Hemant Sharawat & Ors. v. University of Delhi & Ors.

5. It is averred that on 25.08.2023, without any reason or basis, University issued the impugned Notification putting on hold the joining of the candidates, with the approval of the Competent Authority. By a further Notification dated 29.08.2023, University notified that matter of selection of candidates to whom offer of appointment were issued vide letters dated 18.08.2023, was under examination and therefore, University had decided to put the joining of all the candidates on hold, including of those who had reported. Candidates were advised to check the University website regularly for further updates in this regard. No reason was spelt out even in this Notification as to why the joining process was put on hold. Many of the candidates had by this time resigned from their existing jobs or received NOCs from previous employers and/or had become over age.

6. Aggrieved by the unjust and arbitrary action of the University, some of the Petitioners preferred representations to the University, but seeing no positive outcome, these petitions were filed.

COMMON CONTENTIONS ON BEHALF OF THE PETITIONERS:

7. Petitioners were eligible and duly qualified for the posts and on clearing the written examination on their own merit, they were offered appointments. There was no justifiable and plausible reason for placing the joining of the Petitioners on hold in the absence of any complaint with regard to the selection process. The impugned Notifications do not spell out a single reason which prompted the University to place the process under examination. Having taken an erroneous decision, University attempted to create material by constituting Committees to justify the said decision by a process of reverse engineering. Even before this Court, no material is placed which even remotely points to the alleged unfair means used by the Petitioners during the examination.

8. Admittedly, only 9 candidates had reported for joining in the category of Laboratory Attendant and 15 candidates in the category of Library Attendant when the impugned Notifications were issued at the initial stage. The justification given by the University in the affidavits filed before this Court is that informal interaction was held with these candidates and during the interaction, it was noticed that their competency did not match the high marks scored by them in the written examination, raising a suspicion of use of unfair means and thus the Competent Authority decided to put the process on hold. Till date, it is not disclosed as to who were the officials involved in the informal interaction and what was the nature of questions asked to test the knowledge, competency or intellect of these few candidates. In any event, neither the Scheme of Examination nor any Rules or Regulations of the University, permitted the University to have an informal interaction with the selected candidates to assess their knowledge or competency, as a somewhat second tier of selection process. This was clearly a motivated action to oust the selected candidates at the instance of the employees who were working on contract basis against these posts and were insecure of losing their jobs pursuant to regular appointments being made.

9. From the affidavits filed by the University, it comes forth that a Committee constituted on 06.04.2023 for the purpose of providing information about additional credits for employees working on contract was tasked to look into the matter of processing of results vide letter dated 08.09.2023, in light of some indications of use of unfair means. The Committee gave its report in 2023 on 17.10.2023 (‘2023 Report’) and recommended that candidates who had used unfair means be not allowed to join and NTA/University should take punitive action. The second Committee gave its report in 2024 (‘2024 Report’) concluding that the recruitment examination for both the posts was compromised and use of unfair means was established. It was recommended that the recruitment process and consequent appointments be quashed and process be started afresh. The conclusions by the Committees are erroneous and without actual proof of use of unfair means. Irrelevant considerations have gone into the decision making process such as: visible pattern indicating that toppers of the merit lists were from some specific centres; some candidates with marks in the range of 33-50% in Classes X and XII have scored more than 90% in the written test conducted by NTA; more than 100 candidates secured above 90% marks with 20 candidates obtaining over 95%, which was impossible without use of unfair means since the marking scheme included negative marking and the questions were from a wide range of subjects.

10. None of these factors can lead to a conclusion that the exam was compromised, especially in the absence of iota of evidence showing that unfair means were adopted. There can be infinite number of reasons for selection of more candidates from certain centres, such as higher number of candidates taking examination in such centres as a matter of sheer coincidence or because certain centres may be large to accommodate thousands of candidates whereas smaller centres may have a capacity of 50 or even less. This is not a peculiar phenomenon as in many other competitive examinations also, higher number of selected candidates are invariably seen either in one centre or in one urban area compared to another centre or rural area, especially where the urban areas are a hub of coaching centres. Invariably, students attending a particular coaching centre tend to fill forms at the same time and are allotted the same centre. In any event, in the present case, candidates were only required to select 03 preferred cities while filling up the forms and not examination centres, which were allocated by NTA, a neutral body, with no role of the candidates in making this choice. In fact, a lot of selected candidates were not allotted the examination centre in any of the 03 preferred cities. This could also be merely due to more meritorious candidates appearing in the exam in some centres.

11. The alleged suspicion of the University arising from the fact that candidates have scored higher marks despite negative marking, is completely misplaced. In various other examinations, candidates have scored 100% despite negative marking and this per se cannot lead to a conclusion that unfair means were used, in the absence of cogent and sufficient material to support this theory. Illustratively, the cut-off for the post of Forest Guard examination conducted by GNCTD in 2022 was 200 out of 200 for General Category. The top 3 successful candidates for the post of Canteen Attendant in the examination conducted by SSC in 2024 had scored higher than 95% marks and significantly, in both the examinations, the marking scheme included negative marking.

12. To support its plea of use of unfair means, University has also taken a position that there was 80% similarity in the answer choices inasmuch as not only the correct options but even the incorrect options of selected candidates at suspected centres matched, even though they were not from the difficult portions of the examination. These allegations are purely based on presumptions and conjectures arising from analysis of selective data containing responses of few selected candidates from few centres, which would invariably result in this pattern in most exams. This kind of methodology of ex post facto analysis of correct and incorrect answers was condemned by the Division Bench of this Court in Staff Selection Commission and Another v. Sudesh, 2014 SCC OnLine Del 7534, upholding the order of the Central Administrative Tribunal, wherein the Tribunal observed that show cause notices did not indicate details of malpractices committed by the candidates or the nature of copying indulged in as also the method or modus operandi adopted in the alleged copying. Tribunal had also held that in the absence of details of alleged malpractices committed by the candidates, the mere ipse dixit of the Staff Selection Commission that there was incontrovertible and reliable evidence, which had emerged upon post-examination scrutiny and analysis by outside experts, was insufficient. In the instant case, the 2024 Report is predicated on the analysis of the response sheets of the selected candidates through three methods and conclusions were drawn on probabilities, without a whisper on the unfair means that the candidates adopted at the centres. It was imperative for the University to prove the unfair means adopted in light of the stand of NTA that there was absolutely no breach of the security protocols involving frisking, installation of CCTV cameras, jammers etc.

13. It is preposterous for the University to contend that a candidate who does not secure high percentage in Classes X and XII or is not highly qualified can never perform well in any competitive examination and score high marks. It is not unknown that many students while studying for Classes X and XII Board examinations focus more on preparing for competitive examinations and it is also not unprecedented that many students with average marks in school, do exceptionally well in the competitive examination. As for educational qualifications, the advertisement, which is based on the recruitment rules of the posts, stipulated the minimum eligibility qualifications as 10th Class pass, which the Petitioners possess. The level of the questions in the examination was in keeping with the prescribed qualification and thus it cannot be a matter of surprise that some candidates scored high marks. In fact, as a matter of record, many Petitioners are Graduates or even Post-Graduates. Documents on record show that many Petitioners appeared in other competitive examinations conducted by NTPC, SSC, DSSSB, CET as also the Delhi University and were successful in some, which is a pointer to their high calibre, intellect and competency, contrary to the perception of the University.

14. There is no material, either before this Court or before the two Committees tasked to look into the allegations of unfair means, which evidences that unfair means were used in the centres. Indisputably, no unsuccessful candidate or third party made any complaint that there were any malpractices/copying/cheating/leakage of question paper in the present examination, leave alone at a systemic level. In Vanshika Yadav v. Union of India and Others, (2024) 9 SCC 743, the Supreme Court has delineated the following tests for determining the propriety of cancelling the examination: (a) whether the alleged breach took place at a systemic level; (b) whether the breach is of a nature which affected the integrity of the entire examination process; and (c) whether it is possible to segregate the beneficiaries of the fraud from the untainted students.

15. University is unable to show any breach of examination protocol at a systemic level and to the contrary, NTA’s stand fortifies the position that there was no compromise with the integrity of the examination. NTA has in no uncertain terms stated that there was no breach in security protocols during the conduct of examination. It has brought forth that: (a) question paper preparation and storage was done in a very secured manner; (b) randomisation of candidates was done during scheduling at examination centres; (c) randomisation of questions appearing on the screen of the candidates was ensured; (d) independent observers were deployed by NTA; CCTV cameras and jammers were installed and were functional; (e) multilevel biometric checks were done; (f) candidates were duly frisked; (g) CCTV control rooms were set up to get live feeds from exam centres; (h) virtual observers were deployed; (i) deployment of invigilators was ensured in high numbers; (j) responses of candidates were transferred by secured means from examination centres to NTA; and (k) result validation was done by an independent third party. With such stringent measures in place, there was no scope of any candidate using an unfair mean and which is why there was no complaint even from unsuccessful candidates, a regular feature in almost every competitive examination. University’s case has its basis only in conjectures, presumptions and probabilities and is wholly misconceived. In Roop Singh Negi v. Punjab National Bank and Others, (2009) 2 SCC 570, the Supreme Court held that suspicion or presumption cannot take the place of proof, even in a departmental inquiry, where an employee is facing a charge sheet.

