the voter in a democracy)Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(The first objective of these guidelines is to ensure that polling organi\
sations take all possible )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(16 of 22\)5/7/2005 8:58:55 AM\
)Tj
ET
EMC
endstream
endobj
147 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 754.858 Tm
(technical steps to ensure that polls published close to the vital decisi\
on point for voters are an )Tj
0 -1.2 TD
(objective guide to the state of public opinion and voting intentions. Th\
e process of sampling )Tj
T*
(cannot guarantee highly precise measurement by every single poll. Also, \
the measurement of )Tj
0 -1.2 TD
(stated intentions to vote cannot guarantee that all electors will actual\
ly vote in line with their )Tj
0 -1.2 TD
(earlier stated voting intentions. People do change their mind, some even\
in the second before )Tj
T*
(marking their vote on the ballot slip. Polling organisations have a resp\
onsibility to electors to )Tj
T*
(ensure that polls, especially those polls published in the last few days\
of an election campaign, )Tj
T*
(provide reliable and objective information. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(2. Protecting the reputation of market research)Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(The second objective of these guidelines is to protect the public reputa\
tion of market research )Tj
0 -1.2 TD
(using sample surveys. Pre-election opinion polls which are published in \
the final days of an )Tj
T*
(election campaign have a major influence on this. They are inevitably se\
en as predictions of )Tj
T*
(the election result. While it is true that opinion polls are a snapshot \
of intentions at a specific )Tj
T*
(point of time, the publication of this snapshot in the very late stages \
of a campaign is almost )Tj
T*
(universally treated by the media as a prediction. In general, pollsters \
have not effectively )Tj
T*
(challenged this use of polling data, partly because the track record of \
the polls in \223predicting\224 )Tj
T*
(the result is good. )Tj
0 -1.1 TD
( )Tj
T*
(In some countries where the publication of polls in the final stages of \
a campaign is restricted, )Tj
0 -1.2 TD
(polls based on national samples are often conducted on polling day, or t\
he day before, for )Tj
T*
(publication within minutes of the close of the polling stations. Also ca\
rrying out exit polls )Tj
T*
(\(interviewing voters as they leave the polling station\) has become muc\
h more common. Such )Tj
T*
(polls are even more likely to be seen as prediction polls. Their accurac\
y is equally important to )Tj
T*
(the public image of market research, though they play no part in informi\
ng the voter in the )Tj
T*
(democratic process. )Tj
0 -1.1 TD
( )Tj
T*
(Pre-election polls are a very public test of sampling theory and survey \
research in action. Polls )Tj
0 -1.2 TD
(have a good track record for accuracy but the occasional poll which appe\
ars to be wrong gets )Tj
T*
(extensive media coverage. \223Polls wrong\224 is news and gets major cov\
erage. \223Polls accurate\224 is a )Tj
T*
(headline which will never be written. ESOMAR hopes that these guidelines\
will contribute to the )Tj
T*
(technical education of journalists responsible for poll reporting. Howev\
er, special care must still )Tj
T*
(be taken by polling organisations to minimise the risk of \223getting it\
wrong\224. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(6.2 The Guidelines)Tj
/TT0 1 Tf
( )Tj
0 -1.101 TD
( )Tj
0 -1.1 TD
(In the following sections a number of critical technical issues about th\
e conduct of pre-election )Tj
0 -1.2 TD
(opinion polls are considered and a guide to good practice is given. The \
order of the issues is )Tj
T*
