False Data

Falsifying scientific data is a serious offense.

All or virtually all CANS publications include false data and/or citations.

Here is one example from the Chapin Hall “experts” (Chor, McClelland, Weiner, Jordan, & Lyons, 2013). In the measures section, the authors make these knowingly false statements about the validity and reliability of CANS:

The CANS assessment is embedded across [Illinois] IDCFS programs to track clinical outcomes (Weiner, Schneider, & Lyons, 2009) and in child welfare systems in multiple states and international settings (Lyons, 2009). The CANS has demonstrated strong field reliability, audit reliability, and concurrent validity with the CAFAS (Anderson, Lyons, Giles, Price, & Estle, 2003; Lyons, 2004, 2009). The CANS also has established strong item-level inter-rater reliability among researchers and between researchers and clinicians, and it has achieved on average .81 for the former and .85 for the latter across CANS domains (Anderson et al., 2003).

These research results are fabricated:

  • Anderson et al., 2003, does not support the claim that the inter-rater reliability is “strong” or more specifically that “between researchers and clinicians” the “inter-item inter-rater reliability” is 0.85. Review of Table 2 in the referenced publication shows average item-level percent agreement of only 0.72. When converted to Cohen’s Kappa the average inter-rater reliability is only 0.63. These Defendants have consistently misreported results from this study. Further, the researchers know that the Anderson et al., 2003 study does not test the reliability of CANS as used in Illinois. The correct study would have required testing the inter-rater reliability of two clinicians, not by using chart reviews. Since Lyons claims that inter-rater reliability is essential at the item level, CANS should not be used, and certainly not promoted with “strong reliability and validity.”

  • Lyons (2004, 2009) nor Anderson et al.,( 2003) provide any support for concurrent validity. Anderson studied chart audit reliability not concurrent validity and did NOT use the CAFAS. Lyons (2009) does not provide any primary-source research and instead cites conference presentations given by parents of Illinois children and cites back to Lyons (2004). Lyons (2004) provides only the following text related to validity.

Why is the University of Chicago and Chapin Hall hiring doctors who falsify data?