In the theoretical frameworks of English Language Teaching (ELT) and language assessment, evaluating receptive and productive language proficiency across the four modalities requires rigorous construct definition and methodological alignment. Consider the following four complex propositions regarding the evaluation of speaking, listening, reading, and writing: I. In assessing reading comprehension, macro-skills such as skimming (rapidly surveying text to establish global gist, overarching themes, and macro-structure) and scanning (searching text meticulously for specific, isolated information or discrete data points) engage fundamentally different cognitive search strategies and top-down versus bottom-up processing loads. II. When evaluating writing proficiency through direct performance assessments, holistic scoring evaluates a composition against a single, specialized, analytic rubric featuring multiple sub-traits (such as grammar, vocabulary, mechanics, and task fulfillment), whereas primary-trait scoring assesses the overall quality of a text as an undifferentiated, single global impression. III. In oral proficiency testing (speaking), direct interviews and semi-direct elicitation tasks (such as computer-based speaking tests) are completely identical in eliciting naturalistic conversational repair, backchanneling cues, overlapping speech, and socio-pragmatic interactional competence. IV. Diagnostic listening tests designed to evaluate micro-skills typically isolate discrete acoustic competencies, such as discriminating between minimal pairs, recognizing phonological assimilation, parsing intonation contours, and identifying word boundaries, before testing macro-level inferential listening. Identify the option that correctly evaluates the pedagogical and psycholinguistic validity of propositions I, II, III, and IV.