16. During the course of hearing before this Court, University has been improving its stand by filing affidavits after affidavits, with a view to justify the unjust and unfair decision. In the first affidavit dated 21.03.2024, filed by the University, there was not a whisper of any suspicion that unfair means were used by candidates. The affidavit only refers to the power of the University to withdraw the offer of appointment at any stage, if any discrepancy is noted. Reference is made to the contents of the offer letters stating that the appointment was provisional and subject to further scrutiny by the University. In the second affidavit dated 22.07.2024, the stand in paragraph 10 is that when candidates reported for verification of records, certain issues were noted, which raised a suspicion of use of unfair means since the competency of the selected candidates judged on informal interaction by some officials did not match their respective scores in the written examination. It was stated that a Committee tasked to look into the issue of use of unfair means had on scrutiny of available data given an observation that there was concentration of toppers in few centres; nonselection of candidates from majority of the centres; nil or minimal representation from bigger States; non-appearance of names of candidates in the select list from major part of the country; and a wide gap between the knowledge of the candidates and their scores in Class X or XII and the marks scored in the examination. This affidavit was a marked improvement over the first affidavit dated 21.03.2024 in which there was absolutely no mention of the 2023 Report, which had given its recommendations by this time and possibly with which even the University was not satisfied.

17. Moreover, in the 2023 Report while there is an observation in para 4.1(i) that a visible pattern was noted that candidates in the select list were from few specific centres, there is no reference to the said specific centres. In para 4.3, the Committee clearly stated that details of the candidates such as their admit cards, centres, number of candidates in a centre, qualifications etc. were not provided. In the absence of all these details and on the basis of interaction with a miniscule number of candidates who had reported, it is not understood how the Committee came to a conclusion of a visible pattern in some centres and consequently a conclusion that unfair means were used in specific centres. While NTA has completely discredited the reports of the Committees that there was any malpractice in any centre, University has been unable to dent the stand of NTA.

CONTENTIONS ON BEHALF OF UNIVERSITY:

18. Executive Council of the University in its meeting dated 21.10.2020 decided to approach NTA for conduct of computer based written tests for recruitment of non-teaching posts of the University and after consent of NTA, University decided to engage the agency. NTA advertised the vacant posts vide Advertisement dated 23.02.2021 followed by subsequent corrigenda. Written test for post of Library Attendant was scheduled for 16.03.2023 and rescheduled to 20.03.2023 while that of Laboratory Attendant was scheduled for 18.03.2023. The entire process was carried out by NTA which included setting of question papers, selection of centres, deputing invigilators, providing computers, logistic support etc. with no intervention by the University. After conduct of written tests, on 18.04.2023, NTA sought the data of candidates engaged on contract/temporary/ad hoc basis to finalise the result, as additional marks were to be awarded in terms of the judgment of the Supreme Court in Civil Appeal Nos.1007-1008/2021. Required data was provided by the University and NTA was requested to expedite the results, which shows their bona fide attempt to finalise the result at the earliest.

19. As per NTA’s data, 7058 candidates had applied, out of which 2655 appeared in the written test and list of 108 candidates, selected for the post of Library Attendant was provided by NTA. Similarly, list of 151 candidates selected for post of Laboratory Attendant was also sent. However, details of academic qualifications and other particulars including contact details of the candidates etc. were not provided and University vide e-mail dated 10.07.2023 sought these details.

20. After lot of chasing up, NTA provided minimal data on 19.07.2023, pertaining to candidates selected in the merit list and when University compared each candidate’s particulars with confirmation sheet provided by NTA, it emerged that few candidates placed in the merit list under unreserved category were placed in their reserved category, as a result of which candidates who were declared successful in the reserved categories could not find place in the final merit select list. After compiling the merit list, a Notification was issued by the University on 18.08.2023 with the approval of the Competent Authority and offers of appointment were sent to 147 candidates in the category of Laboratory Attendant and 108 candidates in the category of Library Attendant, through speed post.

21. Laboratory Attendants who were offered appointments were to be posted in different departments of the University for deployment in different laboratories and in order to understand their interests and aptitude for deciding their postings, officials of the University informally interacted with the candidates who reported on 21.08.2023. This interaction was also necessitated for the reason that the offer letters were issued without documentary evidence of their educational credentials, dates of birth etc. Informal interaction revealed a huge gap in the basic understanding and knowledge of these candidates in the respective fields vis-à-vis their performance in the written test, wherein they had obtained high scores i.e. more than 90%. On further scrutiny of the documents provided at the time of joining, it was also observed that performance of these candidates in Classes X and XII was average or below average and there was a substantial time gap between the period when their schooling ended and the dates of written tests. This raised a serious apprehension that unfair means were used to obtain high scores. The matter was brought to the notice of senior officials of the University and looking at the gravity of the situation, the Vice Chancellor directed that joining of the candidates be put on hold and thus Notification was issued to this effect on 25.08.2023.

22. University of Delhi is a Public Funded Premier Central Educational Institution and is duty bound to recruit employees in a fair and transparent manner, keeping intact the highest standards of education. Laboratory and Library Attendants interact with and assist the students on regular basis in the colleges and if appointed, would have been employed at least for the next three decades. In these circumstances, it was imperative that their selection is on merit and fair and without recourse to using unfair means.

23. Later, with the approval of the Competent Authority, another Notification was issued on 29.08.2023 and uploaded on the University’s website for wider publicity, intimating that University had decided to put on hold joining of all the candidates including those who had reported in the University and the candidates were advised to check the University website regularly for further updates in this regard.

24. Petitioners cannot claim a vested right to appointment merely on the basis of offer letters. Clause 12 of Section ‘E’ of the Advertisement-cum- ‘Information Brochure & Guidelines for filling of Online Application form for recruitment of Non-Teaching Positions in Delhi University’, dealing with filling of online application forms clearly notified that in case of any inadvertent mistake in the process of selection, which may be detected at any stage, even after issuing an appointment letter, University will have the right to modify/withdraw/cancel any communication made to the applicant. Therefore, once it was found that candidates had used unfair means in the examination, University was well within its right to put on hold the joining of the candidates till the issues were resolved.

25. Committee constituted on 06.04.2023 for the purpose of providing information about additional credits to the employees working on contract, was asked to look into the issue of unfair means. Since NTA did not provide enough data regarding the categories, eligibility, centres of examination etc. of the candidates, information was collected from the selected candidates and collated by the Committee. Preliminary report of 2023 Committee revealed a visible pattern indicating that candidates in the select list and toppers of the merit list were from some specific centres and candidates scoring 33-50% marks in Classes X and XII scored more than 90% in the written examination. Committee also found that in the merit list more than 100 candidates scored more than 90% with almost 20 candidates scoring 95%, which was impossible, considering that the scheme of examination involved negative marking with questions from wide range of subjects. Cumulatively, these findings raised a suspicion of use of unfair means during the written test, which would in turn lead to the integrity of the examination being compromised.

26. NTA was the examination conducting agency and therefore, these observations of the Committee were forwarded to NTA, however, response of NTA vide letter dated 26.10.2023 was very cursory and without even adverting to the data furnished by the University, NTA denied the use of unfair means in the examination in a mechanical manner by merely harping that all standard procedures and protocols were followed. NTA’s communication disclosed that examination was conducted across 87 centres on pan India basis. Finding the response of NTA unsatisfactory, University constituted another Committee vide Notification dated 28.11.2023 and sought data from NTA vide e-mail dated 09.02.2024, which was provided as a link vide e-mail dated 16.08.2024 including question papers, response sheets of candidates, exam centre details and confirmation page of qualified candidates. The link could be accessed only on 22.08.2024 and on receipt of the data and other documents from NTA, Committee co-opted Dean, Faculty of Technology and Dean, Rankings of the University, for proper analysis and assessment of the data.

27. The data analysis was done on the basis of response sheets of the candidates provided by NTA, using standard technologies such as Microsoft Excel and Python and industry standard data analysis programming libraries such as Pandas, Matplotlib and Numpy. Additionally, visual inspection was also employed as a technique to manually confirm the pattern(s) inferred by data analytic techniques. No external data was used by the Committee. The suspicion was not based purely on better performance of the candidates in some centres but was also on account of the fact that there were matching patterns of responses of candidates from the suspected centres. The data from NTA was objectively analysed in respect of particular centres also not based on probabilities and likelihoods but based on direct calculation of matching responses of candidates.

