Regression in colors
Locked
 
|
A student of mine is producing a simple linear regression with several predictors, including a scatterplot of actual versus predicted values of the dependent variable. He wants that the plot points are identified by colors corresponding to a categorical variable (not included in the regression) which has four values from 0 to 3. Labeling in the REGRESSION procedure prints values or value labels into the plot. How could one have cases identified in colors or different markers, with no labels or values printed? Thanks for your help Hector |
|
|
Just save the predicted values to the dataset
in the usual way and then use the Chart Builder to create the plot with
the desired effects.
Jon Peck Senior Software Engineer, IBM [hidden email] new phone: 720-342-5621 From: Hector Maletta <[hidden email]> To: [hidden email] Date: 06/16/2011 09:24 AM Subject: [SPSSX-L] Regression in colors Sent by: "SPSSX(r) Discussion" <[hidden email]> A student of mine is producing a simple linear regression with several predictors, including a scatterplot of actual versus predicted values of the dependent variable. He wants that the plot points are identified by colors corresponding to a categorical variable (not included in the regression) which has four values from 0 to 3. Labeling in the REGRESSION procedure prints values or value labels into the plot. How could one have cases identified in colors or different markers, with no labels or values printed? Thanks for your help Hector |
|
|
They know how to go all round the houses saving and then using graphics. it is a waste of our valuable time I have been making this request since the GUI 1st appeared It is INSANE to have option to plot every kind of residual, but not the main results: the dependet variable as a function of the independent variable But spss listen to users? You cannot be serious Best Diana On 16/06/2011 20:00, "Jon K Peck" <peck@...> wrote: Just save the predicted values to the dataset in the usual way and then use the Chart Builder to create the plot with the desired effects. Professor Diana Kornbrot email: d.e.kornbrot@... web: http://web.me.com/kornbrot/KornbrotHome.html Work School of Psychology University of Hertfordshire College Lane, Hatfield, Hertfordshire AL10 9AB, UK voice: +44 (0) 170 728 4626 fax: +44 (0) 170 728 5073 Home 19 Elmhurst Avenue London N2 0LT, UK voice: +44 (0) 208 883 3657 mobile: +44 (0) 796 890 2102 fax: +44 (0) 870 706 4997 |
|
|
I always enjoy seeing helpful and polite
comments from users.
Plotting the dependent variable against the independents has been in the REGRESSION command for generations. /PARTIALPLOT ALL. will do all partial plots, and /PARTIALPLOT a b c will do listed variables. But the request in this case was for a plot colored by a variable that was NOT in the regression. For that the solution proposed is straightforward and appropriate IMO. Jon Peck Senior Software Engineer, IBM [hidden email] new phone: 720-342-5621 From: "Kornbrot, Diana" <[hidden email]> To: Jon K Peck/Chicago/IBM@IBMUS, "[hidden email]" <[hidden email]> Date: 06/16/2011 01:24 PM Subject: Re: Regression in colors What spss user are requesting is that the graphics option be included in REGRESSION They know how to go all round the houses saving and then using graphics. it is a waste of our valuable time I have been making this request since the GUI 1st appeared It is INSANE to have option to plot every kind of residual, but not the main results: the dependet variable as a function of the independent variable But spss listen to users? You cannot be serious Best Diana On 16/06/2011 20:00, "Jon K Peck" <peck@...> wrote: Just save the predicted values to the dataset in the usual way and then use the Chart Builder to create the plot with the desired effects. Jon Peck Senior Software Engineer, IBM peck@... new phone: 720-342-5621 From: Hector Maletta <hmaletta@...> To: SPSSX-L@... Date: 06/16/2011 09:24 AM Subject: [SPSSX-L] Regression in colors Sent by: "SPSSX(r) Discussion" <SPSSX-L@...> A student of mine is producing a simple linear regression with several predictors, including a scatterplot of actual versus predicted values of the dependent variable. He wants that the plot points are identified by colors corresponding to a categorical variable (not included in the regression) which has four values from 0 to 3. Labeling in the REGRESSION procedure prints values or value labels into the plot. How could one have cases identified in colors or different markers, with no labels or values printed? Thanks for your help Hector Professor Diana Kornbrot email: d.e.kornbrot@... web: http://web.me.com/kornbrot/KornbrotHome.html Work School of Psychology University of Hertfordshire College Lane, Hatfield, Hertfordshire AL10 9AB, UK voice: +44 (0) 170 728 4626 fax: +44 (0) 170 728 5073 Home 19 Elmhurst Avenue London N2 0LT, UK voice: +44 (0) 208 883 3657 mobile: +44 (0) 796 890 2102 fax: +44 (0) 870 706 4997 |