(not intended to indicate relative importance or priority. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(1. General design)Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(A major issue in the design of a pre-election opinion poll is reconcilin\
g the rather contradictory )Tj
0 -1.2 TD
(needs for large sample sizes and also for fieldwork dates as close as po\
ssible to polling day. )Tj
T*
(Larger samples produce more reliable measurements but also take longer t\
o carry out. This in )Tj
T*
(turn makes them less up to date than a later survey with a smaller sampl\
e. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
T*
(Guideline)Tj
/TT2 1 Tf
( Historically, in elections with volatile electorates, the closeness of \
fieldwork to )Tj
0 -1.201 TD
(polling day has been more important than sample size or sampling purity.\
If very large )Tj
0 -1.2 TD
(samples are collected, the events of the campaign may affect voters afte\
r the completion of )Tj
T*
(much of the fieldwork but before the analysis and publication of the pol\
l findings.)Tj
/TT0 1 Tf
( )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(2. Timing of fieldwork)Tj
/TT0 1 Tf
( )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(17 of 22\)5/7/2005 8:58:55 AM\
)Tj
ET
EMC
endstream
endobj
148 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 754.858 Tm
(Anyone attempting to prepare a critical case about polls would certainly\
decide that the date of )Tj
0 -1.2 TD
(publication of a poll is a key date. Regardless of when the interviewing\
took place, the )Tj
T*
(publication date is the important fact in judging the contribution of th\
e poll to the electoral )Tj
0 -1.2 TD
(process. )Tj
0 -1.1 TD
( )Tj
T*
(Polling organisations must be responsible for ensuring that polls publis\
hed in the very late )Tj
0 -1.2 TD
(stages of an election are likely to be a fair representation of public o\
pinion as close as possible )Tj
T*
(to the end of the campaign. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
T*
(Guideline)Tj
/TT0 1 Tf
( )Tj
/TT2 1 Tf
(Polling companies should try to reduce the risk of \223getting it wrong\224\
by minimising )Tj
0 -1.201 TD
(the time elapsed between fieldwork and publication. A poll is more likel\
y to achieve a good )Tj
0 -1.2 TD
(representative sample if the fieldwork period includes some time in the \
evening when electors )Tj
T*
(in full-time employment are available for interview.)Tj
/TT0 1 Tf
( )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(3. Sample size)Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(The measurement of the share of the vote for a party is subject to the n\
ormal statistical )Tj
0 -1.2 TD
(confidence limits for sample surveys. Two factors affect the size of the\
confidence limit of any )Tj
T*
(party share. The first is the absolute level of support for a party. The\
closer this gets to 50%, )Tj
T*
(the wider will be the confidence limit around the share estimate. The se\
cond is the size of the )Tj
T*
(sample interviewed in order to produce the estimate. In most pre-electio\
n polls the size of the )Tj
T*
(sample is the more important factor. )Tj
0 -1.1 TD
( )Tj
T*
(In countries with a simple proportional representation system, the pre-e\
lection poll )Tj
0 -1.2 TD
(measurement of the share of the vote to each party is a fair indicator o\
f the election outcome. )Tj
T*
(In other electoral systems this may not be the case. In the UK, for exam\
ple, the winner in )Tj
T*
(each constituency is the candidate with the most votes. However, the bes\
t that the polls can )Tj
T*
(do is estimate the share of the vote to each party at a national level. \
The key statistic reported )Tj
T*
(in the media is the gap in share of vote between the leading parties, an\
d the measurement of )Tj
0 -1.2 TD
(the gap has much larger confidence limits than those for an individual p\
arty\222s share. )Tj
0 -1.1 TD
( )Tj
T*
(Polling organisations frequently represent the margin of error of their \
polls as \261 3%. This may )Tj
