## GEOSTATS: Summary of Cost Distance Calculation

Expand Messages
• Dear friends, Here is a summary of the responses I received to my question about cost distance calculation. I would like to thank to Dr. Duane Marble, Dr Roger
Message 1 of 1 , Sep 8, 2000
Dear friends,

Here is a summary of the responses I received to my question about cost distance
calculation.
I would like to thank to Dr. Duane Marble, Dr Roger Bivand and Dr Ned Levin for
their help and kindness

1. From Dr. Duane Marble:

A GIS?based approach to estimating the human effort involved in movement over
natural terrain,
(with Brett Machovina), in Junger and Lavento (eds.), Iskos 11, pp. 117?126.
[Keynote address given at the VII Nordic Conference on the Applications of
Scientific Methods in Archaeology,  Savonlinna, Finland, September 1996.]
You will find some useful information there on energy expenditure vs slope,
terrain type, etc.
I am now working on a paper dealing with movement in urban areas.
Dr. Duane F. Marble
Professor Emeritus of Geography              Telephone:  614-292-4419
Center for Mapping                                   Fax:  614-292-8062
The Ohio State University
1216 Kinnear                                            Email:  marble.1@...

Columbus, Ohio  43212

2. From Dr Roger Bivand:

Elsa,
In general I think it would be OK to use a "cost" calculation to measure the
effort required to get to and from school. A possible reservation is that this
is beginning to feel like what is known as "multilevel" modelling. For instance,
all the children from one village
going to one school would have very similar "costs" - varying a bit more if
spatial resolution is high, but probably not much. There is a book -Introducing
Multilevel Modeling by Kreft and de Leeuw (Sage, 1998) that explains the issues
if this is new to you - the idea is that some of the individual variance in
response is due to individual-level factors, while some is due to group level
factors (which school the children go to, which village they come from, etc.).
The method tries to take into account dependencies between observations that are
grouped at a higher level, say at the school level. There's a big literature in
the use of the method in education studies, and it has also been used in spatial
contexts:

Regional variations in mortality rates in England and Wales: An analysis
using multi-level modelling. Langford, IH / Bentham, G; SOCIAL SCIENCE &
MEDICINE Vol. 0042, Nr. 6, Mar 1996, side 897-908.

Multilevel modelling of the geographical distributions of diseases
Langford, IH / Leyland, AH / Rasbash, J / Goldstein, H; JOURNAL OF THE
ROYAL STATISTICAL SOCIETY SERIES C-APPLIED STATISTICS Vol. 48, 1999, side
253-268.

(this was looking for Ian Langford, one of the key people, but in medical
geography rather than education geography)

Good luck!

Roger
3. From Dr. Ned Levine:
Check out Luc Anselin's SpaceStat program. It can create a spatial weights
(contiguity) matrix and can conduct spatial lag regression. I've used it
extensively to model the effects of nearby zones on a dependent variable.

Ned Levine, PhD

Thanks
Elsa

--
*To post a message to the list, send it to ai-geostats@....
*As a general service to list users, please remember to post a summary
of any useful responses to your questions.
*To unsubscribe, send email to majordomo@... with no subject and
"unsubscribe ai-geostats" in the message body.
DO NOT SEND Subscribe/Unsubscribe requests to the list!
Your message has been successfully submitted and would be delivered to recipients shortly.