|
|
In reply to this post by Hector Maletta
Have him save the predictors from regression and use any one of the various graphing procedures. GRAPH, IGRAPH, XGRAPH, and GGRAPH all provide grouping via a categorical variable. From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Hector Maletta A student of mine is producing a simple linear regression with several predictors, including a scatterplot of actual versus predicted values of the dependent variable. He wants that the plot points are identified by colors corresponding to a categorical variable (not included in the regression) which has four values from 0 to 3. Labeling in the REGRESSION procedure prints values or value labels into the plot. How could one have cases identified in colors or different markers, with no labels or values printed? Thanks for your help Hector |
|
|
In reply to this post by Jon K Peck
Thanks, Jon. I knew about that. What I meant is to have it in one step, as a standard outcome of REGRESSION. Since this procedure does produce a scatterplot with text labels for each point according to values of a specified variable, it may well have colors or markers according to the values of the external variable. In fact, the text labels in a regression output is seldom useful: a scatterplot with text attached to every point is usually a mess unless the points are really few and quite apart from each other, in which case REGRESSION would be seldom be advisable (too few points, not close to each other). Hector De: SPSSX(r) Discussion [mailto:[hidden email]] En nombre de Jon K Peck Just save the predicted values to the dataset in the usual way and then use the Chart Builder to create the plot with the desired effects.
No virus found in this message. |
|
|
Those point labels are really most useful
when the points are isolated, because you generally want to know who the
outliers are. But since the labelling variable is often a case id,
just coloring automatically by the id variable would probably not be wise.
You can turn off all the labels with one gesture and then turn them
on selectively in the Chart Editor. That works for outliers, but
it isn't much help for the example in question.
The trouble with this is feature creep. As soon as you start enhancing the chart, a million other improvement come up, so soon you wind up with an entire graphics interface inside Regression - and why just regression? So we thought it best to pass responsibility for more elaborate charting to the procedures that are actually dedicated to graphics such as the Chart Builder or GraphBoard Template Chooser. If someone wanted to automate a particular style of charting just for regression, though, it would be easy to do with a macro or programmability extension command. And that could have its own custom dialog box with just the features one wanted there. When I teach a course on this, I do an example of an alternative interface to Summarize that, in fact, includes charts. Regards, Jon Peck Senior Software Engineer, IBM [hidden email] new phone: 720-342-5621 From: "Hector Maletta" <[hidden email]> To: Jon K Peck/Chicago/IBM@IBMUS, <[hidden email]> Date: 06/16/2011 02:57 PM Subject: RE: Regression in colors Thanks, Jon. I knew about that. What I meant is to have it in one step, as a standard outcome of REGRESSION. Since this procedure does produce a scatterplot with text labels for each point according to values of a specified variable, it may well have colors or markers according to the values of the external variable. In fact, the text labels in a regression output is seldom useful: a scatterplot with text attached to every point is usually a mess unless the points are really few and quite apart from each other, in which case REGRESSION would be seldom be advisable (too few points, not close to each other). Hector De: SPSSX(r) Discussion [[hidden email]] En nombre de Jon K Peck Enviado el: Thursday, June 16, 2011 16:00 Para: [hidden email] Asunto: Re: Regression in colors Just save the predicted values to the dataset in the usual way and then use the Chart Builder to create the plot with the desired effects. Jon Peck Senior Software Engineer, IBM [hidden email] new phone: 720-342-5621 From: Hector Maletta <[hidden email]> To: [hidden email] Date: 06/16/2011 09:24 AM Subject: [SPSSX-L] Regression in colors Sent by: "SPSSX(r) Discussion" <[hidden email]> A student of mine is producing a simple linear regression with several predictors, including a scatterplot of actual versus predicted values of the dependent variable. He wants that the plot points are identified by colors corresponding to a categorical variable (not included in the regression) which has four values from 0 to 3. Labeling in the REGRESSION procedure prints values or value labels into the plot. How could one have cases identified in colors or different markers, with no labels or values printed? Thanks for your help Hector No virus found in this message.
