Upload
lunorip
View
221
Download
0
Embed Size (px)
Citation preview
8/9/2019 Technical Report Ch 08
1/10
192
First European Survey on Language Competences: Technical Report
!"#$%&' )*
+$&'#%,-./ 0
12#3,%4 5-.,%-',.6
8/9/2019 Technical Report Ch 08
2/10
193
First European Survey on Language Competences: Technical Report
8 Operations - Quality monitoring
This chapter provides an overview of the quality procedures employed for the ESLC.Note, the discussion focuses on the Main Study processes only, Field Trial processes
are not discussed unless relevant.
8.1 An introduction
It is essential that users of the ESLC data have confidence that the data collection
activities were undertaken to a high standard and for the purpose of creating an
international dataset of a quality that will enable valid comparisons across participating
educational systems. There were various methods, detailed further below, used to
ensure this confidence.
The ESLC Technical Standards (which can be seen in Appendix 4) provided the set of
standards upon which the data collection activities were based and were fundamental
to the quality control methods employed by SurveyLang for the ESLC. There are three
types of standards; each with a specific purpose. Data Standards ensure that all
collected data can be added to the final ESLC 2011 dataset that will be released by
the Commission. Management Standardsensure that all ESLC operational objectives
are met in a timely and coordinated manner. National Involvement Standardsensure
that the internationally developed instruments meet the highest standards of cross-
national, cross-cultural and cross-linguistic validity and equivalence and that the ESLC
results have the greatest possible meaning for national stakeholders.
The Data Standardsoutlined the standards for the following areas:
target population and sampling
translation
test administration
security of materials
quality monitoring
printing of material
marking, coding and data entry
data submission
All SurveyLang procedures were carefully developed and documented to ensure data
of the desired quality. Quality monitoring played an important role and the
implementation of the operational procedures documented in the guidelines and
manuals made available to NRCs were continually monitored. In any cases where the
documented operational processes were not fully implemented, these were logged and
discussed with the NRC to understand the likely impact for the data. Quality monitoring
was, therefore, the process of systematically observing and recording the extent to
which data were collected and stored according to the procedures described in the
8/9/2019 Technical Report Ch 08
3/10
194
First European Survey on Language Competences: Technical Report
ESLC field operations documentation. Quality monitoring was a continuous process
and was a shared responsibility between the NRC and SurveyLang.
The main elements of the quality monitoring procedures were:
project milestones and signed off documentation were managed
Central Training: NRC attendance at central SurveyLang training
Credential sheets: SurveyLang provided role definitions and criteria for the
appointment of Test Administrators, Quality Monitors, Data Entry staff,
Markers of Writing, and Coders for the open responses in the Student
Questionnaire
SurveyLang team: SurveyLang assisted NRCs in the planning and
implementation of key processes. SurveyLang systematically monitored the
key processes of sampling, translation, field operations and data entry, the
marking of writing and the coding of the open responses in the Student
Questionnaire
Central Issues Log: all risks and issues relating to the implementation of
operational processes were stored on a central register and regularly
reviewed
Quality Monitors: Quality Monitors observed the implementation of ESLC field
operations at the educational system level, NRCs appointed a Quality Monitor
to make unannounced visits to, typically, 10 schools in each educational
system, interviewing both the Test Administrator and School Coordinator. The
Quality Monitor also visited each National Research Centre and interviewed
the NRC and five data entry staff. The lead Quality Monitor for eacheducational system wrote a full report on the results of their visits and
interviews. Further details about their role are available below in section 8.4.
ESLC Administration Report forms: ESLC Test Administrators completed a
report after each ESLC test administration, thus providing an overview of the
test administration across participating educational systems.
ESLC Administration Issues Log: ESLC Test Administrators completed a log
of all administration issues, thus providing an overview of any arising issues
across participating educational systems.
NRC report: SurveyLang developed a template report which allowed NRCs to
systematically self-report on the implementation of key processes in their
educational system.
