In this set of recordings, CORC Director Miranda Wolpert talks through the one slide below which summarises the key findings in terms of reliable change and recovery and explains the key terms used in our analysis of data from child mental health services (2011 - 2015).  


                                                                            Click here for full report

                                                                                                   Click here for full presentation 

Key term


% of paired clinical sample (95% Margin of Error)

(Number in sample)



Refers to scores having moved from above the threshold on that scale to below the threshold on that scale.

It should be noted that different scales in this report have different thresholds, determined in different ways. Please note that “recovery” is referred to in inverted commas in this report when used to refer to this movement across threshold/s, as opposed to the broader concept of recovery as reflected in the lived experience of the individual concerned (e.g. Leamy, Bird, Le Boutillier, Williams & Slade, 2011).


(35% - 37%)

(N = 2117 out of 5896)

Reliable Improvement

Refers to a score on at least one measure changed enough for it to be considered statistically reliable and no other score reliably deteriorated.


(51% - 53%)

(N = 3056 out of 5896)

Reliable deterioration

Refers to a score on one or more measures changed enough in a negative direction for it to be considered statistically reliable.


(9% – 12%)

(N = 617 out of 5896)

Reliable recovery

Refers to change in scores on a scale that indicates both reliable change and “recovery” have occurred.


(25% - 28%)

(N = 1569 out of 5896)


No reliable change

Refers to a score on all completed measures did not show any statistically reliable change. This does not mean they did not show any change at all, only that the amount of change couldn't be differentiated from measurement error/noise in the measure.


(36% - 39%)

(N = 2223 out of 5896)


It is really important when thinking about the results we found from our analysis of data from child mental health services (2011-2015) to remember that this data definitely falls into the category of FUPS data, as defined by CORC - data that is flawed, uncertain, proximate and sparse. In the following recording, Miranda explains what we mean by FUPS and the reasons behind it. 


                                                                                   Click here for full report on using FUPS data