0 -1.2 TD
(be accurate for a single party but is rarely accurate for the key media \
figure - the gap between )Tj
T*
(leading parties. A poll that produces a 95% confidence limit of \261 3% \
on the share for one of )Tj
T*
(the leading parties, could produce an equivalent confidence limit of \261\
5.7% on the gap )Tj
T*
(between the two leading parties. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
T*
(Guideline)Tj
/TT2 1 Tf
( Pre-election polls should not have a sample of less than 1,000 responde\
nts. In )Tj
0 -1.201 TD
(circumstances where the gap between leading parties is expected to be sm\
all, the sample size )Tj
0 -1.2 TD
(should be larger and samples of 1,500 to 2,000 should be used.)Tj
/TT0 1 Tf
( )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(4. Sample distribution)Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(In European countries there are two main approaches to selecting samples\
for face to face )Tj
0 -1.2 TD
(interviews. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(Method 1)Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(Select sampling points in proportion to the number of electors. Then, in\
each selected area )Tj
0 -1.2 TD
(take an equal number of interviews. The rule for samples using this meth\
odology should be to )Tj
/TT1 1 Tf
T*
(maximise)Tj
/TT0 1 Tf
( the number of sampling points and minimise the number of interviews con\
ducted at )Tj
0 -1.201 TD
(any one sampling point. This implies )Tj
/TT1 1 Tf
(minimising)Tj
/TT0 1 Tf
( the number of interviews conducted by any )Tj
T*
(single interviewer. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
T*
(Guideline )Tj
/TT2 1 Tf
(There is empirical evidence that if the number of interviews per samplin\
g point )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(18 of 22\)5/7/2005 8:58:55 AM\
)Tj
ET
EMC
endstream
endobj
149 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 754.858 Tm
(exceeds 20, then the influence of increased interviewer variation outwei\
ghs increased sample )Tj
0 -1.2 TD
(size.)Tj
/TT1 1 Tf
( )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
0 -1.099 TD
(Method 2)Tj
/TT1 1 Tf
( )Tj
0 -1.201 TD
(Select sampling points with equal probability but then to take a differe\
nt number of interviews )Tj
0 -1.2 TD
(in each area determined by the size of the electorate in the area. )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
0 -1.1 TD
(Guideline)Tj
/TT0 1 Tf
( Polls using this method should still aim to minimise the number of inte\
rviews )Tj
0 -1.201 TD
(conducted by any single interviewer and stick to the maximum of 20.)Tj
/TT1 1 Tf
( )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
T*
(Guideline)Tj
/TT0 1 Tf
( For both methods, the sample design should give priority to maximising \
the )Tj
0 -1.201 TD
(number of sampling points and minimising the number of interviews conduc\
ted by each )Tj
0 -1.2 TD
(interviewer.)Tj
/TT1 1 Tf
( )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
0 -1.099 TD
(5. Telephone interviewing)Tj
/TT1 1 Tf
( )Tj
0 -1.201 TD
(There has been technical debate about the use of telephone interviewing \
for opinion polls. In )Tj
0 -1.2 TD
(principle telephone surveys offer high quality, unclustered, random samp\
les, with fast )Tj
T*
(completion of fieldwork. However, in most EU countries, telephone owners\
hip is not 100%. )Tj
T*
(Telephone ownership or availability is frequently correlated with voting\
intention, i.e. those )Tj
T*
(who cannot be contacted by phone are more likely to support one rather t\
han another of the )Tj
T*
(parties in the election. This may also be the case for those telephone o\
wners who are not )Tj
T*
(listed in the telephone directory. )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
T*
(Guideline)Tj
/TT0 1 Tf
( If telephone ownership is not extremely high \(85%+\) and is by definit\
ion likely to )Tj
0 -1.201 TD
