Ref. Ares(2015)4498011 - 22/10/2015
From:
(SG)
Sent:
10 February 2013 13:42
To:
(EPSO);
(EPSO); EPSO ACCES
DOCUMENTS
Cc:
(EPSO);
(SG); SG DOSSIERS ACCES
Subject:
FW: New confirmatory request for access to documents
Attachments:
FINAL 29.01.2012 .pdf; EU
draft 8.2.2013.doc
Categories:
Dear
I hope you are well. I have been myself absent during January for health reasons.
I have examined the reply given by EPSO to the new request of Mr
.He introduced his
application because of the lack of initial reply . He has contacted us today to confirm his interest in his
application after receiving EPSO position.
Consequently, please find enclosed a draft reply including SG questions as well as a copy of the definitive
reply given by SG to his previous confirmatory request. The reply to his previous request has received
EPSO agreement but it could not been sent before since we have only received the approval of the LS last
week. Although EPSO reply confirms our position regarding his previous request, there are some issues
that I would like to discuss with you before launching the LS consultation.
Surprisingly, he stated in his message (see copy enclosed) that he understood that the document
requested did not exist, which is precisely the Commission conclusion that he did not want to accept. He
also recognizes that his confirmatory application modified his initial request. Both are the main points we
address in our reply to his previous request.
SG will like to close this new request as soon as possible. Therefore,I would appreciate to have your views
by Tuesday 12.02 2013 (end of business)
Thanks in advance
From: EPSO ACCES DOCUMENTS
Sent: Tuesday, February 05, 2013 5:18 PM
To: SG DOSSIERS ACCES
Cc:
(EPSO)
Subject: RE: New confirmatory request for access to documents
Cher M.
Afin de répondre à votre demande, je vous informe que :
1) En ce qui concerne les documents refusés, la gestionnaire du dossier étant en congé de
maladie, voudriez-vous la contacter lundi 11/02.
2) Mme
(
) est gestionnaire du dossier au stade de la demande initiale.
3) La réponse à la demande initiale est attachée dans Gestdem.
Unité 06 : Communication et attractivité
www.eu-careers.eu
QUESTIONS JURIDIQUES
Facebook: EU Careers
Secrétariat du secteur juridique
Twitter: EU_Careers
Tel.
European Personnel Selection Office
From: SG DOSSIERS ACCES
Sent: Wednesday, January 30, 2013 6:36 PM
To: EPSO ACCES DOCUMENTS
Subject: FW: New confirmatory request for access to documents
Cher(e) Collègue,
Je vous prie de bien vouloir trouver ci-joint, copie de la lettre que M.
a envoyé à la Secrétaire Générale, par laquelle un
recours est introduit pour
absence de réponse à certains documents demandés à la DG EPSO (
GESTDEM n°
2013/68).
Je me permets de vous rappeler que le délai impératif de réponse est de
15 jours
ouvrables à partir de l'enregistrement de la demande.
Afin de permettre à la Secrétaire générale de répondre à ce recours avant
l'échéance du 20/02/2013 prochain, je vous remercie d'avance de bien vouloir me
communiquer (via l'adresse électronique SG DOSSIERS ACCES) immédiatement (dans les
24 heures –
délai:01/02/2013) :
1) les documents refusés lors de la demande initiale, afin de nous permettre une
évaluation indépendante de la pertinence du refus (
si les documents sont trop
nombreux, nous vous demandons de nous faire parvenir au moins
une liste des
documents concernés);
2) le nom du (de la) gestionnaire du dossier au stade initial.
Dans les jours suivants, un(e) gestionnaire (administrateur) de l'équipe "accès aux
documents" du SG prendra contact avec le (la) gestionnaire du dossier au niveau de la
DG pour préciser
les éléments complémentaires de contribution dont le SG aura
besoin. Une
prompte réponse à cette démarche sera attendue
.
Pour toute information complémentaire, vous pouvez prendre contact avec moi.
Merci d'avance de votre collaboration.
SG.B.5.
Transparence.
Berl.
From: SG ACCES DOCUMENTS
Sent: Wednesday, January 30, 2013 6:33 PM
To:
Subject: RE: New confirmatory request for access to documents
Dear Sir,
Thank you for your e-mail dated 30/01/2013, registered on 30/01/2013
[Ares(2013) 119295].
I hereby acknowledge receipt of your confirmatory request for access to
documents - Gestdem 2013/68.
In accordance with Regulation 1049/2001 regarding public access to
European Parliament, Council and Commission documents, you will receive a
response to your request within 15 working days.
Yours sincerely,
SG.B.5. Transparence.
Berl.
