Mohamed Abdirahman Omar <br/><br/>
Estimating parameters in some ANCOVA Models with non- normal error distributions /  
 تقدير المعالم لبعض نماذج التغاير مع توزيعات الخطأ الغير معتدلة 
Mohamed Abdirahman Omar ; Supervised Ahmed Amin El-sheikh , Naglaa AbdelMoneim Morad  
 - Cairo :   Mohamed Abdirahman Omar ,   2014 
 - 98 P.  ;   30cm 
<br/><br/>Thesis (M.Sc.) - Cairo University - Institute of Statistical Studies and Research - Department of Statistics and Econometrics 
<br/><br/> According to the results of the simulation study the following points will  be introduced:  1) The model (3.14) just valid for and 2 while model (3.1) is valid  for any positive integer of .  2) For , and the estimates of the parameters of model (3.1) is better than those of model (3.14), compared to three different values for.  3) For , and the estimates of the parameters of model (3.1) is  better than those of model (3.14), compared to three different values for .  4) For , and the estimates of the parameters of model (3.14)  is better than those of model (3.1), compared to three different values for .  5) As sample size increase the Bias of the estimates of model (3.14) is  better than model (3.1), while the MSE of the estimates of model (3.1)  has better than model (3.14).  6) In model (3.14) the maximum value of r is 2, so in all cases the value of  Bias and MSE for r=4 not applicable.     
<br/><br/>
<br/><br/><label>Subjects--Index Terms: </label>ANCOVA  Estimating parameters  Normal error distributions