(produce an unrepresentative sample, this method of interview should not \
be used for pre-)Tj
0 -1.2 TD
(election polls. )Tj
/TT1 1 Tf
( )Tj
0 -1.1 TD
( )Tj
T*
(In countries with high phone ownership, it may be possible to establish \
weighting systems )Tj
0 -1.2 TD
(which largely compensate for the unrepresentativeness of telephone sampl\
es. )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
T*
(Guideline)Tj
/TT0 1 Tf
( If telephone samples are used for polls, the polling organisation must \
correct for )Tj
0 -1.201 TD
(any under-representation of supporters of particular political parties. \
Simple demographic )Tj
0 -1.2 TD
(profile adjustments will generally not be adequate. )Tj
/TT1 1 Tf
( )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
0 -1.099 TD
(6. Weighting)Tj
/TT1 1 Tf
( )Tj
0 -1.201 TD
(In order to conduct fast surveys with large samples, most pre-election o\
pinion polls based on )Tj
0 -1.2 TD
(face to face interviews will use quota sampling methods. The application\
of simple )Tj
T*
(demographic weights to ensure an accurate sample balance is normal good \
practice. If some )Tj
T*
(parts of the electorate have been deliberately over-sampled, weighting s\
hould be used to re-)Tj
T*
(establish the correct balance \(see section 6.2.11 for additional inform\
ation\). )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
T*
(Guideline)Tj
/TT0 1 Tf
( The demographic profile of pre-election polls should be checked for )Tj
0 -1.201 TD
(representativeness and, if necessary, weighting should be applied to rep\
resent correctly the )Tj
0 -1.2 TD
(electorate. Polling companies should ensure that the population profile \
used is that of electors )Tj
T*
(eligible to vote rather than the more normal all adults profile used in \
commercial market )Tj
T*
(research.)Tj
/TT1 1 Tf
( )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
0 -1.099 TD
(7. Adjustments)Tj
/TT1 1 Tf
( )Tj
0 -1.201 TD
(In some circumstances polling organisations may feel that the result of \
their pre-election poll )Tj
0 -1.2 TD
(is not an accurate guide to the likely outcome. The most obvious example\
is where the voting )Tj
T*
(intention of those with a high likelihood of casting their vote is diffe\
rent from the total sample. )Tj
0 -1.1 TD
( )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(19 of 22\)5/7/2005 8:58:55 AM\
)Tj
ET
EMC
endstream
endobj
150 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 754.858 Tm
(In some countries it is common to measure claimed voting at the previous\
election and use )Tj
0 -1.2 TD
(this to adjust the voting estimate from the current survey. In Denmark a\
nd France this is an )Tj
T*
(important quality improving procedure. However, this approach has also b\
een shown not to )Tj
0 -1.2 TD
(work well in a number of other countries. )Tj
0 -1.1 TD
( )Tj
T*
(Experience has shown that voting intention measurements in some countrie\
s need adjustment )Tj
0 -1.2 TD
(in order to provide a reliable guide to public opinion. In these countri\
es the weighting or )Tj
T*
(adjustment strategy of the polling organisation may be confidential to t\
hat company for )Tj
0 -1.2 TD
(competitive reasons. In such circumstances, where the adjustment is made\
via a stable and )Tj
T*
(systematic procedure and not subject to a "gut feeling", the polling com\
pany may wish to )Tj
T*
(withhold full details of its method. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
T*
(Guideline)Tj
/TT2 1 Tf
( Polling organisations should not make any adjustments to the simple pol\
l result )Tj
0 -1.201 TD
(which cannot be documented and defined in advance of seeing the results.\
Any adjustments )Tj
0 -1.2 TD
(must be capable of being repeated and justified. Adjustments made on "gu\
t feeling" are )Tj
T*
(unacceptable. )Tj
/TT0 1 Tf
( )Tj
0 -1.1 TD
( )Tj
T*
(It is good practice for pre-election polls to measure key variables such\