8.2 Support for NRCs in quality monitoring
The documentation that formed the basis for the quality monitoring procedures were:
ESLC Technical Standards
NRC Field Operations Manual
8/9/2019 Technical Report Ch 08
4/10
195
First European Survey on Language Competences: Technical Report
Test Administrator Manuals (paper-based and computer-based including a
script for test administration)
School Coordinator Guidelines (paper-based and computer-based)
Sampling Manual Translation Guidelines and Manuals
Data Entry Guidelines (Listening, Reading and Writing and Student
Questionnaires)
Marking of Writing documentation
Coding Guidelines
Quality Plan for Quality Monitors and Quality Plan Report.
The quality monitoring instruments developed from these manuals and guidelines
included role credential sheets, a range of sampling forms, WebTrans for the
translation and verification work, a Test Administrator interview protocol for the QualityMonitor, a School Coordinator interview protocol for the Quality Monitor, a Data Entry
interview protocol for the Quality Monitor, NRC interview protocol for the Quality
Monitor, NRC feedback report templates, an ESLC Administration Report Form and an
ESLC Administration Issues Log.
Credential sheets: as outlined above, SurveyLang provided role definitions and criteria
for the appointment of Test Administrators, Quality Monitors, Data Entry staff, Markers
of Writing, and Coders for the open responses in the Student Questionnaire
Sampling forms: SurveyLang developed a series of forms for collecting key data and
for monitoring school and student sampling outcomes. The NRC and SurveyLang
experts negotiated agreement on sampling plans and outcomes (see chapter 4).
WebTrans for translation and verification: this system managed the quality monitoring
activities for translation of all documentation at the national level (see Chapter 5).
Test Administrator quality monitor interview protocol: a standard schedule was
prepared by SurveyLang to systematically record the outcomes of the Quality
See Task 1 below in section 8.4 for further information.
School Coordinator quality monitor interview protocol: a standard schedule was
prepared by SurveyLang to systematically record the outcomes of the Quality
See Task 1 below in section 8.4 for further information.
Data Entry quality monitor interview protocol: a standard schedule was prepared by
.
See Task 2 below in section 8.4 for further information.
NRC quality monitor interview protocol: a standard schedule was prepared by
te visit.
The interview protocol recorded information on:
the general organisation of the ESLC in that educational system
8/9/2019 Technical Report Ch 08
5/10
8/9/2019 Technical Report Ch 08
6/10
8/9/2019 Technical Report Ch 08
7/10
198
First European Survey on Language Competences: Technical Report
SurveyLang provided a full credential sheet for Quality Monitors. In brief, the Quality
Monitor:
should have past experience of acting as a Quality Monitor on a similar project
or in a similar role should have fluency in English and the questionnaire language
should not be an employee at the same organisation as the NRC
must not be an immediate relative of an employee at the NRC
must not be line managed by the NRC
must send their report directly to SurveyLang in electronic format
must be able to and have the capacity to independently and effectively
communicate with SurveyLang using email and telephone.
However any feedback that could be used by the NRC to correct the way tasks were
managed duringthe Main Study were discussed with the NRC so the NRC could takequick action. Such instances were to be documented in the report.
The NRC had to meet with the Quality Monitor to:
Train the Quality Monitor in the background of the ESLC.
Make all operational documentation available particularly the School
Coordinator Guidelines, Test Administrator Manual and the Data Entry
Guidelines.
Make background information about the project available, e.g. the Inception
and Interim Reports.
Make a plan of schools and data entry staff for the Quality Monitor to visit. Inform the Quality Monitor of variations agreed to standard SurveyLang
procedures.
Be available to respond to questions raised by the Quality Monitor. Where
there are several Quality Monitors, one person should be appointed as the
lead person and should be the key liaison with the NRC.
Inform the Quality Monitor that they can ask questions directly of SurveyLang
if they wish. Where there are several Quality Monitors, the lead person should
be the contact with SurveyLang.
SurveyLang suggested that the Quality Monitor attended -country Test
Administrator training.
The appointed Quality Monitor (or Quality Monitors in the case of Task 1 below) made
unannounced visits to assess the quality of the implementation of SurveyLang
processes and procedures carried out within the educational system.
