Skip to main content

Table 8 Results of McFadden-Hausman IIA test / East African Countries’ sample.

From: A study of the factors affecting mobile money penetration rates in the West African Economic and Monetary Union (WAEMU) compared with East Africa

 

Coefficients in Eq. 1 of M1 = Coefficients in Eq. 1 of M2

Coefficients in Eq. 2 of M1 = Coefficients in Eq. 2 of M2

Coefficients in Eq. 3 of M1 = Coefficients in Eq. 3 of M2

M1 versus M2 without alternative 1

Chi 2

 

0.7200

1.3800

Prob > chi2

 

1.0000

0.9999

M1 versus M2 without alternative 2

Chi 2

0.8100

 

4.6200

Prob > chi2

1.0000

 

0.9694

M1 versus M2 without alternative 3

Chi 2

0.3600

0.8600

 

Prob > chi2

1.0000

1.0000

 
  1. According to the McFadden-Hausman specification of the test for IIA, the coefficients of the equations m1 and m2 should be equal. Throughout the table the χ 2 probabilities show strong evidence that we cannot reject the null hypothesis. Therefore, the IIA assumption is not violated. M1 = The full specified model. M2 = The model specified excluding one alternative of choice. Alternative 1 = Using one mobile money service only (Sending remittances or Receiving remittances or Paying utility bills), Alternative 2 = Using two mobile money services (Sending and receiving remittances or Sending remittances and paying utility bills or Receiving remittances and paying utility bills), Alternative 3 = Using all three mobile money services (Sending and receiving remittances and paying utility bills). (The base category is “performing no mobile money service”)