 |
 |
Dear Thom, Thank you ever so much for your response. I will read further into this. Kind Regards, Iain Iain Wilson Teaching Fellow School of Sport, Exercise and Health Sciences ZZ.0.19 Matthew Arnould Building Loughborough University Loughborough, Leicestershire, LE11 3TU From: Baguley, Thomas [mailto:[log in to unmask]] Sent: 11 December 2015 16:28 To: [log in to unmask]; Iain Wilson Subject: Re: Centring techniques for Multilevel Modelling for a longitudinal design A quick glance suggests that you are broadly correct. The question might be what the unit of interest is. In your case you have a repeated measures model which involves participants at L2 and time points at L1. In a classic educational model the individual is at level 1 not level 2. Generally, group mean centering would be used when you have other kinds of nested structures that might show effects of different sampling units (children, schools, countries). It is possible the point being made is the prosaic one that if this were an ANOVA one would just look at the total effect (the sum of the effects at each level) - as decomposing the effects at time point level and participant level might not have a clear theoretical interpretation. In my book I tried to use the terms L1-centering and L2-centering for this reason (as the L2 unit isn't always a good - it could be an individual - as here). Thom ________________________________ From: Research of postgraduate psychologists. <[log in to unmask]<mailto:[log in to unmask]>> on behalf of Iain Wilson <[log in to unmask]<mailto:[log in to unmask]>> Sent: 11 December 2015 16:03 To: [log in to unmask]<mailto:[log in to unmask]> Subject: Centring techniques for Multilevel Modelling for a longitudinal design Dear all, I have recently analysed my data from a longitudinal study using multilevel modelling in MLwiN. I have observations from 3 time points (Level 1) which I have nested within the participants (Level 2). I am interested in the effects of several variables across time on a measure of psychological fatigue. I have therefore used the centring technique ‘Group Mean Centring’ based on information I read in the Nezlek (2001) and Enders and Tofighi (2007) articles. However, I have been informed that my justification for using this technique is not appropriate and was recommended an alternative article (Ludtke et al., 2009). I am confused because the arguments for using specific centring techniques appeared to be similar in all articles so I am unsure how to proceed with amending my methods section. I was wondering if anyone could recommend any sources of information which may clarify this issue? Alternatively, does anyone know of any articles which present a clear methods section on how they have conducted their analyses and provided a clear justification of why a particular centring technique has been used? Many thanks in advance for your time and help! Kind Regards, Iain DISCLAIMER: This email is intended solely for the addressee. It may contain private and confidential information. If you are not the intended addressee, please take no action based on it nor show a copy to anyone. In this case, please reply to this email to highlight the error. Opinions and information in this email that do not relate to the official business of Nottingham Trent University shall be understood as neither given nor endorsed by the University. Nottingham Trent University has taken steps to ensure that this email and any attachments are virus-free, but we do advise that the recipient should check that the email and its attachments are actually virus free. This is in keeping with good computing practice.