From:
[mailto:
]
Sent: Wednesday, January 30, 2013 8:21 AM
To: SG ACCES DOCUMENTS;
(EPSO);
(SG)
Subject: New confirmatory request for access to documents
Dear Secretary General, dear Mr.
,
on December 28, 2012, I submitted a request for access to
documents under Regulation (EC) No. 1049/2001 to EPSO.
On January 8, EPSO acknowledged receipt - EPSO-06/RK D(2013)
20023 - and stated that the time limit for handling the application
expires on January 29, 2013.
EPSO failed to reply within the prescribed time limit.
This entitles me to making a confirmatory application according to
Article 7, Paragraph 4, of Regulation (EC) No. 1049/2001. Please
find this confirmatory request, which is identical to my original
request from December 28, 2012, below.
Please acknowledge receipt of this confirmatory request.
This is a new confirmatory request for access to documents under
Regulation (EC) 1049/2001 which differs from my previous
confirmatory request (from August 21, 2012) both in scope
and because I now do not ask for a single table but for parts of
multiple existing documents.
I realize this request is a very long and complicated document. The
second-to-last paragraph contains 6 simple steps for fulfilling this
request, with two of them probably not even needed (No. 1 and No.
6). So, in case you just want a quick overview of the request and of
how you could easily grant it, I recommend reading the second-to-
last paragraph with its 6 steps first. I had to add all the other
paragraphs to make the request fit Regulation (EC) No 1049/2001.
REQUEST FOR ACCESS TO DOCUMENTS
I request anonymized answer data from the admission test for the
open competition EPSO/AD/230/12 (AD 5) AND
EPSO/AD/231/12 (AD 7).
Please send me, in electronic form, a
set
of documents (existing documents or parts thereof) from which I
will be able to construct a table with one row for every
candidate/question pair in the admission test (if there were 45357
candidates and 60 questions, this table should contain 2,721,420
rows), and with the following columns:
• Candidate ID: an identifier of the candidate. This identifier
should not give me any indication of the identity of the
candidate who gave this answer, but it should be the same
for all answers by the same candidate, and no two
candidates should share the same candidate ID. For
anonymization purposes, it is completely acceptable for me
if you hash the Candidate ID, for instance by using the
MD5 function, as long as you do it consistently in all
released documents in which it appears so that the links
across documents are not destroyed.
• Question ID: an identifier of the question, not the question
itself. This identifier should allow you to identify the
question, but it should not give me any indication of the
question's contents. The same question in two different
languages should be listed under the same question ID in
the table I need to be able to construct. In the documents
you give me, the information which questions are
translations of the same original question can of course be
present in whatever form in which you currently store it, not
necessarily in the form of a common question ID for all
translations of the same original question.
• Question Type: an indicator of whether this question is a
verbal reasoning, abstract reasoning, numerical reasoning,
or situational judgement question.
• Language: the language in which this question was
presented to this candidate.
• Neutralized: the information whether the answers to this
question were "neutralized" or not. By "neutralized", I mean
that any answers to this question were disregarded and
instead scores from the non-"neutralized" questions were
extrapolated. I do not imply here that a retroactive
neutralization which deserves this name is actually possible.
• Expected answer: an identifier of the answer that is
expected. This identifier should not give me any indication
of the contents of the answer, but for the same
question/answer pair, it should stay the same. If the answer
options were always presented in the same order, the letter
A/B/C/D would be sufficient. If answer options were not
presented in the same order to all candidates, the same
identifier should always be used for the same answer
option. For situational judgement questions, the entire
expected answer should be indicated, that is, the best and
the worst option. Again, since I cannot know how you store
this information, I request it in whatever form you have it.
• Given answer: an identifier of the answer chosen by this
candidate for this question. Again, I do not request any
indication of the contents of the answer, just an identifier. It
should be specified in the same fashion as for the expected
answer, so that correct answers can be recognized by the
identity of identifiers. If the candidate did not answer a
question, this information should be included as well. For
situational judgement questions, the complete answer (that
is, best and worst option).
• Time spent: the time (for instance, in seconds) the
candidate spent on this question.
• Difficulty: the level of difficulty of this question.
I do not request that you create a new document by merging
information from existing documents, but instead I request that you
provide me with the existing documents which are stored in
electronic form in your various IT systems, possibly after removing
some information from them which I did not request and which is
covered by one or more exceptions from Article 4.
It is, as far as I know, not publicly known how many documents
you are using to store the information contained in the table which I
want to be construct from your documents, so I cannot specifically
ask for individual documents about whose existence I can be
certain. However, you have already confirmed that the information
I requested on July 2, 2012 exists (EPSO/RK D(2012) 1130795:
"EPSO does indeed possess the information referred to in your
request"). That is, the existence of all requested information except
for the difficulty column has already been confirmed.
