Dear All,
When I posted my query, I think I was hoping for some nice webpages,
from which I could download images. Alas no luck!
Many thanks to all who replied, in particular those who sent curves
I append the replies below, although I can't include the attachments
in MAILBASE.
The replies split into 3:
1)Those who wouldn't use Microsoft anyway
2a) Those who would use Excel and cut and paste
2b) Those who would use MINITAB and cut and paste
Mike
PS There are 3 types of statistician: those who can count and those
who can't.
This was the original query
>
> I am sure many statisticians have attempted to draw Normal curves
> free hand on Word, or Power point for teaching purposes. From the
> appearance of some text books, some authors have this problem also
> (I've even seen semi-circular Normal curves). I wondered if anyone
> could point me to some copyright free figures that could be
> incorporated into a word processor and edited. Other distributions
> would be handy as well.
From: [log in to unmask]
An easy way to produce curves of the common distributions is to use
Minitab to generate the probability densities corresponding to equally
spaced x-values. The x-values can be chosen to give a smooth a curve
as is desired. The two columns of data (x-values and densities) can
then be cut and pasted into a graph plotting function within a word
processing package. In Word, for example, the graph can be produced
using Microsoft Graph and then edited by using Paintbrush to add text
or shade in particular areas etc.
From: [log in to unmask] (Danny Kaye)
it would be fairly easy to produce postscript figures of most
distributions in SPlus and then incorporate those in word or any other
WP. If your printer cannot handle postscript then the shareware RIP
GhostScript could be used.
If you want to discuss this please feel free to email me.
From: C Wallace <[log in to unmask]>
If you insist on using windows software, try using excel to generate
random normal (or poisson etc) data and graphing it. This can then be
imported straight into word or powerpoint.
If you really want a clip art resource, there's bound to be some
freeware on the web somewhere for converting excel-style graphs to
clip art.
From: Jeremy Miles <[log in to unmask]>
I have an excel file that I use to draw normal curves, and other
distributions.
I could send you a copy if you like.
From: "David Tyrrell" <[log in to unmask]>
Mike,
I have attached a Word document with a normal distribution curve I
drew = for a presentation. Unfortunately I have no others but feel
free to use = this. If it is any easier I have it as a PowerPoint
slide as well.
From: Alexander Donev <[log in to unmask]>
Dear Mike,
Some thoughts...
I have successfully used First Bayes, a free package, to obtain
various useful probability plots. You can get it from Tony O'Hagan.
However, for more specific tasks, this may not be the best option. For
example, in order to illustrate power of a test I found MINITAB better
as it allows you to overlap plots. I don't think there is a copyright
issue when the plot is obtained using a licensed package.
Of course, if there is anything readily available, I would be glad to
know too.
From: "Dr. Hans-Christian Waldmann"
<[log in to unmask]> Maybe it would be a good idea to
turn to non-MS-Software.
Generate ANY statistical graph using your stats package
(like SAS / SPLUS or whatever) or a vector-oriented
drawing tool like UNIX-xfig and include it in your LaTeX-document as
an encapsulated postscript file. There is virtually _nothing_ you
can't do with LaTeX and postscript, including printer/camera-ready
ps-output at any format or pdf and html-output.
If you are faced with heavy formula-typesetting, you're in for a
_very_ hard time using OFFICE-stuff, anyway. Partly, LaTeX was
developped for maths publications, and still is unbeat in this domain.
Christian Waldmann
From: Elise Whitley <[log in to unmask]>
I have managed to produce reasonable normal curves in the past using a
convoluted route of stata, powerpoint and Word (example attached).
(1) Produce a histogram of reasonably normal data in stata and fit a
normal curve over it using the normal option. (A similar approach may
work for other stats packages too) (2) Use the copy graph option in
stata and paste it into powerpoint (3) Double click on the graph and
convert it to a Microsoft Office drawing - this will split the graph
into its componant parts. (4) Delete the bars of the histogram and any
unwanted labels. (5) Import into Word
However, I'm sure there must be an easier way(!) so I'd be interested
to hear what other responses you get.
>From Jean M. Russell M.A. M.Sc. [log in to unmask]
Dear Mike
I have a series of Power point talks that I did a year ago that have a
number of drawings in them. Including a normal distribution,
expoential distribution. I can either show you how I created them
(Excel, SPSS and Word) though now in Powerpoint or let you have a copy
of the slides that have drawings on them.
Jean
From: Diana Kornbrot <[log in to unmask]>
i always produce the normal in excel & then transfer.
if you are using power power point then probably best way is to get
co-ordinates as 2 variables in excel or orther package (eg if excel
does not have all the distributions you want) & then draw from
powerpoint's graph module.
of course if you are lecturing with your computer you can show how
changing parameters changes distribution shape etc
diana
From: Shen Xiaoxiang <[log in to unmask]>
Dear Professor Campbell,
You may use Excel to draw the normal curve (other curves are similar).
Here's one way (taking standard normal distribution as an example):
In Column A, input -3.0, -2.95, ..., 0, ..., 2.95, 3.0 (it could be
from -4 to +4 or others, the step could also be other rather than
0.05; up to you).
In Column B, use function NORMDIST(). For each row, e.g., Row 1, put
A1 as X, then Mean and Standard_dev, (e.g, 0, 1), put False as
Cumulative ('cos you need probability mass function). A shortcut: copy
B1, then paste it in B2, B3, ..., Bn; in this way, there is no need to
set the function for each row.
Then draw the normal curve based on these two culumn data, using
XY(scatter).
Hope this helps you.
Happy New Year!
Xiaoxiang Shen
--
Department of Industrial & Systems Engineering
National University of Singapore
Mike Campbell
Professor of Medical Statistics with an Interest in Primary Care
Institute of Primary Care
Community Sciences Building
Northern General Hospital
Sheffield S5 7AU
Tel 0114 271 5919
FAX 0114 242 2136
email [log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|