[Blindmath] analyze of experimental results

[Jonathan Godfrey]

Hello,

As a practicing statistician I find the suggestions promoted by Susan 
and Christine the most compelling.

Perhaps my only addition of import is that we need to know where the 
data comes from to give really useful feedback about a specific 
approach. You said experimental data. Is this one experiment or many? 
Is the experiment what a chemist or physicist would call an 
experiment or and experiment in terms of comparative investigation as 
I think of experiments.

Certainly, my automatic suggestion for a set of 100 points would be 
to model it and see what comes out. I would never trace through this 
many points let alone a thousand of them.

The thing that surprises me the most about this discussion is that 
the offered suggestions are all about being a blind user aiming to 
emulate the sighted user. I teach my students to graph data and look 
at it, but that is because they can see. I do not have any access to 
tactile methods for viewing data so I find other ways of 
understanding the data I am given. I would argue that for example 
calculating correlation coefficients on the raw data and the ranked 
data (called Spearman's correlation) and summary statistics of the 
two marginal distributions to give me some idea of what relationship 
might exist. Of course, I need to also consider that a parabola might 
be needed to fully investigate the relationship so polynomial 
regression is required.

The advantage is for me that in going ahead and doing some modelling 
is what is usually required if the sighted user finds a relationship 
visually anyway. My investigation of the data is probably quicker 
than making a tactile image and assessing it, hence the lack of 
hardware at this time. Having automated may of these tasks via 
programming, I am as efficient as a sighted person in completing 
tasks like this.

The question for me is what tool you use for fitting suitable models 
and getting the summary statistics you need. For me that is R. The 
simple analyses suggested here are possible in many software solutions though.

Jonathan






_____
Dr A. Jonathan R. Godfrey
Lecturer in Statistics
Institute of Fundamental Sciences
Massey University
Palmerston North

Office: Science Tower B Room 3.15
Phone: +64-6-356 9099 ext 7705
Mobile: +64-29-538-9814
Home Address: 22 Bond St, Palm. Nth.
Home Phone: +64-6-353 2224 (Just think FLEABAG)