However, it is clear that CPS records alone are an imperfect source of data for measuring child maltreatment and these conclusions may not apply to states with different types of child welfare agency structures (e.g. non-centralized) (Fallon et al. 2010).
Birth certificates are a useful, if imperfect, source of data on pregnancy outcome: Birth weight is generally accurate, more so than gestational age (DiGiuseppe et al. 2002), with pregnancy-induced hypertension the most fallible outcome.
Systematic errors, on the other hand, may be identified by comparing one imperfect source of information (such as self-report data) with another imperfect but probably more accurate source of information (such as administrative data), and looking for directional trends when there are discrepancies.
"Impact of Imperfect Source-to-Relay CSI in Amplify-and-Forward Relay Networks," (with Jia-Chin Lin, HandKui Chang and Meng-Lin Ku).
Before his imperfect sources, the historian is powerless: "He must take what they choose to tell, the broken fragments, and the scattered ruins of evidence".
The Norwegian principal, as we have seen, declined to give any information to the agents of the Commission and left them to make up their estimate of the foreign costs from indirect and imperfect sources.
While both the self-report and the chart records were imperfect sources for verifying the absolute numbers of client referrals and receipt of treatment, significant changes over time from either or both sources in the intervention group relative to the control group sites would indicate that the intervention had the intended downstream effects on client flow (Aim 2).
The result further confirms that a faceted boundary, either fully or partially faceted, is an imperfect atom source for densification; this idea is in contrast to the conventional understanding.
Researchers will also recognise that recommendations are an imperfect data source because they typically focus on what can feasibly be done in a project, in a given period of time, often with limited resources – not what should be done (in an ideal world).
But all real polls have many sources of error that we cannot quantify and that keep us awake at night — imperfect sampling sources, interviewer bias, question wording and question order effects, inaccurate responses, weighting effects and (potentially huge) nonresponse bias, if Republicans are even slightly more, or less, likely to refuse to be interviewed than Democrats.
For real-world imperfect information sources (i.e. noisy training and/or test data), C2 is able to deliver more accurate and reliable prediction models than its other peers can offer.
