Do the above factors’ effects combine in a linear way to predict ELAR latency?

A multiple linear regression analysis was completed for the factors that were found to have a significant effect on ELAR latency with the analysis of variance, that is to say the audiometric data and the duration of deafness. The actual values of both parameters were used (and not simply the groups). The normality test passed for N1 analysis (p = NS), P2 analysis (p = NS), and N1-P2 interval analysis (p = 0.029), while the constant variance test passed for N1 analysis (p = NS) and P2 analysis (p = 0.041), and failed for N1-P2 interval analysis (p = 0.002).

Latency of N1 can be predicted from a linear combination since the equation found (39 + 0.50 x audiogram mean + 0.01 x deafness duration) involved significant contribution from the audiometric data (p < 0.001) and explained 38 % of the variance (adjusted R² = 0.38), even if the duration of deprivation did not contribute significantly to the equation (p = NS). P2 latency was found to be predictable from the following linear combination, which explains 39 % of the variance (adjusted R² = 0.39): 109 + 0.64 x audiogram mean + 1.06 x deafness duration. Contribution to this equation was significant for both parameters (audiometric data: p < 0.001, deafness duration: p = 0.001). Interval N1-P2 can be predicted from the following linear combination, which explains 24 % of the variance (adjusted R² = 0.24): 70 + 0.14 x audiogram mean + 1.05 x deafness duration. Contribution to this equation was significant for deafness duration (p < 0.001) but not for audiometric data (p = 0.38).