In this article, the leaderless and leader-following scaled consensus (SC) of second-order multi-agent systems (SOMASs) with nonlinear dynamics under distributed adaptive control are investigated by non-reduced order method (NROM). Firstly, the NROM is first applied to transform the variable-replaced SOMASs into a pure second-order differential system. Secondly, a novel Lyapunov function involving the error variables and derivative of the error variables is constructed to directly discuss the second-order differential system, which is completely different from the traditional analysis method. Thirdly, two different coupling gains are designed in the light of the position and velocity of the agent to solve the leader-following SC more flexibly. Finally, two numerical examples are cited to verify the practicability of the theoretical derivation.