Abstract:

The main goal of the informatics Bebras challenge is to encourage children to learn informatics concepts, and to support development of computational thinking. Bebras contest takes place in almost sixty countries, but Slovakia is the only country that allows blind students to participate in the competition without the assistance of a sighted person. This is thanks to special categories for the blind. We offer a separate category for lower secondary blind students (aged 11 to 15) and another one for upper secondary blind students (aged 15 to 18). In our paper we will focus on the category of upper secondary blind students. This category was offered for the first time in 2016.

During the existence of this category, we have prepared 63 tasks for the blind. In our article, we focus on tasks that a certain number of contestants did not solve. There were 23 such tasks. We performed a detailed analysis of these tasks.

We used a case study research strategy. All the 23 unsolved tasks were observed cases. We observed the success rate of all blind participants. The average number of blind students who took part in the competition each year ranged from 2 to 10 students.

We were interested in which thematic area most of the unsolved tasks belonged to. In terms of subject area, the tasks could be divided into three categories: (1) data structures and tools,
(2) algorithms and problem solving,
(3) propositional logic. It turned out that most of the unsolved tasks belonged to category (2).

For each task, we found its measured difficulty Q as a percentage of incorrect and unsolved tasks and all responses. Tasks with a Q value less than 30 were easy, tasks with difficulty of 30 to 70 were medium and tasks with difficulty over 70 were hard. Almost 75% of the unsolved tasks were of medium difficulty.
Solutions of tasks are evaluated in such a way that for a correct answer the contestant gets points, for an incorrect answer he loses points and if he does not solve the task, he gets zero points. So, if the contestant is not sure that he can solve the problem, he may decide to skip it. When a competitor does not solve a task, we do not know if the reason is that he did not know how to solve the task or he did not solve it because the time ran out.

Contest participants often solved the tasks sequentially and so, if some of the previous tasks took longer time to solve, they may not have had enough time for the remaining tasks and may thus have not attempted them at all. We found that most of the unsolved tasks were in the second half of the task set. Considering that most of the unsolved tasks were of medium difficulty, it is likely that many tasks remained unsolved due to lack of time.

For each unsolved task, we measured the percentage of contestants who did not solve it. We found that more than half of the tasks were not solved by less than 20% of the competitors. Only one task was not solved by more than 50% of the participants. Its assignment will be given in the article. In the article, we will also indicate what knowledge and skills were needed to solve the largest number of unsolved tasks.

We believe that this analysis will help us to better select tasks, clearly formulate their assignments and estimate their difficulty in the future.

Keywords:

Competition, computing, Beras, unsolved tasks, blind pupils.