SIT finds lapses in evaluation of Group-I Mains answer scripts

The SIT had found serious procedural violations in the evaluation of answer scripts of exams for the Andhra Pradesh Public Service Commission (APPSC) Group-I Mains notification 27/2018, including the evaluation of the same scripts in three stages. It also confirmed violation of evaluation rules and found that answer scripts were evaluated three times at Digital, Haailand Resorts and Vijayawada.

The SIT, headed by the AP CID Director General Ravishankar Ayyanar, submitted its final compliance report in a sealed cover to the AP High Court on August 10, following directions of a Division Bench. The report is a follow-up to the interim report submitted on June 29.

The SIT recorded statements of 173 witnesses and examined forensic reports from seven Central Forensic Science Laboratory (CFSL) centres in Hyderabad, Delhi, Pune, Bhopal, Shimla, Kolkata and Guwahati. It examined 165 selected candidates, 161 unsuccessful candidates close to the cut-off, and 408 writ petitioners.

The investigation agency informed the High Court that changes were detected in the answer scripts of 160 of the 165 selected candidates (96.97%), while 153 of the 161 unsuccessful candidates (95.03%) had alterations.

‘Evaluation done sans APPSC Chairman’s nod’

Among 408 writ petitioners, 405 scripts (99.26%) showed changes. The SIT also found correction-fluid traces on 24.26 per cent of slips in the disputed Haailand Resorts evaluation.

The report said digital evaluation was undertaken without the APPSC Chairman’s approval, and a contract worth `1.38 crore was awarded to TrueGrader/Stone Media Advisory without a tender.

Evaluators were allegedly selected through WhatsApp, while Gmail and Google Drive were used for official work. The SIT also detected discrepancies between marks recorded in Excel sheets and the final database, besides 287 barcode mismatches.

The manual evaluation conducted at Haailand Resorts between December 5, 2021 and February 26, 2022 was also probed.

Answer scripts allegedly remained with a private agency for 84 days without the Chairman’s approval. The SIT said 66 unqualified data-entry operators were used instead of the officially selected 452 operators.

A third manual evaluation was conducted in Vijayawada between March 25 and May 19, 2022.

The SIT found no documented justification from the Chairman for the exercise, and said reprinting new or duplicate OMR barcode sheets effectively resulted in re-evaluation.

The report noted that APPSC rules did not empower chief examiners to alter marks. However, the SIT distinguished genuine corrections arising from totalling errors from other alterations.

The SIT further found that a list of 57 approved examiners was leaked through WhatsApp before evaluation began, with 47 participating later. Though no direct evidence showed that the leak affected evaluation, it was termed a serious breach of confidentiality.

The report said CCTV systems were absent during the initial evaluation period, and mobile phones were used despite restrictions.

The referral rate in the third evaluation rose to 10.2%, compared with 0.45% cent in the 2016 notification.

It also clarified that an earlier suspicion regarding barcode swapping involving two candidates was caused by physical mixing of documents while transferring them from the APPSC to the SIT for forensic examination, rather than manipulation during evaluation.

The SIT concluded that evaluating the same answer scripts three times appeared prima facie contrary to Rule 3(ix) of the APPSC Rules of Procedure, which prohibits re-evaluation and permits only recounting within 15 days.