Although I do not know and cannot know how you store this
information about which we agree that it exists, let me make an
educated guess about which documents exist. If my guess is
correct, please deliver those documents. If it is not correct, please
deliver a set of documents containing equivalent information.
You probably have a document (I will call it "
candidate table")
containing a list of all candidates and several properties of the
candidates. In other documents, you probably have
pointers to records in this candidate table. These pointers use one
column (or a column combination) in the candidate table which I
will call the "Candidate ID". In database terminology, this
Candidate ID would be called a key.
I request the following subset of the candidate table:
All rows with candidates who took part in the competition
EPSO/AD/230/12 (AD 5) and/or EPSO/AD/231/12 (AD 7). Each
row should contain at least these columns:
• Candidate ID (the ID used in other
documents to refer to records of this document; if multiple
such IDs exist, I request all of them; if the IDs would reveal
the identity of candidates, I request that you anonymize this
information without losing the identifying property of the
ID so that the references from the other documents still
work; such anonymization would constitute the removal of
information from a document to arrive at a part of a
document in the sense of Regulation (EC) No 1049/2001.
Art. 4.6: "If only parts of the requested document are
covered by any of the exceptions, the remaining parts of the
document shall be released." The identity of a candidate is
probably covered by one of the exceptions, but the
information which answers have been given by the same
candidate is not. So you should not release the identity of
candidates to me, but an anonymous identifier is still
needed for releasing the information which answers come
from the same candidate).
• language combination of this candidate, in other words,
information which will allow me to determine, in
conjunction with the question type, which language was
used for which question when it was presented tothis
candidate.
Please remove any additional columns which would trigger an
exception according to Article 4.
You most likely have a document (I will call it "
question table")
containing a list of all questions, answer options, expected answers,
and difficulty. In other documents, you probably have
pointers to records in the question table. These pointers use one
column or column combination in the question table which I will
call "Question ID".
I request the following subset of the question table:
All rows pertaining to questions which were used at least once in
the competition EPSO/AD/230/12 (AD 5) and/or EPSO/AD/231/12
(AD 7), with at least the following columns:
• Question ID (the ID used in other
documents to refer to records of this document; if multiple
such IDs exist, I request all of them; if the IDs would reveal
the contents of the question, I request that you anonymize
this information without losing the identifying property of
the ID so that the references from the other documents still
work). The same question in two different languages should
be listed under the same question ID, or if this is not how
you store this information, you should give me this
information (which questions have the same content in
different languages) in whatever format you use.
• Difficulty: the level of difficulty of this question
• Question Type: an indicator of whether this question is a
verbal reasoning, abstract reasoning, numerical reasoning,
or situational judgement question.
• Neutralized: the information whether the answers to this
question were "neutralized" or not. See above for a
definition of what I mean by "neutralized".
• Expected answer: an identifier of the answer that is
expected. This identifier should not give me any indication
of the contents of the answer, but for the same
question/answer pair, it should stay the same. If the answer
options were always presented in the same order, the letter
A/B/C/D would be sufficient. If answer options were not
presented in the same order to all candidates, please ensure
that the same identifier is always shown for the same
answer option. For situational judgement questions, indicate
the entire expected answer, that is, the best and the worst
option.
Please remove any additional columns which would trigger an
exception according to Article 4.
You most likely have a document (I will call it "
answer table")
containing a list of all answers given by candidates in the
competition EPSO/AD/230/12 (AD 5) and/or EPSO/AD/231/12
(AD 7).
I request the following subset of the answer table:
All rows which were entered for the competition EPSO/AD/230/12
(AD 5) and/or EPSO/AD/231/12 (AD 7), with at least the following
columns:
• Candidate ID: an identifier of the candidate. This identifier
should not give me any indication of the identity of the
candidate who gave this answer, but it should be the same
for all answers by the same candidate, and no two
candidates should share the same candidate ID. The
candidate ID should allow me to join this table with the
candidate table. See also my other remarks about Candidate
IDs above.
• Question ID: an identifier of the question, not the question
itself. This identifier should allow you to identify the
question, but it should not give me any indication of the
question's contents. The same question in two different
languages should be listed under the same question ID (also
see my remarks about the question ID above). The question
ID should allow me to join this table with the question
table.
• Given answer: an identifier of the answer chosen by this
candidate for this question. Again, I do not request any
indication of the contents of the answer, just an identifier. It
should be specified in the same fashion as for the expected
answer, so that correct answers can be recognized by the
identity of identifiers. Also, this identifier should not only
contain the information whether the question was correctly
answered or not, but also which wrong answer was selected
in case the answer is wrong. In other words, for any pair of
candidates who chose the same wrong answer option, this
fact should be visible from this identifier, even though the
answer itself should not be released to me. A separate
identifier should be used if the candidate did not answer this
question. For situational judgement questions, indicate the
complete answer (best and worst option).
