Resistive-switching memory (RRAM) devices based on transition metal oxides have emerged as promising candidates for applications including non-volatile memories, random number generation, and neuromorphic computing. Nevertheless, for all these applications, switching variability and degradation processes under different operating conditions remain key issues that must be thoroughly understood. In this work, we investigate cycling variability and degradation in TiN/Ti/HfO₂/W memristive devices by applying controlled voltage sweeps with progressively increasing amplitude within the electrical stress protocols. Using three specifically designed experimental protocols that independently control the positive and negative voltage limits of the voltage ramp stimulus, we assess the evolution of the low-resistance state (LRS), high-resistance state (HRS), and the dynamics of SET and RESET transitions under repeated switching cycles. The resulting I-V characteristics for each protocol are detailed as follows: (i) incremental variation of the positive voltage limit (VMAXPOS) every 100 cycles; (ii) incremental variation of the negative voltage limit (VMAXNEG) every 100 cycles (Fig. 1(c)); and (iii) alternation every 100 cycles between a fixed reference voltage range (V0NEG, V0POS) and a voltage range in which both VMAXNEG and VMAXPOS are progressively increased. These protocols induce degradation in the memristive cells under different operating conditions, enabling the independent identification of the underlying degradation processes during cycling. The memristive cells were modeled as a voltage divider consisting of a switching element in series with an intrinsic resistance (Rseries), which acts as a current limiter. The value of Rseries was determined by sweeping the resistance parameter until the corrected I-V curve exhibited an approximately vertical SET transition. Subsequently, the SET (VTS) and RESET (VRS) thresholds, obtained by subtracting the Rseries contribution were also analyzed. Alternatively, the Rseries was extracted by analyzing the dependence of VRESET on Rseries for different voltage ranges. The crossover point indicates the optimal value for Rseries. These results indicate that VMAXPOS and VMAXNEG affect the resistance window through modifications of the HRS, likely associated with changes in the filament gap and defect distribution while the switching thresholds remain unaffected, suggesting that the underlying switching mechanism is preserved