|
Automatic reply: Regression in colors
|
|
I will be out of the office until Monday, June 27. If you need immediate assistance, please call 812-856-5824. I will respond to your e-mail
as soon as possible. Thank you,
Shimon Sarraf
Center for Postsecondary Research Indiana University Bloomington |
|
|
In reply to this post by Jon K Peck
Apologies for venting my frustrations with the DIALOGUE interface to REGRESSION on the list
Best Diana On 16/06/2011 20:36, "Jon K Peck" <peck@...> wrote: I always enjoy seeing helpful and polite comments from users. Professor Diana Kornbrot email: d.e.kornbrot@... web: http://web.me.com/kornbrot/KornbrotHome.html Work School of Psychology University of Hertfordshire College Lane, Hatfield, Hertfordshire AL10 9AB, UK voice: +44 (0) 170 728 4626 fax: +44 (0) 170 728 5073 Home 19 Elmhurst Avenue London N2 0LT, UK voice: +44 (0) 208 883 3657 mobile: +44 (0) 796 890 2102 fax: +44 (0) 870 706 4997 |
|
|
Partial plots are supported in the Regression
Plots subdialog. There is a checkbox for doing all partial plots
- Produce all partial plots. Specifying a subset of these would
require syntax, although one could just paste and change ALL to the variables
desired.
As for interactions, regression does not address these specifically, but GLM (or General Linear Model>Univariate) does. That includes a number of useful plots involving interactions, though not exactly what you are looking for. Maybe it would be useful to produce an extension command (with dialog box) that did a set of plots of one variable vs each of a specified set with optional coloring by a group. This can, of course, be constructed with the Chart Editor, but that takes several steps, so packaging a set of these would simplify the process. I would be interested in hearing more about exactly what plotting logic would be appropriate. While it would be difficult to get this into the standard releases, and even then it would be subject to the long release cycle, it would be easy to do an extension command. Jon Peck Senior Software Engineer, IBM [hidden email] new phone: 720-342-5621 From: "Kornbrot, Diana" <[hidden email]> To: Jon K Peck/Chicago/IBM@IBMUS Cc: "[hidden email]" <[hidden email]> Date: 06/17/2011 04:58 AM Subject: Re: Regression in colors Thanks Jon Apologies for venting my frustrations with the DIALOGUE interface to REGRESSION on the list 1. if it has been easily available in script for yonks, why not implement in GUI dialogue? I find the GUI mostly very powerful & much easier for students to learn. Makes the glitches more frustrating 2. can the script do interactions? E.g. /partial plot a a*b, where b is categorical. If so it would be possible t run REGRESSIONS once without the categorical for statistics and once with to get plots. There is a nice implementation in survival plots, where one can ask for survival regression plots at overall mean or fro each category [and its in the GUI] Obviously most people on this list are completely comfortable with scripts. BUT many of us advise people who really are only happy with the GUI dialogues, so suggestions on improving the GUI are worth noting Best Diana On 16/06/2011 20:36, "Jon K Peck" <peck@...> wrote: I always enjoy seeing helpful and polite comments from users. Plotting the dependent variable against the independents has been in the REGRESSION command for generations. /PARTIALPLOT ALL. will do all partial plots, and /PARTIALPLOT a b c will do listed variables. But the request in this case was for a plot colored by a variable that was NOT in the regression. For that the solution proposed is straightforward and appropriate IMO. Jon Peck Senior Software Engineer, IBM peck@... new phone: 720-342-5621 From: "Kornbrot, Diana" <d.e.kornbrot@...