28. Second Committee gave the 2024 Report rendering findings that: (a) similarity index for selected candidates from the specific centres was 80% and above; and (b) a pattern in correct as well as incorrect responses was discernible such that both correct and incorrect responses of selected candidates at suspect centres matched. This strengthened the belief of the University that the examination was compromised by use of unfair means.

29. NTA’s stand that the mechanism adopted by the University for data analysis is flawed, is baseless. A detailed analysis was carried out by the Committee based on centre-wise data, number of candidates selected, total number of candidates in the suspected centres and relative percentage selection with respect to total number of candidates who gave the examination at these centres. 13 centres where examination of Laboratory Attendant was conducted and 10 centres where examination of Library Attendant was conducted, were shortlisted basis this analysis. Response sheets of selected candidates were analysed using three methods i.e. visual inspection of colour coded responses; similarity calculation using response sheets; and visual similarity in incorrect answers. Under method-1, response sheets were visually analysed by colour coding their responses. Analysis revealed a very high amount of visual similarity in the responses of many selected candidates when listed centre-wise. Besides this, at least two suspected groups of candidates within at least 8 out of 12 centres were shortlisted as suspect.

30. In method-2, response sheets were analysed using ‘similarity metric’ calculated on the basis of responses against a question ID and groups of candidates within each centre were formed on the basis of a similarity threshold of 80%. It was found that the groups formed after percentage similarity calculations were exactly the same as the ones formed after the visual inspection of colour coded responses. In method-3, response sheets of all selected candidates were analysed using answer keys provided by NTA, basis the responses against a question ID and the analysis revealed a prominently visible pattern in the incorrect answers of the candidates from the suspected centres.

31. In light of these glaring revelations, University was well justified in putting on hold the joining of the selected candidates and no legal infirmity can be found in this decision. Selected candidates were given the option to undergo a fresh written test to prove their merit and suitability and that too only to undergo a qualifying test, but they refused, knowing that they will not clear the test on merit. University relied on the following judgments:

CONTENTIONS ON BEHALF OF NTA:

32. NTA conducted the recruitment examination for six different nonteaching posts of the University in CBT mode from 18.03.2023 to 21.03.2023 in two different shifts (09:00 AM to 12:00 Noon and 03:00 PM to 06:00 PM) at 87 examination centres in 36 cities across the country for 1,15,997 candidates. Out of 87 examination centres, 64 centres were used for conducting examination for the Post of Laboratory Attendant and 37 centres were used for the examination for appointments to the Post of Library Attendant. Eligibility criteria for the post of Laboratory Attendant was Class X pass or an equivalent examination with Science subjects from recognized Board and for Library Attendant was Class X pass or equivalent examination from any State Education Board or Government recognized Institution with certificate in Library Science/Library & Information Science from a recognized Institution. As per the marking scheme, the written test was objective type with MCQs and total marks were 300, allocated to different test components. As per the communication of the University dated 12.05.2023, additional marks were to be granted to employees of the University working on contract/daily wages/ad hoc basis in the University or its constituent Colleges.

33. Meeting of the Result Committee, comprising three Professors from the University and three officers from NTA, which included Director, NTA an IRS officer; Joint Director, NTA; and Deputy Director, NTA, was held on 01.07.2023 and result of Ist stage examination prepared by NTA was approved. The result was as follows:-

34. Significantly, prior to finalisation of results, the Unfair Means Committee (‘UFM’), comprising of one Professor being the Senior Advisor and one Professor from IIT, Delhi and Director, NTA, found only two cases of malpractices through impersonation and the Committee recommended cancelling of their candidatures. Result Committee accepted the recommendations of UFM Committee and candidature of these two candidates was cancelled. There were no allegations by the University at this stage of any unfair means being adopted in the centres and/or any possibility of remote access, hacking, malpractices at the suspected centres etc., the allegations which are now sought to be levelled, as an afterthought. Result of stage-I examination was declared on 04.07.2023 along with list of finally Selected Candidates in order of merit for the posts of Library Attendant and Laboratory Attendant and list of candidates shortlisted for Stage-II examination for the remaining four posts.

35. The allegations levelled by the University with respect to use of unfair means during the process of examination are absolutely baseless, misconceived and not supported with any material on record and it is evident that the desperate attempt of the University is to prove the nonexistent, by creating a cloud over the examination process. The examination was conducted in an environment where it was impossible for any candidate to adopt any unfair means or indulge in malpractices and none has been shown by the University.

36. The 2024 Committee constituted by the University used inappropriate tools of analysis and relied upon incomplete and selected data rendering the conclusions erroneous. The analysis and conclusions are based on probabilities, without any specific and tangible factors or evidence and the Report does not inspire confidence. Conclusions are extremely general in nature and there is complete mismatch with the data analysis carried out by NTA, which is an sexpert examination conducting body. NTA works on a zero tolerance policy to malpractices or use of unfair means etc. while conducting the examinations. NTA has successfully conducted large number of examinations in CBT mode such as JEE (Main), UGC Net, CSIR UGC Net, CUET (UG), CUET (PG), in a large number of cities across India and outside India for lakhs of candidates.

37. NTA adopted stringent and rigorous measures to ensure that integrity and sanctity of the examination is maintained and for this purpose, several security protocols were put in place before, during and after the examination in question in these petitions. There was no reported breach of a single security protocol in the conduct of the examination. Illustratively, the safety and security measures adopted by NTA were as follows:- • Randomised allocation of Exam Centres to the Candidates; • Randomised allocation of Seating/Computer Systems to the Candidates at Exam Centres; • Conduct of the Exam/CBT strictly through LAN in the Computer Labs, with no internet access thereto and only client-toserver communication was allowed, blocking connection with any other external third party connections; • Delivery of Question Paper (QP) to the Exam Centre from NTA HQ to the Local Servers in the Exam Centres through the Central Server of the CBT Delivery Agency in encrypted mode only, with Questions Paper/Response Sheet decrypted only at the time of display on the candidate's machine; • The sequence of the questions in the QP/Response Sheet was jumbled to make it different from candidate to candidate; • Secured transfer of Recorded Responses of the candidates in their respective QPs/Response Sheets to NTA through the Server of the CBT Delivery Agency; • Installation of password-protected software firewall on each client machine/computer system used by the Candidates for the exam to prevent any unauthorized application/remote access thereto and also to block unwanted ports; • Deployment of Invigilators at the Exam Centres by the CBT Delivery Agency in a ratio of 1:30 Candidates; • Deployment of Independent Observers (mostly from Academics, Assistant/Associate Professors from Universities/ Colleges), including Faculties from DU and its Constituent Colleges as well as from JNU, Punjab, Chandigarh, Himachal Pradesh and Pondicherry Universities; • Installation of CCTV Surveillance System at all Exam Centres for the close monitoring of the examination through the Virtual Observers, deployed physically at the Control Room in NTA HQ; • Installations of Jammers at all Examination Centres to prevent any use of mobile/communication device for malpractice; • Frisking of Candidates at the Exam Centres through third party Service Providers; • Biometric checks of candidates including Thumb Impressions and Face Recognition, • Scrutiny and evaluation of Answer Key Challenges by Subject Experts; and • Validation of Results by Independent Third-Party Service Provider for accuracy.

38. Significantly, University has neither been able to establish any breach in the aforesaid protocols nor placed any cogent material on record which would establish either that these protocols were not followed or there was any kind of lapse at the examination centres or during transfer of response sheets or during their evaluation. In fact, adherence to these protocols and set procedures is duly reflected in the reports given by independent observers deployed at the examination centres. No complaints of any kind least of all of malpractices or use of unfair means were received by the NTA during or post the examination, save and except, the convoluted stand of the University, desperately taken to prove that its arbitrary decision was flawless. Examination Delivery Agency for the examination verified and certified through a letter dated 30.11.2024, authored by the Chief Technology Officer and addressed to Director, NTA, that measures were undertaken to avoid any access to the systems through hacking or remote access, which belies the stand taken by the University at the fag end of the hearing that there was a possibility that the test computers were hacked.

