Suppressing an Inconvenient Finding

The data-exclusion request is a research-integrity problem: deleting a legitimate stratum because it produces an inconvenient result is selective reporting and falsification. The relevant principles are honesty, objectivity, and transparency in data analysis. The sampling frame shows the two districts were part of the stratified random sample, so removing them would bias estimates toward better-resourced schools and inflate the ABL figure. The authorship demand raises a separate issue: authorship must be based on substantial intellectual contribution, not supervisory status alone. If the supervisor did not design, analyze, or write the paper, first authorship is not justified. UGC regulations require accurate reporting and fair authorship. The scholar also has a duty to participants who consented to the original sampling design. The 10-day deadline increases pressure but does not change the principles. The scholar should separate the two issues: data integrity and authorship.

First, document the sampling design and save a copy of the raw data and analysis script in a secure, timestamped location. Second, write a short memo to the supervisor explaining that the two districts were part of the stratified sample and that exclusion would create bias; propose instead to report the 42% figure with district-level breakdowns and a limitations paragraph. Third, propose a sensitivity analysis showing results with and without the two districts, so the supervisor sees the effect transparently. Fourth, discuss authorship in a separate meeting: ask what contribution the supervisor will make to the conference paper; if it is only supervision, propose acknowledgement or second authorship based on written contribution. Fifth, if the supervisor insists on exclusion, escalate to the head of department or the university's research integrity officer, keeping the memo trail. Do not delete data. The scholar can also consult the UGC regulations and a senior colleague for support.

If the supervisor insists, the scholar should not comply with data deletion. Recourse includes: (a) a written complaint to the head of department or dean of research, with the memo trail and sampling documentation; (b) contacting the university's research integrity officer or equivalent, even if the ethics committee is inactive; (c) using UGC's research integrity guidelines as the basis for the complaint; and (d) if the conference paper is submitted without consent, asking the conference organizers to investigate. The scholar should also protect the thesis by keeping the original analysis in the thesis and not allowing a biased conference paper to become a chapter. Possible consequences include delayed approval, but capitulating risks retraction, loss of credibility, and harm to participants. The scholar should seek advice from a mentor outside the supervisory relationship and keep all communication in writing. The goal is to resolve internally first, then escalate.