Evidence-Based Frameworks for University Major Selection: A Five-Step Operational Model for Vietnamese 17-Year-Olds, with the Psychometric Displacement of MBTI

The dominant career-guidance instruments used in Vietnamese secondary-school content — Myers-Briggs Type Indicator (MBTI), 16Personalities, horoscope or zodiac matching, and folk kinship-pattern advice — are not supported by the vocational-psychology evidence base. This paper synthesises the empirical literature on how 17-year-olds should actually choose a college major and produces a five-step operational framework that displaces these popular but unsupported approaches with instruments that have psychometric backing. The framework integrates: (1) interest measurement using Holland's RIASEC hexagon and Big Five personality assessment, both with documented test-retest reliability and predictive validity for satisfaction and persistence [1], [2], [3], [4]; (2) ability selfassessment grounded honestly in g-factor research [5] and the deliberate-practice meta-analytic literature [6] which constrains overclaiming of innate-talent narratives; (3) value clarification using Schwartz's circumplex of basic human values [7]; (4) labour-market reality check (Vietnam-specific data deferred to companion node I3.2); and (5) exploration heuristics for the genuinely uncertain, drawing on Krumboltz's planned-happenstance theory [8], [9], Schwartz's paradox-of-choice satisficing prescription [7], and reversible-choice prioritisation. The paper documents the psychometric failure of MBTI without strawmanning — citing test-retest unreliability, factor-structure non-replication, the Forer effect, and the absence of incremental validity for occupational outcomes [10], [11] — while acknowledging MBTI's pedagogical-language utility for self-reflection conversations. The framework is operationally usable by a 17-year-old without paid instruments: free RIASEC short-forms (O*NET Interest Profiler, the public-domain version of the original Holland SelfDirected Search items) and free Big Five short-forms (BFI-10, IPIP-50) are recommended. Three cultural-integration considerations specific to Vietnamese family decision dynamics are flagged for follow-on node I3.3. Failure modes — over-fitting interest to current high-school subjects, premature foreclosure [12], and the over-promise of any single test — are surfaced as design constraints on guidance content.