Cohen's d and Hedges' g shouldn't be interpreted as the size of the change in scores, but rather how the distribution is shifted. A score of .35 is considered a "low to moderate" effect.

https://rpsychologist.com/cohend/