A few important notes about raw data: 1. Stimuli names are derived from Serbian: - TXT are the continuous units - SLK is the noncontinuous unit - N1-N3 are L1-L3 difficulty levels - informative reading screens are marked with "-info", with the exception of L2 continous item (TXTN2), which was not marked with "-info", but it can be differentiated from the question screen by using the number of trial (Trial 005 is the informative reading screen) 2. A software bug caused the programme to freeze during calibration after the first stimulus while recording two participants, with codes "011" and "nema sifre". Because the freezing happened during calibration, there was no data loss, and a new session was launched to collect answers to other reading units (sessions "011 II" and "nema sifre IV"). For these participants, one should look at the data from the original session for the first stimulus, and the data from the new recording sessions for the other stimuli. 3. The participant with the code "BATMAN" mistakenly pressed SPACE earlier on the L2 continuous item, and the participant "neki lik 2002" did the same with the L1 continous item - data from their replies to these stimuli was not looked into