as likelihood to vote, )Tj
0 -1.2 TD
(and to consider whether the simple poll result should be adjusted. )Tj
0 -1.1 TD
( )Tj
T*
(Where adjustments to the "simple" poll findings are made, this should be\
noted in the )Tj
0 -1.2 TD
(publication of the poll findings. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(8. Survey content)Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(Pre-election opinion polls will have greater political and social value \
if they do not confine )Tj
0 -1.2 TD
(themselves only to measuring voting intention but also explore the reaso\
ns for party choice )Tj
T*
(and opinions on important campaign issues. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
T*
(Guideline)Tj
/TT2 1 Tf
( Wherever possible, pre-election polls should measure reasons for party \
choice or )Tj
0 -1.201 TD
(attitudes on issues or other aspects of the campaign.)Tj
/TT0 1 Tf
( )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(9. Time series)Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(Polling organisations use different methodological designs. The meaning \
of a final pre-election )Tj
0 -1.2 TD
(poll is far easier to determine if it is the latest of a series of polls\
conducted by the same )Tj
T*
(organisation during the campaign. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
T*
(Guideline)Tj
/TT2 1 Tf
( The validity of the methods used by an organisation can be judged bette\
r if they )Tj
0 -1.201 TD
(produce a series of voting intention estimates during the campaign. Any \
obvious biases will )Tj
0 -1.2 TD
(become apparent by comparison with the published polls of other organisa\
tions.)Tj
/TT0 1 Tf
( )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
0 -1.099 TD
(10. Consistent design)Tj
/TT0 1 Tf
( )Tj
0 -1.201 TD
(The ability to judge a final pre-election poll by comparison with previo\
us polls from the same )Tj
0 -1.2 TD
(organisation is weakened if the organisation changes key aspects of its \
methodology for the )Tj
T*
(final poll. It could be argued that there is a benefit if polling organi\
sations improve the )Tj
T*
(methodological quality of the design they use for final polls. This is a\
weak position to take. )Tj
T*
(First, it reduces the comparability of a series of polls from the same p\
olling organisation. )Tj
T*
(Second, it suggests that some campaign polls can be of poorer quality. H\
owever, if polls are to )Tj
T*
(make a valid contribution to informing the political process, they must \
all be of high quality. A )Tj
T*
(two tier quality policy undermines the value of polls. )Tj
0 -1.1 TD
( )Tj
/TT1 1 Tf
T*
(Guideline)Tj
/TT2 1 Tf
( Polling organisations should attempt to keep key elements of methodolog\
y )Tj
0 -1.201 TD
(consistent throughout the election campaign. This particularly applies t\
o sampling method, )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(20 of 22\)5/7/2005 8:58:55 AM\
)Tj
ET
EMC
endstream
endobj
151 0 obj<>stream
/Artifact <>BDC
0 0 0 rg
0 i
BT
/T1_0 1 Tf
0 Tc 0 Tw 0 Ts 100 Tz 0 Tr 9 0 0 9 18 780.17 Tm
(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)Tj
ET
EMC
/Article <>BDC
q
0 18 612 756 re
W* n
BT
/TT0 1 Tf
12 0 0 12 15 754.858 Tm
(question wording and the positioning of voting intention questions. It d\
oes not apply to sample )Tj
0 -1.2 TD
(size.)Tj
/TT1 1 Tf
( )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
0 -1.099 TD
(11. Indicator variables)Tj
/TT1 1 Tf
( )Tj
0 -1.201 TD
(It is possible to interview a sample representative as far as age, sex a\
nd social grade are )Tj
0 -1.2 TD
(concerned, but still have a politically unrepresentative sample. For exa\
mple, section 6.2.2 )Tj
T*
(above referred to the issue of representing working electors by ensuring\
that the fieldwork )Tj
0 -1.2 TD
(period includes some evening interviewing. )Tj
0 -1.1 TD
( )Tj
T*
(It would be good practice in a pre-election poll to include the collecti\
