Turkish premier Recep Tayyip Erdoğan threatened a journalist over an exam row. Credit: Reuters/M. Azakir

Imagine sitting your university entrance examinations, only to learn the next day that the correct answers to many of the multiple-choice questions were signalled by a secret code in the exam paper itself.

For 1.7 million Turkish high-school students competing for a place at university, the bizarre scenario came true in March. Newspaper accounts of the cipher, and rumours that a privileged few had been told how to decode it, sparked student demonstrations across the country. Although the cipher may have been an accident of the test's design rather than a conspiracy, outrage has persisted and the debacle has become an issue in the run-up to the country's general election on 12 June.

Last week, prime minister Recep Tayyip Erdoğan accused a journalist investigating the case of campaigning to discredit the government. "He will pay a heavy price, not now but in the future," said Erdoğan. This is no empty threat — several investigative journalists were arrested in March as part of an inquiry into an alleged plot to overthrow the government.

Ankara's Chief Public Prosecutor's Office declared on 11 May that the cipher indisputably exists. It said it had found no evidence of cheating, but requested permission to investigate Ali Demir, head of the Student Selection and Placement Center (ÖSYM), the Council of Higher Education office responsible for the examinations. On 25 May, the council's president Yusuf Ziya Özcan said that it would probably not consider the investigation request until after the election.

In a country where corruption is rife, some educators see benefit in multiple-choice examinations because they can be graded by computer. In previous years, up to ten different versions of the exam paper had been printed with the same questions in a different order, to reduce the possibility of copying. Crucially, the easiest questions had always appeared first. This year, however, ÖSYM decided to create a unique version for each student by changing the orders of both questions and answers, commissioning an external company to create an algorithm for the task.

According to some academic experts, that algorithm may have inadvertently generated the cipher that eventually emerged for those questions with numerical answers. If the answers were rearranged in ascending order and placed next to the original list, the correct answer occupied the same position in each list. Enough questions followed the pattern that students could have passed the exam by relying entirely on this decryption method.

"It's not possible to tell" whether the cipher was accidental or deliberate, says Ersan Akyıldız, a cryptologist at the Middle East Technical University in Ankara. Either way, he is critical of the test design, noting that the reshuffling of questions put students who encountered more daunting questions at the beginning of the exam at a disadvantage. The ÖSYM team should have consulted academic experts in education and cryptology before trying such an ambitious experiment, Akyıldız says.

Many Turkish academics also complain that multiple-choice exams do not test reasoning skills well enough to be suitable for selecting university students. "We need to have real university entrance exams in Turkey, like the [international] baccalaureate," says chemical engineer Kemal Gürüz, a former president of the Higher Education Council, at Middle East Technical University.

The stakes are high. Turkey has a policy of widening access to higher education but has only 600,000 university places available this year. As a consequence, profitable cram schools have flourished — and have lobbied against change.