On 2015 May 13, Tim Smits commented:

As already commented by James Coyne, there are serious issues with the statistics in this paper as well as with the overall level of academic scrutiny in the write-down of the methodology and findings. These have indeed been discussed in two Letters to the Editor Smits T, 2014 & Smits T, 2015. Though published, the journal has not yet warned its readers of these mistakes in the original article. Next to these two published letters, more concerns about the article are described in this blogpost http://persuasivemark.blogspot.be/2014/02/dont-get-all-psychotic-on-this-paper.html

On 2015 May 12, James C Coyne commented:

This study compares a nurse-delivered cognitive behavior therapy to treatment as usual in a small, convenience sample of older psychotic patients from community care who averaged over 20 years of previous treatment. The treatment as usual is not adequately described or quantified in a way that allows interpretation of any differences between conditions. It cannot be determined whether any effects are due to specific elements of the intervention or to inadequacies in professional contact time, support, or expectations likely to be occurring in the community psychiatric care of patients who have been there over 20 years.

But there are more basic problems. As detailed in letters to the editor Smits T, 2014 Smits T, 2015 and numerous blog posts, there are substantial discrepancies between findings reported in the abstract, in the narrative Results section, and in the tables and figures of this article. Just a sampling:

For Table 2. The confidence intervals were suspiciously wide. The effect sizes seemed too large for what the modest sample size should yield. The table was inconsistent with information in the abstract. Neither the table nor the accompanying text had any test of significance nor reporting of means and standard deviations. Confidence intervals for two different outcomes were identical, yet one had the same value for its effect size as its lower bound.

Figure 2 is mislabeled as a histogram and if results are to be believed, there are no significant pre-post differences between intervention and control group on any of the outcome variables. This is discrepant with what is reported in the abstract and narrative of the Results section.

Figure 5 lacks any metric for the vertical lines.

Their responses to letters about this article clearly indicate that the author Turkington D, 2014 and the statistical editor of the Journal < PMID:25816048 >are now aware of these problems.

It is unclear why there has not been a retraction or correction issued.

You can read more about the problems of this article and the response to criticism at my posts at PLOS Mind the Brain.

Sordid tale of a study of cognitive behavioral therapy for schizophrenia gone bad http://blogs.plos.org/mindthebrain/2015/04/14/sordid-tale-of-a-study-of-cognitive-behavioral-therapy-for-schizophrenia-gone-bad/

On 2015 May 12, James C Coyne commented:

This study compares a nurse-delivered cognitive behavior therapy to treatment as usual in a small, convenience sample of older psychotic patients from community care who averaged over 20 years of previous treatment. The treatment as usual is not adequately described or quantified in a way that allows interpretation of any differences between conditions. It cannot be determined whether any effects are due to specific elements of the intervention or to inadequacies in professional contact time, support, or expectations likely to be occurring in the community psychiatric care of patients who have been there over 20 years.

But there are more basic problems. As detailed in letters to the editor Smits T, 2014 Smits T, 2015 and numerous blog posts, there are substantial discrepancies between findings reported in the abstract, in the narrative Results section, and in the tables and figures of this article. Just a sampling:

For Table 2. The confidence intervals were suspiciously wide. The effect sizes seemed too large for what the modest sample size should yield. The table was inconsistent with information in the abstract. Neither the table nor the accompanying text had any test of significance nor reporting of means and standard deviations. Confidence intervals for two different outcomes were identical, yet one had the same value for its effect size as its lower bound.

Figure 2 is mislabeled as a histogram and if results are to be believed, there are no significant pre-post differences between intervention and control group on any of the outcome variables. This is discrepant with what is reported in the abstract and narrative of the Results section.

Figure 5 lacks any metric for the vertical lines.

Their responses to letters about this article clearly indicate that the author Turkington D, 2014 and the statistical editor of the Journal < PMID:25816048 >are now aware of these problems.

It is unclear why there has not been a retraction or correction issued.

You can read more about the problems of this article and the response to criticism at my posts at PLOS Mind the Brain.

Sordid tale of a study of cognitive behavioral therapy for schizophrenia gone bad http://blogs.plos.org/mindthebrain/2015/04/14/sordid-tale-of-a-study-of-cognitive-behavioral-therapy-for-schizophrenia-gone-bad/

On 2015 May 13, Tim Smits commented:

As already commented by James Coyne, there are serious issues with the statistics in this paper as well as with the overall level of academic scrutiny in the write-down of the methodology and findings. These have indeed been discussed in two Letters to the Editor Smits T, 2014 & Smits T, 2015. Though published, the journal has not yet warned its readers of these mistakes in the original article. Next to these two published letters, more concerns about the article are described in this blogpost http://persuasivemark.blogspot.be/2014/02/dont-get-all-psychotic-on-this-paper.html