It is difficult to say whether there is an "effect" of gender, but there is an association between many types of memory important important in many eyewitness situations (for discussion about gender not being causal see http://www.nyu.edu/classes/shrout/G89-2247/statisticscausal.pdf). One of the most discussed types of memory process is memory for faces where women seem to do better than men, particularly with female faces (see attached publication).

I think there are difficulties talking about eyewitness memory as if this is some homogeneous set of situations. For example, some eyewitness situations are highly traumatic, some are not. Similar issues exist for memory with age (and there is lots of research on age and memory).

Article An own sex bias and the importance of hair in face recognition

There are age differences in eyewitness identification both when comparing young adults to older adults and when comparing young adults to children. Further there is an own age bias in which witnesses are best at recognizing people close to their own age

Article Eyewitness Identification Across the Life Span: A Meta-Analy...

Article Eyewitness Identifications by Older and Younger Adults: A Me...

Thank you Professor Daniel Wright and Professor James Michael Lampinen for sending me two very interesting as well as keenly relevant papers for my experiment. I would like to test few hypothesis on Bangladeshi samples having a composition of Gender and Age. Hope I will share the findings with you two at a later time by December this year.