Is there a dominant negative stereotype that matches this? Is there a balancing, positive stereotype, or stereotypes focused on other traits which might also be in play? Are respondents signing on to that negative stereotype? Is there an association between the negative stereotype and the gendered voice? Does the selection of the voice, including pitch and timbre, word selection, and voice inflection invoke the negative stereotype either unintentionally or by design? There's a lot of assumptions bound up in the statement they make - they need to be teased out before the stereotype can reasonably be put forward as the cause of the voice preference. That's all I'm saying.