Why strengthening is unlikely to alter abnormal running biomechanics
The context here is appropriate for the average healthy runner.
It is written with regard to patellofemoral, sore hips, low back pain and lower limb stress injuries from running.
You may well be familiar with the idea of "weak glutes" and of course "weak core muscles", or the "weak vastus medialis".
The average healthy runner is highly unlikely to have strength deficits from an atrophied muscle. These will be strength deficits from pain, trigger points, short muscles and "inefficient skill".
"Inefficient skill" relates to the pattern of movement and muscle activation that results in abnormal injurious movement. Anatomically we might reason that since the hip in single-leg stance phase in running excessively adducts, we should strengthen the abductors.
But clinical and research observation show strengthening this motion doesn't necessarily reduce the hip adduction.
This is because strengthening activates different areas of the motor cortex than skill training. Isolated strengthening does not become integrated into whole body movement planning. The specificity of exercise training has been known for many decades.
What is noticeable is that biofeedback cues which alter the movement pattern in some way can show immediate improvements in abnormal biomechanics like medial collapse, cross-over gait, and hip adduction.
There can be a role in strength training in an isolated manner, but when a muscle has trigger points, is acutely strained, or the joint mechanics is very asymmetrical, it will often appear falsely weak.
The same is true for "inefficient skill patterns" - you can't train your way out of a poor running pattern doing clams, sumos, single leg squats. But you can make use of individualised running cues to change the pattern for the long-term.