School of Psychology - Activities - Tutorials - The reliable change index

Login to the School of Psychology

Tutorial - The reliable change index

Mark Horswill's notes on the reliable change index

Reliable Change Index (RC or RCI)

•This is a statistic that we can use to work outwhether a change in an individual’s score (e.g.before and after some intervention) is statistically significant or not (based on how reliable the measure is).

•It is defined as the change in a client’s scoredivided by the standard error of the difference for the test(s) being used.

•It is a standardized score representing the changein a client’s score on a test.

How to calculate the RC

•In words, the RC is equal to the individual’s scorebefore the intervention minus their score after the intervention then divided by the standard error of the difference of the test.

RCI Image 1


This is how to work out the Standard Error of Measurement (=SEM or SE) for a test–and also how to work out the Standard Error of the Difference (=SEdiff or SDIFF) directly from a test’s standard deviation and reliability.

RCI Image 2
How to use the RCI

•If the RCI is 1.96 or greater then the difference is statistically significant (1.96 equates to the 95% confidence interval).

•If the RCI is less than 1.96 then the difference is not significant.

•This is the same as saying the difference between the two scores has to be at least twice the SEdiff (to be precise, 1.96 times the SEdiff) to be significant.

Extensions of the reliable change index

•There are some extensions of the RCI that take into account practice effects (the fact that people get better the second time they do a test even ifthere’s no change in whatever is being measured).

•The following is taken from Parsons et al (2009)International Journal of Neuroscience:

RCI Image 3

Extensions of the reliable change index
  • All this means is that you work out the average improvement in the scores of people who done the test twice with no intervention (i.e. the improvement is just a practice effect).

  • Then you subtract this number from the change inthe individual’s score.

  • Then you divide by the Standard Error of the Difference.

  • The RCI you end up with is now “corrected” forpractice effects–so any significant differences in your client you find are real changes rather thanjust due to the fact they’re doing the same testtwice (see Heaton et al 2001).

  • Heaton R K, Temkin N, Dikmen S, Avitable N, Taylor M J, Marcotte T D, Grant I, 2001 "Detecting change: A comparison of three neuropsychological methods, using normal and clinical samples"Archives of Clinical Neuropsychology16 75-91

  • Jacobson N S, Truax P, 1991 "Clinical significance: a statistical approach to defining meaningful change in psychotherapy-research"Journal of Consulting and Clinical Psychology59 12-19

  • Parsons T D, Notebaert A J, Shields E W, Guskiewicz K M, 2009 "Application of Reliable Change Indices to Computerized Neuropsychological Measures of Concussion"International Journal of Neuroscience119 492- 507 

Source: Download PDF


Find out more...

Accessed: 1470 times
Created: Wednesday, 16th May 2012 by blake
Modified: Tuesday, 13th January 2015 by blake
Psychology News, Events & Publications RSS 2.0 Feed School of Psychology on Facebook School of Psychology on Twitter School of Psychology on Google Plus School of Psychology on Linkedin School of Psychology on YouTube
Share this page on Facebook Share this page on Twitter Share this page on Google Plus Share this page on Linkedin