I have a dataset of about 1500 different hospitals and about 40 characteristics for each hospital (e.g. floor area, number patients, type of hospital, age of building, etc.). I am interested in finding out which characteristics have the strongest impact on energy consumption in the hospital. The aim of the study is to suggest ways of reducing energy consumption in some of the hospitals. My initial thought was to perform a cluster analysis to cluster hospitals according to some basic characteristics like type/floor area/number of patients. I could then do a regression analysis separately for each of the 3 or 4 clusters identified to determine which of the remaining characteristics are most influential for each cluster. My reasoning behind this is that there are certain characteristics of a hospital which will definitely impact the energy consumption, but also unchangeable (e.g. kicking out a bunch of patients may reduce energy consumption but would generally be frowned-upon!). Does this sound like a reasonable approach? Am I violating any statistical assumptions by first doing a cluster analysis and then following up with regression on each cluster? I realise I could just do a regression in the first place, but I suspect that the effect of any of the less obvious variables will be lost in the presence of the main variables.

  • $\begingroup$ Welcome to CV. What is the link between these 1,500 hospitals and energy consumption? In other words, do you have an actual measure of consumption or are you forced to infer it based on the 40 characteristics? Either way, I don't see the advantages of clustering over a model built using hospital as the unit of analysis. $\endgroup$ – Mike Hunter Nov 20 '15 at 12:49
  • $\begingroup$ Getting meaningful clusters will be really hard, and getting clusters that relate to energy consumption is very unlikely. I doubt your prediction will improve much. Good luck. $\endgroup$ – Has QUIT--Anony-Mousse Nov 21 '15 at 9:02
  • $\begingroup$ Thanks for your comments. We have actual energy consumption data so that isn't a problem. The issue is that energy consumption will differ widely depending on the size of the hospitals so I was looking for a way of splitting the data before doing the analysis. $\endgroup$ – SB17 Nov 24 '15 at 14:16

Your suggestion is close to multi-level regression.

Find more explanation for example here:


The gist is that the population (in your case hospitals) is not homogeneous, but that there are subgroups (levels) that can be identified. Multi-level regression in practice allows for different models per group, and insight into the difference between groups.

The difference is that you will be forming the groups based on a cluster analysis.

  • $\begingroup$ Thanks! That sounds promising. I will dust off my stats books and investigate. $\endgroup$ – SB17 Nov 24 '15 at 14:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.