Personal versus telephone interviews and the effects of telephone reinterviews on reporting of psychiatric symptomatology