39. NTA has carried out an independent analysis of the data, which does not even remotely suggest that the examination at the alleged suspected centres was compromised in any manner and/or that better performance of candidates at the suspected centres was due to adoption of unfair means. 2024 Committee concluded that certain candidates from the suspected centres had greater than 80% similarity in their answer choices, both correct and incorrect. Committee has clearly chosen only a few sample candidates’ responses and did not take into consideration the entire result to calculate the answering pattern. Visible pattern was that wherever selected candidates opted for incorrect options in an attempt to go nearest to the correct answers, the same incorrect options were opted for by a majority of the candidates who appeared in the examination. Illustratively, for the post of Laboratory Attendant (Shift-I) in centre Swami Parmanand Group of College, in respect of Question ID 201018, as per answer key option ‘2’ was the correct answer. Number of candidates who opted for option ‘2’ was 39 while 4 opted for option ‘1’, 6 for option ‘3’ and 29 for option ‘4’. Total responses were 78 and the pattern that emerged was as follows:-

40. In other words, question-wise percentage of incorrect responses by 2655 candidates for the post of Library Attendant, 4604 candidates for the post of Laboratory Attendant (Shift-I) and 3802 candidates for the post of Laboratory Attendant (Shift-II), demonstrates the same pattern. Illustratively in respect of Library Attendant for Question ID 201162, correct option was ‘3’. 495 candidates opted for option ‘1’ making it 18.64%, while 125 opted for option ‘2’ making it 4.71% and 66 opted for option ‘4’ i.e. 2.49%. This analysis shows that number of candidates who attempted incorrect response out of 2655 candidates were 686 i.e. 25.84%. Therefore, the contention of the University that there was a visible pattern of increased similarity in the incorrect answers of selected candidates alone is factually incorrect.

41. The possibility of the examination being compromised is also ruled out by the fact that the seating arrangements in all the centres were completely automated and randomised. All questions were jumbled and same question was not displayed on the computer screen of all candidates at the same time as also that question IDs had no correlation with the seating arrangements. The question papers did not follow the same series or patterns and were prepared after randomisation of questions for different candidates, thereby changing the sequence of the questions.

42. The examination centres were allotted randomly with no set pattern. The invigilators were deployed in the ratio of 1:30 i.e. overall 3867 invigilators were deployed for the entire examination process involving all the posts in question. Additionally, 387 observers were also deployed, of which 18 were faculties from colleges of the University. The examination was conducted over a secured LAN network, blocking all external connections and ensuring encrypted delivery of question papers. Sequence of questions were randomised and candidates’ responses were securely transmitted to the main server. Password protected firewalls were installed on all clients machines and biometric checks with strict monitoring through CCTV ensured transparency. This position is fortified by the reports of independent observers, which do not bring forth any adverse comment on the conduct of the examination.

43. Moreover, contrary to the stand of the University, tabular representation of centre-wise data showcases the number of allotted candidates, attended candidates and the selected candidates in two shifts and when percentage is calculated on the basis of total number of candidates at a particular centre, it does not support the University’s stand that there was an exceptional number of candidates selected from any one centre. University has adopted a flawed method of calculating by taking a number of three selected candidates from each centre and termed it as ‘suspected centre’, irrespective of the total number of candidates who undertook the examination and therefore, the percentages on both counts worked out by the University are erroneous.

44. Looking at the timeline/difference of the time lag in many cases, which spans over 40 minutes in many cases, the allegation of candidates attempting questions by copying are completely ruled out. Analysis by NTA also reflects the time lag in attempting incorrect questions by the candidates and when this is seen holistically with randomised seating arrangement and the fact that question papers had different series or patterns, there can be no doubt that no unfair means could have been adopted by the candidates.

45. University highlighted 13 centres in Laboratory Attendant examination and 10 in Library Attendant examination, wherein at least 3 candidates were finally selected from 64 examination centres for the post of Laboratory Attendant and 35 for the post of Library Attendant. After conducting a thorough examination of the candidate audit logs by NTA, it was observed that in Laboratory Attendant examination, 8406 candidates appeared of which 151 were selected resulting in selection rate of 1.80% and no candidate was selected from 36 centres out of 64. In the 13 centres, a total of 4418 candidates appeared of which 134 were selected resulting in overall selection rate of 3.03%. Selection rate varies significantly across these centres, with the lowest being 0.50% and highest reaching 13.16%. For the Library Attendant examination conducted across 35 centres, 2655 candidates appeared of which 108 were selected leading to selection rate of 4.07%. Also, no candidate was selected from 14 centres. With the sub-set of 10 centres that University flagged as suspicious, 2197 candidates appeared for the examination and 93 were selected, resulting in a slightly higher rate of 4.23%. The rate varied widely amongst these centres with the lowest being 2.05% and the highest at 42.86%. Therefore, it cannot be said that the selection rate at the allegedly suspected centres was abnormally higher compared to a pattern that was visible in respect of all centres.

46. Three methods for analysis are stated to have been followed by the University. Each of these analysis overlooks various crucial issues. Consistency observed in colour coded responses can be attributed to the fact that each question ID is linked to a unique answer. The evaluation process failed to consider the audit logs, which provide insights into the actual duration candidates spent on answering the questions and their seating arrangement. Audit logs for a sample of 8 candidates divided into four groups, from three different centres were scrutinized to validate the timing of their identical same responses and analysis revealed that same answers were provided with a time difference exceeding 40 minutes. The second method of similarity analysis has been done by dividing the total number of matching responses by total number of attempted questions, overlooking that candidates may have attempted a varying number of questions, which could lead to variations in the calculated percentage of similarity.

47. In the third method of analysis pertaining to incorrect same responses, NTA conducted an analysis focusing on the incorrect responses in the case of Laboratory Attendants carried out on the top 5 centres with the highest number of selected candidates, using two approaches. Approach ‘A’ included identifying 5 candidates based on highest number of same incorrect answers and each of the candidates was compared against another set of 10 candidates, who took the examination at the same centre and a huge time difference was found centre-wise. In Approach ‘B’, 5 candidates were identified, who had the highest number of incorrect answers and for these question IDs were examined and the incorrect responses were compared to those given by all candidates in the centre. It was found that roll numbers of no selected candidate were in a consecutive sequence thereby, ruling out any possibility of cheating.

48. The stand of the University that selected candidates had average or below average scores in Classes X and XII leading to a conclusion of unfair means to obtain high marks, is misconceived. Firstly, there is no rule that a candidate scoring average or below average marks in Class X or Class XII cannot perform well in a competitive examination and secondly, even on a factual note, the submission is incorrect inasmuch as 73% of candidates applying for Library Attendant and 67% of candidates applying for Laboratory Attendant achieved scores exceeding 60% in Class XII and most of the candidates were either graduates or post-graduates albeit the required educational qualification was Class X and certainly, these qualifications acquired after school added to the knowledge and calibre of the candidates.

49. University is also not correct in its stand that hardly any question was left un-attempted by candidates from suspected centres which is strange considering that the marking scheme had negative marking. The data indicates that many candidates had left several questions un-attempted. Illustratively, candidate at Roll Number DL01151004 gave 112 correct, 17 incorrect answers and left 21 unattempted. Interestingly, in the extensive hearing before the Court and in exercise undertaken to point out the alleged unfair means used, University has been unable to place iota of material or evidence which even remotely suggests what unfair means were used by selected candidates and if so, how.

50. Heard learned counsels for the parties and examined their respective submissions.

51. Before delving into the merits of the case, it is imperative and useful to look at the law laid down by the Supreme Court in matters pertaining to cancellation of examinations, both from the perspective of the scope of judicial interference as also the parameters and facts and circumstances in which Courts may or may not interfere.

52. On the aspect of judicial review, it will be apposite to refer to the judgment of the Supreme Court in Tata Cellular v. Union of India, (1994) 6 SCC 651, wherein it was held that judicial review is concerned with reviewing not the merit of the decision but the decision making process itself and is thus different from an appeal. Where selection or rejection is arbitrary, certainly the Court would interfere. In Chairman, All India Railway Recruitment Board and Another v. K. Shyam Kumar and Others, 2010 SCC OnLine SC 579, the Supreme Court reiterated this principle as also the law that basis of the judicial review could be highlighted under three principal heads namely, illegality, procedural impropriety and irrationality. Illegality means that decision maker must understand the law correctly that regulates his decision making power and must give effect to it. Grounds such as errors of law and/or fact, taking into consideration irrelevant factors and ignoring the relevant ones, acting in bad faith, fettering discretion etc. fall under ‘illegality’. Procedural impropriety is where mandatory procedures such as breach of natural justice, absence of bias etc. are not followed. Ground of irrationality takes in Wednesbury unreasonableness propounded in Associated Provincial Picture Houses, Limited v. Wednesbury Corporation, [1948] 1 K.B. 223. Wednesbury applies to a decision which is so reprehensible in its defiance of logic such that no sensible person who applied his mind to the issue would have arrived at the decision. Leyland and Anthony in Textbook on Administrative Law, 5th Edn., OUP, 2005, proposed as follows:- “Proportionality works on the assumption that administrative action ought not to go beyond what is necessary to achieve its desired results (in everyday terms, that you should not use a sledgehammer to crack a nut) and in contrast to irrationality is often understood to bring the courts much closer to reviewing the merits of a decision.”

