Why are the + and - states for a Qubit represented as |0>+|1> and |0>-|1> respectively and not the other way around?

Why are the + and - states for a Qubit represented as |0>+|1> and |0>-|1> respectively and not the other way around?

Is this only a matter of convention or is there a formula to arrive at each vector in Hilbert space only from |0> and |1> states?

It's a matter of convention. Remember that what you write inside the ket is just a label. Labels can be anything you want. So it is a convenience that helps us remember that $$|\pm\rangle=|0\rangle\pm|1\rangle.$$ We'd all be making mistakes all the time if it were the other way around!
I've just argued that the notation makes a lot of sense in terms of helping us to remember the correct way of writing the state. It also makes a lot of sense in terms of the Pauli $$X$$ matrix: $$\left(\begin{array}{cc} 0 & 1 \\ 1 & 0 \end{array}\right)$$ This has two eigenvalues $$\pm 1$$ with corresponding eigenvectors $$|\pm\rangle$$, so there's also a convenient correspondence between the label and the eigenvalue.