Pre-Election Polls Watch Group (PEPWAG) -General goal

5 views
Skip to first unread message

Yoav

unread,
Jan 21, 2009, 11:16:24 AM1/21/09
to surveys2009
As you may feel we have re-entered this weekend into the heated
period of the pre election campaigns, and this brings with it the
important role of pre-election surveys. Pre-Election polls are
associated with our profession in the public's eyes - for the better
or the worse.
Can we contribute both to the profession and to the general society at
the same time?

You may not be aware, but the current election laws put tough
requirements on publishing the results of surveys or polls over this
period, mostly in the direction of having to disclose many details
about the survey so that the value of the results can be assessed by
the reader. The relevant part of the law is enclosed below (if the
enclosed is not readable on your machine see
http://www.knesset.gov.il/Laws/Data/law/1869/1869_All.html ). You'll
find it surprising, as so many details are regularly omitted.

The current initiative is to set up a voluntary watch-group (mostly)
of members of the Israeli Statistical Association, who will monitor in
real time what is being published and will notify immediately when the
rules are broken, explaining how was it done and why it is important.

The law has two parts. (A) Information to be disclosed by the
publisher of the results (B) Information to be disclosed by the survey
provider to the Central Election Committee as soon as possible (but
possibly after the election). We'll concentrate on (A) until after the
election.
These are:
________________________________________________________________________
A.
הראשון שמשדר לציבור על תוצאותיו של סקר בחירות ומי שמשדר כאמור בתוך 24
שעות מאז שתוצאות הסקר נמסרו לציבור לראשונה, יציין בצד התוצאות את כל
אלה:
(1) שמו של הגוף שהזמין את הסקר;
(2) שמו של הגורם שערך את הסקר;
(3) התאריך או התקופה שבה בוצע הסקר;
(4) האוכלוסיה שמתוכה נלקח מדגם המשיבים לסקר;
(5) מספר האנשים שהתבקשו להשתתף בסקר ומספר האנשים שהשתתפו בו בפועל;
(6) מרווח הטעות לגבי הנתונים שהושגו.
(ג) המפרסם בכתב לציבור על תוצאותיו של סקר בחירות כאמור בסעיף קטן
(ב), יפרט, נוסף על האמור באותו סעיף קטן, גם את השאלות שנשאלו בסקר.
B.
על אף האמור בסעיפים קטנים (ב) ו-(ג), מי שמשדר או מפרסם לציבור בכתב, על
תוצאות של סקר בחירות שלא מבוססות על שיטות סטטיסטיות מוכרות לעריכת
סקרים, יציין בהבלטה שהסקר לא התבסס על שיטות סטטיסטיות מוכרות ולפיכך אי
אפשר להסיק ממנו מסקנות לגבי דפוסי הצבעה או עמדות של הציבור, וכן יציין
את הנתונים האמורים בפסקאות (1) ו-()2
_________________________________________________________________________
I ask every member of Israel Statistical Association who encounter's a
breech of the rules to email this group a "RedAlert". Please note the
nature of the problem according to the above. If you can, add the
scanned piece or a web address, but do not hesitate to report merely
date/time/where information. If you can get into the group see if
anyone reported it before you add. If someone did report, you may
wish to add a note about how important is seems to you, or what should
be done. Otherwise, if you are first, please open a new topic. A few
members have agreed to look and discuss each such breech. This will be
done here, in the group site. Action will be taken in a timely way.

Finally please be precise but brief. We are all busy people even
though we wish to contribute to the public's good.

Reply all
Reply to author
Forward
0 new messages