53. The Supreme Court has from time to time elucidated principles delineating the scope of judicial interference in matters relating to examination and selection processes. In Sachin Kumar (supra), Supreme Court referred to the earlier decisions on the subject over the last five decades wherein the question was whether the examination process was vitiated and the resultant consequences and/or the reliefs that can be given in a particular case. It was held that the answer to the question whether process of an examination stands vitiated or not essentially lies in finding out whether the irregularities in the process have taken place at a systemic level so as to vitiate the sanctity of the process. One end of the spectrum are cases which border upon or cross over into the domain of fraud as a result of which credibility and legitimacy of the process is denuded and in such a case the authority concerned may well take a decision to cancel the examination in entirety as the process loses its legitimacy. Here, the decision need not turn upon a fact finding exercise into individual acts involving use of malpractices or unfair means and it may be difficult to segregate tainted from untainted participants. However, at the other end of the spectrum are cases where some participants are guilty of irregularities and segregation of tainted and untainted may be possible as this exercise would find embodiment in the Constitutional duty by which public bodies have to act fairly and reasonably.

54. In Anamica Mishra and Others v. U.P. Public Service Commission, Allahabad and Others, 1989 SCC OnLine SC 414, the Supreme Court was examining the issue of cancellation of the examination process involving recruitment to various posts in the educational services of the State of Uttar Pradesh. It was held that the cancellation of the entire process was not justified as the case was representative of a situation where there was no systemic flaw in the written test. In Union of India and Others v. Rajesh P.U., Puthuvalnikathu and Another, (2003) 7 SCC 285, the Supreme Court held that in the absence of any specific or categorical finding supported by any concrete and relevant material that widespread infirmities of an allpervasive nature which could be really said to have undermined the very process itself in its entirety or as a whole, there was hardly any justification in law to deny appointments to selected candidates whose selection were not found to be vitiated.

55. In Inderpreet Singh Kahlon and Others v. State of Punjab and Others, (2006) 11 SCC 356, the Supreme Court emphasised compliance of three principles at the hands of the State: (a) to establish satisfaction in regard to sufficiency of material collected so as to enable the State to arrive at its satisfaction that the selection process was tainted; (b) to determine the question that illegalities committed go to the root of the matter which vitiate the entire selection process and such satisfaction as also sufficiency of material were to be gathered by reason of a thorough investigation in a fair and transparent manner; and (c) whether the sufficient material present enable the State to arrive at a satisfaction that majority were found to be part of fraudulent purpose.

56. In Vanshika Yadav (supra), the Supreme Court held that cancellation of an examination is justified only in cases where the sanctity of the exam is found to be compromised at a systemic level i.e. the standard of evidence on record should indicate systemic malaise. It was also held that Court may direct cancellation of an exam or approve such cancellation by the Competent Authority only if it is not possible to separate the tainted candidates from the untainted ones.

57. I may now refer to the judgments where examination process was held to be vitiated and the facts and circumstances obtaining therein as this would shed light on the gross facts in which Courts have interfered and held that the examination was vitiated. In Sachin Kumar (supra), the Supreme Court noticed that the Committee had found deficiencies of a systemic nature which casts serious doubts on the legitimacy of the entire recruitment process. The starting point of the case was receipt of serious complaints in the office of Chief Minister, GNCTD regarding conduct of examination by DSSSB for the post of Grade-II DASS and a Committee was constituted comprising of Director, Vigilance and District Magistrate to inquire into the matter. The Committee found on preliminary examination that: (a) as against 62056 applicants, only 8224 had appeared in Tier-I examination as adequate information had not been furnished to the candidates; (b) candidates had secured high marks in Tier-I but extremely low marks in Tier-II;

(c) concentration of candidates from a particular area of Delhi and a domination by a particular section of the society based on their surnames;

(d) members of same family were found to be sitting in close proximity both in Tier-I and Tier-II exam; (e) absence of randomisation in the seating arrangement; (f) inability of the candidates to access the internet to download the e-admit cards; (g) racket leading to impersonation of candidates involving a person who was identified as Chief Invigilator at a particular centre and who was connected with a coaching centre involved in leakage of question papers; (h) this very person had repeatedly fixed his duties in a choice of his own centres with the help of DSSSB staff;

(i) videography was blurred; (j) thumb impressions were unrecognisable;

(k) jammers were not working properly; (l) candidates were allowed to appear without looking into their educational qualifications; and (m) prima facie evidence that flying squad members were passing answers to candidates. It is in this backdrop that the Supreme Court held that recruitment to public services must command public confidence and where entire process is found to be flawed, the decision to cancel the examination cannot be faulted with.

58. In K. Shyam Kumar (supra), the Supreme Court took note of the report of the Vigilance Department that several candidates were suspected to have obtained answers to the questions a few hours before the exam through a middleman who had accepted bribe. 62 candidates were alleged to have impersonated and at least 6 candidates had adopted unfair means. Investigation prima facie revealed leakage of question paper to a sizeable number of candidates. Decision of the Railway Recruitment Board was upheld on the ground of the process being vitiated due to leakage of question papers, large scale impersonation and mass copying.

59. In State of Tamil Nadu and Another v. A. Kalaimani and Others, (2021) 16 SCC 217, the allegations were of large scale malpractices in the written examination involving tampering of OMR sheets. After reevaluation, discrepancies were found in entries pertaining to 196 candidates, who were beneficiaries of fraudulent alteration of marks. Division Bench of the High Court held that fabrication of record pertained only to 196 candidates and since segregation was possible, entire examination need not be cancelled. Supreme Court reversed the decision on the ground that many people had benefitted due to tampering of OMR answer sheets and on a deeper scrutiny, sufficient material was found against 196 persons who were clearly beneficiaries of the fraud in alteration of marks. More material was being unearthed as the investigation was in progress and several people had been arrested.

60. From a reading of the aforesaid judgments, it is clear as day that purity of an examination process is unquestionable and where there are allegations of occurrence of large scale malpractices, fraud, leakage of paper, mass copying, exchange of bribe etc., in the course of conduct of any examination process, the State or its instrumentalities are certainly entitled to cancel the examination. It is equally settled that the view taken by the recruiting authority must be a bona fide view based on sufficient material before it, leading to a conclusion that the entire process stands vitiated. Of-course, there may be situations, as held by the Supreme Court in a number of judgments, where candidates who have indulged in irregularities can be identified and there is a possibility of segregating tainted and untainted candidates. This too must be a considered and bona fide decision of the recruiting agency. Pithily put, there must be material on record justifying cancellation of the entire selection/examination process.

61. Before delving into the rival contentions of the parties, it is useful to refer to the trajectory of this case. Genesis of this case lies in Notifications dated 25.08.2023 and 29.08.2023, whereby joining of selected candidates, who had offer letters in their favour for appointments, was put on hold. The first Notification was issued basis an ‘informal interaction’ by University officials with candidates who had reported before this date. The stand in affidavit dated 22.07.2024 is that their competency did not match the respective marks scored by them in the written examination. In the next affidavit dated 25.11.2024, it is stated by the University that candidates who received their offers of appointment started reporting to the University from 21.08.2024. Laboratory Attendants who were offered appointments were required to be posted in different Departments for deployment in the laboratories and therefore to understand their interest and aptitude to decide their postings, officials of the University informally interacted with these candidates and it was found that there was a huge gap in the basic understanding and knowledge in the respective fields of these candidates vis-à-vis their performance in the written test where they scored as high as more than 90% marks. There was also a gap in the marks obtained by them in Class X and Class XII and the written examination. This procedure, virtually amounting to introducing a second step in the selection process, to say the least, is untenable in law, for multiple reasons.

62. Recruitment to the posts in question was regulated by a Scheme of Examination clearly spelt out in the Advertisement-cum-‘Information Brochure & Guidelines for filling of Online Application form for recruitment of Non-Teaching Positions in Delhi University’. The eligibility conditions with respect to age, educational qualifications etc. were also prescribed. For both the posts, the minimum educational qualification was Class X pass with additional certificates as per the requirement of the posts in question. Perusal of the advertisement shows that for the posts of Laboratory Attendant and Library Attendant, only an objective written test with MCQs was prescribed. Manner and mode for drawing up the final merit list was also provided in the Advertisement, as per which minimum overall qualifying marks for the written test were 45% for unreserved posts, 40% for posts reserved for OBC and 35% for SC/ST/PwBD categories. The marking scheme did not envisage a second tier in the selection process in the form of interview/interaction/personality test/skill test etc. for the two posts in question. Wherever required or contemplated, it was specifically mentioned in the Advertisement, against the specific posts that written test will be followed by interview/interaction/personality test. The question that begs an answer is whether University was entitled to resort to an informal interaction with 9 candidates in the category of Laboratory Attendant, who joined pursuant to receipt of offers for appointment, not just for postings but for assessing their knowledge and competency and the answer can only be a clear ‘No’. This process is completely in the teeth of the Advertisement and the marking scheme postulated therein. It was not open to the University to introduce a second tier of selection, which was not mentioned in the Advertisement and which finally became the cause of putting the whole selection process to a halt.