• Time spent: the time (for instance, in seconds) the
candidate spent on this question.
Please remove any additional columns which would trigger an
exception according to Article 4.
If you have split any of the requested documents in multiple
documents (either horizontally, that is, by maintaining separate sets
of rows, or vertically, that is, by maintaining separate sets of
columns, or both), I request all documents which I
need to construct the table which I have described first in my
request.
In the unlikely case that in one or more of the requested documents,
there are no ID columns which identify records without giving
away important content such as the name of a candidate, his
passport number, or the wording of a question or answer, and you
try touse this as an argument for not disclosing requested data, I
request that you apply a hash function such as MD5 to the affected
IDs in order to remove the information which should not be shared
while retaining the information that must be shared (the "remaining
parts" of the document include the information which records from
one document need to be joined with what records from another
document).
Please note that I do not request any processing of data beyond
what is required by Regulation (EC) 1049/2001. I only request the
releasable parts of existing documents. Any data processing I ask
for is only needed for separating the releasable information from
information which would trigger an exception according to Article
4, and Article 4 obliges you to perform this minimal amount of data
processing. Art. 4.6 says that "If only parts of the requested
document are covered by any of the exceptions, the remaining parts
of the document shall be released." In other words, you are
obliged to separate the remaining parts of the documents from the
non-releasable parts of the documents if the documents contain
non-releasable parts. If you store any requested information in a
format different from what I describe here, I do not request re-
formatting – giving it to me in the existing format would satisfy my
request, as long as equivalent information can be extracted from
what you deliver.
QUICK OVERVIEW OF HOW TO FULFIL THE REQUEST
WITH MINIMAL EFFORT
Please note that you can fulfil my request quite easily
by slightly changing your existing mechanism for reporting the
candidates’ answers back to them. The table I want to ultimately
construct (see above) differs only slightly from the concatenation
of tables which you already routinely send to every candidate as
part of the application process, similar to the message you sent me
via my EPSO account on 28/06/2012 (Candidate number:
2575906). I realize that following these 6 steps could be construed
as creating a new document which does not yet exist. Because of
this, I do not request that you follow these 6 steps in particular. I
merely point out that this would probably be the most efficient and
economical way of fulfilling my request for parts of existing
multiple documents outlined above, and I would also accept such a
new, not yet existing document instead of the requested parts of
existing documents in case you would prefer delivering the
requested information in this equivalent form.
You could deliver all requested information by modifying your
existing data export mechanism with the following simple steps:
1. Optional: anonymize the existing Candidate ID in case you
consider it too revealing. This can be trivially achieved by
supplying the ID in hashed form, for instance, by supplying
md5(candidate ID) instead of the raw candidate ID.
See http://en.wikipedia.org/wiki/MD5 and http://php.net/manual/en
/function.md5.php
2. Combine the language combination and the table with
expected/given answers. Since you already send them both within
the same message (e.g., in your message to me from 28/6/2012),
this should not be much effort.
3. Add the difficulty. This information must be stored together with
the question in the database of questions (or at least linked to it)
and should therefore be easily retrievable.
4. The question ID would have to be given to me as a globally
unique ID (i.e., as an identifier that stays the same across all
candidates), not just a candidate-specific question number. Since
your mechanism for generating the tables you send to candidates
clearly has a link between the global question ID and the candidate-
specific number (which must be true because you report the
expected answer), reporting this global question ID in
addition to the candidate-specific question number is a trivial
change.
5. Add the situational judgement questions to the mechanism
already in place for all the other question types.
6. Only if the answer options are presented to different candidates
in different orders, you might have to compensate for this.
WHY AM I ENTITLED TO RECEIVING THIS FILE FROM
YOU?
• Regulation 1049/2011 applies because the file specified
above fits the definition of a document in Art. 3a, I am a
citizen of the Union residing in a Member State, and the
document is held by an institution.
• No exceptions according to Art. 4 apply:
Art. 4.1(a) does not apply because public security, defence
and military matters, international relations, and
financial/monetary/economic policy are not affected.
Art. 4.1 (b) does not apply because the data I request are
anonymized.
Art. 4.2 does not apply because the contents of questions
and answers are not included in the data I request.
Art. 4.3 does not justify an exception because disclosure
would not undermine the ongoing decisionmaking process,
and in any case there is an overriding public interest in
disclosure (transparency, ensuring fairness, the public
needs to be able to check whether, and to what degree,
language-based discrimination occurs).
The rest of Art. 4 obviously does not justify an exception
either.
With best regards,
--
Tel
Fax
Document Outline