PUBLISHED ONAugust 05, 2026 9:32 AMBYLim KeweiWas the Singapore Examinations and Assessment Board (SEAB) suggesting there had been an occasion of "collective hallucination" among hundreds of candidates who reported discrepancies during the O-level English oral examination on July 15?
That was the question raised by Workers' Party MP Jamus Lim in Parliament on Wednesday (Aug 5), prompting a sharp response from Senior Minister of State for Education Janil Puthucheary.
"I would ask him to perhaps not put words in my mouth," responded Dr Janil.
Dr Janil had earlier told Parliament that SEAB had conducted a thorough investigation into allegations made by 432 candidates from 102 schools that there was a discrepancy between the text prompt they were shown during the preparation stage and the actual prompt read by examiners on July 15.
@asiaone "I would ask him to perhaps not put words in my mouth about collective hallucinations, nor imply a privileging of one set of data over the other,' said Senior Minister of State for Education Janil Puthucheary. More on Dr Janil's exchange with Workers' Party MP Jamus Lim. #sgnews #Singapore #Parliament #OLevel ♬ original sound - AsiaOne
SEAB reviewed its e-system, digital logs, and conducted cybersecurity checks in investigating the matter. It found no technical fault.
The senior minister of state also said there was "significant variation" of the supposed discrepancy recalled by different students.
SEAB ultimately concluded that all oral examiners and candidates had accessed the same examination material on July 15 and found no evidence of a system malfunction, and that the incidents were "not technical in nature, nor the result of a deliberate or malicious act".
Picking up on SEAB's conclusion that there were no technical issues, Assoc Prof Lim asked whether the board was implicitly suggesting that the students who reported discrepancies had experienced a form of "collective hallucination".
"If so, does this mean that the investigation would have perhaps implicitly privileged computer forensic evidence over the contemporary testimony of a significant number of human witnesses?" he asked.
In response, Dr Janil said the investigation had examined the candidates' testimonies as well, noting that students had recalled a variety of different prompts.
The topic for the O-level oral exam that day was regarding street barbers. The prompt asked candidates: "Who do you think would go to a place like this for a haircut? Why do you say so?"
Different phrases recalled by students in their accounts include "Would you visit?", "Would you be interested?" and "Would you cut your hair?"
Dr Janil said to Assoc Prof Lim: "We're looking at both the testimony, as well as the technical assessment. And the fact that the testimonies covered a wide variation in the phrases in itself is a very significant data point."
"So I would ask him to perhaps not put words in my mouth about collective hallucinations, nor imply a privileging one set of data over the other."
WP's Kenneth Tiong then probed if SEAB examined the candidates' contemporaneous planning sheets and if an external investigation into the issue will be conducted.
Dr Janil revealed that in some cases, the information from the contemporaneous planning sheets indicated that the student had misread the prompt and later corrected themselves.
"And so, on one sheet, we can see having the student having written down what they thought originally was the prompt, and then crossing that out, and then writing the correct prompt, and then planning from there," he said.
According to him, the Government Technology Agency (GovTech) has also conducted an audit of the technical platform and processes used in the exam and reached the same conclusion as SEAB.
Multiple users on Reddit forum SGExams who claimed they sat for the oral exam on July 15 had earlier claimed to encountering an "error" which showed them a different text prompt when they were preparing their oral response.
After reports emerged that SEAB had found no discrepancy in the examination, some forum users questioned how multiple students could have apparently experienced the same "hallucination".
Some also alleged that the prompt they were shown on the screen had "changed" midway through their 10-minute preparation time.
@asiaone Preliminary data shows that the mean score for the English O-level oral examination on July 15 is comparable to that of the other four sessions in 2026, said Senior Minister of State for Education Dr Janil Puthucheary. #sgnews #Singapore #Parliament #OLevel ♬ original sound - AsiaOne
[[nid:741964]]
lim.kewei@asiaone.com
