Transcript STANAG 6001 Testing
STANAG 6001 OPI Testing
Julie J. Dubeau Bucharest BILC 2008
Bill Who???
Julie J. Dubeau
Julie J. Dubeau
Are We All On the Same Page?
An Exploratory Study of OPI Ratings Across NATO Countries Using the NATO STANAG 6001 Scale*
*This research was completed in 2006 as part of a M.A. Thesis in Applied Linguistics
Julie J. Dubeau
Presentation Outline
Context Research Questions Literature Review Methodology Results – Ratings – Raters – Scale Conclusion Julie J. Dubeau
NATO Language Testing Context
Standardized Language Profile (SLP) based on the NATO STANDARDIZATION AGREEMENT (NATO STANAG) 6001 Language Proficiency Levels (Ed 1? Ed 2?) – 26 NATO countries, 20 Partnership for Peace (PfP) countries & others … Julie J. Dubeau
Interoperability Problem ?
Language training is central within armed forces due to the increasing number of peace-support operations, and is considered as having an important role in achieving
interoperability
among the various players. “ The single most important problem identified by almost all partners as an impediment to developing interoperability with the Alliance has been shortcomings in communications ” (EAPC (PARP) D, 1997, 1, p.10). Julie J. Dubeau
Overarching Research Question
Since no known study had investigated inter-rater reliability in this context, the main research question was:
How comparable or consistent are ratings across NATO raters and countries?
Julie J. Dubeau
Research Questions
Research questions pertaining to the ratings RQ1
Research questions pertaining raters’ training and background RQ2
Research questions pertaining to the rating process and to the scale RQ3
Julie J. Dubeau
Research Questions
RQ1-Ratings:
How do ratings of the same oral proficiency interviews (OPIs) compare from rater to rater?
Would the use of plus levels increase rater agreement?
How do the ratings of the OPIs compare from country to country?
Are there differences in scores within the same country?
Julie J. Dubeau
Research Questions
RQ2-Raters’ training and background:
Are there differences in ratings between raters who have received varying degrees of tester/rater training and STANAG training?
Did very experienced raters score more reliably than lesser experienced ones? Are experienced raters scoring as reliably as trained raters?
Are there differences in ratings between participants who test part-time versus full-time, are native or non-native speakers of English, and are from ‘Older’ and ‘Newer’ NATO countries?
Julie J. Dubeau
Research Questions
RQ3-Rating process and scale use:
Do differing rating practices affect ratings?
Do raters appear to use the scale in similar ways?
What are the raters’ comments regarding the use and application of the scale?
Julie J. Dubeau
Literature Review
Testing Constructs – What are we testing?
• General proficiency & Why • Rating scales Rater Variance – How do raters vary?
• Rater/scale interaction • Rater training & background Julie J. Dubeau
Methodology
Design of study:
Exploratory survey – 2 Oral Proficiency Interviews (OPIs A & B) – Rater data questionnaire – Questionnaire accompanying each sample OPI Participants : Countries recruited at BILC Seminar in Sofia 2005 103 raters from 18 countries and 2 NATO units Julie J. Dubeau
Analysis:
– Rating comparisons – Original ratings – ‘Plus’ ratings – Rater comparisons – Training – Background – Country to country comparisons • Within country dispersion – Rating process • Rating factors – Rater/scale interaction • Scale user-friendliness Julie J. Dubeau
Results RQ1-
Summary Ratings : To compare OPI ratings and to explore the efficacy of ‘ plus ratings ’ .
