I need an expert with Rasch analysis.

I have short data with 100 cases only. I have done reliability analysis using alpha coefficient and correlation using Pearson correlation. then I had to do correction for attenuation. however the correction lead to overestimation of the correlation because my reliability was below 0.70. when the paper was submitted to publication the editors recommend using other way to calculate for correlation and correct it and he suggest reading the reference. (http://pareonline.net/getvn.asp?v=8&n=11). I think we need to do structural equation models or Rasch analysis as suggested by this reference.

I will explain to you breifly the study and what i have done. Basically we are comparing the scores of students in Multiple choice questions and fill in the blank question.
Here what i have done
1- i had to caculate the relibility of each exam using Apha coefficient .
2- to caculate the corrlation i used have used the pearson correlation formula
3- to correct to correlation i had to divid the obesrved correlation by the relibilty of the two exams

I have submited the paper for publication but the editors asked me to look at this refernce and use other method for calculating my relibility and corrlations becuase i have relibility below 0.6