ICC - Item Characteristic Curve / item responce function (IRF)
PCC - Person Characteristic Curve
T - person proficiency (model range [-5; +5] logits)
B - item difficulty (model range [-5; +5] logits)
A - item discrimination (model range [0; 3])
C - item pseudo-guessing (model range [0; 1]) -- the lower asymptote of IRF
DD - attentiveness, as oppose to careless error probability (model range [0; 1]) -- the upper asymptote of IRF
4PL - Four parameter (B, A, C, DD) logistic model
P - p-value, "item difficulty" in CTT
Didx - item discrimination index (dichotomous model) where N is total number of persons, Ng is group size (both low and up);
Nup - number of correct answer in group of Ng best test scores;
Nlow - number of correct answer in group of Ng worst test scores
D - raw test score, that corresponds to proficiency, equal to B
Rir - correlation of item’s score with total of all other items scores where xi - item "i";
yi - total test score minus xi (total of all items but item number "i");
N - sample size (number of test takers),
Sx, Sy - standard deviations of x and y respectively
dA - Chronbach’s alpha delta for item "i". Reflects contribution of item "i" into test reliability.
dA = (A - Ai) where A - Chronbach’s alpha of complete test (all items);
Ai - Chronbach’s alpha of test made of all items but item number “i”
B0 - initial (#0) iteration difficulty
B - centered Initial difficulty (B0 corrected by average difficulty and average proficiency)
