I woud like to use the outer weight to form a equation so that i can use the equation to claculate the value of latent variables, all i want to do is to normalize the weights . As usual, I adopt the formula as the following:

**Xi/(sum all the XI)**

But now , one the oueter weight is

**negative**,can i still use it ?

For example, THE latent variable LV1 has 4 explicit variables, their

**Outer weights**are as following:

V1 0.3205

V2 0.3235

V3 0.3138

V4 -0.1844

so , if i use the formula

**Xi/(sum all the XI)**

The coefficient of V1 is 0.3205/（0.3205+0.3235+0.3138-0.1844）=0.4144

The coefficient of V2 is 0.3235/（0.3205+0.3235+0.3138-0.1844）=0.4183

The coefficient of V3 is 0.3138/（0.3205+0.3235+0.3138-0.1844）=0.4057

The coefficient of V4 is -0.1844/（0.3205+0.3235+0.3138-0.1844）=-2308

so the euation I got is

**LV1=0.4144V1+0.4183V2+0.4057V3-2308V4**

But because there some value is negative ,i don't know it is correct or not.

Pls hlep me deal with this problem, I will really appreciate your help .Thank you so much .