At the start of the semester（学期）/si'mestә/, every freshman（大一新生）/ˈfrɛʃmən/ and undergraduate（大学本科生）/.ʌndә'grædjuit/ in Mira's program had to take part in（参加）/teɪk pɑrt ɪn/ a compulsory（必修的）/kәm'pʌlsәri/ research workshop in the main auditorium（礼堂）/.ɒ:di'tɒ:riәm/. Dean（院长）/di:n/ Wu explained that the curriculum（课程）/kә'rikjulәm/ should build intermediate（中级的）/.intә'mi:diәt/ skills before students began a thesis（论文）/'θi:sis/. Yet faculty（教师）/'fækәlti/ feedback（反馈）/'fi:dbæk/ showed that many learners could repeat a formula（公式）/'fɒ:mjulә/ without understanding its logic（逻辑）/'lɒdʒik/. The principal（关键的）/'prinsipәl/ question came from Mira. She asked, “Could the workshop measure aptitude（能力倾向）/'æptitju:d/ rather than memory?” Dean Wu replied, “Yes, but every task must reveal how students reach an answer.”

In response to（回应）/ɪn rɪˈspɑns tɪ/ that concern, Mira's team rewrote the language section of the manual（手册）/'mænjuәl/. Students had to classify（分类）/'klæsifai/ each noun（名词）/naun/ and pronoun（代词）/'prәunaun/, choose a conjunction（连接词）/kәn'dʒʌŋkʃәn/ or preposition（介词）/preposition/, correct a plural（复数）/'pluәrәl/ form, and repair a clause（从句）/klɒ:z/ with a missing comma（逗号）/'kɒmә/. Then they gave every paragraph a useful heading（标题）/'hediŋ/. Mira, who edited the first transcript（文字记录）/'trænskript/, noticed a missing heading. The team found that clear labels improved cohesion（结合）/kәu'hi:ʒәn/, but clarity（清楚）/'klærәti/ also depended on a coherent（清晰的）/kәu'hiәrәnt/ line of thought. If the sequence（序列）/'si:kwәns/ remains unclear, readers will not comprehend（理解）/.kɒmpri'hend/ the argument.

The second task combined astronomy（天文学）/ә'strɒnәmi/ with arithmetic（算术）/ә'riθmәtik/. A small apparatus（装置）/.æpә'reitәs/ produced audio（音频）/'ɒ:diou/ signals from a model telescope. Students had to keep track of（持续记录）/kip træk əv/ each digit（数字）/'didʒit/, work out（算出）/wərk aʊt/ a proportion（比例）/prә'pɒ:ʃәn/, place the results in sequence, and turn them into a graph（图表）/græf/. They used an index（索引）/'indeks/ to filter（筛选）/'filtә/ the records, compute（计算）/kәm'pju:t/ an average, and explain the code behind the display. Every claim was verified（核实）/'verifai/ against the original graph. This task helped the class define（定义）/dɪˈfaɪn/ evidence and derive（得出）/di'raiv/ a conclusion from observable results.

The final task used a microphone（麦克风）/'maikrәfәun/ and a student journalism（新闻业）/'dʒә:nәlizm/ forum（论坛）/'fɒ:rәm/. Each group had to make use of（利用）/meɪk juz əv/ an interview, edit its transcript, and formulate（拟写）/'fɒ:mjuleit/ a short editorial（短评）/.edi'tɒ:riәl/. The faculty has already collected feedback from two classes. According to（根据）/əˈkɔrdɪŋ tɪ/ the responses, students preferred a collective（集体的）/kә'leiktiv/ task in which they could collaborate（合作）/kә'læbәreit/ and be responsible for（负责）/bi riˈspɑnsəbəl fər/ different stages. The forum also revealed whether speakers were proficient（熟练的）/prә'fiʃәnt/ enough to explain a difficult idea without hiding behind technical language.

At the end, the dean asked students to be familiar with（熟悉）/bi fəˈmɪljər wɪθ/ the complete process: identify a problem, select a method, verify the evidence, and present the result. The revised curriculum did not promise that every undergraduate would immediately produce a perfect thesis or earn a diploma（文凭）/di'plәumә/. It did something more practical. It showed whether students could connect language, data, and public explanation, then improve their work after precise feedback.