Low item separation with item reliability below 0.90 indicate that the sample is not large enough to reproduce the item difficulty hierarchy, while negative or zero PTMEA correlations imply that the scale items function in an opposite direction to the Rasch dimension (Fan & Bond, 2019; Green, 2013). The evaluation of the response categories was based on the guidelines by Fan and Bond (2019) and Linacre (2002a): a) at least 10 observed counts for each category, b) monotonic increase in the average category measures, c) outfit MNSQ lower than 2, d) category thresholds advance monotonically and in the range of 1.4–5.0 logits, and e) distinct peak for each category in the category probability curve.