– Some rater-to-rater differences – ‘Plus’ levels brought ratings closer to the mean – Some country-to-country differences – Greater ‘within-country’ dispersion in some countries Julie J. Dubeau
View of OPI ratings sample A
10 0 60 50 40 30 20 60 32 within level 1 10 within level 2
Stacked view of A
1 within level 3 Julie J. Dubeau Adjusted + range A Within L1 range Within L2 range Within L3 range
Results
Sample A (L1) All Ratings (with +)
Levels Numbers Within Level 1 range Within Level 2 range Within Level 3 range Total 70 32 1 103 Julie J. Dubeau % 68.0
31.1
1.0
100.0
All Countries’ Means for Sample A 17 19 20 18 15 16 13 11 9 14 12 7 8 10 5 4 6 3 2 1
1.00
1.20
1.40
1.60
1.80
2.00
Overall Country Mean
Julie J. Dubeau 2.20
2.40
All Ratings for Sample B
(level 2)
Levels 1 1+ 2 2+ 3 3+ 4 Total Numbers 2 1 47 8 34 2 2 96 % 1.9
1.0
45.6
7.8
33.0
1.9
1.9
93.2
Julie J. Dubeau
View of OPI ratings sample B
60 50 40 30 20 55 31 10 0 2 1 within level 1 5 within level 2 within level 3
Stacked view of B
1 1 within level 4 Julie J. Dubeau Adjusted + range B Within L1 range Within L2 range Within L3 range Within L4 range
All Countries’ Means for Sample B
25.00
20.00
15.00
10.00
5.00
0.00
1.80
12 15 13 9 20 17 4 5 11 14 7 10 8 6 18 3 2.10
2.40
2.70
2 1 3.00
3.30
19
Results RQ2-
Summary Raters: To investigate rater training and scale training and see how (or if) they impacted the ratings, and to explore how various background characteristics impacted the ratings – Trained raters scored within the mean, especially for sample B – Experienced raters did not do as well as scale trained raters – Full-time raters scored closer to mean – ‘New’ NATO raters scored slightly closer to mean – NNS raters scored slightly closer to mean Julie J. Dubeau
70 60 50 40 30 20 10 0
Tester (Rater) Training
63.27% 36.73%
none to little Julie J. Dubeau substantial to lots
Years of Experience
10 0 50 40 30 20
49.5% 19.8% 14.85% 15.84%
0 to 1 year 2 to 3 years 4 to 5 years 5 years + Julie J. Dubeau
STANAG Scale Training
60 50 40 30 20 10 0
60.0% 40.0%
none to little Julie J. Dubeau substantial to lots
‘Old’ vs. ‘New’ NATO Countries
Newer NATO member?
Yes
No Total Summary of Tester Trg Little Lots 14 23 2 29
30
28 58 Julie J. Dubeau Total 36 51 87
‘Old’ vs. ‘New’ NATO Countries
Newer NATO member?
Yes No Total
Rating OPI B Correct?
Yes
No
Other/Missing 14 27 2 54
6 26
32 4 2 6 Total 37 55 92 Julie J. Dubeau
Results Raters’ Background
Conducts Testing Full-time?
• Yes 34 (33.0 %) • No 67 (65.0 %) • Full-time testers more reliable (accurate) – NNS (60%) raters better trained?
– ‘New’ raters better trained?
Julie J. Dubeau
Results RQ3-
Summary Scale: To explore the ways in which raters used the various STANAG statements and rating factors to arrive at their ratings.
– Rating process did not affect ratings significantly – 3 main ‘types’ of raters emerged: • Evidence-based • Intuitive • Extra-contextual Julie J. Dubeau
Results
An ‘ evidenced-based ’ rating for Sample B (level 2): I compared the candidate ’ s performance with the STANAG criteria (levels 2 and 3) and decided that he did not meet the requirements for level 3 with regard to flexibility and the use of structural devices. Errors were frequent not only in low frequency structures, but in some high frequency areas as well.
(Rater 90
–
rated 2)
Julie J. Dubeau
Results
An ‘ intuitive ’ rating for Sample A (level 1): I would say that just about every single sentence in the interpretation of the level 2 speaking could be applied to this man. And because of that I would say that he is literally at the top of level 2. He is on the verge of level 3 literally. So I would automatically up him to a low 3.
(Rater 1- rated 3)
Julie J. Dubeau
Results
An ‘extra-contextual’ rating for Sample A (level 1): Level 3 is the basic level needed for officers in (my country). I think the candidate could perform the tasks required of him. He could easily be bulldozed by native speakers in a meeting, but would hold his own with non-native speakers. He makes mistakes that very rarely distort meaning and are rarely disturbing.
(Rater 95 – rated 2)
Julie J. Dubeau
Implications
Training not equal in all countries Scale interpretation Plus levels useful Different grids, speaking tests Institutional perspectives Julie J. Dubeau
Limitations & Future Research
Participants may not have rated this way in their own countries OPIs new to some participants Future research could – Get participants to test – Investigate rating grids – Look at other skills Julie J. Dubeau
Conclusion of Research
So, are we all on the same page?
YES!
BUT…
Plus levels were instrumental in bridging gap Training was found to be key to reliability More in-country training should be the first step toward international benchmarking.
Julie J. Dubeau
Thank You!
Are We All On the Same Page?
An Exploratory Study of OPI Ratings Across NATO Countries Using the NATO STANAG 6001 Scale
The full thesis is available on the CDA website http://www.cda-acd.forces.gc.ca
Or google Dubeau thesis