A 2011 CDC Survey about sexual violence covered in the Chicago Tribune (9/8/14, sec. 1, p. 10) does not provide reliable information. The survey researchers, themselves, admit as much when they say “it’s likely their results underestimate the prevalence of sexual violence”. Likely??–but your not sure?, don’t really know? Why underestimate, rather then overestimate? So-o-o-o, what we have to conclude from this waffling is that the numbers reported are more or less accurate, guesstimates, not scientifically determined estimates.
There are many reasons why the results of this survey are not reliable:
1. Words in the questions–such as, “rape”, “stalked”, “sexual violence”, “sexual coercion”, and how about, “unwanted non-contact sexual experience”–have different meanings for different respondents. Moreover, these words may have different meanings for the survey researchers than for one or more respondents.
2. Sex, sexual violence, rape, and the like are sensitive topics and respondents tend to answer questions about sensitive matter less accurately than when the topic is nonsensitive (e.g., age, gender)–although it needs to be recognized that virtually any topic can be a sensitive matter, depending upon who asking whom, the situation, and numerous other factors.
The response rate “wasn’t high”–that is, the response rate was low; thus, “bias resulting from nonresponse”.
The survey was administered via cell and land line phones. Research indicates that these modes of asking bias/skew answers differently than other modes; e.g., face-to-face interviews.
If you want to find out what’s really going on, don’t ask! If you want the only complete statement of the weaknesses of survey research, see my book, The Problem with Survey Research.