Aspirants who failed to get the jobs they wanted in the forest department have banded together to demand that the Gujarat Subordinate Service Selection Board dismantle the long-standing Computer-Based Recruitment Test (CBRT).

The CBRT was held in February to fill 823 Forest Guard vacancies. Its results were announced July 31.

The protesters, who gathered in Gandhinagar on Aug 1, cited issues with the language of the test, technical glitches, variations in difficulty levels and the method used to process scores as flaws in the system.

They said that the CBRT contains a high number of language errors. “It seems like no one at the private agency hired to programme the test, which is in Gujarati, knows Gujarati. So they use computer applications to translate the questions and don’t cross check them,” one aspirant said, adding, “Meaning often gets lost in translation.”

Candidates are also not happy with the varying difficulty levels across different shifts of the same exam. “Some papers are very easy and some are very difficult,” the aspirant said. Another blamed technical glitches for costing them time and morale, leading to frustration and an inability to “justice to each question in the allotted time”.

However, the biggest thorn in their side is the Mean Standard Deviation Method of Normalization  used to calibrate the merit list.

Notably, a statement from GSSSB Secretary Hasmukh Patel, dated July 3, explained that multi-session exams would be held in cases where the number of candidates is very high. Since each of these sessions uses a unique question paper, the Normalization method is used to know how each candidate’s score relates to the average score, using a common scale.

How Normalization works

Say, for instance, five candidates score 70, 80, 90, 85, and 75, respectively. The mean, or average, in this case is 80. So Candidate A, with 70 points, is -10 points away from the average. Candidate C is +10 points away from the average; Candidates D and E are +5 and -5 away from the mean, respectively. So the typical difference, or standard deviation, is + or - 7.5 points from the mean of 80 marks.

To get the normalized score, you subtract the average score and divide the result by the standard deviation. So the normalized score for the student with 90 points would be +1.33, while it would be -1.33 for the student with 70 points.

Critique and ‘solution’

Protesting candidates say that Normalization does not give an “accurate” idea of scores.

“Normalization does not allow for merit to be measured accurately. It also does not maintain benchmarks, making it very harmful and unfair,” the aspirant said.  “We want to know the actual marks of each candidate before and after Normalization, as seen in other exams like SSC, CGL, IBPS, RRB. But the GSSSB only publishes only a list of names. The Board should publish the merit list along with all the information about how many marks have been added, how many marks have been reduced and every information should be published category wise and marks wise.”

The candidates have a simple solution: Have a single paper for each examination and conduct it offline.

“If offline exams are good enough for organizations such as the Gujarat Public Service Commission and Police Recruitment Board, why aren’t they good enough for GSSSB?”