A binary code is said to be a disjunctive list-decoding $s_L$-code (LD $s_L$-code), $s \ge 2$, $L \ge 1$, if the code is identified by the incidence matrix of a family of finite sets in which the union (or disjunctive sum) of any $s$ sets can cover not more than $L-1$ other sets of the family. In this paper, we consider a similar class of binary codes which are based on a {\em symmetric disjunctive sum} (SDS) of binary symbols. By definition, the symmetric disjunctive sum (SDS) takes values from the ternary alphabet $\{0, 1, *\}$, where the symbol~$*$ denotes "erasure". Namely: SDS is equal to $0$ ($1$) if all its binary symbols are equal to $0$ ($1$), otherwise SDS is equal to~$*$. List decoding codes for symmetric disjunctive sum are said to be {\em symmetric disjunctive list-decoding $s_L$-codes} (SLD $s_L$-codes). In the given paper, we remind some applications of SLD $s_L$-codes which motivate the concept of symmetric disjunctive sum. We refine the known relations between parameters of LD $s_L$-codes and SLD $s_L$-codes. For the ensemble of binary constant-weight codes we develop a random coding method to obtain lower bounds on the rate of these codes. Our lower bounds improve the known random coding bounds obtained up to now using the ensemble with independent symbols of codewords.