How to design such a secure multiparty computation scheme with the players using a majority rule

Suppose that $y$ is a uniform random variable that is defined over the field (or group or abelian group) $Y$. Let us suppose that there are $N=\{1,2,\cdots,i\cdots,N\}$ agents and only one of them, say $i$, knows the random variable $y$. She wants to share the secret with the other $J=N-\{i\}$ players. Could someone provide a secure scheme for player $i$ to share her secret with players $j\in J$ in the following way:

Assume that player $i$ takes the role of a dealer. We could split the space $J$ in three groups of agents $J_1$, $J_2$ and $J_3$ mutually disjoint and $J=J_1\cup J_2\cup J_3$. How could player $i$ split $y$ giving different parts of information of this to each group of agents and if these three different parts could communicate and then using any calculation $\oplus$ and $\otimes$ they could obtain $s$ based on a majority rule for at the end of the process where the players at the endo of the process should agree that indeed $s$ is implied by a majority rule. Namely of the majority of $J$ learns $s$ then all of them will learn it.