on of information which is )Tj
0 -1.2 TD
(correlated with voting behaviour but which is not part of the quota cont\
rol. The item should )Tj
T*
(have a known penetration from alternative sources. )Tj
0 -1.1 TD
( )Tj
T*
(For example, in the UK membership of a trade union can serve this purpos\
e. It is correlated )Tj
0 -1.2 TD
(with party support and should not change its penetration substantially f\
rom survey to survey. )Tj
T*
(If a demographically representative poll produces an estimate of trade u\
nion membership )Tj
T*
(which is too high, the poll will probably overstate the vote for Labour.\
In other countries )Tj
T*
(variables such as religion or language spoken may be possible indicator \
variables if they have )Tj
T*
(not been used in the design of the survey quota controls. )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
T*
(Guideline)Tj
/TT0 1 Tf
( Polling companies should be encouraged to develop an indicator variable\
for sample )Tj
0 -1.201 TD
(weighting purposes. In the period between elections it is possible to ca\
librate the electoral )Tj
0 -1.2 TD
(implication of an over or under representation of the indicator variable\
.)Tj
/TT1 1 Tf
( )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
0 -1.099 TD
(12. \223Rogue\224 Polls)Tj
/TT1 1 Tf
( )Tj
0 -1.201 TD
(Probability theory suggests that one poll in twenty may give results out\
side the normal 95% )Tj
0 -1.2 TD
(confidence limits. Polling organisations generally do better than probab\
ility theory suggests )Tj
T*
(because they design their surveys to reduce the risk of error. Neverthel\
ess, a polling )Tj
0 -1.2 TD
(organisation may find itself with a pre-election poll that is out of lin\
e with all the campaign )Tj
T*
(evidence available up to that point. Usually there will be little time a\
vailable between getting )Tj
T*
(the final result and writing the copy for publication. )Tj
0 -1.1 TD
( )Tj
T*
(It may be possible to recontact some respondents from earlier surveys by\
telephone to check )Tj
0 -1.2 TD
(whether they have changed their mind in a way that would support the unu\
sual finding in the )Tj
T*
(final poll \(this however, would be subject to national data protection \
requirements\). )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
T*
(Guideline)Tj
/TT0 1 Tf
( It is unacceptable to suppress a pre-election poll that looks out of li\
ne with earlier )Tj
0 -1.201 TD
(polls unless a valid technical reason why the poll is wrong has been est\
ablished. It is also )Tj
0 -1.2 TD
(unacceptable to make a \223gut feeling\224 adjustment. The poll should b\
e published with )Tj
T*
(appropriate warning about the unusual poll result. The odds are 20:1 in \
favour of the poll )Tj
T*
(being correct and voters do change their minds even in the polling booth\
.)Tj
/TT1 1 Tf
( )Tj
0 -1.1 TD
( )Tj
/TT2 1 Tf
0 -1.099 TD
(13. Exit polls)Tj
/TT1 1 Tf
( )Tj
0 -1.201 TD
(An increasingly popular component of the television coverage of election\
nights is the exit poll. )Tj
0 -1.2 TD
(These are not based on nationally representative samples of voters asked\
their voting )Tj
T*
(intention. They are based on selections of electors leaving samples of v\
oting places. The )Tj
T*
(technical requirements of this sort of poll are very different from pre-\
election opinion polls and )Tj
T*
(are outside the scope of these guidelines. )Tj
0 -1.1 TD
( )Tj
T*
(The methodology for exit polls varies from country to country and is sti\
ll being developed. )Tj
0 -1.2 TD
(Elections are quite rare events and it is impossible to test exit poll m\
ethods except at )Tj
T*
(elections. Some years will be needed before it is possible to provide gu\
idelines for the conduct )Tj
ET
EMC
/Artifact <>BDC
Q
BT
/T1_0 1 Tf
9 0 0 9 18 7.17 Tm
(http://www.esomar.org/web/show/id=103897 \(21 of 22\)5/7/2005 8:58:55 AM\
)Tj
ET
EMC
endstream
endobj