> To: Jon K Peck/Chicago/IBM@IBMUS, "SPSSX-L@..." <SPSSX-L@...> Date: 06/16/2011 01:24 PM Subject: Re: Regression in colors What spss user are requesting is that the graphics option be included in REGRESSION They know how to go all round the houses saving and then using graphics. it is a waste of our valuable time I have been making this request since the GUI 1st appeared It is INSANE to have option to plot every kind of residual, but not the main results: the dependet variable as a function of the independent variable But spss listen to users? You cannot be serious Best Diana On 16/06/2011 20:00, "Jon K Peck" <peck@... <peck@...> > wrote: Just save the predicted values to the dataset in the usual way and then use the Chart Builder to create the plot with the desired effects. Jon Peck Senior Software Engineer, IBM peck@... <peck@...> new phone: 720-342-5621 From: Hector Maletta <hmaletta@... <hmaletta@...> > To: SPSSX-L@... <SPSSX-L@...> Date: 06/16/2011 09:24 AM Subject: [SPSSX-L] Regression in colors Sent by: "SPSSX(r) Discussion" <SPSSX-L@... <SPSSX-L@...> > A student of mine is producing a simple linear regression with several predictors, including a scatterplot of actual versus predicted values of the dependent variable. He wants that the plot points are identified by colors corresponding to a categorical variable (not included in the regression) which has four values from 0 to 3. Labeling in the REGRESSION procedure prints values or value labels into the plot. How could one have cases identified in colors or different markers, with no labels or values printed? Thanks for your help Hector Professor Diana Kornbrot email: d.e.kornbrot@... <d.e.kornbrot@...> web: http://web.me.com/kornbrot/KornbrotHome.html <http://web.me.com/kornbrot/KornbrotHome.html> Work School of Psychology University of Hertfordshire College Lane, Hatfield, Hertfordshire AL10 9AB, UK voice: +44 (0) 170 728 4626 fax: +44 (0) 170 728 5073 Home 19 Elmhurst Avenue London N2 0LT, UK voice: +44 (0) 208 883 3657 mobile: +44 (0) 796 890 2102 fax: +44 (0) 870 706 4997 Professor Diana Kornbrot email: d.e.kornbrot@... web: http://web.me.com/kornbrot/KornbrotHome.html Work School of Psychology University of Hertfordshire College Lane, Hatfield, Hertfordshire AL10 9AB, UK voice: +44 (0) 170 728 4626 fax: +44 (0) 170 728 5073 Home 19 Elmhurst Avenue London N2 0LT, UK voice: +44 (0) 208 883 3657 mobile: +44 (0) 796 890 2102 fax: +44 (0) 870 706 4997 |
|
In reply to this post by Kornbrot, Diana
improvement on the GUI are certainly worthwhile.
However, I believe that it is a disservice to not teach students
that the GUI is is great way to write a _draft_ of the analysis
process. Analysis is almost always iterative in the sense of
continuously improving the output as communication. Also
using the syntax has many advantages: one does not have to rely on
short term memory to know what one did when
refining/polishing/fixing an analysis. Having students "reference"
(review) each others syntax, helps them establish the habit of
having their work cross checked. Syntax helps with the fact that one
will inevitably be interrupted while setting up an analysis.
Showing syntax facilitates getting help from other students,
instructors, help desks, and lists like this one. Using syntax
instills the habit of being sure that there is an "audit trail".
Now that many disciplines' ethics codes and grant-giving agencies
require sharing data and metadata syntax is often needed. etc. etc.
Art Kendall Social Research Consultants On 6/17/2011 6:57 AM, Kornbrot, Diana wrote: ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD
Art Kendall
Social Research Consultants |
|
|
One reason for having the graphing separate is that
the same kinds of plots would be applicable to many kinds of data
not just regression.
