Scoring Regional Standard German on an Oral Exam
At a German IV oral exam, 25 students do a five-minute paired discussion on the prompt 'Sollte man Handys in der Schule verbieten?' Two raters score with a four-criterion rubric: comprehensibility, vocabulary range, grammatical accuracy, and interaction. One student, Jonas, spent last year in Graz and consistently uses Austrian standard forms: 'Jänner,' 'Erdäpfel,' 'Sackerl,' 'heuer,' and 'ich bin gesessen' for 'I was sitting.' His pronunciation includes Austrian vowel qualities, and he is among the most fluent speakers in the class. A second student, Priya, says 'Ich bin größer wie du' and 'das Buch, was ich gestern gekauft habe' in an otherwise strong performance. The first rater gives Jonas 3/4 on grammatical accuracy and notes 'nonstandard German'; the second gives him 4/4. The department has no policy on regional variation, and the exam counts for 20% of the course grade.
1. How should the rubric treat Jonas's Austrian forms, and what would you tell the two raters?
Read the worked answer
The rubric should distinguish recognized regional standard variants from errors. 'Jänner,' 'Erdäpfel,' 'Sackerl,' and 'heuer' are standard Austrian vocabulary, not dialect slang. 'Ich bin gesessen' is grammatical in Austrian and southern standard German, where 'sitzen' selects 'sein' in the perfect; it is not a case error or an auxiliary mistake. Austrian vowel qualities are a standard pronunciation, not a defect, and the comprehensibility criterion should reward intelligibility rather than proximity to a German-German norm. The first rater's note reflects a norm bias, not evidence about Jonas's grammar. The fix is textual and procedural: rewrite the accuracy descriptor to read 'control of a standard variety of German, including recognized regional standard forms,' and attach an anchor list of accepted variants (Jänner/Januar, Erdäpfel/Kartoffeln, Sackerl/Tüte, heuer/dieses Jahr, bin gesessen/habe gesessen). Then run a short calibration in which both raters score Jonas's recording against the revised descriptor. His grammar score should be 4/4; the Austrian forms are a vocabulary and register feature, not an accuracy deficit.
2. How should Priya's forms be scored, and is that consistent with your treatment of Jonas?
Read the worked answer
Priya's forms are nonstandard in the written and formal standard: 'größer als' is prescribed, and the relative pronoun after 'das Buch' should be 'das,' not 'was.' Unlike Jonas's Austrian forms, these are not regionally standard variants with their own accepted norm; they are colloquial or prescriptive violations that would be flagged in formal writing. At German IV, the appropriate response is a minor accuracy deduction, not a severe one: 3/4 on grammatical accuracy with a note on register, because her meaning is clear and the errors are isolated. This is consistent with the treatment of Jonas. The principle is not 'be lenient with variation' but 'distinguish standard regional variants from nonstandard usage.' The rubric should state that principle explicitly so raters do not have to improvise. If the department prefers to tolerate colloquial forms at this level, it should say so in the descriptor and apply it to all students, rather than letting one rater's intuitions decide.
3. Design a 30-minute calibration procedure to improve inter-rater agreement before the next exam.
Read the worked answer
Select four recorded anchor performances spanning the scale, including one with regional standard features and one with isolated colloquial errors. Both raters score all four independently against the current descriptors, then compare scores criterion by criterion. Where they diverge, they identify the linguistic feature that drove the difference and decide whether it belongs on a 'do not penalize' list (regional standard lexicon and pronunciation, dialect-influenced vowels that remain intelligible) or a 'do penalize' list (case errors, nonstandard relative pronouns, verb-second in subordinate clauses). Rewrite vague descriptors: replace 'nonstandard German' with 'a recognized standard variety,' and replace 'native-like pronunciation' with 'intelligible pronunciation that does not impede the listener.' During the live exam, double-score every fifth student and calculate adjacent agreement; if agreement falls below 80%, pause and recalibrate. Keep the anchor list attached to the rubric so the policy survives staff turnover.