The study aimed to reveal the effect of the maximum odds method to treat missing values and the percentage of loss (10%, 15%, 20%) on the suitability of items and individual coefficients according to the parameter quadrant model. To achieve the objectives of the study, the researcher used the non-verbal reasoning test prepared by Mik Bryon (2012). And their fit according to the parameter quadrant model, and then using the addition of kutool for excel program, responses were deleted (10%, 15%, 20%) and then compensated for the missing values according to the method of maximum likelihood and extracting the individual coefficients, their accuracy and suitability according to the parameter quadrant model. The results showed that:
 1- That the assumptions of the response theory are fulfilled, and the indicators of one-dimensional assumptions (explained variance, vocabulary saturation and indicators of local independence) are increasing.
 2- There are no statistically significant differences in the proportion of appropriate vocabulary between the complete data on the one hand and the complementary data at the percentage of loss (20%) and (15%), and the results also showed the absence of statistically significant differences in the proportion of appropriate vocabulary between the complementary data between the proportion of Loss (10%) and (15%), there are statistically significant differences in the percentage of appropriate vocabulary between the complete data on the one hand and the complementary data on the percentage of loss (10%), and the results showed that there are statistically significant differences in the percentage of appropriate vocabulary between the complementary data Between the percentage of loss (10%) and (20%), and between the percentage of appropriate vocabulary between the percentage of loss (15%) and (20%).
 3- There are no statistically significant differences in the value of the coefficient of difficulty and vocabulary guesswork and its accuracy according to the percentage of loss, and there are statistically significant differences in the value of the coefficient of discrimination and the coefficient of lack of interest and its accuracy according to the percentage of loss, and to find out the sources of the differences, the researcher conducted an examination test, and the results showed that there are differences Statistically significant in the discrimination coefficients between the complete data on the one hand and the complementary data at the loss percentage (10%) and (20%) on the other hand. Distinguishing the complementary data at the percentage of loss (15%) and (20%), and there are statistically significant differences in the disinterestedness transactions between the complete data from the complementary data at the loss percentage (10%). Attention to the percentage of loss (10%) on the one hand, and the discrimination coefficients for the complementary data at the percentage of loss (15%) and (20%), and there are statistically significant differences in the accuracy of the discrimination coefficients between the complete data and the complementary data at the loss ratio (10%). The results also showed that there were statistically significant differences in the accuracy of the coefficients Discrimination at the percentage of loss (10%) on the one hand and the accuracy of the discrimination coefficients for the complementary data at the percentage of loss (15%) and (20%). The results also showed that there were statistically significant differences in the accuracy of the disinterest coefficients at the loss percentage (10%) on the one hand, and the discrimination coefficients for the complementary data at the loss rate (15%) and (20%).