63. Strangely, despite multiple affidavits filed by the University, it has not come to light as to who were the officials who interacted and where. There is no disclosure on the nature of interaction and/or the questions put to the candidates to assess their knowledge, competency and calibre and how the answers to the questions led to the assumption that unfair means were allegedly used by these 9 candidates in the examination. Moreover, it is not understood how the alleged incompetency of merely 9 candidates could lead to a conclusion that the other selected candidates, who had not even joined, were incompetent or lacked knowledge enough to obtain higher scores in the examination. Petitioners are right in flagging that if this interaction became the basis of the serious decision to put on hold the joining of selected candidates, the process of interaction ought to have been docketed in the records of the University with the names of the officials who interacted, but as a matter of fact, University never placed any material on record to establish this interaction, assuming it was a valid exercise. Therefore, whether or not there was any such informal interaction is itself suspect.

64. The entire process of the alleged informal interaction, besides being alien to the scheme of examination, also gets shrouded in a mystery for the reason that till the date of the first Notification dated 25.08.2023, there was no complaint pertaining to the conduct of examination by NTA, either by any unsuccessful candidate or a third party, during or after the examination. The selected candidates had cleared the written test and were in receipt of offers of appointment. On joining, the only exercise that remained was verification of their documents pertaining to educational qualifications, age proof, caste certificates etc. There was neither any occasion nor reason for the University officials to enter into the exercise of informal interaction to assess the competency of selected candidates. The stand that the interaction was for deciding the postings in various laboratories, is also an afterthought. Petitioners were qualified and selected candidates and there was no requirement of testing anyone’s interest or aptitude to post in any laboratory. This procedure gives strength to the allegation levelled by the Petitioners in writing as also during the course of hearing repeatedly, that the decision to put on hold the joining of selected candidates was motivated at the instance of contract employees working in the University who were naturally insecure with regular appointments being made and were behind this motivated action.

65. There is, in fact, a danger in endorsing this kind of an informal interaction, assuming there was one. If it is left open to employers to informally interact with selected candidates and judge their calibre, knowledge or competence, at the time of reporting for joining, the selection process will become subjective and open doors to arbitrariness, pick and choose, favouritism, making the examination meaningless, besides breeding corruption. Any selected candidate could then be thrown out on the basis of an informal interaction.

66. Coming back to the chronology of events, moving ahead with this weak foundation, University proceeded to task an existing Committee constituted on 06.04.2023 to look into the issue of use of unfair means by the candidates in the examination. Admittedly, at this stage, the Committee was not privy to complete data from NTA. The analysis of the Committee was based on information received from the candidates to whom offer letters were issued. The 2023 Report shows that the analysis was Centre-based. Committee looked into the number of centres at Delhi, Chandigarh and Shimla and the number of selected candidates and concluded that concentration of toppers in the merit list was from few centres and there was no or minimal representation from bigger States like U.P., Bihar and Rajasthan. As the report shows, the Committee was also influenced by the results of the selected candidates in the Classes X and XII finding a wide gap between the marks obtained by the candidates in these classes and those scored in the written examination. To the Committee it appeared to be an impossible proposition that more than 100 candidates had secured more than 90% marks with almost 20 obtaining 95% marks especially when there was negative marking. All this led to a conclusion that unfair means had been used in the written examination and a recommendation was made that the joining of the selected candidates be put on hold.

67. As the chronology goes, University constituted another Committee on 28.11.2023 and after NTA provided the requisite data, Committee rendered the 2024 Report, relevant part of which is as follows:-

68. Broadly understood, according to the University, the analysis of the reports of two Committees revealed that: (a) there was concentration of the toppers in the merit list from few specific centres; (b) non-selection of candidates from majority of the centres; (c) nil or minimal representation from bigger States such as U.P., Bihar, Rajasthan; (d) wide gap between knowledge of the candidates and marks obtained by them in the written tests; (e) wide gap in the percentage of marks secured in Classes X or XII ranging between 33 to 50% and marks in the examination in question i.e. 90% or more; (f) more than 100 candidates secured more than 90% in the examination with almost 20 candidates obtaining almost 95% marks; (g) impossibility of obtaining high scores in an examination with negative marking consisting of questions from a wide range of subjects; (h) similarity index for selected candidates from specific centres being 80% and above;

(i) visible pattern of correct and incorrect responses; (j) same questions answered incorrectly being not from the Sections where the questions are generally considered difficult; and (k) no question being left unattempted by candidates from suspected centres. This, according to the University, raised serious suspicion of unfair means being used by the candidates during the conduct of the examination.

69. The entire case of the University is predicated on data analysis by the 2024 Committee, constituted by the University to look into various facets of the examination process, which allegedly gave rise to a suspicion of use of unfair means. To demonstrate the data analysis, University has placed on record the data in the form of excel sheets colour coded to indicate the marks scored, percentages, ranks etc. In the common additional affidavit filed towards the fag end of the hearing, it is explained that data analysis was done using standard technologies such as Microsoft Excel and Python and industry standard data analysis programming such as Pandas, Matplotlib and Numpy including visual inspections. University categorises ‘suspected centres’ as those centres from where at least three candidates were finally selected with a rationale of choosing three as the least number of candidates to highlight that systemic cheating/unfair means at a centre must at least affect the result of three candidates belonging to the centre. As per the affidavit, there were 21 suspected centres for both the posts, which have been detailed in the common affidavit. As per the University, the analysis outcome of the visual inspection of the response sheets of all selected candidates by colour coding their responses was that there was a very high amount of visual similarity in the responses when listed centre-wise and at least two suspected groups of candidates within at least eight out of twelve centres had similar responses.

70. By the second method, the Committee analysed the response sheets of selected candidates using a similarity metric calculated on the basis of responses against a question ID and groups of candidates within each centre were formed on the basis of similarity threshold of 80% and the outcome was that groups formed after percentage similarity calculations were exactly the same as those formed after visual inspection and multiple response sheets from the same centre had even more similarities. The third method adopted by the University was by taking the response sheets of all selected candidates and analysing them using answer keys provided by the NTA basis the responses against a question ID and coming to a conclusion that there was a visible pattern in the incorrect answers of the candidates in the suspected centres in terms of similarity which did not appear to be in the unsuspected centres. The conclusion drawn was that an extremely high majority of the highest scoring candidates came from the pool of candidates who had given their exams in the suspected exam centres while pool of candidates who had given their exams in the unsuspected centres had considerably law values of grand totals and percentages as also that candidates from the suspected centre had greater than 80% similarity in their answer choices, both correct and incorrect.

71. As rightly flagged by NTA, the data analysis by the Committee was restricted to what the University terms as ‘suspected centres’ and ‘selected candidates’ and therefore, by travelling on a wrong path, University reached the wrong destination. Suspected centres, as explained in the affidavit dated 17.02.2025, were centres where at least 03 candidates were finally selected, the rationale of choosing three as the least number of candidates being to show a systemic cheating/unfair means pattern. As further stated in the affidavit, 13 centres for the examination of Laboratory Attendant and 10 centres for the examination of Library Attendant were shortlisted, basis this analysis. With this restricted data, University proceeded to analyse the same by three methods, as aforementioned. The data analysis exercise reveals that having taken an erroneous decision at the initial stage to put on hold joining of the selected candidates, possibly under the extraneous influence of the Karamchari Union, University resorted to a process of reverse engineering to create material to support its plea of use of unfair means.

72. It is pertinent to mention at this stage that neither from the data analysis, which is placed on record nor from the affidavits filed from time to time by the University, it is discernible as to what unfair means were adopted by the selected candidates and/or what was their modus operandi to obtain the high scores. The conclusions drawn are purely hypothetical, based on percentage analysis of number of suspected centres, number of selected candidates, their scores in Classes X and XII, geographical locations of the centres etc. The onus was clearly on the University to place on record some material to reach a conclusion that the process of examination was compromised, which it has failed to discharge. This missing link becomes significant when one looks at the categorical stand of NTA that security protocols were scrupulously observed, both during the conduct of the examination and subsequent thereto, for transferring and evaluating the responses of the candidates. It is equally pertinent to note that University has not succeeded in denting the case of NTA that there was absolutely no breach of any protocol at any centre.