Task 1: Test Administration: This task required the Quality Monitor to visit 10 schools;
a mix across administration modes (computer and paper-based where both were
used), regions and Test Administrators was preferred. The Quality Monitor was
required to:
8/9/2019 Technical Report Ch 08
8/10
199
First European Survey on Language Competences: Technical Report
Be at the school from 1 hour before the start of the administration to the end of
administration when the Test Administrator and School Coordinator had
packaged completed test materials up and completed the student tracking
form.
Interview the Test Administrator and School Coordinator separately for
approximately 15 minutes. A template list of questions was provided for this
purpose covering all aspects before, during and after the test administration
focusing particularly on the logistical arrangements, materials management
and clarity of documentation and processes.
Write all responses down from the Test Administrator and School Coordinator
interviews.
Summarise the key findings from the interview in terms of what worked well
and what did not work well for the Test Administrators and the School
Coordinators. Were the procedures followed as specified? What problems
were encountered? How were these resolved?
Quality Monitors may also have wished to review some of the Administration
Report Forms for the Test Administrations that they observed.
Task 2: Data Entry staff: This task required a quality check of the data entry work. The
Quality Monitor had to review a sample of the data entry work from each data entry
person employed by the NRC. The Quality Monitor was required to:
Review a sample of five booklets per skill (Reading, Listening, Writing and the
questionnaires) from the work each data entry person had performed.
Check that the data was entered correctly for each data entry person. In the
report, the number of mistakes found had to be entered. The data had to be
corrected and the NRC informed so that they could take corrective measures
if necessary.
Interview each data entry person for approximately 5 minutes. A template list
of questions was provided for this purpose, covering aspects such as the
confidence of their accuracy.
Write all responses down and document in the report what worked well and
well for each data entry staff member. Were the procedures
followed as specified? What problems were encountered? How were these
resolved?
Task 3: NRC report: The Quality Monitor was required to:
Talk through each step in the period from mid January (including printing and
the receipt of materials) to data submission for approximately 30 minutes. A
template list of questions was provided for this purpose.
Write all responses down and document in the report what worked well and
well for the NRC. Were the procedures followed as specified?
What problems were encountered? How were these resolved?
8/9/2019 Technical Report Ch 08
9/10
200
First European Survey on Language Competences: Technical Report
8.5 Quality monitoring data
The quality monitoring data collected from all of the documents and mechanisms in
paragraph 0 were carefully reviewed and analysed both after the Field Trial and theMain Study.
Feedback was assessed by SurveyLang in order to improve all central and educational
system processes and documentation after the Field Trial. All feedback and data was
assessed again after the Main Study.
ESLC Quality monitor reports: each of the 16 educational systems submitted a Quality
Monitor report on the conduct of testing sessions. The report consisted of a summary
of their general observations together with a summary of the main findings from each
of the different types of interviews conducted.
ESLC NRC Reports: each of the 16 NRCs submitted feedback reports on the overall
processes and documentation for the Main Study.
In general, the quality monitoring reports and NRC feedback reports suggested a
strong organisational base existed within educational systems for the conduct of the
ESLC. The Quality Monitor reports indicated that, overall, NRC staff had a very good
understanding of the operational aspects of the ESLC. The reports indicated that the
ESLC administrations were conducted in a manner that was largely consistent with the
documented procedures in the ESLC operations manuals.
Issues that were reported from the Main Study include:
The required school sample size was not reached for one or both test
languages (England, Greece). Note: to minimise the impact of this issue,
additional students were drawn in the sample across participating schools.
Some schools did not reach the required participation rate but did not inform
the NRC and did not organise follow-up sessions (France).
Some schools were unable to administer the Listening test in the required
classroom setting because there were not enough rooms or test
administrators available (German Community of Belgium).
Data entry deadlines could not be met due to staff shortages. The team leader
monitoring data input had to undertake data entry himself and therefore could
not supervise others properly (Malta).
Data entry deadlines could not be met due to staff shortages and financial
restrictions which prevented the NRC from employing experienced staff
(Greece).
Data entry files became corrupted and could not be recovered requiring re-
entry of data which resulted in delays (German Community of Belgium).
There were complaints about the lack of information to parents before the
administration (German Community of Belgium).
8/9/2019 Technical Report Ch 08
10/10