Art Kendall On 6/21/2011 10:29 AM, Art Kendall wrote: improvement on the GUI are certainly worthwhile. However, I believe that it is a disservice to not teach students that the GUI is is great way to write a _draft_ of the analysis process. Analysis is almost always iterative in the sense of continuously improving the output as communication. Also using the syntax has many advantages: one does not have to rely on short term memory to know what one did when refining/polishing/fixing an analysis. Having students "reference" (review) each others syntax, helps them establish the habit of having their work cross checked. Syntax helps with the fact that one will inevitably be interrupted while setting up an analysis. Showing syntax facilitates getting help from other students, instructors, help desks, and lists like this one. Using syntax instills the habit of being sure that there is an "audit trail". Now that many disciplines' ethics codes and grant-giving agencies require sharing data and metadata syntax is often needed. etc. etc.===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD
Art Kendall
Social Research Consultants |
|
I agree about syntax, not just for teaching, but for file management and transformation operations and for all basic (and even some complex) analyses. Using PASTE isn’t the same thing. When I occasionally offer advice and assistance if an interesting problem appears on the list, I usually ask the sender to let me have, in confidence, a copy of (a subset of) the *.sav file to investigate. I like to understand the data (and the research question) as well as analyse them. Some things just can’t be done in syntax (or take 50 times as long). A recent example is in the thread Recoding for more than one result sent in by Judy Harman, who was trying to generate a set of unique combinations of diagnoses from an original list of dozens of separate diagnoses entered as string variables, and with different numbers of diagnoses for each case . There's no substitute for looking at the original data and checking all transformations and analyses step by step. Even with syntax it took a while to solve, and some syntax offered by listers didn’t actually work, either because they hadn’t understood the problem or because they hadn’t seen or checked the data. Incidentally, trying to generate a set of dummy variables for MULT RESPONSE using 10**n I learned that SPSS can’t handle integers greater than 10**16, so I redid it with 2**n (which made for a not easily interpreted set of values) but I also discovered that AUTOCODE (as suggested by one lister) can’t spot typos and can yield misleading values if they are present. If anyone wants to see the final version of the syntax (sent to Judy, not the list) please ask Judy for a copy. John F Hall From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Art Kendall One reason for having the graphing separate is that the same kinds of plots would be applicable to many kinds of data not just regression. improvement on the GUI are certainly worthwhile. However, I believe that it is a disservice to not teach students that the GUI is is great way to write a _draft_ of the analysis process. Analysis is almost always iterative in the sense of continuously improving the output as communication. Also using the syntax has many advantages: one does not have to rely on short term memory to know what one did when refining/polishing/fixing an analysis. Having students "reference" (review) each others syntax, helps them establish the habit of having their work cross checked. Syntax helps with the fact that one will inevitably be interrupted while setting up an analysis. Showing syntax facilitates getting help from other students, instructors, help desks, and lists like this one. Using syntax instills the habit of being sure that there is an "audit trail". Now that many disciplines' ethics codes and grant-giving agencies require sharing data and metadata syntax is often needed. etc. etc. Thanks Jon
Obviously most people on this list are completely comfortable with scripts. BUT many of us advise people who really are only happy with the GUI dialogues, so suggestions on improving the GUI are worth noting I always enjoy seeing helpful and polite comments from users.
Professor Diana Kornbrot Professor Diana Kornbrot ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
|
|
Jon Peck Senior Software Engineer, IBM [hidden email] new phone: 720-342-5621 From: John F Hall <[hidden email]> To: [hidden email] Date: 06/21/2011 10:54 AM Subject: Re: [SPSSX-L] Regression in colors Sent by: "SPSSX(r) Discussion" <[hidden email]> [snip] Incidentally, trying to generate a set of dummy variables for MULT RESPONSE using 10**n I learned that SPSS can’t handle integers greater than 10**16, so I redid it with 2**n John F Hall >>>The 10**n issue has nothing really to do with SPSS. Numbers are represented in Statistics using IEEE double precision floating point, and that gives you only 53 or so bits for a mantissa. Using powers of 2 gives you the ability to hold up to 53 dichotomies in a one SPSS variable. In fact, the functions packDummies and extractDummies in the extendedTransforms.py programmability module are designed to let you store large numbers of dummy variables in this more compact representation. One of these days I intend to package that in a friendlier facility with its own little codebook to make it easy to mange such data. Regards, Jon Peck johnfhall@... www.surveyresearch.weebly.com From: