Duplicates, Impossible Values, and a Five-Day Deadline
A faculty member in a funded project discovers problems in a dataset of 240 respondents collected by a PhD scholar. Fourteen rows are exact duplicates of earlier rows, and nine records contain impossible values: ages of 12 and 96 in a sample of college students, and monthly income recorded as ₹0 for respondents who reported spending ₹8,000 on a device. The PhD scholar says the duplicates are “just data entry” and wants to delete them, keep the nine records, and submit to a conference whose deadline is in five days. The senior co-author, who is also the department head, says the funder will not release the next instalment unless a paper is submitted, and proposes adding the funder’s programme officer as a co-author “for support.” The faculty member must decide what to do before the deadline and how to document it.
1. Are the duplicates and impossible values research misconduct? What should be done before any submission?
Read the worked answer
The duplicates and impossible values are data-quality problems on their face, not automatically misconduct. Exact duplicates may arise from repeated form submissions or data-entry errors; impossible ages and income values may be entry errors or unit errors. Misconduct would arise if the researcher knowingly retained fabricated or falsified data, or deleted or altered records to deceive. The scholar’s proposal to delete duplicates without a record and to “keep” impossible values is therefore unacceptable as a process even if the underlying errors are innocent.
Before submission, preserve the original file, assign unique IDs, verify suspect records against source forms or respondent contact where possible, and maintain an audit log of every correction or exclusion. If verification is impossible, exclude affected records with a documented rule and run sensitivity analyses. The paper must report the cleaning procedure and any resulting limitations. A five-day deadline does not justify skipping verification; it may justify submitting later or to a different outlet.
2. Evaluate the proposal to add the funder’s programme officer as a co-author and the pressure created by the funding instalment.
Read the worked answer
Authorship requires substantial contribution to conception, design, analysis, drafting, or critical revision, plus approval of the final version and accountability. “Support” by a programme officer, without evidence of such contribution, does not meet the criteria. Adding an honorary author is a breach of publication ethics, and the fact that the funder controls the next instalment makes the proposal coercive rather than collegial. Funding should be acknowledged in a funding statement, not converted into authorship.
The senior co-author’s pressure is also not a valid justification. The faculty member should ask for a written account of the officer’s contribution; if none exists, refuse the authorship offer. If the head insists, the faculty member should escalate the matter rather than accept a compromised author list.
3. What documentation and escalation steps should the faculty member take, and how should the PhD scholar be treated?
Read the worked answer
The faculty member should write a contemporaneous memo describing the anomalies, the verification steps taken, the proposed corrections, and the authorship request. Keep the raw data and the cleaned file separate, with version control and a data-cleaning log. Communicate in writing with the senior co-author and, if necessary, the institutional research-integrity or ethics committee. If the funder’s officer is being added improperly, the funding agency’s research-integrity contact may also need to be informed. The aim is not to punish the scholar but to establish what happened and prevent a false submission.
The PhD scholar should be treated as a developing researcher: explain the difference between error, questionable practice, and misconduct; involve her in verification; and record her response. If she was pressured by the senior co-author, that context matters. The immediate decision, however, is clear: no submission until the data are verified and the authorship issue is resolved.