73. It is crucial to capture the security protocols put in place by NTA, at the cost of repetition, as follows:- • Randomised allocation of Exam Centres to the Candidates; • Randomised allocation of Seating/Computer Systems to the Candidates at Exam Centres; • Conduct of the Exam/CBT was strictly through LAN in the Computer Labs, with no internet access thereto and only client-toserver communication was allowed, blocking connection with any other external third party connections; • Delivery of Question Paper (QP) to the Exam Centre from NTA HQ to the Local Servers in the Exam Centres through the Central Server of the CBT Delivery Agency in encrypted mode only, with Questions Paper/Response Sheet decrypted only at the time of display on the candidate's machine; • The sequence of the questions in the QP/Response Sheet was jumbled to make it different from candidate to candidate; • Secured transfer of Recorded Responses of the candidates in their respective QPs/Response Sheets to NTA through the Server of the CBT Delivery Agency; • Installation of password-protected software firewall on each client machine/computer system used by the Candidates for the exam to prevent any unauthorized application/remote access thereto and also to block unwanted ports • Deployment of Invigilators at the Exam Centres by the CBT Delivery Agency in a ratio of 1:30 Candidates; • Deployment of Independent Observers (mostly from Academics, Assistant/Associate Professors from Universities/ Colleges), including Faculties from DU and its Constituent Colleges as well as from JNU and several other Universities; • Installation of CCTV Surveillance System at all Exam Centres for the close monitoring of the examination through the Virtual Observers physically deployed at the Control Room in NTA HQ; • Installation of Jammers at all Examination Centres to prevent any use of mobile/communication device for malpractice; • Frisking of Candidates at the Exam Centres through Third Party Service Providers; • Biometric checks of candidates including Thumb Impressions and Face Recognition; • Scrutiny and evaluation of Answer Key Challenges by Subject Experts; and • Validation of Results by Independent Third-Party Service Provider for accuracy.

74. To support this plea, NTA has shown to the Court the reports of the independent observers deployed at the examination centres and has been able to demonstrate that there were no adverse reports of any malpractice/ use of unfair means at any centre. The format of the observer’s report shows that a detailed questionnaire was required to be filled by the observers by answering ‘Yes’, ‘No’ or ‘Satisfactory’. Illustratively, in one of the observer’s report, the relevant questions were: (a) whether the candidates were randomly allocated to their nodes; (b) was there proper partitioning between each computer system to avoid looking into the screen of the other;

(c) whether clear instructions were given for not carrying mobile phone or any other electronic gadget at the main entrance; (d) whether candidates were frisked at the entrance as per prescribed norms; and (e) whether only one gate was used for entry and other gates, if any, were locked.

75. As rightly flagged by NTA, no complaints of any kind were received in respect of the present examination. NTA also relied on communication from Chief Technology Officer of the Firm which was technical partner of NTA for the conduct of examination certifying that there was no evidence of hacking/remote access in the examination bringing to light that: (a) an EXE was installed on all test computers individually, with a view to disable any third party applications including remote desktop applications; (b) Hardware Firewalls were used to create a Virtual Private Network (VPN) to secure the network traffic, block unauthorized access to test computers from outside and to ensure that only authorized systems access the exam server via MACbinding; (c) Software Firewalls were used on each Test computer to block all the ports other than the port required to communicate to the local server so that other computers cannot talk to the test computers; and (d) continuous monitoring of the examination process was done by the Technical Team of the said Firm from its headquarter in Bengaluru as well as NTA control room in Delhi. This position is completely uncontroverted by the University, save and except, subtly stating that the possibility of hacking/remote access cannot be ruled out.

76. With these rigorous security protocols in place, NTA rightly urges that the examination in question was not susceptible to any unfair or malpractices by the candidates by emphasising on the factum of randomised allocation of seating/computer systems to the candidates and also highlighting that it is not the case of the University that the candidates in any centre were sitting in a sequence where any of them were either known to each other or related in any manner. Once there was no breach of the aforementioned protocols and the candidates had no opportunity to adopt any unfair means, it is not understood on what basis these serious allegations were levelled against the selected candidates leading to putting their joining on hold. As noted above, University is not able to demonstrate the modus operandi of the candidates to cheat, with all the stringent security protocols of NTA in place.

77. The data analysis exercise is nothing but a guise to cover up the illegal and arbitrary decision of the University to stall the joining of selected candidates and does not inspire any confidence. For the sake of completeness, I may briefly deal with the contentions of the University on this aspect, conscious of the fact that in judicial review this Court cannot go into the merit of the decision as an appellant Court but equally conscious of the fact that the decision making process and its rationality, is open to examination to prevent miscarriage of justice to the individuals involved and affected by the decision.

78. The contention that there was concentration of toppers in the merit list from few specific centres or that there was non-selection of candidates from majority of the centres albeit may be correct on a factual note but cannot per se lead to an inference of use of unfair means, sans sufficient material on record pointing to a compromise at systemic level in these centres. Moreover, it is not uncommon in any competitive examination that number of selected candidates may be more from some centres as compared to other centres, which may be co-related to the seating capacity of the centres, number of candidates appearing in those centres, location of the centres in rural or urban areas, where the latter may be hubs of coaching centres. This phenomenon could also be a result of candidates attending a particular coaching centre in which case ordinarily they tend to fill up the forms at the same time and are allotted the same centres.

79. This plea of the University is completely baseless in the present case since Petitioners have taken a categorical position that they had only selected three preferred cities while filling up the forms and were not entitled to select examination centres which were allocated by NTA and this is fortified by NTA and not traversed by the University. In fact, Petitioners have also stated that a lot of selected candidates were not allotted centres in any of the three cities preferred by them, a fact again uncontroverted. Moreover, NTA has brought forth that after conducting a thorough examination of the candidate audit logs by NTA, it was observed that in Laboratory Attendant examination, 8406 candidates appeared of which 151 were selected resulting in selection rate of 1.80% and no candidate was selected from 36 centres out of 64. In the 13 centres, a total of 4418 candidates appeared of which 134 were selected resulting in overall selection rate of 3.03%. Selection rate varied significantly across these centres, with the lowest being 0.50% and highest reaching 13.16%. For the Library Attendant examination conducted across 35 centres, 2655 candidates appeared of which 108 were selected leading to selection rate of 4.07%. Also, no candidate was selected from 14 centres. With the sub-set of 10 centres that University flagged as suspicious, 2197 candidates appeared for the examination and 93 were selected, resulting in a slightly higher rate of 4.23%. The rate varied widely amongst these centres with the lowest being 2.05% and the highest at 42.86%. Therefore, it cannot be said that the selection rate at the allegedly suspected centres was abnormally higher compared to a pattern that was visible in respect of all centres.

80. The submission that there was nil or minimal representation from bigger States such as U.P., Bihar and Rajasthan is negated by NTA by placing before the Court data, which indicates that candidates were selected from Bihar, Rajasthan, U.P., Delhi, Haryana, Kerala, Madhya Pradesh, Manipur, Odisha, Punjab, Tripura, West Bengal, Maharashtra, Jharkhand and Andhra Pradesh.

81. It was contended on behalf of the University that there is a wide gap between the knowledge of the selected candidates and marks in Classes X and XII on one side and marks obtained in the written examination by NTA. This contention only deserves to be rejected. It is not uncommon that many candidates may not have secured exceptionally high marks in Classes X and XII but have performed exceptionally well in the competitive examinations. The minimum eligibility educational qualification for the present examination for the posts of Laboratory Attendant and Library Attendant was Class X and many Petitioners are Graduates or even Post-Graduates. The examination was notified in 2021 but was held in 2023 i.e. after one and a half year. Seen holistically, the time available to the candidates for preparing the examination together with their educational qualifications, one cannot reach a conclusion that having scored 60 to 70% marks in Class X years earlier, candidates had no scope of improvement. The argument is merely presumptive.

82. There is also no merit in the plea of the University that it was impossible to obtain high scores over 90% in an examination with negative marking consisting of questions from a wide range of subjects. Petitioners have rightly flagged that in many other examinations, candidates have scored 100% despite the marking scheme including negative marking. Illustratively, it was pointed out that cut-off for the post of Forest Guard examination conducted by GNCTD in 2022 was 200 out of 200 for the general category and the top 3 successful for the posts of Canteen Attendant in an examination conducted by SSC, 2024 had scored higher than 95% marks. Once the University has failed to prove use of any unfair means, this convoluted process of questioning the calibre of selected candidates to obtain high scores, is unacceptable.

83. Much emphasis was laid by the University on the similarity index for selected candidates from specific centres being 80% and above. The plea is that there is a visible pattern of correct and incorrect responses being the same for selected candidates even though questions answered incorrectly are not from sections which were difficult to answer. The very foundation of this data analysis and conclusion is shaky for the reason that University has only taken into consideration selected candidates and that too, from limited suspected centres to form the data base. NTA, on the other hand, has analysed with a broader data base. The data sheets filed on record show that analysis is made question-wise first by taking number of candidates attempting incorrect response out of 108 selected candidates for the post of Library Attendant and then comparing the percentage of incorrect responses out of 108 with percentage of incorrect responses out of 2655 candidates. Similar exercise is done for 86 selected candidates for the post of Laboratory Attendant by comparing the percentage of incorrect responses out of 86 and percentage of incorrect responses out of 4604. Thereafter, the analysis is of question-wise percentage of incorrect responses by 2655 candidates by taking the number of candidates who attempted incorrect responses and working out the percentage with a similar exercise for question-wise percentage of incorrect responses by 4604 candidates for the post of Laboratory Attendant and working out the percentage of incorrect responses. This exercise was in respect of Shift-I and with different data base, similar exercise was carried out for Shift-II as well.