152 0 obj(Esomar - ESOMAR/WAPOR Guide to Opinion Polls)
endobj
153 0 obj<>
endobj
154 0 obj<>
endobj
155 0 obj<>
endobj
156 0 obj<>
endobj
157 0 obj[154 0 R]
endobj
158 0 obj(http://www.esomar.org/web/show/id=103897)
endobj
159 0 obj(C-RNҝ?,)
endobj
160 0 obj<>
endobj
161 0 obj<>
endobj
162 0 obj(L2B^Ӱ0Ѻ)
endobj
165 0 obj 3
endobj
166 0 obj<>
endobj
167 0 obj<>
endobj
168 0 obj<>stream
2005-05-07T08:58:55+07:00
2005-05-07T08:58:02+07:00
2005-05-07T08:58:55+07:00
application/pdf
Esomar - ESOMAR/WAPOR Guide to Opinion Polls
uuid:5ad427f7-597c-4e73-b8b1-420c4c12256c
uuid:426bc9aa-9c53-4da9-9632-e0be6b3d37b8
Acrobat Web Capture 7.0
endstream
endobj
xref
0 169
0000000004 00000 f
0000000016 00000 n
0000000146 00000 n
0000000220 00000 n
0000000006 00000 f
0000000396 00000 n
0000000070 00001 f
0000000464 00000 n
0000000507 00000 n
0000000560 00000 n
0000000752 00000 n
0000000840 00000 n
0000000875 00000 n
0000000962 00000 n
0000001013 00000 n
0000001069 00000 n
0000001126 00000 n
0000001183 00000 n
0000001391 00000 n
0000001987 00000 n
0000002028 00000 n
0000002132 00000 n
0000002177 00000 n
0000002226 00000 n
0000002927 00000 n
0000003167 00000 n
0000003191 00000 n
0000003234 00000 n
0000003277 00000 n
0000003503 00000 n
0000003569 00000 n
0000003795 00000 n
0000004021 00000 n
0000004247 00000 n
0000004473 00000 n
0000004699 00000 n
0000004923 00000 n
0000004966 00000 n
0000005009 00000 n
0000005192 00000 n
0000005216 00000 n
0000005257 00000 n
0000005440 00000 n
0000005464 00000 n
0000005647 00000 n
0000005671 00000 n
0000005712 00000 n
0000005906 00000 n
0000005930 00000 n
0000005971 00000 n
0000006154 00000 n
0000006178 00000 n
0000006372 00000 n
0000006396 00000 n
0000006590 00000 n
0000006614 00000 n
0000006819 00000 n
0000006843 00000 n
0000006884 00000 n
0000007139 00000 n
0000007163 00000 n
0000007204 00000 n
0000007245 00000 n
0000007440 00000 n
0000007533 00000 n
0000007626 00000 n
0000008844 00000 n
0000009073 00000 n
0000009163 00000 n
0000010373 00000 n
0000000073 00001 f
0000010596 00000 n
0000010620 00000 n
0000000077 00001 f
0000010826 00000 n
0000010850 00000 n
0000010891 00000 n
0000000080 00001 f
0000011098 00000 n
0000011122 00000 n
0000000083 00001 f
0000011317 00000 n
0000011341 00000 n
0000000088 00001 f
0000011610 00000 n
0000011634 00000 n
0000011674 00000 n
0000011944 00000 n
0000000093 00001 f
0000012037 00000 n
0000012061 00000 n
0000012102 00000 n
0000012145 00000 n
0000000096 00001 f
0000012353 00000 n
0000012377 00000 n
0000000101 00001 f
0000012585 00000 n
0000012609 00000 n
0000012817 00000 n
0000014042 00000 n
0000000104 00001 f
0000014281 00000 n
0000014306 00000 n
0000000108 00001 f
0000014515 00000 n
0000014540 00000 n
0000014749 00000 n
0000000111 00001 f
0000014860 00000 n
0000014885 00000 n
0000000120 00001 f
0000015109 00000 n
0000015134 00000 n
0000015177 00000 n
0000015326 00000 n
0000015368 00000 n
0000015516 00000 n
0000015724 00000 n
0000016943 00000 n
0000000121 00001 f
0000000123 00001 f
0000017177 00000 n
0000000124 00001 f
0000000125 00001 f
0000000126 00001 f
0000000127 00001 f
0000000128 00001 f
0000000129 00001 f
0000000163 00001 f
0000017267 00000 n
0000021557 00000 n
0000026275 00000 n
0000031368 00000 n
0000036433 00000 n
0000041386 00000 n
0000046369 00000 n
0000050974 00000 n
0000055638 00000 n
0000059688 00000 n
0000064841 00000 n
0000069058 00000 n
0000073713 00000 n
0000076287 00000 n
0000080926 00000 n
0000086109 00000 n
0000090727 00000 n
0000095569 00000 n
0000100605 00000 n
0000105415 00000 n
0000109984 00000 n
0000114655 00000 n
0000119532 00000 n
0000119595 00000 n
0000119633 00000 n
0000119710 00000 n
0000119741 00000 n
0000119843 00000 n
0000119869 00000 n
0000119928 00000 n
0000119963 00000 n
0000120229 00000 n
0000120282 00000 n
0000000164 00001 f
0000000000 00001 f
0000120317 00000 n
0000120336 00000 n
0000120423 00000 n
0000120549 00000 n
trailer
<]>>
startxref
124045
%%EOF