Comprehending conflicting science-related texts: graphs as plausibility cues