84. NTA has demonstrated that question-wise percentage of incorrect responses by 2655 candidates for the post of Library Attendant, 4604 candidates for the post of Laboratory Attendant (Shift-I) and 3802 candidates for the post of Laboratory Attendant (Shift-II), reveals the same pattern. Illustratively in respect of Library Attendant for Question ID 201162, correct option was ‘3’. 495 candidates opted for option ‘1’ making it 18.64%, while 125 opted for option ‘2’ making it 4.71% and 66 opted for option ‘4’ i.e. 2.49%. This analysis shows that number of candidates who attempted incorrect response out of 2655 candidates were 686 i.e. 25.84%. Therefore, the contention of the University that there was a visible pattern of increased similarity in the incorrect answers of selected candidates alone is factually incorrect.

85. NTA has shown the incorrect responses did not match only amongst the selected candidates at the suspected centre but also matched with the selected candidates in the non-suspected centres as also that the percentage of incorrect responses within the same centre and within the same shift was fairly close, which demolishes the star argument of the University that the incorrect responses of the selected candidates in the suspect centres were matching and therefore the probability of use of unfair means. The two tables used by NTA to explain its stand, prepared question wise, are extracted hereunder, for ready reference:-

86. The analysis by NTA completely defeats the theory of University that the incorrect responses of the selected candidates at the suspected centres only were matching, leading to an inference of use of unfair means by them. NTA has also negated the plea of the University that no questions were left un-attempted. It has placed on record a detailed data showing a number of questions left un-attempted. Illustratively, candidate at Roll Number DL01151004 gave 112 correct, 17 incorrect answers and left 21 un-attempted.

87. Another important facet of this case, rightly highlighted by NTA is that a Result Committee was constituted, comprising of three Professors from the University and three Officers from NTA. The University members included Director, DUCC, OSD (Examination) and Joint Dean (Students Welfare) and members from NTA were Director, NTA, an IRS officer, Joint Director, NTA and Deputy Director, NTA. The Result Committee held its meeting on 01.07.2023 and results of first stage examination were declared wherein 108 candidates were selected for the post of Library Attendant and 151 for Laboratory Attendant. 2324 candidates were shortlisted for Stage-II examination for the post of Assistant and 5870 + 30 for the post of Junior Assistant/Junior Assistant (Store). In addition, an UFM Committee was also constituted comprising two Professors, one of whom was from IIT, Delhi and Director, NTA. This Committee found only two cases of impersonation and recommended cancellation of the candidature, which was accepted by the Result Committee. Even the recommendations of these two Committees have been given a go by the University, proceeding forward on the basis an informal interaction on 21.08.2023.

88. NTA has also demonstrated through log details that there was a time lag between same questions answered by different candidates and in some cases, the time lag spanned over 40 minutes ruling out any use of unfair means and this crucial fact is completely overlooked by the Committee constituted by the University. It is brought out that University’s evaluation process failed to consider the audit logs, which provide insights into the actual duration candidates spent on answering the questions and their seating arrangement. Audit logs for a sample of 8 candidates divided into four groups, from three different centres were scrutinized to validate the timing of their identical same responses and analysis revealed that same answers were provided with a time difference exceeding 40 minutes. The second method of similarity analysis has been done by dividing the total number of matching responses by total number of attempted questions, overlooking that candidates may have attempted a varying number of questions, which could lead to variations in the calculated percentage of similarity.

89. In the third method of analysis pertaining to incorrect same responses, NTA conducted an analysis focusing on the incorrect responses in the case of Laboratory Attendants carried out on the top 5 centres with the highest number of selected candidates, using two approaches. Approach ‘A’ included identifying 5 candidates based on highest number of same incorrect answers and each of the candidates was compared against another set of 10 candidates, who took the examination at the same centre and a huge time difference was found centre-wise. In Approach ‘B’, 5 candidates were identified, who had the highest number of incorrect answers and for these question IDs were examined and the incorrect responses were compared to those given by all candidates in the centre. To illustrate this position, several charts were handed over by NTA showing the log details. Pertinently, University did not even attempt to respond to this contention during the course of hearing.

90. This Court is unable to find any flaw in the conduct of the examination on any aspect which could lead to a conclusion that the process is vitiated. No doubt, Clause 12 of Section ‘E’ of the Advertisement-cum- ‘Information Brochure & Guidelines for filling of Online Application form for recruitment of Non-Teaching Positions in Delhi University’, for filling the online application forms notified that in case of any inadvertent mistake in the process of selection, which may be detected at any stage, even after issuing an appointment letter, University reserved the right to modify/ withdraw/cancel any communication made to the applicant and/or that the offer letters stipulated that the offers were provisional, but the facts and circumstances obtaining in these cases, as noted above, do not warrant cancellation of the examination and selected candidates deserve to be appointed.

91. Heavy reliance was placed by the University on the judgment of the Supreme Court in Sachin Kumar (supra). The judgment is wholly inapplicable to the present cases on facts. In the said case, as noted above, there were serious complaints in the office of Chief Minister, GNCTD of irregularities. Candidates securing high marks in Tier-I examination scored extremely low marks in Tier-II examination. As a matter of fact, Committee found that the mark list was dominated by particular section of the society based on their surnames. There was no randomisation of seating arrangements resulting in factually members of the same family sitting in close proximity to each other in both Tier-I and Tier-II examination. Committee also noted a racket leading to impersonation of candidates, which involved an identified person who was the Chief Invigilator at a particular centre connected with a coaching centre and was involved in leakage of question paper. He repeatedly fixed his duty in a choice of centre with the help of DSSSB staff. Besides, as a matter of fact, the Committee found that videography was blurred, thumb impressions were unrecognisable, jammers were not working properly and possibly flying squad members were passing answers to candidates. None of these factors even remotely exist in the present cases.

92. For the same reason, the judgment in Varun Bhardwaj (supra) is distinguishable. In the said case, as a matter of fact, candidates were caught using mobile phones while attempting to answer the question papers and many of them confessed to using mobile phones/SMS texting the answers even though they were sitting in different centres. In the said case, there was no discussion of the security protocols during the conduct of examination as is in the present case, which is a very crucial aspect of this case. Insofar as the judgment of the Division Bench in Manish Dabas (supra) is concerned, it does not aid the University. As a matter of fact, in the said case, use of unfair means was found where two candidates were caught by invigilators at the respective examination centres and mobile phones were recovered. One of the two candidates had confessed that the last SMS message received on his mobile phone consisted a set of numerals and it was proved that the modus operandi was that keys of the answers to questions in the QPB were received clandestinely in the form of sets of five digit numbers. The Scrutiny Committee revealed that there was a unique pattern of bunching of results/ranks. It was observed that in case of bunched cases not only the total marks were the same but the marks secured in different segments also tallied and the occurrence of such bunching could occur only when there was a common source of information. In the present cases, it is not even the stand of University that any gadget/mobile phone/blue tooth etc. was recovered from any candidate in any suspected centre or that the jammers or other security protocols were non-functional.

93. Having given a thoughtful consideration, this Court is of the view that University has miserably failed in establishing use of any unfair means in the examination in question and what comes to light is that by a process of reverse engineering for motivated reasoning, by analysing a skewed data the University is seeking to illegally justify its erroneous decision to shelve the process of selection, which cannot be accepted. The process of hindsight analysis adopted by the University by picking up a limited data of selected candidates and attempting to match the incorrect responses is a dangerous path to follow as by this process, any and every competitive examination will become vulnerable, even in the absence of any material showing use of unfair means and/or the modus operandi adopted by the candidates.

94. Accordingly, impugned Notifications dated 25.08.2023 and 29.08.2023 are quashed and set aside. University is directed to complete the remaining formalities of document verification etc. and take the selection process to its logical end. Upon appointment, Petitioners will be permitted to join their respective places of postings forthwith and will be entitled to all consequential benefits. W.P. (C) No. 73/2024

95. Petitioners in this writ petition were shortlisted for Tier-II examination for the posts of Assistant and Junior Assistant (Store), which they were precluded from doing because the process was put on hold. In view of the findings above, University is directed to proceed with the Tier-II examination and permit the Petitioners to participate in the same. Further course of action for these Petitioners will depend on the result of the Tier-II examination.

96. All the writ petitions are allowed and disposed of in the aforesaid terms along with pending applications.

97. Before drawing the curtains, I may pen down that because of the arbitrary and illegal action of the University, Petitioners have lost nearly two crucial years of their lives and careers. Some of the Petitioners had in fact resigned from their earlier jobs when offer letters were received from the University and many have become overage for appearing in any other examination. This is a classic case of scant regard for fairness in action and ‘motivated reasoning” and the University must introspect!!! JYOTI SINGH, J MAY 30, 2025/KA/Shivam