Prevalence and harm of sexual violence
Limitations of the data
2.5It is difficult to paint an accurate picture of sexual violence in New Zealand (including prevalence, reporting and attrition) using statistics. In addition to a lack of data, there are issues including:
(a) methodology (for example relating to the standard errors or point estimate used making it troublesome to reach a strong estimate figure);
(b) limitations, including privacy restraints that can have implications for the figures presented (for example small self-selecting survey samples may lead to cohort bias);
(c) the use of prevalence rates that fail to take into account that individuals might be victimised more than once; and
(d) in victimisation studies, the sensitive nature of the questions may mean people do not accurately self-identify with a category or select an inconclusive statement such as “don’t know/can’t remember/don’t wish to answer”.
2.6Police statistics, for example, do not reflect minor offences where the victim did not wish to pursue the matter or multiple instances of the same type of offence that are reported concurrently and may be recorded as a single offence. Changes in 2014 have meant that Police are seeking to record data that will permit a more victim-focused record of crime.
2.7Victimisation surveys, such as the NZCASS, are surveys of segments of the general public asking about their experience of crime victimisation. Such surveys may provide a more accurate picture of prevalence, as people are directly asked about their experiences and may reveal offending that has not been reported to Police. However, these studies also have limitations, such as the fact that they do not include victims who have been murdered, children and victims who do not reside in private households (such as the homeless, prisoners and those in hospitals, psychiatric institutions, retirement homes or boarding houses). This can result in under-estimation of the extent of victimisation, yet the NZCASS are also at risk of over-estimating victimisation because they accept what participants report at face value.
2.8Much of New Zealand’s data on sexual violence is dated, infrequent or inconsistent in its focus. For example, lifetime prevalence of sexual violence was not included in the 2006 or 2009 NZCASS. Those surveys state that the figures on sexual violence are statistically unreliable due to the small survey sample. Similarly, the 2014 NZCASS involved just over 6,000 respondents.
2.9We also note the difficulty of gathering data in this field due to potential ethical difficulties (which informed the NZCASS methodology) and that this might lead to bias arising in sample cohorts that do respond to surveys, in turn making it difficult to accurately measure sexual violence that occurs in society.