SPSSX(r) Discussion [[hidden email]] On Behalf Of Art Kendall Sent: 21 June 2011 16:59 To: [hidden email] Subject: Re: Regression in colors One reason for having the graphing separate is that the same kinds of plots would be applicable to many kinds of data not just regression. Art Kendall On 6/21/2011 10:29 AM, Art Kendall wrote: improvement on the GUI are certainly worthwhile. However, I believe that it is a disservice to not teach students that the GUI is is great way to write a _draft_ of the analysis process. Analysis is almost always iterative in the sense of continuously improving the output as communication. Also using the syntax has many advantages: one does not have to rely on short term memory to know what one did when refining/polishing/fixing an analysis. Having students "reference" (review) each others syntax, helps them establish the habit of having their work cross checked. Syntax helps with the fact that one will inevitably be interrupted while setting up an analysis. Showing syntax facilitates getting help from other students, instructors, help desks, and lists like this one. Using syntax instills the habit of being sure that there is an "audit trail". Now that many disciplines' ethics codes and grant-giving agencies require sharing data and metadata syntax is often needed. etc. etc. Art Kendall Social Research Consultants On 6/17/2011 6:57 AM, Kornbrot, Diana wrote: Thanks Jon Apologies for venting my frustrations with the DIALOGUE interface to REGRESSION on the list 1. if it has been easily available in script for yonks, why not implement in GUI dialogue? I find the GUI mostly very powerful & much easier for students to learn. Makes the glitches more frustrating 2. can the script do interactions? E.g. /partial plot a a*b, where b is categorical. If so it would be possible t run REGRESSIONS once without the categorical for statistics and once with to get plots. There is a nice implementation in survival plots, where one can ask for survival regression plots at overall mean or fro each category [and its in the GUI] Obviously most people on this list are completely comfortable with scripts. BUT many of us advise people who really are only happy with the GUI dialogues, so suggestions on improving the GUI are worth noting Best Diana On 16/06/2011 20:36, "Jon K Peck" <peck@...> wrote: I always enjoy seeing helpful and polite comments from users. Plotting the dependent variable against the independents has been in the REGRESSION command for generations. /PARTIALPLOT ALL. will do all partial plots, and /PARTIALPLOT a b c will do listed variables. But the request in this case was for a plot colored by a variable that was NOT in the regression. For that the solution proposed is straightforward and appropriate IMO. Jon Peck Senior Software Engineer, IBM peck@... new phone: 720-342-5621 From: "Kornbrot, Diana" <d.e.kornbrot@...> To: Jon K Peck/Chicago/IBM@IBMUS, "SPSSX-L@..." <SPSSX-L@...> Date: 06/16/2011 01:24 PM Subject: Re: Regression in colors What spss user are requesting is that the graphics option be included in REGRESSION They know how to go all round the houses saving and then using graphics. it is a waste of our valuable time I have been making this request since the GUI 1st appeared It is INSANE to have option to plot every kind of residual, but not the main results: the dependet variable as a function of the independent variable But spss listen to users? You cannot be serious Best Diana On 16/06/2011 20:00, "Jon K Peck" <peck@... <peck@...> > wrote: Just save the predicted values to the dataset in the usual way and then use the Chart Builder to create the plot with the desired effects. Jon Peck Senior Software Engineer, IBM peck@... <peck@...> new phone: 720-342-5621 From: Hector Maletta <hmaletta@... <hmaletta@...> > To: SPSSX-L@... <SPSSX-L@...> Date: 06/16/2011 09:24 AM Subject: [SPSSX-L] Regression in colors Sent by: "SPSSX(r) Discussion" <SPSSX-L@... <SPSSX-L@...> > A student of mine is producing a simple linear regression with several predictors, including a scatterplot of actual versus predicted values of the dependent variable. He wants that the plot points are identified by colors corresponding to a categorical variable (not included in the regression) which has four values from 0 to 3. Labeling in the REGRESSION procedure prints values or value labels into the plot. How could one have cases identified in colors or different markers, with no labels or values printed? Thanks for your help Hector Professor Diana Kornbrot email: d.e.kornbrot@... <d.e.kornbrot@...> web: http://web.me.com/kornbrot/KornbrotHome.html <http://web.me.com/kornbrot/KornbrotHome.html> Work School of Psychology University of Hertfordshire College Lane, Hatfield, Hertfordshire AL10 9AB, UK voice: +44 (0) 170 728 4626 fax: +44 (0) 170 728 5073 Home 19 Elmhurst Avenue London N2 0LT, UK voice: +44 (0) 208 883 3657 mobile: +44 (0) 796 890 2102 fax: +44 (0) 870 706 4997 Professor Diana Kornbrot email: d.e.kornbrot@... web: http://web.me.com/kornbrot/KornbrotHome.html Work School of Psychology University of Hertfordshire College Lane, Hatfield, Hertfordshire AL10 9AB, UK voice: +44 (0) 170 728 4626 fax: +44 (0) 170 728 5073 Home 19 Elmhurst Avenue London N2 0LT, UK voice: +44 (0) 208 883 3657 mobile: +44 (0) 796 890 2102 fax: +44 (0) 870 706 4997 ===================== To manage your subscription to SPSSX-L, send a message to LISTSERV@... (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
| Free forum by Nabble | Edit this page |